category stringclasses 1 value | source stringclasses 1 value | url stringlengths 74 77 | text stringlengths 16 2.15M |
|---|---|---|---|
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12676282&blobtype=pdf | # Diagnostic performance of a RHAM-based point-of-care test for Mycobacterium tuberculosis
Xiang Chen, Vincenzo Lai, Leonardo Sechi, Paola Molicotti
## Abstract
Tuberculosis (TB) remains a global health crisis, hampered by significant diagnostic delays, particularly for extrapulmonary TB and in resource-limited settings. The development of point-of-care tests (POCTs) meeting the WHO's ASSURED criteria is crucial. This prospective laboratory-based study evaluated the diagnostic performance of a novel, affordable POCT based on RNase Hybridization-Assisted Amplification (RHAM) technology for detecting Mycobacterium tuberculosis complex. The test was evaluated using a variety of clinical specimens collected consecutively from suspected TB patients, compared against standard methods (PCR, Microscopy, culture). The RHAM-based POCT demonstrated promising sensitivity of 83.3% (10/12; 95% CI: 50.9-97.1%) and a specificity of 100% (25/25; 95% CI: 83.4-100%). All five non-tuberculous mycobacteria samples were correctly identified as negative. The two false-negative results occurred in samples with very high PCR cycle threshold values (>36), suggesting detection challenges in paucibacillary specimens. The test exhibited a rapid average turnaround time of 18 min and requires minimal infrastructure, operating via a portable, low-power consumption device, even compatible with mobile phone or car chargers. Its closed-cartridge system enhances biosafety by minimizing aerosol generation. Furthermore, the estimated cost per test is substantially lower than leading commercial molecular assays. This study indicates that the RHAM-based POCT is a rapid, user-friendly, and cost-effective diagnostic tool with high specificity. Its ability to function with diverse specimen types positions it as a potential game-changer for TB diagnosis in field and resource-poor environments, though larger-scale studies are warranted to confirm sensitivity, especially in low-bacterial-load scenarios.
Tuberculosis (TB) remains one of the most pressing global public health threats, with the World Health Organization (WHO) reporting an annual incidence of over 11 million cases worldwide (1). As the leading infectious disease killer, TB claims approximately 1.3 million lives each year, surpassing even HIV/AIDS and malaria in mortality rates (2). A particularly alarming issue is that nearly 30% of TB cases are undiagnosed, which makes them a potential source of infection on the move (1,2). This diagnostic gap not only leads to treatment delay but also increases the risk of transmission, especially in high-burden regions with limited The only exclusion criterion was an insufficient sample volume for performing both the POCT and all reference standard tests. The samples consisted of sputum (n = 9), tongue swab (n = 1), bronchial aspiration (n = 11), bronchoalveolar lavage (n = 3), gastric aspiration (n = 2), feces (n = 2), cerebrospinal fluid (n = 1), urine (n = 1), and biopsy (n = 1). Five extra non-tuberculous mycobacteria (NTM) samples were included for further confirmation of specificity. All the subjects were informed that their samples would be used for Mycobacterium identification. The POCT kit used for TB identification is from Pluslife Biotech Co., LTD. (Guangzhou, China). The test kit is based on the RNase hybridization-assisted amplification (RHAM) technology, a nucleic acid amplification method that utilizes specific probes for the detection of Mycobacterium tuberculosis complex DNA. The analyzer is a compact device (dimensions: 101 mm * 91 mm * 65 mm; weight: 210 g). Just a mobile phone charger or a 12 V car charger can power the portable device, making it suitable for use in field settings with unstable or no grid electricity. The entire process, from sample loading to result generation, is fully integrated into a single-use, closed test card. As shown in Figure 1A, sputum and tongue swabs are the manufacturer's recommended specimens. The sample underwent a 5-min thermostatic vortex mixing before being transferred to a card. The result was generated in 10 to 25 min after inserting the test card into the device, and wireless output to a personal computer or mobile device would be available. Traditional TB identification methods are performed as well, including microscopy smear examination after acid-fast staining, culturing, and polymerase chain reaction (PCR). Lowenstein-Jensen (L-J) solid medium and Mycobacterium growth indicator tube (MGIT) liquid medium were used for culturing. The Anyplex™ MTB/NTMe Real-time Detection Kit from Seegene Inc. (Seoul, Korea) was applied to PCR.
The POCT showed strong performance in TB diagnosis, across nine sample types beyond manufacturer-recommended specimens (sputum or tongue swab), with only two false negatives derived from sputum and urine (Figure 1B). As for specificity, it is 100% using sputum (4/4), bronchial aspiration (10/10), bronchoalveolar lavage (3/3), gastric aspiration (1/1), feces (1/1), biopsy (1/1), and NTM samples (5/5). The overall sensitivity and specificity of the POCT were 83.3% (10/12, 95% CI, 50.9-97.1%) and 100% (25/25, 95% CI, 83.4-100%), respectively. Statistical analysis using McNemar's test showed no significant difference in detection rates between the POCT and PCR, or microscopy, or culture(p > 0.05). The average turnaround time (TAT) is 18 min from sample pretreatment to results output. About the two false negative samples, they exhibited high PCR Cycle threshold (Ct) values (36.94 and 39.34), while the average Ct value of the other positive samples was 25.65. Notably, the average Ct value of using colonies from the culture medium (18.96) is significantly lower than clinical specimens (27.13) in our daily tests (p < 0.01). Thus, we suspected that a low pathogen load in sample underlie detection failure and the pathogen count might be below the limit of detection specified by the manufacturer (50 CFU/mL).
This study demonstrates the promising potential of RHAMbased POCT in TB diagnosis, showing promising sensitivity and high observed specificity across a broad range of clinical specimens, aligning with WHO's ASSURED criteria for sensitive, specific, and user-friendly. Its rapid TAT (18 min average) and minimal infrastructure requirements position it as a potential solution for resource-limited settings where diagnostic delays perpetuate transmission. Two false-negative results were both associated with very high Ct values, which indicates the correlation between pathogen load and the accuracy of POCT results. This also highlights a common challenge in TB diagnostics, particularly for paucibacillary disease, which is more frequent in children, people living with HIV, and cases of extrapulmonary TB (4). To enhance the sensitivity of the POCT in such scenarios, strategies such as pre-treatment of samples with centrifugation or filtration to concentrate bacilli, or technological refinement of the amplification method to lower the limit of detection, could be explored. While sensitivity requires optimization for paucibacillary samples, the test's robustness across varied specimens is encouraging. The POCT's versatility beyond sputum samples expands its practical applicability and also addresses the critical need for extrapulmonary TB detection. Our cost-analysis indicates that the list price of this POCT system (~3 USD per test) is more affordable than that of widely used commercial molecular tests, such as Xpert MTB/RIF (~10 USD per test), potentially enhancing its accessibility in resource-limited settings (5,6). In terms of biosafety, the RHAM-based POCT offers a notable advantage. Similar to the GeneXpert system, the test utilizes a closed-cartridge system that minimizes the need for manual sample manipulation after initial loading, thereby reducing the risk of generating infectious aerosols compared to open-batch processing such as traditional PCR or smear microscopy. Furthermore, the device's low power consumption and flexibility in power sources align with the "Equipment-free" and "Deliverable" aspects of the ASSURED criteria, addressing a critical challenge for diagnostic deployment in remote areas. This RHAM-based POCT has demonstrated its advantages in the detection of SARS-CoV-2 and other pathogen (7,8). While this study also demonstrates the promising performance of the RHAM-based POCT across a variety of sample types, a key limitation is the relatively small sample size, particularly for certain specimen types. To further validate and generalize these findings, future studies involving larger cohorts from tuberculosis-endemic regions are warranted. In summary, our findings support the integration of this POCT into TB diagnostic workflows, especially in field settings where conventional methods are impractical.
## References
1. Who (2024)
2. Trajman, Campbell, Kunor et al. (2025) *Tuberculosis. Lancet*
3. Hong, Lee, Menon et al. (2022) "Point-of-care diagnostic tests for tuberculosis disease" *Sci Transl Med*
4. Basu, Chakraborty (2025) "A comprehensive review of the diagnostics for pediatric tuberculosis based on assay time, ease of operation, and performance. Microorganisms"
5. Andama, Whitman, Crowder et al. (2022) "Accuracy of tongue swab testing using Xpert MTB-RIF ultra for tuberculosis diagnosis" *J Clin Microbiol*
6. Mokaddas, Ahmad, Eldeen (2019) "GeneXpert MTB/RIF is superior to BBD max MDR-TB for diagnosis of tuberculosis (TB) in a country with low incidence of multidrug-resistant TB (MDR-TB)" *J Clin Microbiol*
7. Herrmann, Breuer, Duc et al. (2024) "Comparison of the diagnostic accuracy of the Pluslife Mini dock RHAM technology with Abbott ID now and Cepheid GenXpert: a retrospective evaluation study" *Sci Rep*
8. Charfi, Guyonnet, Untrau et al. (2024) "Performances of two rapid LAMP-based techniques for the intrapartum detection of group B Streptococcus vaginal colonization" *Ann Clin Microbiol Antimicrob* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12673353&blobtype=pdf | # Geographic Expansion of Oropouche Fever Into Non-Endemic Regions of Brazil: Implications for Health Surveillance
Rafael Pedro De Souza Nascimento, Paula Esbaltar De Oliveira, Adeilton Gonçalves Da Silva Junior, Riana Oliveira, Fernandes Amorim, Cleison Keulys, Santos Silva, Daniela Santos, Lethícia Lohane, Almeida Barros, Hernany Fabricio De Novaes Menezes, | Anderson Da Costa, Rodrigo Feliciano Do Carmo, | Carlos, Dornels Freire De Souza, Carlos Dornels, Freire De Souza
## Abstract
Oropouche fever is an arboviral disease caused by the Oropouche virus (OROV; Peribunyaviridae family), first identified in Brazil in 1960. This study aims to describe the epidemiological profile and spatio-temporal dispersion of Oropouche fever cases across Brazil in 2024. The variables analyzed were sex, age, transmission area, macroregion, and federative unit. A simple descriptive analysis (absolute frequencies and proportions) was performed. Data were extracted from the Oropouche Fever Epidemiological Panel and the Brazilian Institute of Geography and Statistics (IBGE). The analysis revealed 13,786 reported cases (6.80/100,000 inhabitants), with a predominance of males (52.6%; n = 7247). The North region presented the highest incidence (41.60%; n = 5732; 33.03/100,000 inhabitants). Incidence peaks during early (weeks 5-6) and late 2024 (week 52) were identified, with heterogeneous regional spread reflecting a possible outbreak progression from the North to Northeast and subsequently Southeast. Espírito Santo state recorded the highest national incidence (151.60/100,000 inhabitants; 42.20% of cases; n = 5812). The expansion of OROV into non-endemic regions highlights an emerging public health threat. Urgent, systematic measures are required to strengthen Brazil's health surveillance system, ensuring timely and effective responses to mitigate further geographic spread.
| IntroductionOropouche fever is an arboviral disease caused by the Oropouche virus (OROV; Peribunyaviridae family), first isolated in Trinidad and Tobago in 1955 [1] and subsequently identified in Brazil in 1960 [2]. Since its initial detection, recurrent outbreaks have been documented across northern and northeastern Brazilian states. The first major epidemic occurred in 1961 in Pará, with ~11,000 reported cases, followed by subsequent outbreaks in Acre, Amazonas, Maranhão, and other northern states [3]. Oropouche fever infection outbreaks have been also reported in other Latin American countries, including Peru, Colombia, Ecuador and Venezuela [4]. Its occurrence often coincides with outbreaks of other clinically similar arboviral diseases, complicating differential diagnosis [1].Globally, over 500,000 OROV infections have been reported, most of them in the Americas region [1]. In 2024 alone,This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
approximately 16,000 cases were recorded across the American continent, with Brazil accounting for more than 80% of these infections; the remaining cases were distributed among Bolivia, Colombia, Cuba, Peru and Dominican Republic [5]. Recent evidence indicates that the current expansion of OROV into extra-Amazonian regions is being driven predominantly by the reassorted M1L2S2 lineage, which differs substantially from the historical Amazonian genotypes. This lineage has been associated with large-scale transmission in the Atlantic Forest biome, particularly in Espírito Santo, where strains belonging to the 2022-2024 Amazon lineage were recently identified [6].
As a notifiable disease in Brazil [2], surveillance data revealed approximately 1,000 confirmed human cases occurring between 2023 and early 2024 in Amazonas state. In February 2024, the Pan American Health Organization (PAHO) issued an alert urging member states to enhance surveillance of Oropouche fever cases [7]. Importantly, the emergence of severe clinical presentations coincided with this genomic shift: in 2024, the first confirmed fatal cases of OROV infection were documented in non-endemic areas of Brazil [8], and subsequent investigations have provided further insight into the pathogenesis of severe OROV disease caused by contemporary reassortant strains [9]. By the end of 2024, Brazil reported over 13,000 cases nationwide [10].
Moreover, the first neuroinvasive infection attributed to the M1L2S2 lineage was recently reported during the country's largest recorded epidemic [11]. Together, these findings underscore the epidemiological and clinical relevance of the emergent reassorted lineage and highlight the need for updated surveillance and genomic monitoring across Brazil.
## 2 | Materials and Methods
## 2.1 | Study Design, Population and Period
The present study is an ecologic study involving all notified cases of Oropouche fever in Brazil during 2024.
## 2.2 | Study Setting
Brazil comprises five macroregions (North, Northeast, Southeast, South, and Central-West) and 27 federative units (including the Federal District). Brazil has a total area of around 8.5 million km², the fifth-largest globally [12].
## 2.3 | Data Source and Variables
Data were extracted from the publicly available Oropouche Fever Epidemiological Panel (https://www.gov.br/saude/pt-br/assuntos/ saude-de-a-a-z/o/oropouche/painel-epidemiologico) [10]. Population data used for incidence calculations were sourced from the Brazilian Institute of Geography and Statistics (IBGE).
The variables included were: sex (male and female), age (0-4; 5-9; 10-19; 20-59; ≥ 60 years), transmission area (endemic and non-endemic regions), macroregion (North, Northeast, Southeast, Central-West and South), state and epidemiologic week of diagnosis (weeks 1 to 52). The primary morbidity indicator used was the incidence rate (IR), calculated as: It was assumed that the population of each location remained constant throughout the epidemiologic weeks. Age categories were defined as 0-4, 5-9, 10-19, 20-59, and ≥ 60 years. The broad 20-59 year category was selected to encompass the entire young adult and middle-aged population.
## 2.4 | Statistical Analysis
Descriptive analyses were performed using absolute frequencies and proportions. Temporal trends and spatial distributions were visualized via time-series graphs and maps generated using: JASP (version 0.16.1.0, University of Amsterdam/Amsterdam, The Netherlands) and QGIS (version 2.14.11, Open Source Geospatial Foundation (OSGeo), Beaverton, OR, USA).
## 2.5 | Ethical Aspects
As this study used aggregated, publicly available data, ethics approval was waived per Resolution 510/2016 of Brazil's National Research Ethics Committee (CONEP).
## 3 | Results
The analysis revealed 13,786 reported cases (6.80/100,000 inhabitants) in 2024, with a predominance of males (52.6%; n = 7247; 7.36/100,000 inhabitants). The North region presented the highest incidence (41.60%; n = 5732; 33.03/100,000 inhabitants), followed by the Southeast (45.40%; n = 6258; 7.40/ 100,000 inhabitants). The Central-West had the lowest incidence (0.4%; n = 53; 0.33/100,000 inhabitants) (Table 1).
Cases occurred year-round, with two distinct peaks: early (weeks 5 and 6) and late 2024 (weeks 49 to 52), with heterogeneities among regions. The North region peaked earlier, during weeks 4 (n = 537) and 8 (n = 469), then declined. Its annual incidence (33.03/100,000) was 4.5 times higher than the Southeast's (7.40/100,000). After the North's decline, cases rose in the Northeast (weeks 15 to 33), shifted to the Southeast (week 41 onward). The Central-West and South had only sporadic and isolated cases of the disease (Figures 1 and2).
Six states exceeded Brazil's incidence rate (6.80/100,000), five in the North. Espírito Santo, despite being in the Southeast, was the most affected state (151.60/100,000; 42.20%; n = 5812), followed by Rondônia (105.40/100,000; 12.10%; n = 1667). Other states with high burdens included Amazonas (80.50/100,000), Roraima (45.40/100,000), and Acre (34.50/100,000). Outside endemic areas, only Espírito Santo surpassed the national rate (Figure 3). Despite bordering northern states, the Central-West region did not show high incidence rates, with 5, 39 and 8 cases reported in Goiás, Mato Grosso and Mato Grosso do Sul, respectively, and only one case in the Federal District. Conversely, Maranhão (0.60/100,000 inhabitants), Paraíba (0.33/100,000 inhabitants) and Rio Grande do Norte (0.10/100,000 inhabitants) presented lower rates compared to neighboring states. In the South, Santa Catarina (2.40/100,000) showed a marked contrast with Rio Grande do Sul (0.02/100,000) and Paraná (0.10/100,000) (Figure 3).
## 4 | Discussion
This study aimed to describe the epidemiological profile and spatio-temporal dispersion of Oropouche fever cases across Brazil in 2024. Cases predominated among males and in the North region (notably Amazonas and Rondônia). Non-endemic areas, particularly the Southeast (carried by Espírito Santo), also exhibited significant incidence, with distinct regional peaks in case notifications.
The observed epidemiological patterns are consistent with a 2022 cohort study conducted in border regions of the Amazon, in which 59.26% of cases occurred among young adults [13]. The geographic findings corroborate a 2023 review on the epidemiology and molecular biology of OROV, which highlights the North and Northeast as historical epicenters of OROV transmission, reporting the highest incidences since the virus was first described [14]. More recently, however, in 2024, the Southeast has emerged as a region of incresing epidemiological relevance.
Urbanization, deforestation [15,16], and the effects of climate change [15] are known to amplify the proliferation of the biting midge vector Culicoides paraensis [17], thereby increasing the risk of re-emergence and expansion of the virus in the Amazon region [15]. Genetic factors should also be considered; a review study published in 2017 suggests that genetic shifts in OROV strains may have enhanced virulence or host adaptability [18]. These factors collectively contribute to the rising case burden in Brazil.
Beyond ecological and demographic drivers, the contemporary expansion of OROV appears to be strongly influenced by the dissemination of the reassorted M1L2S2 genotype, which has rapidly replaced earlier circulating strains in several regions. This lineage has been implicated not only in widespread transmission but also in the emergence of more severe disease phenotypes. In 2024, Brazil documented the first fatal OROV infections in a non-endemic region, a clinical development not previously observed in over six decades of virus circulation [8]. Subsequent analyses have begun to unravel possible mechanisms underlying these severe outcomes, suggesting altered viral-host interactions and enhanced pathogenic potential in reassortant strains [9]. Such findings reinforce the likelihood that genomic evolution is contributing materially to the observed changes in epidemiological dynamics.
This investigation recorded an increase in OROV cases in the Northeast, Central-West and Southeast, which are considered non-endemic. Regarding the occurrence of cases in the Central-West and Southeast, a 2023 review on the Oropouche Fever dynamics pointed out that, since the early 2000s, OROV has been detected in nonhuman primates in the Southeast. Additionally, intensified human migration, along with improvements in transportation and tourism infrastructure, may have facilitated the spread of the virus to non-endemic regions [14].
The Northeast region and key states-notably Espírito Santo (Southeast) and Santa Catarina (South)-emerged as significant foci of extra-Amazonian transmission. These areas were reported in another study [19], and likely became hotspots due to a combination of immunologically susceptible populations and demographic factors (e.g., high population density, which facilitates efficient viral spread-a mechanism consistent with prior findings) [14]. Additionally, avian-mediated viral dispersal via migratory birds has been proposed as another potential dissemination pathway [15].
The clinical spectrum associated with the current epidemic has also broadened. Most notably, a neuroinvasive OROV infection associated with the M1L2S2 lineage was recently confirmed in a patient living with HIV in extra-Amazonian Brazil, marking the first documented case of central nervous system involvement by this emergent reassortant strain [11]. This case occurred amid the largest OROV epidemic ever recorded in the country and raises important questions regarding host susceptibility, viral determinants of neurotropism, and the potential for more severe disease in high-risk populations. As such, these observations further emphasize the significance of incorporating genomic and clinical data into routine surveillance frameworks, particularly in regions undergoing recent viral introduction.
Since May 2024, PAHO has issued alerts regarding unprecedented autochthonous transmission in extra-Amazonian regions [20]. Given the virus's demonstrated adaptability [14], these warnings emphasized the urgent need for enhanced health surveillance and vector control measures [20]. A subsequent PAHO update in September 2024 further reinforced this position, highlighting the critical importance of preventive surveillance, improved laboratory diagnostics, and optimized clinical management [5]. There are, so far, no approved vaccines or specific therapeutics for OROV infection. In response, the Brazilian government has adopted compulsory disease notification policies to enable rapid investigation and transmission chain interruption [21]. Complementing these measures, environmental preservation strategies targeting the disease's ecological determinants [15,16] should be prioritized to control the current outbreak and mitigate future epidemic risks.
Despite rigorous methodological care, this study has limitations. First, secondary surveillance data may be subject to underreporting, particularly given: (1) symptom overlap with cocirculating arboviruses [20,22,23], and (2) healthcare system strain during peak transmission periods [23]. Second, the lack of epidemiological studies on OROV restricts our ability to conduct robust temporal or spatial comparisons of disease evolution.
## 5 | Conclusion
A high incidence was observed in 2024, concentrated in the North region among adult males, followed by a spatial shift characterized by declining incidence in the North and increasing transmission in the Southeast and Northeast, along with sporadic cases in the South and Central-West.
These findings underscore the urgent need for a tiered public health response: (1) strengthening integrated surveillance networks with real-time data sharing across all government levels (municipality, state and federation); (2) expanding diagnostic capacity through decentralized testing and active case-finding; and (3) optimizing resource allocation to improve outbreak containment. These measures are required to strengthen Brazil's health surveillance system, ensuring timely and effective responses to mitigate further geographic spread.
## References
1. (2024) "The Lancet Infectious Diseases Oropouche Fever, the Mysterious Threat" *Lancet Infectious Diseases*
2. Brazil (2024) "Ministry of Health. Oropouche"
3. Dias, Santos, Pauvolid-Corrêa (2022) "An Overview of Neglected Orthobunyaviruses in Brazil" *Viruses*
4. Mohapatra, Mishra, Satapathy et al. (2024) "Surging Oropouche Virus (OROV) Cases in the Americas: A Public Health Challenge" *New Microbes and New Infections*
5. (2024) "Epidemiological Update Oropouche in the Americas Region -6 2024"
6. Delatorre, De Mendonça, Gatti (2025) "Emergence of Oropouche Virus in Espírito Santo State, Brazil, 2024" *Emerging Infectious Diseases*
7. (2024) "Epidemiological Alert -Oropouche in the Region of the Americas"
8. Bandeira, Pereira, Leal (2024) "Fatal Oropouche Virus Infections in Nonendemic Region, Brazil, 2024"
9. Có, De Mendonça, Gatti (2025) "Unravelling the Pathogenesis of Oropouche Virus" *Lancet Infectious Diseases*
10. Brazil (2024) "Painel Epidemiológico Oropouche"
11. Gatti, Nodari, Sousa (2025) "Neuroinvasive Oropouche Virus in a Patient With HIV From Extra-Amazonian Brazil" *Lancet Infectious Diseases*
12. Brazil (2022) "Ministry of Foreign Affairs"
13. Moreira, Sgorlon, Queiroz (2024) "Outbreak of Oropouche Virus in Frontier Regions In Western Amazon" *Microbiology Spectrum*
14. Sakkas, Bozidis, Franks et al. (2018) "Oropouche Fever: A Review" *Viruses*
15. Lorenz, Chiaravalloti-Neto (2024) "Brazil Reports an Increased Incidence of Oropouche and Mayaro Fever in the Amazon Region" *Travel Medicine and Infectious Disease*
16. Romero-Alvarez, Escobar, Auguste et al. (2023) "Transmission Risk of Oropouche Fever Across the Americas" *Infectious Diseases of Poverty*
17. Sah, Srivastava, Kumar (2024) "Oropouche Fever Outbreak In Brazil: an Emerging Concern in Latin America" *Lancet Microbe*
18. Da Rosa, De Souza, De (2017) "Oropouche Virus: Clinical, Epidemiological, and Molecular Aspects of a Neglected Orthobunyavirus" *American Journal of Tropical Medicine and Hygiene*
19. Fujita, Salvador, Da et al. (2024) "Oropouche in Brazil in 2024" *Journal of Travel Medicine*
20. (2024) "Epidemiological Alert -Oropouche in the Region of the Americas -9"
21. Brazil (2024) "Nota Técnica no 6/ 2024-CGARB"
22. De Lima, Dias, De Souza (2024) "Oropouche Virus Exposure in Febrile Patients During Chikungunya Virus Introduction in the State of Amapá, Amazon Region, Brazil" *Pathogens*
23. Martins-Filho, Soares-Neto, De Oliveira-Júnior et al. (2024) "The Underdiagnosed Threat of Oropouche Fever Amidst Dengue Epidemics In Brazil" *Lancet Regional Health -Americas* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12400139&blobtype=pdf | # Key lessons from the COVID-19 pandemic: Role of intensive care, politics and science communication (Review)
Ioannis Mammas, Michalis Agrafiotis, Kazani, Chryssie Koutsaftiki, Alexia Papatheodoropoulou, Simon Drysdale, Maria Theodoridou, Demetrios Spandidos, Spandidos
## Abstract
the post-coronavirus disease 2019 (CoVID-19) era calls for a comprehensive analysis of the recent CoVID-19 pandemic to extract important lessons for the international scientific community for the improvement of its readiness towards future pandemic threats and challenges. the present review article presents key aspects of the CoVID-19 pandemic, with the main topics covered being the following: i) the recent advances in intensive care, focusing particularly on high flow nasal oxygen therapy; ii) COVID-19 and politics; and iii) CoVID-19 and science communication. Both medical aspects of the CoVID-19 pandemic as well as non-medical issues, including politics and science communication, should be further evaluated and be definitely included in future medical educational programs, worldwide. Contents 1. Introduction 2. High flow nasal oxygen therapy: Not just another oxygen delivering modality 3. CoVID-19 and politics: Challenges, dilemmas and lessons 4. CoVID-19 and science communication 5. Conclusions
## 1. Introduction
the level of preparedness of the scientific community for the next pandemic remains a critical concern. the ways in which the international scientific community can contribute to minimizing the public health impact of a new pandemic require careful consideration. the evaluation of the recent coronavirus disease 2019 (CoVID-19) pandemic is indeed crucial (1)(2)(3)(4)(5)(6). Belonging to the broad family of coronaviruses, a well-known family of viruses to the paediatric population, severe acute respiratory syndrome coronavirus 2 (SarS-CoV-2), a positive-sense single-stranded rNa (+ssrNa) virus, emerged as one of the most dangerous pathogens in human history (2). the simple structure of the virus, typical of rNa viruses, such as influenza viruses, human immunodeficiency virus (HIV) and cancer-associated viruses, hindered the ability of the immune system to identify its invasion (2,7). In addition, global genetic variations influenced morbidity and mortality rates related to CoVID-19 (2). the global distribution of SarS-CoV-2, which since the end of 2019 spread immediately around the globe, causing an enormous health and economic catastrophe (https://ourworld-indata.org/grapher/cumulative-covid-cases-region and https://ourworldindata.org/grapher/cumulative-covid-deaths-region) is presented in Fig. 1.
From the very beginning of the pandemic, there was a critical need to develop secure, reliable and effective vaccines and therapeutic agents against SarS-CoV-2 (2,8). lockdown and social distancing significantly influenced social and behavioral aspects of human life. Concurrently, CoVID-19 affected the prevalence of other diseases, including cardiovascular diseases and cancer, while in tropical and subtropical areas of the world, COVID-19 lockdown resulted in a significant reduction in the rates of other infections, such as the Dengue fever (9). the mutations of SarS-CoV-2 resulted in the continuing emergence of new CoVID-19 cases in all age groups, including children, for several months (7). SarS-CoV-2 infection remains prevalent, while efforts to fund research on CoVID-19 have continued to the present day. research has also focused on post-CoVID-19 syndrome and its management (10).
Medical advances and improvements, as well as management limitations, weaknesses and challenges, encountered during the CoVID-19 pandemic, require an up-to-date evaluation. In addition to the medical aspects of the pandemic and issues pertaining to strategic preparedness and response planning (1)(2)(3)(4)(5), it is also important to systematically analyze non-medical issues, including politics and science communication. the lessons derived from this evaluation will help the prioritisation of research and strategic planning in the event of a future pandemic. Moreover, this analysis will guide the development of up-to-date educational programs to be integrated in both undergraduate and postgraduate medical training, worldwide.
the purpose of the present review article is to summarize the key messages on the lessons learnt from the recent CoVID-19 pandemic, one of the most critical and disruptive events of modern times. the main topics on CoVID-19 that are discussed herein are: i) advances in intensive medicine during the COVID-19 pandemic, focusing on high flow nasal oxygen therapy (HFNOT); ii) COVID-19 and politics; and iii) CoVID-19 and science communication (table I).
## 2. High flow nasal oxygen therapy: Not just another oxygen delivering modality
During the recent CoVID-19 pandemic, intensive care medicine came to the forefront of the fight against SARS-CoV-2. thus far, the learning experience has been intense and has affected every aspect of this medical specialty, from therapeutic tools to management strategies and protocols. HFNot is a relatively novel method for delivering warm humidified oxygen at high flows to patients with acute hypoxaemic respiratory failure (11). the interest in HFNot and its potential has been further increased during the CoVID-19 pandemic, which imposed significant demands on hospital resources, necessitating prudent patient prioritization and careful allocation of respiratory care equipment and intensive care unit (ICU) beds (12).
the HFNot system setup is simple: It requires only a flow generator, an active heated humidifier, a single heated circuit with a servo-controlled heating wire, and a silicone nasal cannula (11). HFNot has emerged as an effective and well-tolerated respiratory support technique in various clinical scenarios, although the optimal method for managing acute hypoxaemic respiratory failure remains under debate. Physiological studies have demonstrated that HFNot, apart from being an effective oxygenator, reduces the work of breathing and respiratory resistance, increases positive end-expiratory pressure and end-inspiratory lung volume, washes off anatomic dead space and improves secretion clearance (12). a common practice is to start HFNot with a fraction of inspired oxygen (Fio 2 ) of 100% and a flow of 60 l/min and then adjust FiO 2 and the flow to achieve an oxygen saturation (Spo 2 ) >88-90% and an age-appropriate respiratory rate (rr) (13). a roX index (calculated as the Spo 2 /Fio 2 ratio divided by the rr of the patient) >4.88 at 12 h has a high positive predictive value (89.4%) in predicting treatment success (14). the only absolute contraindication for HFNot is any indication for invasive mechanical ventilation including shock, respiratory and cardiac arrest, bradycardia, severe arrhythmias and an impaired level of consciousness. Facial erythema, skin breakdown and barotrauma may occur in HFNot users, although these represent less common complications compared with non-invasive ventilation (NIV). overall, HFNot is better tolerated than NIV (15).
Prior to the CoVID-19 era, Hernández et al (16,17) demonstrated that HFNot compared with conventional oxygen therapy (Cot) decreases the risk of reintubation and post-extubation respiratory failure in 'low-risk' ICU patients, while in 'high-risk' patients HFNot was not inferior to NIV in averting reintubation and post-extubation respiratory failure. However, neither of these two studies (16,17) noted any benefit in terms of mortality rates. In a meta-analysis by Zhu et al (18) that followed, HFNot reduced the risk of post-extubation respiratory failure, improved oxygenation and reduced respiratory rates in post-extubated ICU patients. Moreover, in another meta-analysis by Granton et al (19), HFNot reduced re-intubation rates compared with Cot, but not when compared with NIV. However, other researchers have failed to duplicate these findings (20). thus, in another meta-analysis by Maitra et al (21) comparing HFNot with NIV and Cot in patients with acute hypoxaemic respiratory failure, no benefit was shown for HFNOT in decreasing requirements for higher respiratory support. Nevertheless, more recently Seow et al (22) reviewed a total of 63 studies [including 23 randomized controlled trials (rCts)], which compared HFNot with Cot and showed that HFNot decreased the risk for escalating to NIV or invasive respiratory support. In the paediatric population, several rCts have suggested that compared with Cot, HFNot reduced the rates of intubation and mechanical ventilation in children with moderate-to-severe bronchiolitis and hypoxaemic respiratory failure (23)(24)(25)(26). HFNot is a growing respiratory treatment for children, particularly for those with respiratory distress, bronchiolitis, or other respiratory illnesses.
Focusing on acute hypoxaemic respiratory failure in patients with CoVID-19, a meta-analysis of 40 studies including two rCts by arruda et al (27), suggested that HFNot reduced the risk of intubation compared with Cot, but showed no additional benefit when compared with NIV. In another meta-analysis by li et al (28), again focusing on patients with CoVID-19, HFNot was demonstrated to reduce the rate of intubation, 28-day mortality and ventilator-free days compared with Cot. However, these results were not reproduced by a recent meta-analysis by Pisciotta et al (29) involving patients with CoVID-19-induced hypoxaemic respiratory failure, which showed no benefit in terms of treatment failure for HFNot compared with NIV and Cot.
recent guidelines issued by the European respiratory Society (ErS) suggest HFNot over NIV or Cot for the management of acute hypoxaemic respiratory failure. However, although they favor HFNot over Cot for post-extubated ICU patients with 'low-' or 'moderate-risk' for re-intubation, they suggest NIV over HFNot for 'high-risk' patients (30). During the CoVID-19 pandemic, HFNC was also widely applied in the early management of hypoxemia and respiratory distress in children with CoVID-19 requiring paediatric intensive care (31,32).
## 3. COVID-19 and politics: Challenges, dilemmas and lessons
Politics was one of the most significant, non-medical issues of the recent CoVID-19 pandemic, which demonstrated the interactions between science, society and politics (33). Since the onset of this unprecedented global health challenge, numerous countries designed and implemented various and controversial policies against SarS-CoV-2 (34). For example, the 'zero-CoVID-19 policy', which was adopted by China as well as other countries, tried strictly to eliminate local transmission of the virus (35)(36)(37). on the other hand, the 'Swedish CoVID-19 approach', which did not enforce strict lockdown measures, was based on voluntary recommendations and guidelines (38). Concurrently, latin american countries appeared to struggle with implementing specific COVID-19 pandemic policies for their citizens (39).
Politics influenced the development, distribution and access of vaccines and therapeutic agents against SarS-CoV-2, as well as public health management and social reaction. the accomplishment of this task would not have been possible without the close collaboration between scientists, scientific institutions and governments. Funding through state resources and support from international organizations, such as the World Health organization (WHo), also played a fundamental role (40). although the scientific society responded promptly by developing and approving novel vaccines and therapeutic agents against SarS-CoV-2, the global community faced deep inequalities in their access. For example, the European Union countries, including Greece, succeeded in achieving timely access to a sufficient amount of vaccine doses against SarS-CoV-2 (2). the European Union prioritized the introduction of vaccination programs against SarS-CoV-2 as its principal political strategy against CoVID-19 (https://health. ec.europa.eu/vaccination/overview_en). Developed countries gained privileged access to the first batches of vaccines, securing deals with pharmaceutical companies long before their release (41). the European Union countries responded quickly and prioritized solidarity in order to provide access to vaccines against SarS-CoV-2 to all the European citizens (https://commission.europa.eu/strategy-and-policy/coronavirus-response/coronavirus-european-solidarity-action_en). Moreover, international efforts, such as the global initiative CoVaX, co-led by Gavi, the Vaccine alliance, the Coalition for Epidemic Preparedness Innovations (CEPI), the WHo and UNICEF, sought to ensure equitable distribution of vaccines to developed and developing countries (41). However, all these efforts failed to meet their goals adequately. 'Vaccine nationalism' prominently affected the allocation of resources, as numerous governments chose to secure the needs of their populations neglecting international commitments. 'Vaccine diplomacy', as a political tool for foreign policy and international influence was also used as leverage to promote political and economic interests. these inequalities highlighted the gap between rich and poor countries and raised ethical and practical issues that may resonate and affect healthcare management of future crises.
the pandemic exposed the reciprocal relationship between politics and public health (42)(43)(44). Political leaders worldwide were challenged to make decisions that directly affected the spread of the virus, healthcare provision and the public perception of the pandemic. In numerous countries, decisions to impose lockdowns or lift restrictions were based on political calculations, such as the need to stabilize the economy or to respond to social pressure, rather than solely on scientific data and advice. The conflict between science and politics proved particularly harmful in cases where politicians downplayed the threat of the pandemic or spread misinformation, as witnessed in some countries with strong populist movements. these decisions undermined public trust in scientific authorities and challenged the implementation of necessary health measures. In several countries, political polarization and misinformation about vaccine safety increased vaccine hesitancy (45).
the need for updated training of healthcare professionals was another clear message from the recent CoVID-19 pandemic (46,47). Political fora are expected to support the adjustment of medical educational programs to new realities and organize targeted actions involving the institutions responsible for providing ongoing medical education. Continuing medical education is critical as this could HFNOT is supported by strong physiological evidence; HFNOT has been shown to improve oxygenation, reduce the work of breathing and enhance lung function (12)(13)(14)(15). During the recent CoVID-19 pandemic, HFNot has been increasingly used in ICU and PICU patients due to its effectiveness and particularly its tolerability (27)(28)(29)(30)(31)(32). Meta-analyses on the use of HFNot in acute hypoxaemic respiratory failure and CoVID-19 infection have generally shown that HFNot is more effective than Cot in reducing the need for higher respiratory support and comparable to NIV (27)(28)(29)(30).
HFNot can be applied in the early management of hypoxemia and respiratory distress in children admitted to the PICU (23)(24)(25)(26)31,32).
## COVID-19 and
The COVID-19 pandemic highlighted the deep interactions between science, society, and politics; politics the rapid development of vaccines was a significant scientific achievement, driven by the collaboration of scientists, governments and international organizations (33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46)(47). Despite various international efforts, inequalities in vaccine access underscored the economic and geopolitical aspects of the CoVID-19 pandemic (41).
Politics played a crucial role in managing the pandemic, with decisions regarding lockdowns and vaccination programs often influenced by political calculations, such as economic stability and social pressure, rather than solely by scientific data (33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46)(47).
The conflict between science and politics, as well as misinformation, undermined public trust and complicated the implementation of necessary measures (44). the need for political focus on the updated training of healthcare professionals was another key message from the recent CoVID-19 pandemic (45)(46)(47).
## COVID-19 and
During the COVID-19 pandemic, frontline researchers around the globe, along with their official science institutions and scientific societies, had the principal role in transparently communicating communication information about SarS-CoV-2 to the public (50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62). However, throughout the pandemic, non-specialist scientists and others also assumed a key role in public communication, while an unprecedented surge of information, disinformation and misinformation about SarS-CoV-2 was spread especially via the social media platforms (52)(53)(54). Science communication on CoVID-19 was another example that required multidisciplinary collaboration with communication experts leading to the urgent development and usage of innovative communication strategies (50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62). the post-CoVID-19 pandemic period provides a valuable opportunity to evaluate the relationship between science communication and society to improve the preparedness of the international scientific community for the next pandemic (61,62).
COVID-19, coronavirus disease 2019; HFNOT, high flow nasal oxygen therapy; ICU, intensive care unit, PICU, paediatric intensive care unit; COT, conventional oxygen therapy; NIV, non-invasive ventilation; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
promote the value of medical education in paediatric viral infections as well, including CoVID-19 (48,49). If the next pandemic disproportionally affects the paediatric population, this effort will play a key role for the preparedness of the paediatric personnel and healthcare system of each country.
For the post-CoVID-19 era, long-term policies are required to prepare humanity for future health crises. the international community must ensure equal access to vaccines and therapeutic agents, regardless of the economic strength of a country. Governments need to collaborate with international organizations and the private sector to create a more resilient global public health system that can respond quickly and effectively to new threats and challenges (47). our experience from the CoVID-19 pandemic has taught us that health cannot be separated from politics and that protecting human life should be the highest priority, beyond economic or political calculations.
## 4. COVID-19 and science communication
Science communication-is a highly demanding process, which deals with complex information, dynamic uncertainty and diverse audiences, with varying educational levels, cultural beliefs, attitudes and behaviours, that impact the understanding of science (50). Effective science communication is now established as an important tool that provides accurate scientific knowledge to the public and helps them identify false information. Ineffective communication, on the other hand, can be detrimental to both science and society in general.
During the recent CoVID-19 pandemic, science communication demonstrated its critical influence on public health. Since the beginning of the CoVID-19 pandemic threat, healthcare professionals strived to translate science and communicate its ongoing findings in a timely and accessible manner to various audiences (51). Frontline researchers, alongside organizations such as the WHo, played the principal role to transparently communicate their findings and explain them to the public in a meaningful and understandable way. there was an unprecedented demand for scientific knowledge; in fact, the overwhelming requests from journalists for epidemiological and research updates threatened to shift focus and resources from viral research to media demands.
However, throughout the pandemic, non-specialist scientists, academics, journalists, and others played a key role in the communication of SarS-CoV-2 and CoVID-19 advances, recommendations and challenges. Despite newspapers and press websites typically being reliable sources, the CoVID-19 pandemic witnessed an unparalleled surge of both accurate and inaccurate information, largely spread via digital channels and platforms, such as twitter/X, Facebook, Instagram and TikTok (52)(53)(54). Official scientific institutions and societies had to address issues, such as 'fake news' and uncertainty, the latter being a typical characteristic of scientific research, which was however misinterpreted and perceived as inaccuracy or even unreliability. Misinformation and disinformation regarding CoVID-19 vaccine safety were strongly related to increased vaccine hesitancy (55)(56)(57).
Science communication with the aid of reliable and accessible official social media platforms was also encouraged (58). Innovative new communication strategies were proposed and used, including social media and podcasts (59). these tools were more effective at targeting specific audiences, such as adolescents and the young population. Science communication on COVID-19 pandemic required multidisciplinary scientific collaboration. Collaboration with visual communicators and design experts produced digital illustrations and demonstrations of SarS-CoV-2, which improved the understanding of the virus and health safety measures and improved vaccine confidence (60).
the post CoVID-19 era offers a chance to assess the social impact of science communication and improve its future effectiveness. Researchers and scientific institutions need to design and develop novel communication strategies in order to respond effectively to future potential crises. Scientists with communication skills, passion and training should be motivated. Moreover, scientific societies should create improved links with the media and ensure that healthcare journalists are well informed and trained. Despite its devastating health, social, and financial ramifications, the COVID-19 pandemic presents a genuine opportunity to improve pandemic preparedness (61,62).
## 5. Conclusions
Pandemic evaluation and planning perspectives towards future infectious threats remain challenging. HFNot, a non-invasive ventilation modality increasingly used prior to the CoVID-19 era in both ward-based and critical care management of respiratory failure (11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25), represents an excellent clinical example of how the CoVD-19 pandemic enriched medical knowledge and experience (26)(27)(28)(29)(30)(31)(32). the medical experience gained from the treatment of critically ill patients with CoVID-19, should be further evaluated for the establishment of state-of-the-art, evidence-based medical consensuses and protocols. these tools are essential for the effective and precise management of adults and paediatric patients and should be integrated in current clinical practice.
the CoVID-19 pandemic was a pivotal moment for global health and politics (33)(34)(35)(36)(37)(38)(39)(40)(41)(42)(43)(44)(45)(46)(47). the collaboration between science and politics contributed to the rapid development of vaccines and therapeutic agents against SarS-CoV-2, however global distribution was uneven due to national policies and geopolitical tensions. Different political agendas influenced not only the distribution of vaccines but also the public perception of their safety and efficacy. Therefore, the international community must be taught from these errors and work towards a more equitable and resilient approach to future health crises. Public health safety necessitates collaboration and impartiality, prioritizing global solidarity and equality above national and political agendas. Political decisions focusing on increasing the financial health resources in primary health care and advancing secondary and tertiary hospital-based care should be encouraged. Health policies should also focus on enhancing specialized as well as continuing medical education (46)(47)(48)(49).
Science communication also demonstrated its potential usefulness and effectiveness during the recent pandemic (50)(51)(52)(53)(54)(55)(56)(57)(58)(59)(60)(61)(62). This burgeoning scientific discipline should be further developed and integrated into both undergraduate and postgraduate medical education. Health professionals must develop effective communication skills and become adept in providing accurate and useful information to their patients and the general population, as well. In the unfortunate event of a future pandemic, effective science communication will depend on multi-disciplinary collaboration between clinical and research scientists and communication experts; this task requires improved digital tools and innovative strategies to address public misinformation and disinformation.
the aim of the present review article was to stimulate further discussion within the international scientific community on the evaluation of the management of the recent CoVID-19 pandemic. a careful interpretation of the lessons learned may help promote strategic planning and preparedness, and advance public health, translational research and future medicine.
## References
1. (2020) "World Health organization: CoVID-19 strategic preparedness and response plan: operational planning guidelines to support country preparedness and response"
2. Zoumpourlis, Goulielmaki, Baliou et al. (2020) "The COVID-19 pandemic as a scientific and social challenge in the 21st century" *Mol Med rep*
3. Jentsch, Anand, Bauch (2021) "Prioritising COVID-19 vaccination in changing social and epidemiological landscapes: a mathematical modelling study" *lancet Infect Dis*
4. O'callaghan, Green, Brierley (2023) "Pandemic preparedness: Is the UK ready for a pandemic that affects children?" *BMJ*
5. Qin, Sun, Zhang et al. (2022) "Lessons from SarS-CoV-2 and its variants (review)" *Mol Med rep*
6. Ga, Óhaiseadha, Hynds et al. (2025) "What lessons can be learned from the management of the COVID-19 pandemic?" *Int J Public Health*
7. Markiewicz, Drazkowska, Sikorski (2021) "Tricks and threats of rNa viruses-towards understanding the fate of viral rNa" *rNa Biol*
8. Singh, Jayant, Singh et al. (2022) "Withania somnifera (l.) Dunal (ashwagandha) for the possible therapeutics and clinical management of SarS-CoV-2 infection: Plant-based drug discovery and targeted therapy" *Front Cell Infect Microbiol*
9. Sharma, Poddar, Chaudhary Aa et al. (2022) "Does CoVID-19 lockdowns have impacted on global dengue burden? a special focus to India" *BMC Public Health*
10. (2022) "No authors listed: Post-CoVID-19 Condition treatment and Management rapid Scoping review: Scoping review. CaDtH Health technology review. Canadian agency for Drugs and technologies in Health"
11. Nishimura (2019) "High-flow nasal cannula oxygen therapy devices" *respir Care*
12. Crimi, Pierucci, Renda et al. (2022) "High-flow nasal cannula and CoVID-19: a clinical review" *respir Care*
13. Ischaki, Pantazopoulos, Zakynthinos (2017) "Nasal high flow therapy: a novel treatment rather than a more expensive oxygen device" *Eur respir rev*
14. Roca, Messika, Caralt et al. (2016) "Predicting success of high-flow nasal cannula in pneumonia patients with hypoxemic respiratory failure: The utility of the ROX index" *J Crit Care*
15. D'cruz Rf, Kaltsakas (2020) "High-flow therapy: Physiological effects and clinical applications" *Breathe (Sheff)*
16. Hernández, Vaquero, González et al. (2016) "Colinas l, Cuena r and Fernández r: Effect of postextubation high-flow nasal cannula vs conventional oxygen therapy on reintubation in low-risk patients: a randomized clinical trial" *JAMA*
17. Hernández, Vaquero, Colinas L, Cuena R et al. (2016) "Effect of postextubation high-flow nasal cannula vs noninvasive ventilation on reintubation and postextubation respiratory failure in high-risk patients: a randomized clinical trial" *JAMA*
18. Zhu, Yin, Zhang et al. (2019) "High-flow nasal cannula oxygen therapy versus conventional oxygen therapy in patients after planned extubation: a systematic review and meta-analysis" *Crit Care*
19. Granton, Chaudhuri, Wang et al. (2020) "High-flow nasal cannula compared with conventional oxygen therapy or noninvasive ventilation immediately postextubation: a systematic review and meta-analysis" *Crit Care Med*
20. Douglas, Ng, Nazeem et al. (2018) "Irving l and Segal r: a randomised controlled trial comparing high-flow nasal oxygen with standard management for conscious sedation during bronchoscopy" *anaesthesia*
21. Maitra, Som, Bhattacharjee et al. (2016) "Comparison of high-flow nasal oxygen therapy with conventional oxygen therapy and noninvasive ventilation in adult patients with acute hypoxemic respiratory failure: a meta-analysis and systematic review" *J Crit Care*
22. Seow, Khor, Khung et al. (2024) "High-flow nasal oxygen therapy compared with conventional oxygen therapy in hospitalised patients with respiratory illness: A systematic review and meta-analysis" *BMJ open respir res*
23. Kepreotes, Whitehead, Attia et al. (2017) "High-flow warm humidified oxygen versus standard low-flow nasal cannula oxygen for moderate bronchiolitis (HFWHo rCt): an open, phase 4, randomised controlled trial" *lancet*
24. Franklin, Babl, Schlapbach et al. (2018) "a randomized trial of high-flow oxygen therapy in infants with bronchiolitis" *N Engl J Med*
25. Kawaguchi A, Yasui, Garros (2017) "the clinical impact of heated humidified high-flow nasal cannula on pediatric respiratory distress" *Pediatr Crit Care Med*
26. Kwon (2020) "High-flow nasal cannula oxygen therapy in children: a clinical review" *Clin Exp Pediatr*
27. Arruda, Ga, Ll (2023) "Effectiveness of high-flow nasal cannula therapy on clinical outcomes in adults with COVID-19: A systematic review" *Can J Respir Ther*
28. Li, Li, Liu (2023) "High-flow nasal cannula reduces intubation rate in patients with CoVID-19 with acute respiratory failure: A meta-analysis and systematic review" *BMJ Open*
29. Pisciotta, Arulkumaran (2024) "High-flow nasal oxygen versus conventional oxygen therapy and noninvasive ventilation in CoVID-19 respiratory failure: a systematic review and network meta-analysis of randomised controlled trials" *Br J Anaesth*
30. Oczkowski, Ergan, Bos L, Chatwin et al. (2022) "ErS clinical practice guidelines: High-flow nasal cannula in acute respiratory failure" *Eur Respir J*
31. Loomba, Villarreal, Farias et al. (2020) "Pediatric intensive care unit admissions for CoVID-19: Insights using state-level data" *Int J Pediatr*
32. Wu, Wang, Lu et al. (2025) "The applications of high-flow nasal cannulas in pediatric intensive care units in Taiwan" *J Formos Med Assoc*
33. (2022) "Vaccination, politics and CoVID-19 impacts" *BMC Public Health*
34. Chung, Heo, Han et al. (2021) "Effects of government policies on the spread of CoVID-19 worldwide" *Sci rep*
35. Bai, Sha, Cheung et al. (2022) "Optimizing the dynamic zero-COVID policy in China" *Int J Biol Sci*
36. Chen, Chen (2022) "China can prepare to end its zero-COVID policy" *Nat Med*
37. (2023) "the lancet regional Health-Western Pacific: the end of zero-CoVID-19 policy is not the end of CoVID-19 for China" *lancet reg Health West Pac*
38. Petridou (2020) "Politics and administration in times of crisis: Explaining the Swedish response to the CoVID-19 crisis" *Eur Policy anal*
39. Martinez-Valle A (2021) "Public health matters: Why is latin america struggling in addressing the pandemic?" *J Public Health Policy*
40. Emanuel, Luna, Schaefer et al. (2021) "Enhancing the WHo's proposed framework for distributing COVID-19 vaccines among countries" *Am J Public Health*
41. Yoo, Mehta, Mak et al. (2022) "CoVaX and equitable access to CoVID-19 vaccines" *Bull World Health organ*
42. Gonsalves, Yamey (2020) "Political interference in public health science during the COVID-19" *BMJ*
43. Fedson (2021) "CoVID-19, host response treatment, and the need for political leadership" *J Publ Health Policy*
44. Greer, King, Da Fonseca (2021) "Coronavirus politics: the comparative politics and policy of CoVID-19"
45. Lazarus, Wyka, White et al. (2022) "Revisiting CoVID-19 vaccine hesitancy around the world using data from 23 countries in 2021" *Nat Commun*
46. Lucey, Johnston (2020) "The transformational effects of COVID-19 on medical education" *JAMA*
47. Frenk, Chen, Chandran et al. (2022) "Challenges and opportunities for educating health professionals after the CoVID-19 pandemic" *lancet*
48. Mammas, Liston, Koletsi et al. (2022) "Insights in paediatric virology during the CoVID-19 era (review)" *Med Int (lond)*
49. Mammas, Drysdale, Charalampous et al. (2023) "Navigating paediatric virology through the COVID-19 era (Review)" *Int J Mol Med*
50. Fischhoff, Da (2013) "the science of science communication. Introduction"
51. (2020) "Coronavirus conversations: Science communication during a pandemic" *Nature*
52. Rubin (2022) "When physicians spread unscientific information about COVID-19" *JAMA*
53. Joseph, Fernandez, Kritzman et al. (2022) "CoVID-19 Misinformation on social media: a scoping review" *Cureus*
54. Lurie, Adams, Lynas et al. (2022) "CoVID-19 vaccine misinformation in English-language news media: Retrospective cohort study" *BMJ open*
55. Sl, Wiysonge (2020) "Social media and vaccine hesitancy" *BMJ Glob Health*
56. Schwarzinger (2021) "addressing CoVID-19 vaccine hesitancy: Is official communication the key?" *lancet Public Health*
57. Karafilakis, Van Damme, Hendrickx et al. (2022) "CoVID-19 in Europe: New challenges for addressing vaccine hesitancy" *lancet*
58. Heuss, Zachlod, Bt (2023) "Social' media? How Swiss hospitals used social media platforms during the early months of the CoVID-19 pandemic crisis" *Public Health*
59. Rubin, Baden, Morrissey (2022) "Audio interview: Covid-19 takeaways at podcast 100" *N Engl J Med*
60. Boender Ts, Louis-Ferdinand, Duschek (2022) "Digital visual communication for public health: Design proposal for a vaccinated emoji" *J Med Internet Res*
61. Davies, Cuddihy (2021) "leveraging data and new digital tools to prepare for the next pandemic" *lancet*
62. Sachs, Karim, Aknin et al. (2022) "the lancet commission on lessons for the future from the CoVID-19 pandemic" *lancet* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12845450&blobtype=pdf | # UV-C Irradiation Effectiveness on Mpox-Virus-Contaminated Surfaces
Anna Gidari, Samuele Sabbatini, Carlo Pallotto, Sabrina Bastianelli, Sara Pierucci, Chiara Busti, Giulia Proietti, Alessia Lai, Giuseppe De Socio, Daniela Francisci
## Abstract
Introduction: Mpox virus (MpoxV), an emerging zoonotic pathogen, has recently caused global concern due to increasing outbreaks beyond its traditional endemic regions. While transmission primarily occurs via close contact, fomites are also suspected of contributing. This study aims to evaluate the effectiveness of UV-C irradiation on MpoxV-contaminated surfaces. Methods: the virucidal activity of UV-C (254 nm) irradiation on MpoxV applied to plastic, glass, and stainless-steel surfaces was assessed. Using a viral stock of 2.49 × 10 5 TCID 50 /mL, the samples were exposed to increasing UV-C doses. Viral titers were quantified through TCID 50 and plaque assays. Results: A UV-C dose of 6.34 mJ/cm 2 achieved a >2-log reduction of viral load, below the detection limit (31.6 TCID50/mL), on all tested surfaces. EC 90 values were determined as 3.33 mJ/cm 2 (plastic), 0.81 mJ/cm 2 (stainless steel), and 1.98 mJ/cm 2 (glass). No viable virus was detectable post-treatment at these doses on plastic and stainless steel while the titer was significantly reduced on glass. Conclusions: UV-C irradiation at low doses effectively inactivated MpoxV on various fomites. These findings support UV-C as a rapid and effective environmental disinfection strategy in healthcare and community settings to prevent indirect transmission of MpoxV.
## 1. Introduction
Mpox is a zoonotic infection caused by Mpox virus (MpoxV), family Poxviridae, subfamily Chordopoxvirinae, genus Orthopoxvirus, species Orthopoxvirus monkeypox, with two distinct clades (Clade 1 and clade 2) [1]. MpoxV, since the detection of the first human case in a child in the Democratic Republic of the Congo in 1970, historically caused sporadic outbreaks primarily confined to Central and West Africa [2]. On 23 July 2022, due to the global spread of Clade 2 MpoxV outside its usual geographic distribution in African countries, involving more than 100 countries across Europe and America, World Health Organization (WHO) declared it a Public Health Emergency of International Concern [3,4]. In particular, since the identification of MpoxV transmission outside endemic regions in May 2022, a large multi-country outbreak has been ongoing globally, with 153,961 cases and 380 deaths reported across 137 countries in all six (WHO) regions as of 30 June 2025 [5].
After the 2022 outbreak, the incidence of the infection outside Africa was significantly reduced. However, despite the decline in cases in Africa during 2023 and 2024, the first half of 2025 saw an approximately 50% increase in reported cases compared to the previous year [5].
Clinically, Mpox manifests with fever, lymphadenopathy, and distinctive vesiculopustular lesions resembling smallpox, albeit generally with lower mortality [6].
Transmission primarily occurs through direct contact with infected individuals or bodily fluids. In particular, the 2022 global outbreak has been associated with close intimate contact like during sexual activity, and most cases have been diagnosed among men who have sex with men. It is also reported a human-to-human transmission of mpox with respiratory droplets through a prolonged face-to-face contact [6].
However, indirect transmission via contaminated surfaces (fomites) plays a potentially critical role in virus spread, influenced by environmental factors such as temperature, humidity, and pH [6]. Prior studies demonstrated prolonged viral stability under low humidity and moderate temperatures, emphasizing the importance of environmental hygiene measures [7,8].
Ultraviolet-C (UV-C) irradiation at approximately 254 nm wavelength is widely recognized for its virucidal properties, effectively disrupting viral nucleic acids and halting replication processes [9]. UV-C has been successfully employed against various pathogens, including SARS-CoV-2, highlighting its potential utility in environmental disinfection [9,10].
This study aims to evaluate UV-C efficacy against MpoxV on different surfaces, determining the minimum UV-C dose needed to reduce viral titer below detectable limits.
## 2. Materials and Methods
## 2.1. Mpox Strain Isolation and Stock Preparation
All experiments were conducted using an Orthopoxvirus monkeypox (MpoxV) strain isolated in the Biosafety Level 3 (BSL-3) Virology Laboratory at the Clinic of Infectious Diseases, University of Perugia-Santa Maria della Misericordia Hospital, Perugia, Italy.
For all the experiment, a strain isolated from a symptomatic patient admitted to the Clinic of Infectious Diseases of the same hospital was used, as described previously [11]. Briefly, following the incision of a vesicle, a swab was collected, and the diagnosis was confirmed by a PCR test. Subsequently, transport medium (UTM) was incubated with a 1:1 Eagle's minimum essential medium (MEM) supplemented with penicillin-streptomycin (1%) and left to react for 1 h at 4 • C to reduce bacterial contamination. The resulting suspension was inoculated onto a monolayer of Vero E6 cells and incubated for 2 h at 37 • C with 5% CO 2 atmosphere. Following this initial incubation, the medium was replaced with MEM supplemented with 1% fetal bovine serum (FBS) at 37 • C with 5% CO 2 . The plates were checked every 24 h to detect the cytopathic effect (CPE). After CPE appearance, supernatant was recovered, filtered (filter 0.45 µm) and viral titer was determined by Median Tissue Culture Infectious Dose (TCID 50 ) endpoint dilution assay [12] and stock aliquots were stored at -80 • C. The viral stock had a titer of 2.49 × 10 5 TCID 50 /mL and aliquots stored frozen were thawed immediately before use in each experiment.
## 2.2. MpoxV Strain Sequencing
The MpoxV strain was sequenced at the Laboratorio di Malattie Infettive, Department of Biomedical and Clinical Sciences, University of Milan. The strain was identified as Clade IIb lineage C.1.1.
Briefly, DNA was extracted from sample using the QIAamp DNA Blood kit (QIAGEN, Hilden, Germany), fragmented using a Covaris M220 ultrasonicator (Covaris, Woburn, MA, USA) and checked using a 4200 TapeStation System (Agilent Technologies, Santa Clara, CA, USA) to verify dimensions. The libraries were prepared using the Nextera XT DNA Library Preparation Kit (Illumina, San Diego, CA, USA) and sequencing was performed using a Miseq sequencer (Illumina, San Diego, CA, USA) with 300 cycles.
Reads were mapped to a reference genome sequences (MT903343.1) using Geneious software V.11 (Biomatters, Auckland, New Zealand) (http://www.geneious.com) obtaining an average depth of 18.5× (min 1-max 1731). Clade assignment was performed using NextClade (https://clades.nextstrain.org/).
## 2.3. Inanimate Surfaces
The materials selected for testing were plastic (polystyrene, 24-well plates; Corning, Falcon ® , New York, NY, USA), glass (sterile disks, 12 mm diameter, Corning, Falcon ® , New York, NY, USA), and stainless steel (AISI 304 sterile disks, 12 mm diameter, Promagroup, Umbertide, Perugia, Italy). The glass and stainless-steel disks were sterilized by autoclaving and subsequently placed into 24-well plates [10].
## 2.4. UV-C Irradiation Assay
All the experiments were conducted in a BSL-3 laboratory as previously described [10]. Each experiment was performed in triplicate and independently repeated at least two or three times for each type of material. Ambient temperature and relative humidity were continuously monitored and maintained at approximately 23-25 • C and 40-50%, respectively.
To assess the virus recovery efficiency from tested surfaces, two separate aliquots of 10 µL virus suspension were used: one aliquot was immediately processed to determine the viral titer by TCID 50 assay, while the second aliquot was deposited onto the surface and recovered by washing after a 30 min incubation period (T 0 ) through washing. Recovery efficiency was then calculated using the following formula: [TCID 50 /mL recovered virus (T 0 )/TCID 50 /mL virus aliquot] × 100 [10].
Given the viral stock (2.49 × 10 5 TCID 50 /mL), 10 or 50 µL of frozen viral stock were placed on different surfaces of the materials by a sterile pipet tip.
A monochromatic UV-C lamp emitting at 254 nm, with an irradiance of 0.82 mW/cm 2 , was placed at 30 cm from the different surfaces. The UV light dose was measured by the manufacturer (Bazzica Engineering ® , Trevi, Italy), who also produced and provided the lamp. Measurements were performed using a photometer (RMD Sensor UVC 200-280 nm, 0-10 W/cm 2 ; Opsytec Dr. Gröbel GmbH, Ettlingen, Germany) at the same 30 cm distance used in the experiments.
Starting from a dose of 136.81 mJ/cm 2 , corresponding to 180 s of exposure, decreasing doses of UV-C were applied by reducing the exposure time, in order to identify the minimum dose capable of reducing the viral titer below the detection limit of the method (31.6 TCID 50 /mL) corresponding to a >2 Log reduction. Simultaneously, a control plate was maintained under identical conditions but shielded with aluminium foil (shielded plate). Additional plates containing only culture medium were also exposed under the same environmental conditions [10].
Following UV-C exposure, supernatants were recovered and titers were determined as TCID 50 /mL [10]. Once the lowest UV-C dose capable of reducing the viral load below the detection limit was identified, the supernatant from the corresponding sample was further analysed using a plaque assay. Plaque assay has been performed as previously described [13,14].
## 2.5. Statistical Analysis
Statistical analysis was performed using Graphpad Prism 8.31 (San Diego, CA, USA). Kolmogorov-Smirnov test was used to test data normality. Based on this test, mean with the respective standard deviation (SD) or median with interquartile range (IQR) were used https://doi.org/10.3390/pathogens15010078 to present data. EC 50 , and EC 90 concentrations were calculated using three-parameter regression modelling.
## 3. Results
The mean recovery efficiency was 49%. Several experiments were conducted on plastic surfaces using different time points, and thus varying UV-C doses, to generate a dose-response curve. Subsequently, four distinct UV-C doses were tested on each material.
As shown in Figure 1A,B, no significant differences were observed between T 0 and shielded plates, nor among shielded plates across the different exposure times (3-180 s). Similar results were obtained for stainless steel and glass over the range of 0-14 s (Figure 1C,D). UV-C doses tested on plastic were: 1.24 mJ/cm 2 (3 s), 2.07 mJ/cm 2 (5 s), 4.33 mJ/cm 2 (10 s), 6.34 mJ/cm 2 (14 s), 10.25 mJ/cm 2 (21 s), 20.06 mJ/cm 2 (36 s), 30.6 mJ/cm 2 (50 s), 40.95 mJ/cm 2 (63 s), 50.71 mJ/cm 2 (75 s), 63.01 mJ/cm 2 (90 s), 87.61 mJ/cm 2 (120 s), 136.81 mJ/cm 2 (180 s).
As shown in Figure 1A, UV-C exposure was first tested on a low viral load, corresponding to 10 µL of viral stock (2.49 × 10 3 TCID 50 ). All UV-C doses, with the exception of the lowest one (2.07 mJ/cm 2 , 5 s), were effective in reducing the viral titer below the detection limit of the method. The 2.07 mJ/cm 2 dose reduced the titer to 83.1 TCID 50 /mL (SD, 67.8 TCID 50 /mL). In this instance, it was not possible to generate a complete dose-response curve, and therefore the lowest dose (1.24 mJ/cm 2 , 3 s) was not tested.
As shown in Figure 2, UV-C exposure was tested on 1.25 × 10 4 TCID 50 of virus, corresponding to 50 µL of viral stock. A dose of 6.34 mJ/cm 2 was the lowest dose that reduced the viral titer below the detection limit of the method (31.6 TCID 50 /mL) corresponding > 2 Log reduction. Lower doses were subsequently tested to generate a dose-response curve, and the data were analysed using a three-parameter linear regression model. The UV-C treatment showed a half-maximal effective concentration (EC 50 ) of 0.09 mJ/cm 2 (95% confidence interval, CI, lower value not available, NA, to 0.33) and an EC 90 of 3.33 mJ/cm 2 (Figure 2A). Although an EC 50 best-fit value was obtained, it was not possible to calculate a complete confidence interval due to the very low EC 50 value. Comparable results were observed for stainless steel and glass. For both materials, the dose of 6.34 mJ/cm 2 was the minimum required to reduce the viral titer below the detection threshold (31.6 TCID 50 /mL). As shown in Figure 2B, the UV-C treatment of Mpox on stainless steel yielded an EC 50 of 0.37 mJ/cm2 (95% CI NA, to 4.2) and an EC 90 of 0.81 mJ/cm 2 . For glass, the EC 50 of UV-C on MpoxV was 0.22 (95% CI NA-1.17) and the EC 90 was 1.98 mJ/cm 2 (Figure 2C).
Furthermore, to verify viral eradication, samples exposed to a UV-C dose of 6.34 mJ/cm 2 , along with their corresponding control wells, were recovered and titrated using a plaque assay, with results expressed as PFU/mL. As shown in Figure 3, the UV-C dose 6.34 mJ/cm 2 eradicated MpoxV on plastic and stainless steel, while it resulted in a significant reduction of viral titer on glass. A UV-C dose of 6.34 mJ/cm 2 was the lowest dose that reduced viral titers below the detection limit of the assay (31.6 TCID 50 /mL). To assess whether this dose achieved complete viral inactivation, samples treated with 6.34 mJ/cm 2 and corresponding controls were recovered and titrated using plaque assay (PFU/mL). This UV-C dose completely inactivated MpoxV on plastic and stainless steel and significantly reduced viral titers on glass. Data represent the mean ± standard deviation of two or three independent experiments.
## 4. Discussion
MpoxV transmission via contaminated surfaces remains a significant public health concern, particularly given its environmental stability under favorable conditions (low humidity, moderate temperature). MpoxV has been shown to remain stable for more than 49 days at 4 • C, and up to 42 days at 37 • C or at room temperature [15]. A recent study explored viral persistence on various fomites, demonstrating that infectious MpoxV can persist for up to 21 days on non-porous surfaces at low temperatures. In contrast, porous materials such as cotton exhibited a rapid loss of infectivity, particularly at room temperature [8]. The MpoxV persistence in wastewater could also be an important issue, especially in low-income countries [7].
These findings underscore the need for effective disinfection strategies in outbreak control. UV-C irradiation is a well-established method for microbial inactivation and has demonstrated effectiveness against a wide range of bacteria and viruses, including SARS-CoV-2. Raeiszadeh et al. reviewed the application of UV-C for disinfection and highlighted its effectiveness against SARS-CoV-2 and other coronaviruses, with doses ranging from 1.2 to 40 mJ/cm 2 [16]. Buonanno et al. effectively inactivated airborne human coronaviruses (HCoV-229E, HCoV-OC43) using far-UV-C at doses around 1-2 mJ/cm 2 , suitable for continuous disinfection in occupied spaces [17]. Heilingloh et al. reported a 1-log reduction in SARS-CoV-2 titers at an exposure dose of approximately 292 mJ/cm 2 [18]. Our group previously demonstrated effective SARS-CoV-2 inactivation at relatively low UV-C doses (10.25-23.71 mJ/cm 2 ) on plastic, glass, and stainless-steel surfaces [10]. While coronaviruses are among the most extensively studied pathogens in the context of UV-C disinfection due to their global health impact, the efficacy of UV-C has also been tested on a variety of other viruses. Gerba et al. achieved three-log titer reductions of echovirus 1, echovirus 11, coxsackievirus B3, coxsackievirus B5 and poliovirus 1 using doses of 25, 20.5, 24.5, 27, and 23 mW/cm 2 , respectively. In the same study, human adenovirus type 2 was found to be more resistant, requiring a dose of 119 mW/cm 2 for 99.9% inactivation [19]. In addition, a recent research has demonstrated the effectiveness of UV-C irradiation in decontaminating environmental surfaces contaminated with Marburg virus [20].
Collectively, these data underscore the broad-spectrum effectiveness of UV-C light for pathogen inactivation on fomites.
In a recent study, Mariotti et al. specifically investigated UV-C inactivation of MpoxV, demonstrating complete virus inactivation after 15 min of exposure; however, the precise UV-C dose applied was not reported [21]. Our findings significantly expand these observations, by establishing a much lower effective UV-C dose (6.34 mJ/cm 2 ) for MpoxV inactivation, with EC 90 values varying according to surface type. This dose is notably lower than previously suggested, reinforcing the practicality of UV-C as a disinfection method suitable for routine use.
The practical implications of these findings are directly relevant to public and healthcare environments. Mobile or fixed UV-C systems could be employed to rapidly disinfect high-touch surfaces in locations such as public transportation, educational facilities, healthcare settings, and retail environments. Implementation of such systems could play a critical role in limiting MpoxV transmission in community settings.
Nevertheless, potential concerns related to accidental overexposure to UV-C lights must be addressed.
A recent study evaluated the benefits of UV-C-based air disinfection in aircraft in relation to the potential risks of UV-C overexposure for passengers and crew. The authors demonstrate that the risks are significantly lower than the benefits and no-long term effects are expected [22].
Limitations of this study include potential discrepancies between laboratory-controlled experiments and real-world settings, and variability in contamination levels and organic load on surfaces. Further research is required to evaluate UV-C effectiveness under practical conditions.
## 5. Conclusions
UV-C irradiation effectively inactivates MpoxV on various surfaces at low doses. These results advocate broader implementation of UV-C protocols in healthcare and public settings to mitigate MpoxV environmental transmission risks.
## References
1. Likos, Sammons, Olson et al. (2005) "A tale of two clades: Monkeypox viruses" *J. Gen. Virol*
2. Cdc (2025) "Monkeypox: Clinical Recognition"
3. Lu, Xing, Wang et al. (2023) "Mpox (formerly monkeypox): Pathogenesis, prevention and treatment" *Signal Transduct. Target. Ther*
4. Who (2025) "World Health Organization. Multi-Country Monkeypox Outbreak in Non-Endemic Countries"
5. Who (2025) *Global Mpox Trends. Available online*
6. Titanji, Hazra, Zucker (1652) "Mpox Clinical Presentation, Diagnostic Approaches, and Treatment Strategies" *JAMA*
7. Atoui, Jourdain, Mouly et al. (2023) "A review on mpox (monkeypox) virus shedding in wastewater and its persistence evaluation in environmental samples. Case Stud" *Chem. Environ. Eng*
8. Banete, Hurst, Yim et al. (2025) "Stability of Monkeypox Virus on Commonly Contacted Surfaces in Clinical Settings" *Open Forum Infect. Dis*
9. Kowalski (2009) "Ultraviolet Germicidal Irradiation Handbook"
10. Gidari, Sabbatini, Bastianelli et al. (2021) "SARS-CoV-2 Survival on Surfaces and the Effect of UV-C Light" *Viruses*
11. Gidari, Nofri, Saccarelli et al. (2020) "Is recurrence possible in coronavirus disease 2019 (COVID-19)? Case series and systematic review of literature" *Eur. J. Clin. Microbiol. Infect. Dis*
12. Reed, Muench (1938) "A simple method of estimating fifty per cent endpoints" *Am. J. Epidemiol*
13. Gidari, Sabbatini, Schiaroli et al. (1475) "The Combination of Molnupiravir with Nirmatrelvir or GC376 Has a Synergic Role in the Inhibition of SARS-CoV-2 Replication In Vitro" *Microorganisms*
14. Gidari, Sabbatini, Schiaroli et al. (1577) "Synergistic Activity of Remdesivir-Nirmatrelvir Combination on a SARS-CoV-2 In Vitro Model and a Case Report"
15. Li, Lv, Zeng et al. "Evaluation of Stability, Inactivation, and Disinfection Effectiveness of Mpox Virus" *Viruses*
16. Raeiszadeh, Adeli (2020) "A Critical Review on Ultraviolet Disinfection Systems against COVID-19 Outbreak: Applicability, Validation, and Safety Considerations" *ACS Photonics*
17. Buonanno, Welch, Shuryak et al. (2020) "Far-UVC light (222 nm) efficiently and safely inactivates airborne human coronaviruses" *Sci. Rep*
18. Heilingloh, Aufderhorst, Schipper et al. (2020) "Susceptibility of SARS-CoV-2 to UV irradiation" *Am. J. Infect. Control*
19. Gerba, Gramos, Nwachuku (2002) "Comparative Inactivation of Enteroviruses and Adenovirus 2 by UV Light" *Appl. Environ. Microbiol*
20. Uwayo, Gahamanyi, Mugisha et al. (2025) "Use of robot technology for decontaminating Marburg virus-exposed environments in Rwanda: A promising tool for managing outbreaks in resource-limited settings" *Am. J. Infect. Control*
21. Mariotti, Bettini, Meschi et al. "Effect of chemical and physical agents on monkeypox virus infectivity and downstream research applications" *Virology*
22. Allen, Mills, Garcia (2025) "Risk vs. Benefit Analysis of Ultraviolet-C Advanced Aircraft Disinfection" *Aerosp. Med. Hum. Perform*
23. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12697173&blobtype=pdf | # Complete genome sequences of the lytic bacteriophage PBM-3 and the avian pathogenic Escherichia coli strain PE144
Jérémy Cherbuin, Jing Zhang, Muhammad Khan, Hira Niaz, Fazal Adnan
## Abstract
Avian pathogenic Escherichia coli, including strains that are resistant to antimicrobials, negatively impacts the poultry industry worldwide and the provision of animal protein to the people. Here, we report the complete genomes of the lytic bacteriophage PBM-3 and its bacterial host PE144, both isolated in Pakistan.
KEYWORDS bacteriophages, PacBio sequencing, avian pathogenic Escherichia coli, PakistanA vian pathogenic Escherichia coli (APEC) causes colibacilloses (1) associated with major economic losses (2). Phages may assist in the control of infections with resistant APEC strains (3-5). We describe the genome sequences of the lytic bacterio phage PBM-3 and its APEC host PE144.PE144 (O1:H7) was isolated from the liver of a 3-week-old chicken that succumbed to colibacillosis on a commercial farm in Rawalpindi, Pakistan, as described elsewhere (6). Genomic DNA was extracted from cells grown in lysogeny broth LB media at 37°C and 220 rpm overnight using the Wizard Genomic DNA Purification Kit (Promega).Phage PBM-3 was isolated from 20 g of chicken bedding material, which was mixed with 40 mL of ¼ Ringer's buffer, incubated overnight, and centrifuged at 5,000 g for 10 min. The supernatant was collected, passed through a 0.22 µm filter, plated on strain PE144 using LB double agar overlay, and incubated at 37°C overnight (7). Picked plaques were plaque-purified three times and enriched using the double-agar overlay method (7). Genomic DNA was extracted from high-titer stocks (>10 9 pfu/mL) using the Monarch HMW DNA Extraction Kit (New England Biolabs).DNA quantity and quality were evaluated using Qubit fluorometry (Thermo Fisher Scientific) and the Fragment Analyzer system (advanced analytical technologies), respectively. DNA was sheared by tagmentation, and fragments smaller than 3 kb were removed using 35% diluted AMPure beads, following the LongPlex Long Fragment Multiplexing Kit (seqWell). Library preparation was done with the LongPlex Kit and the SMRTbell Prep Kit 3.0 (Pacific Biosciences), followed by fragment analysis using Fragment Analyzer. Sequencing was conducted on the PacBio Revio system using one SMRT Cell 25M (30-hour movie time, Revio chemistry).Default parameters were used for all software tools unless specified. Adapter trimming and sample demultiplexing were done using SMRT Link (v13.01), with adapter removal performed using the LongPlex Demultiplex Nextflow pipeline (seqWell). Raw reads were evaluated using Nanoplot (version 1.44.1) (8). The PE144 genome and the phage PBM-3 genome were assembled using Flye (version 2.9.4-b1799) (9, 10) and using QIAGEN CLC Genomic Workbench (version 25.0.1) (https://digitalinsights.qiagen.com), respectively. Quality of PE144 assembly was assessed by Quast (version 5.2.0) (11) and CheckM (12). Afterward, the PE144 genome was rotated to the first nucleotide of the start codon of the dnaA gene using Circlator (version 1.5.5) (13) and annotated using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP version 6.10) (14). Sequence types December 2025 Volume 14 Issue 12 10.
and serotypes were identified using PubMLST (15) and serotype using SerotypeFinder (v2.0) (16), respectively. The plasmid was assessed using PlasmidFinder (v2.1) (17,18) (Table 1). The phage genome was annotated employing Pharokka (version 1.7.5) ( 19) and verified manually using Artemis (version 18.2.0) (20) and BLASTp (21).
The complete genome of PE144 had a 4,944,904 bp chromosome (374× coverage) harboring 4,716 coding DNA sequences (CDSs) and an IncFIB plasmid (160× coverage) of 125,603 bp harboring 131 CDSs (98.39% identity to GenBank AP001918). The PE144 genome had a GC content of 50.5%. The phage PBM-3 had a genome of 72,933 bp (793× coverage), a GC content of 43.08%, harbored 83 CDSs and three tRNA genes, and had 402 bp direct terminal repeats identified by Phageterm (22) (Fig. 1).
## FIG 1
Genome alignment of phage PBM-3 and closest related phage UE-S5b (PP301341.1). Genome organization of the phage PBM-3 and UE-S5b visualized using clinker (23). PBM-3 had 98.89% genome sequence identity and 95% coverage with UE-S5b using BlastN (18).
## References
1. Nolan, John, Vaillancourt et al.
2. Christensen, Bachmeier, Bisgaard (2021) "New strategies to prevent and control avian pathogenic Escherichia coli (APEC)" *Avian Pathol*
3. Kazibwe, Katami, Alinaitwe et al. (2020) "Bacteriophage activity against and characterisation of avian pathogenic Escherichia coli isolated from colibacillosis cases in Uganda" *PLoS one*
4. Eid, Tolba, Hamed et al. (2022) "Bacteriophage therapy as an alternative biocontrol against emerging multidrug resistant E. coli in broilers" *Saudi J Biol Sci*
5. Yao, Bao, Hu et al. (2023) "A lytic phage to control multidrug-resistant avian pathogenic Escherichia coli (APEC) infection" *Front Cell Infect Microbiol*
6. Jalil, Masood, Ain et al. (2023) "High resistance of fluoroquinolone and macrolide reported in avian pathogenic Escherichia coli isolates from the humid subtropical regions of Pakistan" *J Glob Antimicrob Resist*
7. Nicolas, Trotereau, Culot et al. (2023) "Isolation and characterization of a novel phage collection against avian-pathogenic Escherichia coli" *Microbiol Spectr*
8. De Coster, Rademakers (2023) "NanoPack2: population-scale evaluation of long-read sequencing data" *Bioinformatics*
9. Lin, Yuan, Kolmogorov et al. (2016) "Assembly of long error-prone reads using de Bruijn graphs" *Proc Natl Acad Sci U S A*
10. Kolmogorov, Yuan, Lin et al. (2019) "Assembly of long, errorprone reads using repeat graphs" *Nat Biotechnol*
11. Mikheenko, Prjibelski, Saveliev et al. (2018) "Versatile genome assembly evaluation with QUAST-LG" *Bioinformatics*
12. Parks, Imelfort, Skennerton et al. (2015) "CheckM: assessing the quality of microbial genomes recovered from isolates, single cells, and metagenomes" *Genome Res*
13. Hunt, Silva, Otto et al. (2015) "Circlator: automated circularization of genome assemblies using long sequencing reads" *Genome Biol*
14. Tatusova, Dicuccio, Badretdin et al. (2016) "NCBI prokaryotic genome annotation pipeline" *Nucleic Acids Res*
15. Jolley, Bray, Maiden (2018) "Open-access bacterial population genomics: BIGSdb software, the PubMLST.org website and their applications" *Wellcome Open Res*
16. Joensen, Tetzschner, Iguchi et al. (2015) "Rapid and easy in silico serotyping of Escherichia coli isolates by use of whole-genome sequencing data" *J Clin Microbiol*
17. Carattoli, Zankari, García-Fernández et al. (2014) "In silico detection and typing of plasmids using PlasmidFinder and plasmid multilocus sequence typing" *Antimicrob Agents Chemother*
18. Camacho, Coulouris, Avagyan et al. (2009) "BLAST+: architecture and applications" *BMC Bioinformatics*
19. Bouras, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics*
20. Rutherford, Parkhill, Crook et al. (2000) "Artemis: sequence visualization and annotation" *Bioinformatics*
21. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol*
22. Garneau, Depardieu, Fortier et al. (2017) "PhageTerm: a tool for fast and accurate determination of phage termini and packaging mechanism using next-generation sequencing data" *Sci Rep*
23. Gilchrist, Chooi (2021) "Clinker & clustermap.js: automatic generation of gene cluster comparison figures" *Bioinformatics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12522669&blobtype=pdf | # BMC Infectious Diseases
Angila Ataei-Pirkooh, Atefeh Kachooei, Mahtab Mir-Hosseinian, Farzane Behnezhad, Mahtab Eftekhari, Somayeh-Sadat Hoseini-Fakhr, Somayeh Jalilvand, Tayebeh Latifi, Reza Khouy, Zabihollah Shoja
## Abstract
Background Acute gastroenteritis (AGE) is a prevalent gastrointestinal illness and a leading cause of morbidity and mortality among children younger than 5 years of age. Viral pathogens, group A rotavirus (RVA), norovirus (NoV), adenovirus (AdV), and astrovirus (AstV), account for over 70% of AGE cases in this age group. Co-infections with multiple enteric viruses are frequently observed in developing regions with Limited healthcare infrastructure and may contribute to increased pathogenicity, reduced vaccine efficacy, and viral evolution. This study aimed to investigate the prevalence and clinical features of RVA, NoV, AdV, and AstV infections, including patterns of viral co-infection, among children under 5 years of age with AGE in Tehran, Iran.Methods From January 2021 to January 2022, 200 stool specimens were collected from children under 5 hospitalized with AGE at children's hospitals in Tehran. They were tested for RV, AdV, and AstV using RT-PCR, and for NoV, via real-time PCR.
ResultsThe overall prevalence rate for RV, NoV, AdV, and AstV was 24%, 22%, 14%, and 0.5%, respectively. In total, 72% of the samples tested positive for at least one enteric virus, and co-infections were detected in 11.5% of them. The most common co-infection among this population was RV-AdV (25.6%), and RV was the most frequently co-detected virus with NoV, AdV, and AstV.Conclusions Our findings highlight that RV was the most frequently detected virus, followed by NoV, AdV, and AstV. The notable rate of co-infections underscores the need for multiplex diagnostics and may inform vaccination strategies to better prevent severe pediatric gastroenteritis.
## Introduction
Acute gastroenteritis (AGE) is a globally common gastrointestinal disease. In children under five, AGE is a leading cause of morbidity and mortality, predominantly in developing countries [1,2]. Worldwide, gastroenteritis affects around 3 to 5 billion children under the age of five and is responsible for 12% of annual deaths [3,4]. Although bacteria, parasites, and fungi can cause AGE, most cases in children are caused by viral infections [5]. Viral agents are responsible for over 70% of AGE in children [6]. Among viruses, group A rotavirus (RVA), norovirus (NoV), adenovirus (AdV), and astrovirus (AstV) are well-known as main causative agents of viral AGE [7].
As members of the Sedoreoviridae family, RVs are the primary viral cause of AGE-associated morbidity and death in children under five worldwide [8]. In this study, the abbreviation "RV", as used throughout the manuscript, specifically denotes RVA. RV particles consist of 11 double-stranded (ds) RNA segments that encode five or six nonstructural proteins (NSP1-NSP5/6), as well as six structural proteins (VP1-VP4, VP6, and VP7). Twelve RV groups (A-L) have been identified based on serological cross-reactivity to VP6 protein and genome sequence identities [9][10][11][12][13]. Globally, RV is a major clinical and epidemiological concern [14]. RV-associated mortality in children under five years old decreased from 528,000 to 128,000 deaths as a result of the global RV vaccination program [15,16], and there was also a reported 59% decrease in RV hospitalizations [17]. Moreover, a noticeable discrepancy has been reported in RV detection rate in children with AGE between countries without a RV vaccination program, around 38%, and those that introduced the RV vaccine, around 23% [18].
As a genus of the Caliciviridae family, NoVs are the second most prevalent cause of AGE in children younger than five after RVA, responsible for approximately 70,000 to 200,000 deaths per year, particularly in developing nations [19][20][21]. The major (VP1), minor (VP2), and nonstructural (NS) viral proteins are encoded by the three open reading frames (ORF1-3) that constitute the genome [22]. Based on VP1 amino acid sequences, 10 genogroups (G1-GX) of NoV have been recognized, although only GI and GII are known as the most medically important species worldwide [22]. GII strains have been continuously known as the main globally circulating genogroup in most studies [23][24][25].
As non-enveloped DNA viruses, human AdVs are members of the Adenoviridae family's Mastadenovirus genus [26]. HAdVs are regarded as a serious enteric virus that requires medical attention in children under the age of five. They were been documented as both sporadic and epidemic non-bacterial AGE. HAdVs classification system consist of seven species (A-G) which are divided further into 90 types [26][27][28]. Although, species HAdV-F and HAdV-A are known as enteric HAdVs based on tissue tropism and disease patterns, and associated mainly with AGE [26], children with AGE have also sporadically been reported to have HAdV-B, HAdV-C, HAdV-D, and HAdV-G [29][30][31][32][33][34]. The incidence of enteric HAdVs varies considerably in various studies and geographic locations. Accordingly, enteric HAdVs have been found to be related with 1% to 32% of AGE cases worldwide [35][36][37][38].
The prototype AstVs, as the classic human AstV, belong to Mamastrovirus 1 (MAstV1) genus of the Astroviridae family. The genome contains three ORFs (ORF1a, 1b, and 2) that encodes non-structural and capsid proteins. AstVs consist of 8 genotypes (HAstV1-HAstV8) [39]. Additionally, Melbourne (MLB; MAstV-6) and Virginia/Human-Mink-Ovinelike (VA/HMO; MAstV8 and MAstV9) are two other highly divergent HAstV species that have recently been identified in the stools of AGE patients globally [40]. AstVs are regarded as one of the primary causes of AGE in children, elderly populations, and immunocompromised individuals, though the significance of HAstVs in enteric disease has been less characterized than that of RVs, NoVs, and HAdVs [41,42]. Based on various geographical regions, the HAstV positivity rate was variable from 1 to 40% of AGE cases, with global average of 10% [39,43].
The viral AGE cases were commonly characterized by a single causal agent linked to clinical manifestations. However, co-infections with several enteric viruses may occur in the same host, particularly across developing regions, owing to the poor health state of children [44]. Indeed, viral co-infections may, as a potential contributor, influence synergistically pathogenicity as well as impact vaccine efficacy and several evolutionary mechanisms (homologous or non-homologous recombination, and reassortment) [44]. However, the mechanisms underlying viral co-infection are poorly understood.
Limited number of studies have been reported from Iran determining the viral agents of AGE. In this regard, this study aims to investigate the occurrence of RVAs, NoVs, AdVs, and AstVs along with detecting viral co-infections in cases of AGE in Iran, during the study period of 2021-2022. To further analyze, this study also investigates the association of incidence rate of detected viruses with age distribution, seasonal trends, and clinical characteristics. At the time of this study, RV vaccination had not yet been incorporated into Iran's national immunization program. More recently, since December 2024, the ROTASIIL oral vaccine has been added to the program and is now routinely provided to infants at 2, 4, and 6 months of age [45].
## Materials and methods
## Specimen collection
In the context of a routine epidemiological investigation of viral agents responsible for acute gastroenteritis (AGE), 200 stool specimens were obtained from (unvaccinated) children under five years old who were hospitalized with AGE at pediatric hospitals in Tehran, Iran, between January 2021 and January 2022. The samples were preserved at -20 °C in the laboratory of Molecular Virology Division at Pasteur Institute of Iran. It is noteworthy that the sampling occurred during the severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) pandemic and therefore checked for SARS-CoV-2 infection status based on patient records provided by the hospitals. The study was approved by the ethics committee of the Tehran University of Medical Sciences, Iran University of Medical Sciences (The ethics approval number: IR.TUMS.SPH.REC.1403.062; IR.TUMS.SPH. REC.1401.265; IR.PII.REC.1399.052), according to the Helsinki guidelines.
## Viral DNA/RNA extraction
Viral RNA/DNA was extracted from the supernatant of 10% (w/v) stool suspension prepared in PBS using a commercial viral RNA/DNA extraction kit (GeneAll, South Korea), according to the manufacturer's instructions and employing the spin column method.
## Viral detection
For the detection of RNA viruses, the extracted RNA was first denatured by heating at 97 °C for 5 min, immediately quenched in a dry ice-ethanol bath, and subsequently subjected to complementary DNA (cDNA) synthesis. The cDNA was synthesized from the viral RNA using the SinaClon First Strand cDNA synthesis kit (SINACLON, Iran) in a single cycle with a final volume of 20 µL, as per the manufacturer's protocol. Synthesized cDNA samples were stored at -20 °C and utilized as templates for PCR and real-time PCR assays targeting RVs, AstVs and NoVs, respectively.
## RV detection
The RV detection was performed by RT-PCR, using VP6-specific primers (VP6-F/VP6-R; Table 1) designed to amplify a 379 bp fragment [46,47]. Briefly, amplification reaction was performed with an aliquot of cDNA, TEMPase Hot Start 2× Master Mix (Ampliqon, Odense M, Denmark), 500 nM of each primer, to a final volume of 25 µl. The PCR cycling conditions included an initial denaturation at 95 °C for 15 min, followed by 35 cycles of denaturation at 95 °C for 30 s, annealing at 50 °C for 30 s, and extension at 72 °C for 30 s, with a final elongation step at 72 °C for 7 min. Subsequently, amplified products were visualized via electrophoresis on a 1.5% (w/v) agarose gel.
## NoV detection
NoVs were identified using real-time PCR conducted on a Rotor-Gene™ platform (Corbett Research, Qiagen, USA), employing specific primer and probe sets: COG1F; COG1R; probe Ring 1 for GI; COG2F; COG2R; and probe Ring 2 for GII (Table 1) [48][49][50][51]. Briefly, the amplification reaction was carried out in a final volume of 20 µL containing an aliquot of cDNA, 4× CAPITAL™ qPCR Probe Master Mix (Biotech Rabbit, Germany), 500 nM of each primer, and 100 nM of each probe. The thermal cycling conditions included an initial denaturation at 95 °C for 2-3 min, followed by 40 cycles of amplification at 95 °C for 5 s and 60 °C for 30 s.
## AdV detection
The presence of AdVs was assessed by PCR, using the specific primers Hex1deg (+) and Hex2deg (-) (Table 1), with the expected size of 301 bp [33,54]. All reaction conditions for the PCR amplification were the same as mentioned above. The condition of primer annealing was determined based on previously described studies [33,54]. Amplified products were visualized via electrophoresis on a 1.5% (w/v) agarose gel.
## AstV detection
The screening of HAstV was performed by RT-PCR using specific primers SF0073 and SF0076 (Table 1), with the expected size of 409 bp amplicon [52,53]. All reaction conditions for the PCR amplification were the same as mentioned above. The condition of primer annealing was determined based on previously described studies [52,53]. Amplified products were visualized via electrophoresis on a 1.5% (w/v) agarose gel.
## Statistical analysis
For statistical analysis, χ² and Student's t-tests were applied to evaluate group differences, and results were considered statistically significant at P < 0.05. Multivariate analysis was performed using multiple logistic regression, and all statistical tests were conducted using IBM SPSS Statistics software version 22 (SPSS Inc., Chicago, IL, USA).
## Results
## Demography data
The
## Associations between clinical severity and demographical data with virus infection
The AGE cases in this study included a range of viral infections, notably RVs, NoVs, AdVs, AstVs, and SARS-CoV-2 (Table 2). Common clinical features such as diarrhea, vomiting, and fever were documented, alongside other symptoms including respiratory manifestations, rash, and abdominal pain. Respiratory symptoms were found noticeably associated with NoV and AdV infections (Table 2). Moreover, mucoid type of diarrhea was found significantly linked to RVs, AdVs and AstVs (Table 2). A statistically significant association was observed between RV infection and age groups (p < 0.05), as the highest positivity detected in children aged 6-12 months, followed by those aged 13-24 months. Conversely, no significant association was seen between virus infection and sampling seasons.
## Co-infections rvs, novs, advs, AstVs and SARS-CoV-2
Single viral gastroenteritis infections accounted for 60.4% of the AGE cases, while co-infections were identified in 11.5% (23/200) of the patients. Among the co-infected cases, the most frequent combination was RV-AdV (25.6%), and no co-infections were detected between AstV and AdV/NoV. Moreover, hospitalized children with SARS-CoV-2 co-infected with RV (n = 6), AdV (n = 3), and NoV (n = 2). Interestingly, 3 children were multi-infected with RV, AdV, and SARS-CoV-2. Co-infection with all five enteric viruses has not been detected in this study (Table 3).
## Discussion
Acute gastroenteritis (AGE) remains a major global public health issue, attributed to a broad range of enteric pathogens, including bacteria, viruses, and parasites. This condition leads to substantial rates of illness and death worldwide, particularly affecting children under the age of 5, with a greater impact observed in developing settings. Among human viral agents, AstVs, AdVs, and more importantly RVAs and NoVs have been identified as key contributors to AGE [55]. RVs are recognized as the leading cause of severe gastroenteritis in children under five years of age across both high-and low-income countries, whereas NoVs are recognized for causing illness across all age groups [56].
The presented study delved into the prevalence of various viral pathogens, including RVs, NoVs, AdVs and AstVs among pediatric patients exhibiting gastrointestinal symptoms. Through the analysis of 200 hospitalized children with AGE, the study yielded valuable insights into the distribution, co-infection patterns, and associations of these viruses with demographic and clinical parameters. We detected the presence of viruses in 60.4% of the stool specimens, aligning with findings reported by Elliott et al. and Webb et al. [57,58]. Similar to findings from Nasab et al., [59] and Jalilvand et al., [60] we detected RVs in 24% of patients with AGE. Furthermore, our findings differ from those reported by Shoja et al.
[61], Mohammadi et al. [62] Moradi-Lakeh et al. [63], who documented RV prevalence rates of 11.36%, 16.5%, and 79%, respectively, in various regions of Iran. However, similar variability has been observed globally, with RV prevalence ranging from 13% to 40% in gastroenteritis cases [64][65][66]. Such discrepancies may be attributed to differences in detection methods, geographic and cultural contexts, and sample sizes [62,67]. Additionally, in our study, the highest prevalence of RV infections was observed in children aged 6-12 months, with a statistically significant correlation between RV positivity and age group (p < 0.05). These findings are in agreement with previously published data from Iran [68][69][70].
The NoV geographical prevalence in Iran has been documented from 2006 to 2014 ranging from 4.14% to 21.3% among children with AGE [61,[71][72][73]. In this study, the overall prevalence of NoV infection among children with AGE was found to be 22%, which exceeded previous reports in the country [51]. The AdVs are associated with viral diarrhea worldwide, and have led to considerable mortality rates among children. In our survey, the prevalence rate of AdV was found to be 14%. While these finding aligns with previous studies from different countries [74], they contradict prevalence data conducted in Iran [62,74]. The AstVs are a notable cause of viral gastroenteritis but have received relatively less attention in developing countries, and the results on the prevalence of HAstV among pediatric gastroenteritis patients in Iranian cities have been contradictory [43]. We detected only 1 case of AstV out of 200 cases, and the notably low prevalence rate of HAstV-positive cases (0.5%) observed contrasts with several previous reports in Iran [62,75] and a number of other countries [76][77][78]. In addition, our results indicated that mucoid diarrhea was associated with RV, AdV, and AstV infections. This finding is inconsistent with established clinical distinctions, as viral gastroenteritis is typically characterized by watery diarrhea without mucus, whereas mucoid or bloody stools are more often indicative of bacterial infections. Therefore, the observation of mucoid diarrhea in our cohort may suggest mixed infections, secondary bacterial involvement, or alternative etiologies [79].
In low-and middle-income countries, where diarrheal diseases remain a leading cause of illness and death among children under five years of age, co-infections are frequently encountered. However, current knowledge regarding the biological interplay between co-infecting pathogens is still limited [80]. In the present study, coinfections involving two or more enteric viruses were identified in 11.5% of the tested specimens, accounting for approximately 16% of the virus-positive stool samples. RV was the most prevalent virus found in co-infected cases, often in combination with NoVs, AdVs and AstVs. Previous investigations have reported co-infection rates ranging from 0.3% to 45% among pathogen-positive samples across different regions, using either conventional or molecular diagnostic methods [81][82][83][84]. This variability likely reflects differences in regional epidemiological patterns, socioeconomic status, and sanitation practices. Due to the limited access to clinical symptoms of the pediatric patients investigated in this study, no specific correlation of clinical severity and gastroenteritis viruses' co-infections was detected. However, previous reports have suggested a potential association between viral coinfections and the development of necrotizing enterocolitis in some premature infants [85,86]. Taken together, these clinical findings, along with the epidemiological data from our study, underscore the clinical significance of identifying viral co-infections as potential contributors to severe diarrheal illness, particularly in hospitalized infants under 12 months of age. The presence of co-infections in a significant number of positive cases suggests a need for multiplex testing approaches, as single-virus screening may miss co-existing pathogens that influence disease severity.
According to the results obtained from our screening of enteropathogens along with the patients' data sheet, the rate of co-infections between SARS-CoV-2 and gastroenteritis viruses was remarkably high. Co-infections involving respiratory pathogens and RV can be problematic in certain regions, particularly those with temperate climates. In such settings, the incidence of RV infections typically peaks during the winter and spring months [87]. This overlaps with the seasonal surge in acute viral respiratory tract infections [88]. This temporal concurrence may place a considerable burden on healthcare systems and heighten the risk of nosocomial transmission of RV.
It should be noted that the single-center design of the present study may restrict the extent to which these findings can be generalized to other populations. Moreover, many referenced studies rely on the detection of viral agents in stool specimens, which does not inherently confirm causation of clinical symptoms. Establishing a definitive association requires not only identification of the agent but also evidence of active viral replication and a corresponding host immune response. This comprehensive approach ensures a more accurate understanding of the relationship between the detected organism and the symptoms experienced by individuals.
## Conclusion
Our findings highlight that RVA remains the most frequently detected virus among hospitalized Iranian children with AGE, followed by NoV, AdV, and AstV. The notable rate of co-infections underscores the importance of multiplex diagnostic approaches, as single-virus testing may underestimate the burden of disease. These results have public health relevance, suggesting that national vaccination strategies and broader surveillance efforts should consider the role of co-infections in disease severity. Future multi-center studies are warranted to further elucidate the dynamics of viral gastroenteritis and to inform evidence-based prevention and control measures.
## References
1. Troeger, Forouzanfar, Rao et al. (2015) "Estimates of global, regional, and national morbidity, mortality, and aetiologies of diarrhoeal diseases: a systematic analysis for the global burden of disease study" *Lancet Infect Dis*
2. Wang, Naghavi, Allen et al. (2015) "Global, regional, and national life expectancy, all-cause mortality, and cause-specific mortality for 249 causes of death, 1980-2015: a systematic analysis for the global burden of disease study" *Lancet*
3. Salami, Fakih, Chakkour et al. (2019) "Prevalence, risk factors and seasonal variations of different enteropathogens in Lebanese hospitalized children with acute gastroenteritis" *BMC Pediatr*
4. O'ryan, Ashkenazi-Hoffnung, 'ryan-Soriano et al. (2014) "Management of acute infectious diarrhea for children living in resource-limited settings" *Expert Rev Anti-Infect Ther*
5. Guarino, Aguilar, Berkley et al. (2020) "Acute gastroenteritis in children of the world: what needs to be done?" *J Pediatr Gastroenterol Nutr*
6. Hartman, Brown, Loomis et al. (2019) "Gastroenteritis in children" *Am Fam Physician*
7. Iturriza-Gómara, Cunliffe (2020) "Viral gastroenteritis. Hunter's Tropical Medicine and Emerging Infectious Diseases"
8. Lanata, Fischer-Walker, Olascoaga et al. (2013) "Organization chergotwh, UNICEF: global causes of diarrheal disease mortality in children < 5 years of age: a systematic review" *PLoS ONE*
9. Banyai, Kemenesi, Budinski et al. (2017) "Candidate new rotavirus species in Schreiber's bats, Serbia" *Infect Genet Evol*
10. Matthijnssens, Otto, Ciarlet et al. (2012) "VP6-sequence-based cutoff values as a criterion for rotavirus species demarcation" *Arch Virol*
11. Mihalov-Kovacs, Gellert, Marton et al. (2015) "Candidate new rotavirus species in sheltered dogs" *Hungary. Emerg Infect Dis*
12. Johne, Tausch, Ulrich et al. (2023) "Genome analysis of the novel putative rotavirus species K" *Virus Res*
13. Johne, Tausch, Grützke et al. (2004) "Distantly related rotaviruses in common shrews" *Emerg Infect Dis*
14. Doro, Farkas, Martella et al. (2015) "Zoonotic transmission of rotavirus: surveillance and control" *Expert Rev Anti-infect Ther*
15. Troeger, Khalil, Rao et al. (2018) "Rotavirus vaccination and the global burden of rotavirus diarrhea among children younger than 5 years" *JAMA Pediatr*
16. Tate, Burton, Boschi-Pinto et al. (2016) "World Health Organization-Coordinated Global Rotavirus Surveillance N. Global, regional, and National estimates of rotavirus mortality in children < 5 years of age, 2000-2013" *Clin Infect Diseases: Official Publication Infect Dis Soc Am*
17. Burnett, Parashar, Tate (2020) "Global impact of rotavirus vaccination on diarrhea hospitalizations and deaths among children < 5 years old: 2006-2019" *J Infect Dis*
18. Aliabadi, Antoni, Mwenda et al. (2019) "Global impact of rotavirus vaccine introduction on rotavirus hospitalisations among children under 5 years of age, 2008-16: findings from the global rotavirus surveillance network" *Lancet Glob Health*
19. Pires, Fischer-Walker, Lanata et al. (2015) "Aetiology-specific estimates of the global and regional incidence and mortality of diarrhoeal diseases commonly transmitted through food" *PLoS ONE*
20. Cannon, Lopman, Payne et al. (2019) "Birth cohort studies assessing norovirus infection and immunity in young children: a review" *Clin Infect Dis*
21. Bányai, Estes, Martella et al. (2018) "Viral gastroenteritis" *Lancet*
22. Chhabra, De Graaf, Parra et al. (2019) "Updated classification of Norovirus genogroups and genotypes" *J Gen Virol*
23. Vega, Barclay, Gregoricus et al. (2014) "Genotypic and epidemiologic trends of Norovirus outbreaks in the United States, 2009 to 2013" *J Clin Microbiol*
24. Zheng, Widdowson, Glass et al. (2006) "Molecular epidemiology of genogroup II-genotype 4 Noroviruses in the United States between 1994 and" *J Clin Microbiol*
25. Farahmand, Moghoofei, Dorost et al. (2022) "Global prevalence and genotype distribution of Norovirus infection in children with gastroenteritis: a meta-analysis on 6 years of research from 2015 to 2020" *Rev Med Virol*
26. Berk, Adenoviridae (2013) "Fields virology"
27. Matsushima, Shimizu, Kano et al. (2013) "Genome sequence of a novel virus of the species human adenovirus d associated with acute gastroenteritis" *Genome Announc*
28. Ismail, Lee, Lee et al. (2018) "Adenoviromics: mining the human adenovirus species D genome" *Front Microbiol*
29. Li, Phan, Nguyen et al. (2005) "Molecular epidemiology of adenovirus infection among pediatric population with diarrhea in Asia" *Microbiol Immunol*
30. Portes, Volotao Ede, Rocha et al. (2016) "A non-enteric adenovirus A12 gastroenteritis outbreak in Rio de Janeiro, Brazil" *Mem Inst Oswaldo Cruz*
31. Dey, Shimizu, Phan et al. (2009) "Molecular epidemiology of adenovirus infection among infants and children with acute gastroenteritis in Dhaka city" *Infect Genet Evol*
32. Jones Ms 2nd, Harrach, Ganac et al. (2007) "New adenovirus species found in a patient presenting with gastroenteritis" *J Virol*
33. Arashkia, Bahrami, Farsi et al. (2019) "Molecular analysis of human adenoviruses in hospitalized children < 5 years old with acute gastroenteritis in Tehran, Iran" *J Med Virol*
34. Banyai, Martella, Meleg et al. (2009) "Searching for HAdV-52, the putative gastroenteritis-associated human adenovirus serotype in Southern Hungary" *New Microbiol*
35. Shieh (2022) "Human adenovirus infections in pediatric population-an update on clinico-pathologic correlation" *Biomed J*
36. Akihara, Phan, Nguyen et al. (2005) "Existence of multiple outbreaks of viral gastroenteritis among infants in a day care center in Japan" *Arch Virol*
37. Shimizu, Phan, Nishimura et al. (2007) "An outbreak of adenovirus serotype 41 infection in infants and children with acute gastroenteritis in Maizuru city" *Infect Genet Evol*
38. Van, Wun, 'ryan et al. (1992) "Outbreaks of human enteric adenovirus types 40 and 41 in Houston day care centers" *J Pediatr*
39. Khamrin, Maneekarn, Ushijima (2016) "Molecular epidemiology of astroviruses. Viral gastroenteritis"
40. Jiang, Holtz, Bauer et al. (2013) "Comparison of novel MLB-clade, VA-clade and classic human astroviruses highlights constrained evolution of the classic human astrovirus nonstructural genes" *Virology*
41. Bosch, Pintó, Guix et al. (2014) *Clin Microbiol Rev*
42. Vu, Bosch, Pintó et al. (2017) "Epidemiology of classic and novel human astrovirus: gastroenteritis and beyond" *Viruses*
43. Olortegui, Rouhani, Yori et al. (1326) "Astrovirus infection and diarrhea in 8 countries" *Pediatrics*
44. Makimaa, Ingle, Baldridge (2020) "Enteric viral co-infections: pathogenesis and perspective" *Viruses*
45. Dormanesh (2025) "All about the rotavirus vaccination" *J Med Council Iran*
46. Gómara, Wong, Blome et al. (2002) "Molecular characterization of VP6 genes of human rotavirus isolates: correlation of genogroups with subgroups and evidence of independent segregation" *J Virol*
47. Kachooei, Koli, Tava et al. (2023) "Molecular characterization of rotavirus infections in children less than 5 years of age with acute gastroenteritis in Tehran" *J Med Virol*
48. Eftekhari, Kachooei, Jalilvand et al. (2023) "The predominance of recombinant norovirus GII. 4Sydney [P16] strains in children less than 5 years of age with acute gastroenteritis in Tehran" *Virus Res*
49. Kageyama, Kojima, Shinohara et al. (2003) "Broadly reactive and highly sensitive assay for Norwalk-like viruses based on real-time quantitative reverse transcription-PCR" *J Clin Microbiol*
50. Rolfe, Parmar, Mururi et al. (2007) "An internally controlled, one-step, real-time RT-PCR assay for Norovirus detection and genogrouping" *J Clin Virol*
51. Farsi, Roodbari, Nejati et al. (2018) "Prevalence and genetic diversity of Norovirus genogroup II in children less than 5 years of age with acute gastroenteritis in Tehran" *Iran. Med Microbiol Immunol*
52. Finkbeiner, Le, Holtz et al. (2009) "Detection of newly described astrovirus MLB1 in stool samples from children" *Emerg Infect Dis*
53. Kachooei, Mirhoseinian, Jalilvand et al. (0207) "Molecular characterization of human astrovirus infection in children under 5 years of age with acute gastroenteritis in Tehran, Iran, 2021-2022: coinfection with rotavirus" *Virus Genes*
54. Lennon, Cashman, Lane et al. (2007) "Prevalence and characterization of enteric adenoviruses in the south of Ireland" *J Med Virol*
55. Liu, Xu, -Q et al. (2012) "Simultaneous detection of seven enteric viruses associated with acute gastroenteritis by a multiplexed luminex-based assay" *J Clin Microbiol*
56. Tate, Burton, Boschi-Pinto et al. (2016) "Global, regional, and National estimates of rotavirus mortality in children < 5 years of age, 2000-2013" *Clin Infect Dis*
57. Elliott (2007) "Acute gastroenteritis in children" *BMJ*
58. Webb, Starr (2005) "Acute gastroenteritis in children" *Aus Fam Physician*
59. Nasab, Sabahi, Makvandi et al. (2016) "Epidemiology of rotavirus-norovirus co-infection and determination of Norovirus genogrouping among children with acute gastroenteritis in Tehran, Iran" *Iran Biomed J*
60. Jalilvand, Roohvand, Arashkia et al. (2018) "Update on epidemiology and circulating genotypes of rotavirus in Iranian children with severe diarrhea: 1986-2015" *Int J Travel Med Glob Health*
61. Shoja, Jalilvand, Mollaei-Kandelous et al. (2014) "Epidemiology of viral gastroenteritis in Iran" *Pediatr Infect Dis J*
62. Mohammadi, Amini, Akbari et al. (2020) "Prevalence and seasonal frequency of acute viral gastroenteritis in children less than 5 years in Ilam" *Iran . Entomol Appl Sci Lett*
63. Moradi-Lakeh, Shakerian, Yaghoubi et al. (2014) "Rotavirus infection in children with acute gastroenteritis in Iran: a systematic review and meta-analysis" *Int J Prev Med*
64. Junaid, Umeh, Olabode et al. (2011) "Incidence of rotavirus infection in children with gastroenteritis attending Jos university teaching hospital" *Virol J*
65. Li, Yu, Qi et al. (2010) "Study on the epidemiological of rotavirus diarrhea in Lulong in 2008-2009" *Zhonghua Shi Yan he lin Chuang Bing du Xue Za zhi = zhonghua Shiyan he Linchuang bingduxue zazhi = chinese. J Experimental Clin Virol*
66. Fischer, Rungoe, Jensen et al. (2011) "The burden of rotavirus disease in Denmark 2009-2010" *Pediatr Infect Dis J*
67. Azaran, Makvandi, Samarbafzadeh et al. (2016) "Study on rotavirus infection and its genotyping in children below 5 years in South West Iran" *Iran J Pediatr*
68. Nick, Mohebbi, Ghaemi et al. (2019) "Human rotavirus in Iran; molecular epidemiology, genetic diversity and recent updates on vaccine advances" *Gastroenterol Hepatol Bed Bench*
69. Shoja, Jalilvand, Mokhtari-Azad et al. (2013) "Epidemiology of cocirculating human rotaviruses in Iran" *Pediatr Infect Dis J*
70. Sedighi, Karami, Razzaghi et al. (2024) "The frequency of rotavirus gastroenteritis in children from West of Iran and genotyping of rotavirus isolates: a suggestion for further changes in childhood immunization program" *J Res Health Sci*
71. Motamedi-Rad, Farahmand, Arashkia et al. (2020) "VP7 and VP4 genotypes of rotaviruses cocirculating in Iran, 2015 to 2017: comparison with cogent sequences of Rotarix and RotaTeq vaccine strains before their use for universal mass vaccination" *J Med Virol*
72. Fazeli, Baghaie, Khavarinejad et al. (2010) "Hospital based study of prevalence and genotyping of Noroviruses and sapoviruses isolated from children with acute gastroenteritis referred to Masih Daneshvari hospital" *Gastroenterol Hepatol Bed Bench*
73. Romani, Mohebbi, Hosseini et al. (2008) "Prevalence of Norovirus infection in children and adults with acute gastroenteritis" *Food Environ Virol*
74. Motamedifar, Amini, Shirazi (2013) "Frequency of rotavirus and adenovirus gastroenteritis among children in Shiraz" *Iran. Iran Red Crescent Med J*
75. Hamkar, Yahyapour, Noroozi et al. (2010) "Prevalence of rotavirus, adenovirus, and astrovirus infections among patients with acute gastroenteritis in, Northern Iran" *Iran J Public Health*
76. Oyinloye, Aminu, Nimzing (2018) "Molecular detection of astrovirus in diarrhoeic stools of children in North East Nigeria" *MRJI*
77. Monastiri, Aouni, Guix et al. (2015) "Clinical surveillance for human astrovirus in Monastir region"
78. Arowolo, Ayolabi, Adeleye et al. (2020) "Molecular epidemiology of astrovirus in children with gastroenteritis in Southwestern Nigeria" *Arch Virol*
79. Bresee (2020) "Viral gastroenteritis. Principles and Practice of Pediatric Infectious Disease"
80. Bhavnani, Goldstick, Cevallos et al. (2012) "Synergistic effects between rotavirus and coinfecting pathogens on diarrheal disease: evidence from a community-based study in Northwestern Ecuador" *Am J Epidemiol*
81. Levidiotou, Gartzonika, Papaventsis et al. (2009) "Viral agents of acute gastroenteritis in hospitalized children in Greece" *Clin Microbiol Infect*
82. Cardine, Gourlain, Mouterde et al. (2002) "Epidemiology of acute viral gastroenteritis in children hospitalized in Rouen, France" *Clin Infect Dis*
83. Oh, Gaedicke, Schreier (2003) "Viral agents of acute gastroenteritis in German children: prevalence and molecular diversity" *J Med Virol*
84. Román, Wilhelmi, Colomina et al. (2003) "Acute viral gastroenteritis: proportion and clinical relevance of multiple infections in Spanish children" *J Med Microbiol*
85. Bagci, Eis-Hübinger, Franz et al. (2008) "Detection of astrovirus in premature infants with necrotizing enterocolitis" *Pediatr Infect Dis J*
86. Hunter, Upperman, Ford et al. (2008) "Understanding the susceptibility of the premature infant to necrotizing enterocolitis (NEC)" *Pediatr Res*
87. Levy, Hubbard, Eisenberg (2009) "Seasonality of rotavirus disease in the tropics: a systematic review and meta-analysis" *Int J Epidemiol*
88. Theodoropoulos, Hüsing, Dittmer et al. (2024) "Seasonal patterns of common respiratory viral infections in immunocompetent and immunosuppressed patients" *Pathogens* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12723335&blobtype=pdf | # Targeted Protein Degradation of the Latency-Associated Nuclear Antigen Evicts Kaposi Sarcoma-Associated Herpesvirus Episomes From Infected Cells In Vitro
Sarah Mcmahon, Maha Amer, | Hall, Ritu Shekhar, Rolf Renne
## Abstract
Kaposi sarcoma-associated herpesvirus (KSHV), a gamma-herpesvirus, is the main etiological agent of several tumors, including Kaposi's sarcoma (KS) and primary effusion lymphoma (PEL). The large, double-stranded viral genome of KSHV is maintained as a latent episome in the nucleus of host cells, where a small subset of viral genes is expressed that facilitate evasion of immune responses and promotion of cell survival and proliferation supporting tumorigenesis. The major latency-associated nuclear antigen (LANA) is an essential viral factor that is required for genome replication and segregation to maintain viral genomes within dividing cells. Given the essential role of LANA to maintain viral latency, efforts have focused on targeting LANA's role in replication and segregation as a mechanism to overcome latency and tumorigenesis. However, given the complexity in small molecule targeting to DNA-binding domains, current efforts focused on drugging LANA continue to reach potency milestones for clinical trials. Here, we developed a HaloTag-based PROTAC model as a proof-of-concept for targeted LANA protein degradation. Our findings highlight the power of a PROTAC-based strategy on eliminating LANA and viral persistence in tissue culture demonstrating promise in the on-going challenge for targeting this major viral factor in KSHV latency.
## 1 | Introduction
Kaposi sarcoma-associated herpesvirus (KSHV) is an oncogenic virus responsible for Kaposi's sarcoma (KS), primary effusion lymphoma (PEL), and other lymphoproliferative disorders. A major challenge presented to the KSHV field is the persistence of viral genomes in cells long-term. KSHV establishes a life-long persistence using a singular viral tethering protein, the latencyassociated nuclear antigen (LANA). LANA, identified by its distinctive nuclear speckles, is essential for episomal maintenance during latency. LANA binds to three conserved LANA binding sites (LBS1/2/3) within the terminal repeats (TRs) of viral genomes [1][2][3][4], facilitating recruitment of Origin Recognition Complexes, ensuring replication of viral genomes [5]. To facilitate faithful segregation of viral genomes during mitosis, LANA "tethers" genomes to host chromatin by binding to histone H2A/B interface via its N-terminus and viral TRs via its C-terminal DNA-binding domain [6]. This two-pronged binding mechanism enables the physical bridging of viral episomes to host chromatin to maintain long-term persistence of viral episomes and is conserved among other DNA tumor viruses, including MHV68 (mLANA), Epstein-Barr virus (EBV) (EBNA1), and Human Papillomaviruses (E2) [2,4,7,8].
Since the discovery of KSHV by [9], treatment for KS-associated malignancies remains limited, with no therapies directly targeting the etiological agent. Given the necessity of LANA in maintaining genomes long-term, LANA represents a vulnerability and promising therapeutic target. Although previous efforts have explored small molecules to disrupt LANA's DNA-binding activity, overall lead compounds either failed counter screens or lack sufficient potency [10][11][12]. PROteolysis TArgeting Chimeras (PROTACs), harness the proteasome to degrade proteins of interest, and have emerged as a powerful strategy with over 30 PROTACs entering clinical trials since 2019. However, currently the use of PROTACs is largely unexplored as an antiviral strategy. Here, we develop a PROTAC-mediated model system (HaloTag-LANA) to test the proof-of-concept that targeted HaloTag-LANA protein degradation via a small molecule Halo-PROTAC3. We demonstrate that PROTAC-mediated degradation of LANA using this model system rapidly degrades LANA and clears infected cells demonstrating that development of a LANA PROTAC for degradation is an exciting path to explore for the treatment of KS-associated diseases.
## 2 | Materials and Methods
## 2.1 | Generation of Recombinant HaloTag-LANA Virus
The following primers were designed for PCR and restriction cloning of an insert for amplifying recombination product: Fwd 5′-TACTGCCACCGCCTCCATAATTTTACTTTGGTTGTCAGAC CAGATTTCCCGAGGATGG and CAGAAATCGGTACTGGCTT TCCATTC-3′ Rev 5′-GTTAAGGGCGCGCCGGTGCTCCGTCC CGACCTCAGGCGCATTCCCGGGGGCGCCATGTTATCGCT CTGAAAGTACAGATCCTCAGTGGTTGGCTCGCCGGAAATCT CGAGCGTCGACGCCAGTGTTACAACCAATTAACCAATTCTG ATTAGAAAAACTC-3′. pEP-KanS was used as a template for PCR and subsequently cloned into pHTN-HaloTag-CMV-neo with SalI restriction cloning and validated by Sanger sequencing. This was used to amplify our recombination template for two-step-based BAC recombination in WT-BAC16 as previously described [13,14].
## 2.2 | Immunoprecipitation
Cells were lysed in lysis buffer (7.6 mM NaH 2 PO 4 , 12.4 mM Na 2 HPO 4 , 250 mM NaCl, 30 mM NaPPi, 5 mM EDTA, 10 mM NaF, 0.1% NP-40, pH 7.0, complete protease inhibitor cocktail (Roche), MG132) and protein were normalized by BCA protein quantification (ThermoFisher) to IP 1 mg of protein per IP diluted in IP buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton-X, complete protease inhibitor cocktail (Roche), MG132). Nonspecific Ig antibody (Santa Cruz) or LANA antibody (Advanced Biotechnologies: 13-210-100) were used to IP overnight at 4°C. The following day Protein A/G magnetic beads (ThermoFisher: 88802) were washed three times with IP buffer and protein eluted.
## 2.3 | Western Blotting
Cells were lysed in RIPA and protein was resolved on SDS-PAGE gels. Proteins were transferred to nitrocellulose membranes and membranes were blocked in 5% BSA in TBST. Primary antibodies were used as indicated: LANA (Advanced Biotechnologies: 13-210-100), LANA (Rabbit; gift of Dr. Mike Lagunoff), HaloTag (Promega: G9211), Tubulin (Calbiochem: CP06), polyubiquitin (Cell Signaling: 3933), K-bZIP (Santa Cruz: sc-69797) and incubated overnight at 4°C. After incubation, membranes were washed with TBST, incubated with HRPconjugated secondary antibody (Cell Signaling) for 1 h, followed by washing with TBST. Membranes were incubated with chemiluminescence substrate (ThermoFisher) and imaged on a Biorad Chemidoc.
## 2.4 | Immunofluorescence
Cells were seeded onto coverslips, fixed in 1% paraformaldehyde for 10 min, permeabilized in 0.2% Triton X-100, and blocked in 3% BSA for 1 h. LANA primary antibody was then incubated for 1 h. Coverslips were washed with PBST and incubated with the Alexa594-conjugated anti-mouse secondary antibody in 3% BSA for 1 h protected from light. Coverslips were then washed with PBST and mounted onto slides using Vectashield hard-set mounting media with DAPI. Coverslips were imaged on a Keyence BZ-X series fluorescence microscope.
## 2.5 | Flow Cytometry
Cells were washed with PBS and then filtered in FACS Buffer (PBS, 2% FBS, 1 mM EDTA) for analysis. Gating of GFP + /GFP - was determined based on a GFP -(uninfected iSLK) and GFP + (latently infected iSLK-WT-BAC16) samples. Samples were run on the BD Symphony A3 instrument and analyzed using FlowJo.
## 2.6 | DNA Copy Number Using Quantitative PCR (qPCR)
Genomic DNA (gDNA) was extracted using the NEB kit (T3010). Samples were normalized for 50 ng total gDNA input for qPCR on a Roche LightCycler. Copy number was determined by LANA amplification using a pcDNA3.1-LANA plasmid to create a standard curve with the following primers: Fwd 5′-GCGCCCTTAACGAGAGGAAGTT-3′ and Rev 5′-TTCC TTCGCGGTTGTAGATG-3′.
## 2.7 | Quantitative Reverse Transcription-PCR (RT-qPCR)
HaloTag-iSLKs were lysed in RNA STAT-60 and RNA was purified (Zymo: R2072) with an on-column DNA digest. Purified RNA was treated with Turbo DNase (Thermo: AM2239) and repurified by phenol-chloroform extraction. RNA was reverse transcribed (Biorad: 1725038) and 150 ng of cDNA was used for RT-qPCR on the Roche Lightcycler 96.
## 3 | Results
## 3.1 | Generation of a HaloTag-LANA-BAC16 Recombinant Virus
Targeted protein degradation has emerged as an effective strategy for selectively destroying protein targets across a variety of diseases, however, no PROTAC has been developed that directly targets KSHV LANA. Here, we engineered a KSHV BAC16 clone with a HaloTag at LANA N-terminus using two-step recombination (Figure 1A) [13,14]. After transfection of recombinant BACs into 293T and coculture with uninfected iSLK, we verified the ability of cells to form stable viral latency indicated by GFP fluorescence in iSLK-HaloTag-LANA-BAC16 indistinguishable from iSLK-WT-BAC16 (Figure 1B). Using western blotting, LANA was detected in both WT and HaloTag-LANA expressing cells using a LANA antibody at equivalent levels with shifted size in the recombinant HaloTag-LANA virus due to the epitope tag (~33 kDa). We also detected the fusion HaloTag-LANA protein using an antibody to HaloTag that was specific to our HaloTag recombinant virus (Figure 1C). Lastly, given the importance of N-/C-terminal contacts for supporting viral tethering and LANA speckle formation, we examined iSLK-WT-BAC16 and iSLK-HaloTag-LANA-BAC16 cells for LANA speckles by immunofluorescence (Figure 1D). Together, these data demonstrate the successful generation of a HaloTag-LANA virus that can establish viral latency.
## 3.2 | Targeted Protein Degradation of HaloTag-LANA Using HaloPROTAC3
To test whether LANA is a degradable target we utilized Ha-loPROTAC3 a small molecule that engages Halo-tagged proteins and recruits VHL E3 ligase complexes [15]. First, we examined the selectivity and potency of LANA degradation using HaloPROTAC3 (Figure 2A). Analysis of both LANA and HaloTag protein demonstrated that HaloPROTAC3 treatment for 24 h had no effect on LANA in iSLK-WT-BAC16 cells, however, in iSLK-HaloTag-LANA cells we saw significant degradation of LANA, indicated by LANA and HaloTag signal, into mid-nanomolar range. Importantly, loss of LANA protein was not due to clicking of the HaloPROTAC3 small molecule onto the N-terminus as experiments using the inactive enantiomer of HaloPROTAC3 did not cause LANA protein loss (Figure 2B).
To test if LANA degradation was occurring in a proteosomedependent manner, we performed PROTAC treatment studies in the presence and absence of the proteosome inhibitor, MG132. Cells were either pretreated with MG132 or DMSO for 2 h to allow for proteosome inhibition and either HaloPRO-TAC3 or DMSO was subsequently added (Figure 2C). As previously seen, treatment of cells with HaloPROTAC3 leads to LANA degradation, however, this effect was lost in the presence of MG132 demonstrating that LANA is being degraded via the proteosome. Additionally, to demonstrate if HaloPROTAC3 is engaging HaloTag-LANA and the VHL E3 ligase complex to ubiquitinate LANA, we determined the ubiquitination status of LANA. iSLK-HaloTag-LANA cells were treated with MG132 and HaloPROTAC3 as performed earlier, and protein lysates were used for immunoprecipitating LANA (Figure 2D). Given the broad effects of MG132, immunoprecipitation was necessary to more specifically detect LANA ubiquitination. Immunoprecipitation was validated by the presence of LANA in the LANA IP but not in nonspecific Ig pull-downs. LANA ubiquitination was detected using a poly-ubiquitin antibody whether in the presence or absence of PROTAC treatment. While ubiquitinated LANA was detected under proteosome inhibition, in HaloPROTAC3-treated lysate, we saw increased ubiquitin signal. These data, together with proteosome inhibitor studies, validate that LANA is being degraded in a PROTACmediated manner.
Lastly, to evaluate the kinetics and resynthesis of LANA following HaloPROTAC3, we determined how rapidly LANA is degraded following PROTAC treatment. Time course data show robust LANA degradation as early as 4 h posttreatment (Figure 3A). To determine recovery dynamics, cells were treated with HaloPROTAC3 or DMSO for 24 h, followed by drug washout and expression was monitored over 7 days (Figure 3B). Of note, cells were not kept under hygromycin selection as combined HaloPROTAC3 and hygromycin selection led to rapid cell death. Interestingly, cells treated with HaloPROTAC3 for only 24 h reduced LANA to a significant extent that subsequent rebound was not possible indicating the therapeutic potential and impact this strategy holds for the future of developing PROTACs to treat KSHV-associated malignancies.
## 3.3 | PROTAC-Mediated LANA Degradation Eliminates Viral Genomes From Latently Infected Cells
We next tested the significance of targeted LANA degradation using HaloPROTAC3 on viral genome persistence. For these studies, we took advantage of the BAC16 system which harbors a hygromycin selection cassette and GFP marker. Latently infected iSLKs can survive in the context of viral genome loss and stochastically lose genomes (GFP positivity) in the absence of antibiotic selection. To determine if LANA degradation accelerated genome loss, Day 5 post selection removal or treatment with HaloPROTAC3, we performed flow cytometry for GFP expression (Figure 4A). Hygromycin removal in HaloTag-LANA-BAC16 infected cells led to 32% reduction in GFP by Day 5, in contrast, HaloPROTAC3-treated cultures observed a 72% reduction. On Day 9, we performed viral genome copy number analysis using qPCR (Figure 4B). Removing hygromycin selection reduced viral copies four-fold to 10 737 genome copies/ng of DNA; in contrast, in HaloPROTAC3-treated cells, genome copies were close to the detection limit (170 genome copies/ng of DNA). Lastly, we examined viral gene expression (Figure 4C) and protein (Figure 4D) that demonstrated LANA degradation did not trigger lytic induction, however, within 48 h of PROTAC treatment latency gene expression was significantly reduced suggestive of genome loss and LANA's transcriptional regulation of its own promoter [16]. Overall, these data provide proof-of-concept for developing PROTAC strategies targeting KSHV LANA.
## 4 | Discussion
DNA tumor viruses have evolved viral tethering proteins that are essential for maintaining a life-long viral persistence in host cells and infected individuals. For KSHV, LANA is the viral tethering protein that is sufficient to replicate and segregate viral episomes in dividing cells. Given the necessity of LANA, many efforts over the past two decades have focused on this critical viral protein. However, current strategies tested directly targeting LANA, including developing DNA-binding domain inhibitors, continue to undergo optimization for transition into clinical trials [10,11], prompting alternative approaches to be explored. Here, we generated a recombinant virus (HaloTag-LANA) and demonstrate that this HaloTag PROTAC-based strategy can degrade LANA in a proteosome-dependent manner and efficiently evict viral episomes from infected cells. Our model system highlights the promising potential of a LANAspecific targeted protein degradation strategy using either a ligand-binding domain-based PROTAC or more recently developed oligonucleotide-based PROTACs that have demonstrated successes against the oncogenic transcription factor MYC, where ligand binding has been prohibitive [17]. Additionally, unlike DNA-binding inhibitors, degrading LANA via a PRO-TAC has additional benefits given the pleiotropic nature of LANA's activity in cells that contribute to host cellular deregulation and oncogenesis [18]. Furthermore, it has been shown that KSHV-infected PEL cells are sensitive to pomalidomide, a cereblon E3-ubiquitin ligase immunomodulator [19,20], suggesting increased activity of cereblon supports these lymphomas and hijacking this machinery via a PROTAC could potentially further sensitize infected cells. In summary, we provide proof-of-concept data with our HaloTag-based system for a novel therapeutic strategy to directly target the major latency protein of KSHV, which could also be applied to other DNA tumor viruses with viral tethering proteins such as EBV and HPV.
## References
1. Garber, Hu, Renne (2002) "Latency-Associated Nuclear Antigen (LANA) Cooperatively Binds to Two Sites Within the Terminal Repeat, and Both Sites Contribute to the Ability of LANA to Suppress Transcription and to Facilitate DNA Replication" *Journal of Biological Chemistry*
2. Hellert, Weidner-Glunde, Krausze (2013) "A Structural Basis for BRD2/4-Mediated Host Chromatin Interaction and Oligomer FIGURE 4 | LANA depletion leads to episome loss in latently infected cells without lytic induction. (A) Flow cytometry analysis of GFP+ cells in HaloTag-LANA iSLKs under hygromycin selection or selection removal 5 days after DMSO or HaloPROTAC3 (300 nM) treatment. (B) Genome copy number analysis by qPCR at Day 9 DMSO or HaloPROTAC3 (300 nM) treatment in HaloTag-LANA iSLKs. (C) RT-qPCR of viral latent and lytic gene expression. Cells were treated with DMSO or HaloPROTAC3 (300 nM) for 24 or 48 h. Data are normalized to GAPDH and plotted as relative to DMSO. Data are representative of six replicates and Student's T-tests were used for statistical testing (****p < 0.00005; n.d not detected). (D) Western blot for LANA and K-bZIP with DMSO, HaloPROTAC3 (300 nM), or lytic induction (doxycycline (dox) 2 μg/mL + sodium butyrate (NaB) 2 mM) treatment for 48 h. Tubulin represents loading control. Assembly of Kaposi Sarcoma-Associated Herpesvirus and Murine Gammaherpesvirus LANA Proteins" *PLoS Pathogens*
3. Hellert, Krausze, Schulz et al. (2014) "Crystallization, Room-Temperature X-Ray Diffraction and Preliminary Analysis of Kaposi's Sarcoma Herpesvirus LANA Bound to DNA" *Acta Crystallographica Section F, Structural Biology Communications*
4. Hellert, Weidner-Glunde, Krausze (2015) "The 3D Structure of Kaposi Sarcoma Herpesvirus LANA C-Terminal Domain Bound to DNA" *Proceedings of the National Academy of Sciences of the United States of America*
5. Verma, Choudhuri, Kaul et al. (2006) "Latency-Associated Nuclear Antigen (LANA) of Kaposi's Sarcoma-Associated Herpesvirus Interacts With Origin Recognition Complexes at the LANA Binding Sequence Within the Terminal Repeats" *Journal of Virology*
6. Ballestas, Chatis, Kaye (1999) "Efficient Persistence of Extrachromosomal KSHV DNA Mediated by Latency-Associated Nuclear Antigen" *Science*
7. Bochkarev, Barwell, Pfuetzner et al. (1995) "Crystal Structure of the DNA-Binding Domain of the Epstein-Barr Virus Origin-Binding Protein EBNA" *Cell*
8. Domsic, Chen, Lu et al. (2013) "Molecular Basis for Oligomeric-DNA Binding and Episome Maintenance by KSHV LANA" *PLoS Pathogens*
9. Chang, Cesarman, Pessin (1994) "Identification of Herpesvirus-Like DNA Sequences in AIDS-Associated Kaposi's Sarcoma" *Science*
10. Berwanger, Stein, Kany (2023) "Disrupting Kaposi's Sarcoma-Associated Herpesvirus (KSHV) Latent Replication With a Small Molecule Inhibitor" *Journal of Medicinal Chemistry*
11. Beauchemin, Moerke, Faloon et al. (2014) "Assay Development and High-Throughput Screening for Inhibitors of Kaposi's Sarcoma-Associated Herpesvirus N-Terminal Latency-Associated Nuclear Antigen Binding to Nucleosomes" *SLAS Discovery*
12. Messick, Tolvinski, Zartler (2020) "Biophysical Screens Identify Fragments That Bind to the Viral DNA-Binding Proteins EBNA1 and LANA" *Molecules*
13. Nakajima, Inagaki, Espera et al. (2024) "Kaposi's Sarcoma-Associated Herpesvirus (KSHV) LANA Prevents KSHV Episomes From Degradation" *Journal of Virology*
14. Brulois, Chang, Lee (2012) "Construction and Manipulation of a New Kaposi's Sarcoma-Associated Herpesvirus Bacterial Artificial Chromosome Clone" *Journal of Virology*
15. Buckley, Raina, Darricarrere (2015) "HaloPROTACS: Use of Small Molecule PROTACs to Induce Degradation of HaloTag Fusion Proteins" *ACS Chemical Biology*
16. Renne, Barry, Dittmer et al. (2001) "Modulation of Cellular and Viral Gene Expression by the Latency-Associated Nuclear Antigen of Kaposi's Sarcoma-Associated Herpesvirus" *Journal of Virology*
17. Ai, Ma, He (2024) "Targeting Oncogenic Transcriptional Factor c-myc by Oligonucleotide PROTAC for the Treatment of Hepatocellular Carcinoma" *European Journal of Medicinal Chemistry*
18. Schulz, Freise, Stein (2023) "Kaposi Sarcoma-Associated Herpesvirus Latency-Associated Nuclear Antigen: More Than a Key Mediator of Viral Persistence" *Current Opinion in Virology*
19. Shrestha, Treco, Davis et al. (2025) "Novel Cereblon-Binding Immunomodulators Have Increased Potency Against Gammaherpesvirus-Associated Lymphomas In Vitro" *Journal of Medical Virology*
20. Gopalakrishnan, Matta, Tolani et al. (2016) "Immunomodulatory Drugs Target IKZF1-IRF4-MYC Axis in Primary Effusion Lymphoma in a Cereblon-Dependent Manner and Display Synergistic Cytotoxicity With BRD4 Inhibitors" *Oncogene* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12090773&blobtype=pdf | # Twelfth scientific biennial meeting of the Australasian Virology Society: AVS12 2024
| Virology, | Minireview, Ebony Monson, Robson Loterio, Justin Roby, Anurag Adhikari, Rowena Bull, Jill Carr, Demetra Chatzileontiadou, Colin Cheng, Fasséli Coulibaly, Samantha Davis, Joshua Deerain, Mark Douglas, Heidi Drummer, Nicholas Eyre, Wesley Freppel, Anjali Gowripalan, Emma Grant, Stephanie Gras, Jenna Guthmiller, Lara Herrero, Eva Hesping, Bethany Horsburgh, Jennifer Hyde, Marios Koutsakos, Jason Mackenzie, Jackie Mahar, Laura Mccoullough, Christopher Mcmillan, Naphak Modhiran, Rhys Parry, Damian Purcell, Daniel Rawle, Andrii Slonchak, Peter Speck, Gilda Tachedjian, Thomas Tu, Gregory Moseley, Johanna Fraser, Michelle Tate
## Abstract
The Australasian Virology Society (AVS) holds premier biennial virology meetings that foster multidisciplinary research and collaboration and promote equity and inclusion of early-career researchers. The 12th AVS meeting (AVS12), convened by M. Tate, J. Fraser, and G. Moseley, was held from 2 to 5 December 2024 on Dja Dja Wurrung country at the RACV Goldfields Resort in Creswick, Victoria, Australia. In this report, we give a brief overview of the history of AVS and outline the current and developing priorities for the society. We provide a summary of the insightful panel discussions held to address career development and Indigenous virology, highlight the presentations given by international plenary speakers Joe Grove and Chantal Abergel, and celebrate the recipients of the numerous awards.KEYWORDS innate immunity, virus-host interactions, animal viruses, antivirals, bacteriophages, clinical virology, epidemiology, immunology, vaccines T he Australasian Virology Society (AVS) biennial meeting is highly recognized as the premier virology conference in Australasia, fostering collaboration across diverse fields of virology and promoting multidisciplinary research. It provides an engaging platform for researchers, including early-career researchers (ECRs), to share their findings on human, animal, prokaryotic, and plant viruses. The 12th AVS meeting (AVS12) was held from 2 to 5 December 2024 on Dja Dja Wurrung country, at the RACV Goldfields Resort (Creswick, Victoria, Australia). AVS12 was attended by 280 delegates-a record number for this conference-demonstrating the ongoing growth and success of the virology community in Australasia. This report provides an overview of the history of the AVS, highlights the current focus of the society, and provides insights into ECR and Indigenous virology initiatives.
HISTORY OF AVSThe inaugural AVS meeting, originally called the Australian Virology Group (AVG), was founded by Paul Young in 2001 to provide a dedicated scientific meeting for virolo gists in Australia. The concept was embraced by the virology community with the first scientific meeting held at the Kingfisher Bay Resort on K'gari (Fraser Island, Queensland, Australia). The meeting had two major objectives. The first was to provide an enjoyable and productive forum for researchers to present and discuss their work among their collaborators and other virology researchers. The second was to provide a venue to foster and encourage the participation of our students and ECRs, whereby registration costs
were kept to a minimum and sponsorship was sought to provide travel scholarships. The first meeting was held over 5 days in December 2001, with 230 delegates attending, approximately one-third of whom were undergraduate and PhD students. AVG was a relatively informal grouping, with no official structure and governance and continued to host biennial meetings. However, after nearly 10 years, the option was raised of formalizing the group as a society so that it could fulfill a wider objective as an advocacy and lobby group for our discipline. The AVG was incorporated into the AVS in 2010, formalizing the organization as a legal entity with an appropriate governance structure and constitution. Paul Young served as the first president of AVS until 2011, laying the foundation for its growth and success (1).
From the fledgling AVG, the AVS has evolved into a mature society with gender equity and diversity underpinning the society's vision and activities, under the successive distinguished leadership of Damian Purcell (2011-2015), Nigel McMillan (2015-2017), Gilda Tachedjian (2017-2021), and Heidi Drummer (2021-2023). In recognition of their exceptional contributions to advancing AVS and fostering excellence in the field of virology, at AVS12, they were presented with the inaugural "Fellowship of the AVS (FAVS)" by the current president, Rowena Bull (2023-present) (Fig. 1).
## AVS TODAY
AVS has evolved into a highly productive organization, with over 260 current members spanning eight countries, which promotes, supports, and advocates for the discipline of virology. The Australasian region harbors a diverse and unique virome, shaped by its geographical isolation and distinctive flora and fauna. One of the society's foundational goals is mentoring the next generation of virologists to maintain strong expertise in the field throughout the Australasian region. This commitment is evident in our annual meetings, where early-career members frequently present and hold chair positions.
As global warming accelerates and the world becomes increasingly interconnected, viruses continue to pose significant catastrophic potential. While AVS traditionally focused on virologists in Australia and Aotearoa (New Zealand), the society has recently expanded its reach through online seminars and engagement with other nations in the Australasian region and beyond-a natural progression given our shared virological challenges. We also established a reciprocal program with the American Society of Virology (ASV), sending emerging leaders in virology to each other's meetings to foster collaborative research. As our society matures, we have become increasingly mindful of how virological research in Australia impacts Indigenous communities. We are currently drafting recommendations for virologists conducting research on Indigenous lands.
AVS recognizes the challenging times facing our field. In the wake of the COVID-19 pandemic, we have observed increasing vaccine hesitancy, reflected in the resurgence of highly contagious viruses like measles. In response, AVS is collaborating with our members to create engaging educational videos that address relevant public health issues for the broader community.
## OVERVIEW OF AVS12
In 2024, AVS12 featured two distinguished international keynote speakers, 12 additional plenary speakers, 46 presentations selected from submitted abstracts, 15 rapid-fire talks by ECRs, and over 160 poster presentations. The conference encompassed a comprehen sive and insightful program across a broad spectrum of virology and related disciplines, with sessions covering innate and adaptive immunity, virus-host interactions, animal viruses, antivirals, bacteriophages, clinical virology, epidemiology, Indigenous virology, structural virology, and vaccines. Several initiatives were introduced at AVS12, including a new ECR career development session, as well as a panel which discussed how the society could promote excellence in Indigenous virology.
The conference began with a formal Welcome to Country delivered by Djaara Traditional Owner Mr. Lewis Brown. It was an honor to be Welcomed, to hear of Djaara people's deep connection to their land, and to appreciate the intersection between language, culture, and knowledge that is understood by this country's first scientists. AVS strongly advocates for the respectful acknowledgment of, and meaningful engagement with, the Indigenous peoples of our region, which was further discussed during the Indigenous virology panel session.
The AVS12 Local Organizing Committee (LOC) (Table 1) delivered a highly successful meeting led by Meeting Convenors Johanna Fraser, Michelle Tate, and Gregory Moseley (Fig. 2). AVS12 would not have been possible without the generous support from our sponsors and exhibitors (see Acknowledgments).
## Equity and diversity at AVS12
The AVS has a strong commitment to promoting inclusiveness, equity, and the pre vention of discrimination (2). This includes ensuring appropriate inclusion of different ethnicities, genders, career stages, and researchers from different states and institutions across Australasia. At AVS12, there were several initiatives that supported our policy on equity and diversity, including a travel grant for an Indigenous delegate to attend AVS12 and 4 travel grants for delegates from underrepresented states/regions (outlined below and in Table 2).
There were a record-breaking 280 delegates at AVS12, with 269 delegates from Australasia: Victoria (53%), followed by Queensland (21%), New South Wales (14%), South Australia (5%), Australian Capital Territory (2%), and New Zealand (2%) (Fig. 3A). Eleven delegates attended from the United States, France, South Korea, Germany, and the United Kingdom (6%; Fig. 3B). Notably, approximately 57% of registrants were ECRs. In addition, approximately 50% of attendees identified as female (Fig. 3C). This strong participation underscores the conference's efforts toward fostering an inclusive environment. To support and encourage parents to attend and engage with the meeting, a carers room was available which included activities for children and a large screen which live streamed all presentations.
## KEY PLENARY PRESENTATIONS
## Ruth Bishop oration
The Ruth Bishop oration was provided by Joe Grove (The University of Glasgow, Scotland) on "Scaling protein structure prediction to the virosphere" (Fig. 4). Joe presented an engaging opening talk describing how significant advances in the use of artificial intelligence to predict protein structures can be applied to study the evolution of viral glycoproteins as applied to the Flaviviridae. While most of the glycoproteins of the orthoflaviruses, including divergent jingmenviruses and large genome flaviviruses, display a typical class II fusion protein fold, the E1 and E2 glycoproteins of the hepacivi ruses, pegiviruses, and pestiviruses have a unique
## Rob Webster oration
In keeping with the oration's namesake, Chantal Abergel (CNRS and Aix Marseille University, France) presented a vast virus discovery program spanning over 20 years (4) and reaching across the globe (Fig. 5). This program led to the identification of novel viruses characterized by their unprecedented size and complexity. These findings quickly overturned the initial view that giant viruses are an anomaly in the virosphere. Instead, they were found to be ubiquitous and diverse, lurking in 30,000-year-old permafrost (5), as well as in plain sight in our own backyard (6). These viruses are visible by light microscopy and boast up to 2,500 genes, more than some eukaryotic organisms. Beyond the discovery phase, Chantal described how the field has now moved toward a functional understanding of the unique biology of giant viruses thanks to robust reverse genetic systems and powerful structural biology approaches. These insights led to the identification of several hallmarks, such as membrane-less organelles called viral factories, which coordinate viral replication and assembly (7). Despite their large size, some of these viruses also evolved various strategies to compact and package the genome using virally encoded histones for Melbournevirus (8) or helical fiber-forming proteins for Mimivirus (9). Many of these viruses even have their own virophages, smaller viruses that may "infect" their associated giant virus or remain commensal (10). These two decades of bold research have mapped what used to be terra incognita within the virosphere. In doing so, they have unearthed a trove of unique particle architectures, biosynthetic pathways, and protein machinery that promise to be an exciting area of research for the years to come.
## Paul Young plenary
In their plenary, Paul Young (University of Queensland), the founder of AVS, detailed the society's history and development into a pivotal advocacy group for virology in Australasia (Fig. 6A). Paul recounted personal anecdotes and significant milestones, emphasizing the society's evolution and highlighting the importance of community and shared knowledge. He also provided some reflections on his journey, with advice to both young scientists and his peers. Finally, Paul pointed to the future and the power of partnerships between academia, industry, and government, highlighting the role the society could, and should play in helping drive a strategic national agenda for transforming the sector. In recognition of his leadership, Paul was honored with an AVS life membership (Fig. 6B).
## HIGHLIGHTING INDIGENOUS VIROLOGY AT AVS12
AVS recognizes that Indigenous communities maintain thousands of years of accumula ted knowledge regarding traditional medicines and observations on the patterns of disease transmission and resolution. Since 2019, AVS has held an Indigenous virology session at our biennial meeting (1), aiming to highlight the important research under taken by Indigenous scientists and research concerning viruses that disproportionately affect Indigenous communities. At AVS12, we aimed to reassess how we can better support Indigenous researchers; enhance Indigenous engagement in science, technol ogy, engineering, and mathematics; and respectfully engage with Indigenous communi ties and lands in our research.
To do this, AVS12 introduced a 1-hour panel discussion held during the Indigenous virology session. The panel, eloquently chaired by Justin Roby (Charles Sturt University), focused on "What can we do as a society to better promote Indigenous representation and excellence in Virology?". The discussion aimed to identify barriers to Indigenous student engagement and propose initial strategies to bridge this gap, fostering a positive legacy in the field. The panel featured Natalie Netzler (University of Auckland, New Zealand) of Sāmoan (Moto'otua, Falealili) and Māori (Ngāti Ruanui, Ngāti Hauā) heritage; Trevor Lithgow (Monash University)-who has worked with Indigenous communities to pioneer approaches for ethical research; Lloyd Dolan (Charles Sturt University) of Wiradjuri heritage; and Allison Imrie (University of Western Australia) of Indigenous Tongan heritage, who brought diverse Indigenous perspectives (Fig. 7). Topics included pathways to support Indigenous virology students, reflections on Indigenous experien ces in the workplace, the use of Indigenous species in our research, and the integration of Indigenous knowledge into modern science. The panel concluded that scientific progress benefits from local environmental and contextual interactions, emphasizing the need for greater involvement of and partnership with Traditional Owners and Custodians in future research endeavors.
Immediately following the panel discussion, Trevor Lithgow presented his research ventures into ethical bioprospecting for bacteriophages. Working collaboratively with the Wurundjeri Woi Wurrung Cultural Heritage Aboriginal Corporation, the Lithgow laboratory was able to leverage traditional knowledge and prospect for bacteriophages isolated within the Merri Creek (Melbourne, Australia). This led to the discovery of novel, minimalist phages that infect and inhibit the growth of clinical isolates of Klebsiella pneumoniae (11). Their naming by Wurundjeri elders as Merri-merri-uth nyilam marranatj (phage MMNM; meaning "dangerous Merri lurker" in Woi wurrung language) was presented with a characterization of their activity. The minimalist architecture of the phage is original and particularly attractive for understanding and manipulating virushost interactions. Natasha Jansz (Doherty Institute) then presented their research on human T-cell lymphotropic virus type 1c (HTLV-1c), a retrovirus that is endemic in Central Australian First Nations communities, with a prevalence of up to 40%. Infection with HTLV-1c leads to an increased risk of chronic lung disease and early death. Natasha's humanized mouse model coupled with long-read sequencing allowed in-depth analyses of integrated HTLV-1 genome structure and integration site selection.
## SUPPORTING AND PROMOTING EARLY AND MID-CAREER VIROLOGY RESEARCHERS AT AVS12
A key aspect of the AVS mission is to promote, encourage, support, and foster the development of early career researchers (ECRs; ≤10 years post-PhD), as well as mid-career researchers (≤15 years post-PhD), by providing them with valuable opportunities to present their work and engage with the broader virology community. This commitment is evident through the inclusion of ECRs throughout the program of biennial meet ings and virtual symposia, where ECRs can showcase their research and participate in panel and career development discussions, chair sessions, and network with established professionals.
## Career development sessions
Three career development sessions were held at AVS12 that were very well attended. Unlike previous AVS meetings, two sessions were strategically scheduled before the official conference opening, thereby setting the stage and providing ECRs with additional time to apply the knowledge acquired throughout the conference. Moderated by Ebony Monson (La Trobe University) and Robson Loterio (Burnet Institute), the first session was attended by over 100 ECRs. Presentations by Rowena Bull (University of New South Wales) and David Williams (Commonwealth Scientific and Industrial Research Organisa tion [CSIRO]) (Fig. 8A) focused on grant writing and networking, emphasizing the critical importance of professional networks in the scientific community. This was followed by a panel discussion featuring prominent researchers, Paul Young (University of Queens land), Stephanie Gras (La Trobe University), Fasséli Coulibaly (Monash University), Stacey Lynch (CSIRO), and Byron Shue (National Institute of Health, USA) (Fig. 8B). The panelists discussed various topics, such as navigating career transitions, securing funding, and building collaborative research projects.
The "Meet the Professor's Lunch" brought together over 120 ECRs, providing them with a unique opportunity to engage with 22 Professors or Associate Professors, informally and directly (Fig. 8C). The event facilitated meaningful exchanges, inspired new ideas, and strengthened the sense of community.
## AVS12 travel grants
AVS is committed to enabling the next generation of virology researchers to attend our meetings. AVS12 offered seven travel grants (Table 2) that were generously supported by our sponsors (see Acknowledgements). Of these travel grants, four delegates from underrepresented states/regions outside of Australia's four major cities were sponsored to attend. Two viral hepatitis researchers were also awarded travel grants sponsored by the Australian Centre for Hepatitis Virology (ACHV). Lastly, a new initiative for AVS12 was a travel grant that aimed to support the attendance of a delegate who identifies as Indigenous. Collectively, these travel grants promoted diversity at AVS12 and provided emerging leaders the opportunity to present their findings and foster new collabora tions.
Another new initiative of AVS12 was the AVS-ASV exchange program, which intends to highlight the outstanding virology research from the United States and Australasia at the AVS and ASV meetings, respectively. The ASV and AVS sponsored the travel of the outstanding ECR Jenna Guthmiller (University of Colorado, USA), the recipient of the 2024 ASV Ann Palmenberg Junior Investigator Award. Jenna presented their advances in understanding the role of antibodies in driving antigenic drift of influenza A viruses.
## AVS12 awards
In addition to the travel grants, 15 awards were presented to ECRs at AVS12. Table 3 outlines the award winners. These awards were made possible through generous support from our sponsors (see Acknowledgments). Five ECRs were recognized with prestigious awards for high-quality oral presentations. One of these awards included registration and travel costs to attend the World Society for Virology (WSV) confer ence in Kuala Lumpur in May 2025. AVS12 also provided the opportunity for ECRs to present their novel research findings during two sessions of 3-minute oral presenta tions. The audience was asked to vote on the spot, and the top 2 presenters were selected to receive the "People's Choice" awards. In addition to the main program, two poster sessions were held, providing endless opportunities for networking and gaining constructive feedback. Four outstanding poster presenters were recognized with awards that were kindly sponsored by the American Society for Microbiology, Journal of Virology. Lastly, an industry engagement award, which was selected based on the top-ranked eligible abstract, was a new initiative introduced at AVS12 that acknowl edged and recognized excellence in industry-engaged virology research.
The winner of our most prestigious AVS prize, the "AVS Rising Star Award" for an ECR ≤5 years post-PhD demonstrating potential as a future leader in the virology discipline, was won by Natalee Newton (University of Queensland) (Fig. 9A). This award was selected based on the quality of the oral presentation, contributions to the field, and evidence of emerging leadership. Newton delivered an outstanding presentation on their work on high-resolution cryo-EM analysis and antigenic characterization of diverse pathogenic tick-borne flaviviruses.
Finally, AVS12 presented a new award, "The Young Award. " This award honors the legacy of the AVS founder Paul Young and their outstanding contributions to virology and the effort they put into supporting and developing our society for more than two decades. The award recognizes an outstanding mid-career researcher who has made substantial contributions to AVS and the discipline of virology, either within or beyond the scientific community. Lara Herrero (Griffith University) was the recipient of the inaugural AVS Young Award (Fig. 9B) for significant contributions to the develop ment of antivirals for the treatment of alphavirus-induced arthritis. Professor Herrero has contributed significantly to the AVS, convening the 11th AVS meeting (AV11; Gold Coast, Australia) and serving as an AVS committee member for several years.
## FINAL REFLECTIONS
Overall, the AVS12 meeting showcased the breadth and quality of virology being undertaken in the Australasian region, spanning discovery science, translational, clinical, and epidemiological studies. The AVS consistently promotes the talent of our budding virologists and focuses on equity and diversity in our research and within our meetings. AVS12 promoted ECR career development and Indigenous virology and hopes to take the learnings from our first Indigenous virology panel session forward and generate a series of principles and policy documents to guide how respectful and inclusive Indigenous research should be conducted in Australasia. We will continue to endeavor to make this a priority moving forward. The next AVS meeting (AVS13) will be held in December 2026 in Adelaide, South Australia.
## ACKNOWLEDGMENTS
AVS acknowledges all Indigenous Peoples throughout Australasia and Djaara people in particular, upon whose traditional lands our conference convened. We pay our respects to their elders, past and present, and acknowledge Indigenous Peoples' sovereignty and their enduring connection to the traditional lands, waterways, and communities within which we pursue our virology research. We give thanks for the opportunity to continue the ancient practice of knowledge sharing at our conference, contributing to the unbroken chain of scientific knowledge that has been practiced on Dja Dja Wurrung country since time immemorial.
We thank the LOC for their extensive efforts in organizing a wonderful meeting. We also acknowledge that ASN Events, particularly Claire McParland, was instrumental in running AVS12. We are also grateful for the generous support from our sponsors and
## References
1. Helbig, Bull, Ambrose et al. (2019) "Tenth scientific biennial meeting of the Australasian virology society-AVS10"
2. Mifsud, Lytras, Oliver et al. (2024) "Mapping glycoprotein structure reveals Flaviviridae evolutionary history" *Nature*
3. Raoult, Audic, Robert et al. (2004) "The 1.2-megabase genome sequence of Mimivirus" *Science*
4. Legendre, Bartoli, Shmakova et al. (2014) "Thirty-thousand-year-old distant relative of giant icosahedral DNA viruses with a pandoravirus morphology" *Proc Natl Acad Sci*
5. Philippe, Legendre, Doutre et al. (2013) "Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes" *Science*
6. Rigou, Schmitt, Lartigue et al. (2024) "Nucleocytoviricota viral factories are transient organelles made by phase separation" *bioRxiv*
7. Liu, Bisio, Toner et al. (2021) "Virus-encoded histone doublets are essential and form nucleosome-like structures" *Cell*
8. Villalta, Schmitt, Estrozi et al. (2022) "The giant mimivirus 1.2 Mb genome is elegantly organized into a 30-nm diameter helical protein shield" *Elife*
9. Jeudy, Bertaux, Alempic et al. (2020) "Exploration of the propagation of transpovirons within Mimiviridae reveals a unique example of commensalism in the viral world" *ISME J*
10. Thung, Hall, Jati et al. (2024) "Genetic variation in individuals from a population of the minimalist bacteriophage Merri-merri-uth nyilam marra-natj driving evolution of the virus" *MBio* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12736371&blobtype=pdf | # Cell-Free Genomic DNA Release into Serum of Wild Boar and Domestic Pigs Infected with Highly Virulent African Swine Fever Virus
Ann Olesen, Louise Lohse, Graham Belsham
## Abstract
African swine fever virus (ASFV) is the cause of a severe hemorrhagic disease in domestic pigs and wild boar. Currently, a highly virulent genotype II ASFV is causing massive pig mortality worldwide. In its acute form, the disease is characterized by high fever, a range of non-specific clinical signs and cell death. In this study, we demonstrate a greatly elevated level (>1000-fold) of cell-free DNA (cfDNA), more specifically, fragmented host genomic DNA (gDNA), in serum from both wild boar and domestic pigs infected with a highly virulent genotype II ASFV. Increases were also observed, to a lesser extent, in the serum levels of mitochondrial DNA (between 4-to >500-fold). For comparison, release of the cytoplasmic enzyme, lactate dehydrogenase, which is a commonly used marker for cellular damage, was also found to be elevated in some animals, but with less consistency. These results indicate that gDNA in serum (i.e., cfDNA) can be a useful marker for cell death during infection with highly virulent variants of the virus, and could be a promising biomarker to elucidate the pathogenesis of ASFV infection in both domestic pigs and wild boar in future studies.
## 1. Introduction
African swine fever (ASF) is a severe hemorrhagic disease of domestic pigs and wild boar [1,2]. The disease is caused by African swine fever virus (ASFV), a large doublestranded DNA virus, the only member of the Asfarviridae family [3] and the only known DNA arbovirus [4]. Since the introduction of a highly virulent ASFV (of genotype II) from Africa into Georgia in 2007 [5,6], the distribution of the virus has reached pandemic proportions, with massive transmission within Europe and Asia and introduction into the Americas, resulting in huge losses of swine and major socioeconomic impact worldwide [7,8]. Infection with highly virulent ASFV often leads to a peracute to acute disease progression, with high case fatality in infected domestic pigs and wild boar. Clinical signs are usually nonspecific, with high fever, depression and reduced feed intake [9][10][11][12]. In its acute form, other clinical signs include skin hemorrhages, vomiting and bloody diarrhea. Furthermore, necropsy can reveal an enlarged, fragile spleen, enlarged hemorrhagic lymph nodes and internal bleeding. During peracute disease progression, death can occur without pronounced clinical signs and few pathological findings are sometimes reported [10][11][12]. In infected suids, the virus initially infects cells of the mononuclear phagocyte system, mainly monocytes and macrophages [13], and as the disease progresses severe dysregulation of the immune system (including a cytokine storm) and cell death occurs [14]. During infection with ASFV, cell death has been attributed to both apoptosis and necrosis [15,16].
Cell death (i.e., via apoptosis and necrosis) can lead to the release of fragmented, cellular genomic DNA (cell-free DNA, cfDNA) into the circulation, and this cfDNA has been described as a useful biomarker for cell damage [17]. The presence of these biomarkers can be readily detected in serum or plasma using real-time quantitative PCR (qPCR) assays with a small target size. We have previously shown that the production of cfDNA is an indicator of severe disease in domestic pigs that were infected with a highly virulent genotype II ASFV from Poland [18]. The aim of the current study was to investigate the presence of markers of cellular damage, including cfDNA, in the serum of domestic pigs and, additionally, of wild boar that had been experimentally infected with a highly virulent genotype II ASFV from Armenia, as recently described [19].
## 2. Materials and Methods
## 2.1. Serum Samples
The serum samples were kindly provided from a study previously performed at the Animal and Plant Health Agency (APHA) in the UK. [19]. No new animal experiments were performed for the analyses presented here. Briefly, for the animal experiment, domestic pigs and wild boar were inoculated intranasally with 10 4 HAD 50 of the highly virulent genotype II ASFV, Armenia 2007. In the current analysis, serum samples obtained from 16 domestic pigs and 16 wild boar were included (see Supplementary Table S1). Blood samples were obtained from all 32 animals before ASFV inoculation, and on days 1, 2, 3, 5 and 8 post infection (dpi), as well as postmortem (PM), from pigs randomly assigned to these sampling days prior to the experiment (for specific sampling days, see Supplementary Table S1). Following sampling, the serum samples were stored at -80 • C until further analysis. Separate results from this experiment have been published previously by Sánchez-Cordón et al. [19], but there is no overlap with the analyses performed on the serum samples described here.
## 2.2. Analysis for Host DNA in the Serum Samples
For the current study, nucleic acids were extracted from the serum samples using the MagNA Pure 96 system (Roche, Basel, Switzerland), as described previously [12]. Following purification, the samples were analyzed by qPCR using the CFX Opus Real-Time PCR System (Bio-Rad, Hercules, CA, USA). For all performed qPCR assays, a positive result was defined as a fluorescent emission signal appearing above background within 42 cycles (reported as Ct values). ASFV DNA was detected using the assay described by Tignon et al. [20], and absolute quantification performed as described previously [21], while genomic and mitochondrial DNA (mtDNA) (i.e., as used to determine the level of the Sus scrofa cytoskeletal β-actin gene and Sus scrofa mitochondrial cytochrome b gene) were assayed and absolute quantification (genome copies/mL) performed, as described by Olesen et al. [18], using the assays developed by Tignon et al. [20] and Forth [22], respectively.
In addition, an assay designed to distinguish between samples from domestic pigs and wild boar [23] was performed on selected samples (see Supplementary Table S1) to confirm the source of the serum samples that had been supplied [19] and to investigate if the level of this host gene sequence in serum was also affected during ASFV infection. The assay distinguishes between SNPs in the nuclear receptor subfamily 6 group A (NR6A1) gene, located on chromosome 1 [23].
## 2.3. Analysis for a Protein Marker of Cell Damage, Lactate Dehydrogenase, in the Serum Samples
The activity of lactate dehydrogenase (LDH) in serum samples was quantified using a commercial assay kit (Sigma-Aldrich, St. Louis, MO, USA, catalog number MAK066) according to the manufacturer's instructions, as described previously [18]. In the assay, 1:10 dilutions of the serum in 1× Dulbecco's phosphate-buffered saline (1× DPBS) (Gibco Thermo Fischer Scientific, Waltham, MA, USA) were used, together with a SunriseTM absorbance microplate reader (Tecan, Männedorf, Switzerland) for the measurements. Results were calculated according to the manufacturer's instructions and are presented as milliUnits (mU)/mL.
## 2.4. Data Analyses
Production of graphs and data analyses, including calculations of the Spearman rank correlation coefficients that were used in order to compare non-normally distributed variables, were performed using GraphPad prism 9.0 (GraphPad Software, Boston, MA, USA). Standard curves for calculation of genome copy numbers were produced using Excel (LTSC, Microsoft, Redmond, WA, USA).
## 3. Results
In this study, serum samples collected from domestic pigs and wild boar before and at different time points after infection with a highly virulent ASFV (Armenia 2007) were analyzed for the levels of ASFV DNA, gDNA (beta-actin and NR6A1) and mtDNA (mtCytb), as well as for the presence of a cytoplasmic enzyme (LDH). An overview of the results obtained is presented in Supplementary Table S1. By reference to standard curves, levels of ASFV DNA, beta-actin DNA and mtCytb DNA were converted from Ct values into genome copy numbers/mL (standard curves used for conversion of Ct values to genome copy numbers are shown in Supplementary Table S2). Genome copy numbers/mL are shown in Figure 1 Levels of the beta-actin gene in serum were similar in both domestic pigs and wild boar before inoculation (0 dpi), at around 10 5 genome copies/mL (mean 2.53 × 10 5 genome copies/mL), and remained relatively stable until the sampling at 8 dpi (Figure 1B,F). In ASFV-infected domestic pigs and wild boar, the level of the beta-actin gDNA in the sera increased in samples obtained at the end of the infection, i.e., at 8 dpi and postmortem to around 10 7 genome copies/mL (mean 2.80 × 10 7 genome copies/mL). At this time (8 dpi and postmortem), levels of ASFV DNA in the blood of the infected pigs were at around 8.59 × 10 8 genome copies/mL (Figure 1A,E). Specifically, levels of beta-actin gDNA increased between 0 dpi and euthanasia (postmortem) in serum from the three ASFVinfected domestic pigs (pigs no. 2837, 2839 and 2840). This difference of more than 10 cycles (changes in Ct values from 32.9 to 36.1 to Ct values from 22.8 to 22.9, see Supplementary Table S1) represents a greater than 1000-fold increase (2 10 = 1024-fold) in gDNA levels. A similar level of increase in gDNA in serum was observed for most of the ASFV-infected wild boar sampled prior to, and at the end of, the course of infection (0 dpi and 8 dpi, respectively), namely wild boars no. 2856, 2857 and 2859. In these three wild boar, a difference of 8-10 cycles was observed (changes in Ct values from 32.0 to 33.7 to Ct values from 23.5 to 23.8), i.e., a 256-fold to a 1024-fold increase in cfDNA. One exception was wild boar no. 2858, in which no real increase in gDNA levels in serum occurred between 0 dpi and 8 dpi. This wild boar also had a lower level of ASFV DNA in serum at 8 dpi compared to the other infected animals sampled at the end of the experiment (see Supplementary Table S1). S1. The serum samples were also assayed for the presence of the LDH enzyme, a marker for tissue damage (panel (D) for domestic pigs and panel (H) for wild boar). In panels (B-D,F-H) a darker color and shading of the column indicates that ASFV DNA was detected in serum from the pig or wild boar on that sampling day (also see panels (A,E). DP = domestic pig, WB = wild boar, PM = postmortem. Created using GraphPad prism 9.0 (GraphPad Software).
During the entire course of infection, there was a correlation between the level of ASFV DNA present in serum and the presence of gDNA. Using the Spearman rank correlation coefficient, across all animals at all time points, a correspondence between the presence of ASFV DNA and the beta-actin DNA in serum of r = 0.610 was found, yielding a two tailed p-value of <0.0001. No major difference was observed when analyzing domestic pigs and wild boar separately (r = 0.645, p < 0.0001 for domestic pigs only, r = 0.647, p = 0.0001 for wild boar only).
For mtDNA, a higher initial level was observed in serum from both domestic pigs and wild boar when compared to the levels of actin gDNA. Prior to infection, the mean level was 3.07 × 10 6 genome copies/mL. This remained relatively stable until the samplings on 8 dpi (Figure 1C,G). At 8 dpi, and postmortem, the mean level of mtDNA was elevated to 6.76 × 10 7 genome copies/mL. This increase was, however, mainly driven by an increase in the mtCytb levels in domestic pigs. In these animals, levels at 8 dpi and postmortem reached a mean of 9.12 × 10 7 genome copies/mL (i.e., increased about 30-fold), compared to a mean of 3.22 × 10 7 genome copies/mL in wild boar sampled at 8 dpi (also see Figure 1C,G), which is little changed. Specifically, levels of mtDNA changed less than the gDNA in the same domestic pigs (pigs no. 2837, 2839 and 2840) and even less in the wild boar (wild boar no. 2856, 2857, 2858, 2859) when levels prior to inoculation were compared to those at the end of the course of the infection. Thus, in the three domestic pigs, a difference of around 5-9 cycles in individual pigs (changes in Ct values from 26.1 to 27.7 to Ct values from 17.5 to 21.3), representing a 32-fold to 512-fold increase in mitochondrial DNA levels, was observed. In the four wild boar, the increase in the level of mitochondrial DNA was even less pronounced, with a difference of around 2-3 cycles (changes in Ct values from 22.5 to 24.4 to Ct values from 19.7 to 21.0), representing 4-fold to 32-fold changes (see Supplementary Table S1).
Using the Spearman rank correlation coefficient, across all animals at all time points, a correlation between the levels of ASFV DNA and the mtCytb DNA in serum of r = 0.567 was found, yielding a two tailed p-value of <0.0001. However, when looking at domestic pigs and wild boar separately, this correlation was much stronger for domestic pigs (r = 0.648, p < 0.0001), when compared to wild boar (r = 0.383, p = 0.0369).
An increase in the levels of the cytoplasmic enzyme LDH was observed in serum from some, but not all, of the ASFV-infected suids (Figure 1D,H). A high background level was observed in one wild boar (no. 2849) at 0 dpi (Figure 1H). Most likely this was due to hemolysis of the blood and red-staining of the serum from this animal (see Supplementary Table S1). Again, using the Spearman rank correlation, no major difference was observed when analyzing domestic pigs and wild boar separately (r = 0.646, p < 0.0001 for domestic pigs only, r = 0.657, p < 0.0001 for wild boar only). When considering all animals at all time points, the value of r = 0.560, p < 0.0001 was determined.
Assays designed to distinguish between the SNP g.299084751 C > T in the suid genomic NR6A1 gene were able to show large increases in the levels of this genomic DNA in sera from the ASFV-infected suids during the course of the infection (Table 1), consistent with the changes in beta-actin gDNA (see Figure 1B,F). Domestic pigs are expected to be homozygous for the T allele, while wild boar can be either homozygous for the C allele or heterozygous with both alleles. The latter is demonstrated for two of the wild boar (animals no. 2853 and no. 2854) that tested positive for both the C allele and the T allele on 0 dpi and 5 dpi (Table 1). In ASFV-infected wild boar, elevated levels of the C allele of the NR6A1 gene were observed in three out of four wild boar at 8 dpi, while the same was observed for the T allele of the gene in three ASFV-infected domestic pigs at 8 dpi and again postmortem.
Table 1. Detection of NR6A1 gDNA in serum from domestic pigs and wild boar. Nucleic acids isolated from selected serum samples were assayed for the levels of NR6A1 gDNA, for the T allele (DP) and C allele (WB). Results are presented as Ct values. The colored text indicates that ASFV DNA was detected in serum from the pig or wild boar on that sampling day-with the text in red indicating ASFV DNA levels above 10 8 ASFV DNA genome copies/mL and the text in orange indicating a level of ASFV DNA below 10 8 genome copies/mL (also see Figure 1A,E).
## 4. Discussion
During infection, with highly virulent variants of the genotype II ASFV, a peracute to acute disease progression is most often observed in infected wild boar and domestic pigs. This presents as a rapid disease progression, with high fever and a range of mostly nonspecific clinical signs [9][10][11][12]18,19]. The infection is associated with severe dysregulation of the immune system and cell death [13,14]. Cell death can lead to the release of fragmented cfDNA into the circulation [17], and could thus serve as a marker for acute infection with the currently circulating virulent ASFV variants. We have previously investigated the release of various potential biomarkers for ASFV infection in domestic pigs [18]. Those pigs were infected with a highly virulent genotype II ASFV from Poland. More specifically, we looked into the release of beta-actin gDNA, cytochrome b mtDNA and the cytoplasmic LDH protein into serum obtained from the pigs [18]. In the current study, we investigated the release of the same markers into serum obtained from domestic pigs and also wild boar experimentally infected with a highly virulent genotype II ASFV from Armenia [19].
The levels of ASFV DNA in serum obtained from the infected domestic pigs and wild boar in this study are in accordance with the reported course of infection in the infected animals. In particular, the later detection of ASFV DNA in the infected domestic pigs when compared to wild boar is consistent with the reported shorter incubation period in the infected wild boar (4 days) when compared to the infected domestic pigs (7 dpi) [19]. At the end of the study period, all four infected wild boar reached the humane endpoints, while this was only the case for three out of the four infected domestic pigs [19]. This also corresponds well to ASFV DNA only being present in serum from three out of the four domestic pigs (Figure 1A) but in all four wild boar at 8 dpi (Figure 1D). Detection of ASFV DNA in serum samples in the current study corresponded well with the detection of the viral DNA in EDTA-stabilized blood samples obtained from the same animals on the same sampling days. Viral DNA was detected in EDTA-stabilized blood from wild boar no. 2851 (#51 in the previous study [19]) at 3 dpi and in the same material from three and four wild boar sampled at 5 dpi and 8 dpi, respectively. In domestic pigs, the lack of detection of ASFV DNA in serum obtained at 3 dpi and 5 dpi and the detection of ASFV DNA in only three out of four animals at 8 dpi is also in agreement with the results described for EDTA-blood samples [19].
The large increase (ca. 1000-fold) in the level of gDNA in sera from ASFV-infected domestic pigs and wild boar in this study is consistent with our earlier study, in which a large increase in levels of beta-actin gDNA was observed in sera obtained from ASFVinfected domestic pigs [18]. In the current study, it was shown that levels of gDNA (both beta-actin and NR6A1 DNA) increased in both domestic pigs and wild boar in the late stages of infection with ASFV. Furthermore, using the NR6A1 gene as a target, it was also demonstrated that this assay clearly distinguished between samples of domestic pig and wild boar origin. The assay targets the SNP g.299084751, which is associated with the number of thoracic and lumbar vertebrae in suids [24,25]. Studies have shown that the majority of domestic pigs are homozygous for the g.299084751 T, while wild boar are heterozygous or homozygous for the wild type allele g.299084751 C [24], which was also observed here.
We have previously shown that levels of mtCytb in serum increased after infection of domestic pigs with virulent ASFV, but to a lesser extent compared to gDNA [18]. This was also observed in the current study. Presumably this could be due to the higher baseline of mtDNA in serum when compared to genomic DNA. It should be noted that, as also previously discussed, the target for the mtDNA sequence in the qPCR is relatively large (274 bp) [22], which could be sub-optimal for the detection of small fragments of cfDNA [16]. For comparison, the target in the beta-actin assay is only 114 bp [20]. Interestingly, mtDNA seems to correlate better with ASFV viral loads in domestic pigs than in wild boar, although baseline levels of mtDNA in the serum from the two types of pigs were rather similar. At the end of the study, infected wild boar had viral DNA levels in serum that were comparable to those of the infected domestic pigs, so differences in viral load do not seem to explain the observed difference. Even though comparable levels of viral DNA were detected in domestic pigs and wild boar, the ASFV-infected wild boar had a more rapid and severe disease progression when compared to the infected domestic pigs. Hence, wild boar developed viremia and subsequently more marked clinical signs and elevated temperatures earlier when compared to domestic pigs. The continued course of infection, e.g., the duration of clinical courses to the humane endpoints, was, however, similar for the two species [19], and differences observed in mtDNA levels can probably not be attributed to differences in the severity of the infection in domestic pigs versus wild boar. A difference in other biomarkers of stress physiology, assessed in saliva and serum, has been shown between wild boar and domestic pigs infected with a highly virulent ASFV [26]. As also mentioned by those authors, further studies exploring the differences in stress responses [26], as well as responses to cellular damage, between domestic pigs and wild boar could be warranted.
In this study, we have demonstrated an increase in the cellular enzyme LDH in serum of both infected domestic pigs and wild boar in the later stages of infection (at 8 dpi and postmortem). The same increase has been demonstrated in domestic pigs following infection with another highly virulent genotype II ASFV [18,27]. As observed previously, the increase was less marked when compared to the change in the level of cfDNA in the infected suids [18]. In a recent study [26], LDH levels in serum have been reported to correlate strongly with the clinical scores and viral genome loads of the virus in domestic pigs and wild boar infected with a highly virulent genotype II ASFV. It should be noted that for the current study we applied a non-parametric statistical test for correlation, while the previous study applied a parametric test following logarithmic transformation of their data in order to obtain a normal distribution [26]. In general, non-parametric tests are more conservative compared to parametric tests.
As previously, the current study underlines that the parallel detection of ASFV DNA and cfDNA, using qPCR, can be a convenient way to follow the course of infection within domestic pigs during pathogenesis studies, and the current study shows that cfDNA could also be a marker of acute infection with ASFV in wild boar. For example, using an ASFV qPCR assay in which swine gDNA, e.g. beta-actin gene, is included as an internal control [20] would allow for the easy and simultaneous detection of ASFV DNA and gDNA levels during such studies.
The findings of this study demonstrate that, using the methods described in the current study, of the analyzed markers, cfDNA seems to be the optimal marker for cell death during infection with highly virulent ASFV in both domestic pigs and wild boar. As the current study and the previously reported studies on biomarkers in pigs were conducted in samples obtained from pigs infected with highly virulent genotype II ASFVs [18,26], future studies using attenuated strains of the genotype II ASFVs could be warranted to investigate the role of cfDNA in the pathogenesis of ASFV.
In conclusion, cfDNA can be a useful marker for cell death during acute infection with highly virulent variants of ASFV in both domestic pigs and wild boar, and release of cfDNA could offer one method to elucidate the pathogenesis of ASFV infections further.
## Supplementary Materials:
The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/pathogens14121228/s1, Table S1: Overview of data from pigs; Table S2: Standard curves for calculation of genome copy numbers. Institutional Review Board Statement: The animal study from which the serum samples were provided was reviewed by the APHA Animal Welfare and Ethical Review Board and conducted in accordance with the UK Animals (Scientific Procedures) Act 1986 under project license PF971B5E3, as described previously [17].
Informed Consent Statement: Not applicable.
## References
1. Dixon, Sun, Roberts (2019) "African swine fever" *Antivir. Res*
2. Sauter-Louis, Conraths, Probst et al. (1717) "African swine fever in wild boar in Europe" *Viruses*
3. Alonso, Borca, Dixon et al. (2018) "ICTV virus taxonomy profile: Asfarviridae" *J. Gen. Virol*
4. Woah (2021) "African Swine Fever"
5. Beltrán-Alcrudo, Lubroth, Depner et al. (2008) "African Swine Fever in the Caucasus"
6. Rowlands, Michaud, Heath et al. (2007) "African swine fever virus isolate, Georgia" *Emerg. Infect. Dis*
7. Gaudreault, Madden, Wilson et al. (2020) "African swine fever virus: An Emerging DNA Arbovirus" *Front. Vet. Sci*
8. Woah (2025) "African Swine Fever"
9. Gabriel, Blome, Malogolovkin et al. (2011) "Characterization of African swine fever virus Caucasus isolate in European wild boar" *Emerg. Infect. Dis*
10. Gallardo, Soler, Nieto et al. (2017) "Experimental infection of domestic pigs with African swine fever virus Lithuania 2014 genotype II field isolate" *Transbound. Emerg. Dis*
11. Guinat, Reis, Netherton et al. (2014) "Dynamics of African swine fever virus shedding and excretion in domestic pigs infected by intramuscular inoculation and contact transmission" *Vet. Res*
12. Olesen, Lohse, Boklund et al. (2017) "Transmission of African swine fever virus from infected pigs by direct contact and aerosol routes" *Vet. Microbiol*
13. Gómez-Villamandos, Bautista, Sánchez-Cordón et al. (2013) "Pathology of African swine fever: The role of monocyte-macrophage" *Virus Res*
14. Blome, Gabriel, Beer (2013) "Pathogenesis of African swine fever in domestic pigs and European wild boar" *Virus Res*
15. Ramiro-Ibanez, Ortega, Brun et al. (1996) "Apoptosis: A mechanism of cell killing and lymphoid organ impairment during acute African swine fever virus infection" *J. Gen. Virol*
16. Salguero (2020) "Comparative Pathology and Pathogenesis of African Swine Fever Infection in Swine" *Front. Vet. Sci*
17. Celec, Vlková, Lauková et al. (2018) "Cell-free DNA: The role in pathophysiology and as a biomarker in kidney diseases" *Expert Rev. Mol. Med*
18. Olesen, Lohse, Johnston et al. (2023) "Increased presence of circulating cell-free, fragmented, host DNA in pigs infected with virulent African swine fever virus" *Viruses*
19. Sánchez-Cordón, Fabian, Batten et al. (2007) "Comparative evaluation of disease dynamics in wild boar and domestic pigs experimentally inoculated intranasally with the European highly virulent African swine fever virus genotype II strain "Armenia" *Vet. Res*
20. Tignon, Gallardo, Iscaro et al. (2011) "Development and inter-laboratory validation study of an improved new real-time PCR assay with internal control for detection and laboratory diagnosis of African swine fever virus" *J. Virol. Methods*
21. Olesen, Kodama, Lohse et al. (2021) "Identification of African swine fever virus transcription within peripheral blood mononuclear cells of acutely infected pigs" *Viruses*
22. Forth "Standardisierung Eines Nicht-Invasiven Beprobungssystems zur Infektionsüberwachung bei Wildschweinen"
23. Stelder, Olesen, Belsham et al. (2023) "Potential for introduction of African swine fever virus into high-biosecurity farms by flying hematophagous insects" *Transbound. Emerg. Dis*
24. Kaltenbrunner, Mayer, Kerkhoff et al. (2019) "Differentiation between wild boar and domestic pig in food by targeting two gene loci by real-time PCR" *Sci. Rep*
25. Mikawa, Morozumi, Shimanuki et al. (2007) "Fine mapping of a swine quantitative trait locus for number of vertebrae and analysis of an orphan nuclear receptor, germ cell nuclear factor (NR6A1)" *Genome Res*
26. Carrau, Ortin-Bustillo, Munoz-Prieto et al. (2025) "Changes in saliva and serum analytes in domestic pigs and wild boar experimentally infected with African swine fever virus" *Vet. Res*
27. Karalyan, Zakaryan, Arakelova et al. (2016) "Evidence of hemolysis in pigs infected with highly virulent African swine fever virus" *Vet. World*
28. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12680325&blobtype=pdf | # Genomic characterization of novel orthohepeviruses in shrews and rats from Kenya
Carol Nawenja, Griphin Ochola, Vincent Obanda, Sheila Ommeh, Xing-Lou Yang, Yan Zhu, Bei Li, Jie-Wen Deng, Bernard Agwanda, Ben Hu
## Abstract
Rodents and shrews are two groups of small mammals living in proximity with humans and have been known to harbour a variety of zoonotic pathogens. Cross-species transmission of hepeviruses from animals, particularly the recent sporadic emergence of human infections by rat-borne hepeviruses, has posed a growing threat to human health. Here, we report the full-genome identification of two orthohepeviruses in African giant shrew (Crocidura olivieri) and black rat (Rattus rattus) from Kenya, named Co-KY2016 and Rr-KY2016, respectively, the partial polymerase gene sequences of which were previously described. Co-KY2016 is highly distinct from representative strains of all currently recognized orthohepevirus species, sharing less than 55% overall genome identity and possibly representing at least a novel virus species together with other recently reported shrew hepeviruses. Rr-KY2016 shared higher similarity with other hepeviruses of rat origin in the Rocahepevirus genus, including human-infecting strains. Our results provide more evidence that rats and shrews are reservoir hosts of hepeviruses and support previous findings that different hepeviruses have undergone co-speciation with their hosts during evolution. This study increases our understanding of the distribution and genetic diversity of hepeviruses in wildlife as well as their spillover risk in Africa. It also highlights the importance of identifying hepeviruses in rodents, shrews or other wildlife and investigating possible zoonotic transmission of hepeviruses to mitigate the emergence of future diseases that could threaten public health.
## Impact Statement
This study provides the first report of the complete genomes of rodent and shrew hepeviruses in Kenya. Although different groups have previously documented newly identified shrew hepeviruses in recent years, our study proposes for the first time that these shrew hepeviruses, including the one we discovered in Kenya, may belong to a novel species or even a new orthohepevirus genus. Meanwhile, the Kenyan rat hepevirus that we found shows close similarity to another rocahepevirus strain which caused human hepatitis. Our findings highlight the importance of rodents and shrews as natural reservoir hosts of hepeviruses and the potential risk of hepevirus transmission from wildlife in Africa. The study provides reference for prevention and precaution of future emerging infectious diseases associated with novel zoonotic hepeviruses. It also underscores the value of proactive viral discovery and targeted surveillance among those wildlife reservoirs.
## INTRODUCTION
Hepeviruses refer to a group of quasi-enveloped, single-stranded and positive-sense RNA viruses belonging to the Hepeviridae family, including the classic human hepatitis E virus (HEV) and different HEV variants from other animals. These hepatotropic viruses can infect a wide variety of mammalian and avian species and cause acute viral hepatitis in some hosts, including humans [1,2]. HEV annually infects about 20 million human cases worldwide and leads to over 40,000 deaths, posing serious global health threats, especially in low-and middle-income countries with poor sanitation and hygiene conditions where faecal-oral route has been suggested as the main route of transmission [3,4]. Although acute hepatitis E is generally self-limiting in humans, cases of HEV infection with severe symptoms have been reported in different groups of immunocompromised individuals, and chronic HEV infection has become a significant public health problem with increasing concern [5,6].
Members of the Hepeviridae family are highly diverse, with all mammalian and avian hepeviruses belonging to the Orthohepevirinae subfamily. According to the taxonomy of International Commitee on Taxonomy of Viruses (ICTV), orthohepeviruses consist of four genera, namely Paslahepevirus, Rocahepevirus, Avihepevirus and Chirohepevirus, the latter two of which are mainly found in birds and bats, respectively [7]. As these viruses have a broad host spectrum, zoonotic transmission is an important source of hepevirus infection in humans. So far, zoonotic infection of hepeviruses has been primarily associated with the genus Paslahepevirus. Of the eight genotypes within the species Paslahepevirus balayani, genotypes 1 and 2, known as the classic human HEV, are the prototypical causative agents of human hepatitis E and are mainly restricted to the human host, while other genotypes have non-human reservoirs [8,9]. Among them, genotypes 3 and 4 (termed swine HEV) are mainly carried by pigs and wild boars and are responsible for the majority of zoonotic HEV cases, which have been associated with consumption of either raw or undercooked meat [10]. However, more recently, sporadic human infections caused by HEV variants from rats (termed rat HEV) that belong to the genus Rocahepevirus (previously known as HEV-C) have been independently documented worldwide without a clear transmission route, raising new public health concerns [11]. Meanwhile, a growing number of novel HEV variants have been discovered, and more hepevirus reservoir hosts have been identified, including moose, rabbit, tree shrew and shrews (termed moose HEV, shrew HEV, etc.) [12]. The recognition of the extended host range of hepeviruses and the recent frequent occurrence of spillover from rodents necessitate the need for continuous surveillance and identification of novel zoonotic hepeviruses in wildlife.
We previously detected 250 bp partial RNA-dependent RNA polymerase sequences of two novel orthohepeviruses in shrew and rat samples, respectively [13]. Here, we report on full-length genome characterization and phylogenetic analysis of these two viruses. Our findings provide new information for understanding the genetic diversity and evolution of Hepeviridae, as well as for the precaution of zoonotic HEV emergence in Africa.
## METHODS
## Sample collection
Fieldwork was carried out in Kenya during 2016. Rodents and shrews were baited and captured using Sherman live traps. Handling of the captured animals was according to the guidelines by the American Society for Mammalogists and the National Museums of Kenya [14]. Upon retrieval from traps, animals were euthanized, dissected and internal organs were collected in 2-ml collection tubes, which were later temporarily stored in liquid nitrogen for transportation and later in -80 °C for permanent storage.
## Sample processing and RNA extraction
Tissue samples were lysed in Hanks' Balanced Salt Solution using the QIAGEN TissueLyser II as per the manufacturer's instructions (QIAGEN, Germany). Lysed samples were briefly spun in a 4 °C pre-cooled centrifuge, and supernatant was used for RNA extraction using High Pure Viral RNA Kit (Roche, Germany). RNA was stored at -80 °C or subsequently used in downstream experiments.
## Whole viral genome sequencing
RNA of two shrew and rat liver samples positive for hepevirus, as identified in previous screening experiments [13], was used in preparation of sequencing libraries, which were then subjected to next-generation sequencing (NGS) performed on a HiSeq 3000 sequencer (Illumina). The paired-end sequencing was conducted with a read length of 150 bp. Raw sequencing data were processed using fastp (v0.23.4) to remove adaptor sequences and low-quality bases. Quality-controlled reads were aligned to the silva database using Bowtie2 (v2.5.2) to filter out rRNA. The remaining clean reads were assembled into contigs using megahit (v1.2.9), with default parameters and the minimum contig length set to 1,000 bp. Resulting contigs were aligned to the National Center for Biotechnology Information (NCBI) nucleotide database using blastn for identification. To fill gaps in the genome, we employed nested PCR using Platinum Taq DNA Polymerase Kit (Invitrogen, Carlsbad, USA) with primers designed based on known genomic regions. The primer sequences are available in Table S2, available in the online Supplementary Material. PCR thermocycling conditions were set at 94 °C for 2 min followed by 40 cycles of 94 °C for 20 s, 55 °C for 30 s and 72 °C for 30 s. A final extension step was set at 72 °C for 5 min. The annealing temperature was set at 5 °C below the melting temperature of the primer pair used. The two rounds of PCR followed a similar approach, with the PCR product from round one used as a template in round two. PCR products were run on 2% agarose gel and positive samples were sent for Sanger sequencing (Sangon Biotech). 5′ and 3′ ends of the genomes were determined using the HiScript-TS 5′/3′ RACE Kit (Vazyme, China) together with Prime Star High Script Kit (Takara) following the manufacturer's instructions.
## Genomic sequence analysis and phylogenetic analysis
The complete genome sequences were analysed and annotated using BioEdit v.7.2 (Informer Technologies, USA). ORF prediction software tool (NCBI) was used to predict the ORFs. The newly identified orthohepevirus genome sequences were compared and aligned with representative orthohepevirus sequences downloaded from GenBank, and sequence identity was calculated at nt and aa levels. A genome sequence identity plot was created in SimPlot v.3.5.1. Phylogenetic analyses with ORF1 and ORF2 proteins were performed using mega 7 software with 1,000 bootstrap replicates.
## RESULTS
## Host species, sampling location and naming of the novel viruses
We previously detected partial RdRp sequences of two hepeviruses in liver samples from an African giant shrew (Crocidura olivieri) and a black rat (Rattus rattus) collected in Kitale town, western Kenya [13]. Subsequently, in this study, we conducted NGS and obtained full-length genome sequences of these two distinct hepeviruses by de novo assembly, targeted PCR and 5′/3′-RACE. We tentatively rename these two viruses as shrew HEV Co-KY2016 and rat HEV Rr-KY2016, based on the abbreviation of host species, geographic origin and sampling date.
## Genome organization and sequence comparison
The full-length genome of shrew HEV Co-KY2016 comprises 7,019 nt excluding the poly A tail, with its genome ends harbouring the 5′ and 3′ UTRs. The genome is divided into three major ORFs predicted as ORF1, ORF2 and ORF3 that encode the nonstructural protein, capsid protein and phosphoprotein, respectively (Fig. 1a). The genome sequence identity between shrew HEV Co-KY2016 and previously reported hepeviruses is less than 75%. It shows the highest similarity to the other two recently discovered shrew HEVs, namely KS12-1272 from Crocidura russula in Germany and ETH/3402 from C. olivieri in Ethiopia, sharing 74.5% and 74.2% overall genome identity, respectively (Table 1 and Fig. 1b). The ORF1 and ORF2 proteins of Co-KY2016 shared 85.4-87.1% and 88.6-89.4% sequence identities with the above two shrew HEVs, compared to 73.3-73.5% and 76.4-77.0% sequence identities at nucleotide level (Table 1). Co-KY2016 is more varied from another shrew HEV named EcYS16, which was more recently identified in Episoriculus shrew from China. They share only 58% overall genome identity, while their ORF1 and ORF2 gene sequence identities are around only 60%. Furthermore, the Kenyan shrew HEV is highly distinct from other known hepeviruses. The overall sequence identities of Co-KY2016 and representative viruses in the four genera of Orthohepevirinae are all below 55% (Table 1 and Fig. 1b).
Compared with the shrew HEV, rat HEV Rr-KY2016 demonstrates relatively high homology with known hepeviruses. It exhibits above 80% full-length genome sequence identity to a number of previously described Rocahepevirus strains, which were reported in rats or humans (Table S1). Strikingly, it shared the highest overall sequence identity (86.9%) with a rocahepevirus infecting humans in France. It is very similar in size and genomic organization to other rat HEVs. In addition to ORF1, ORF2 and ORF3, it has an additional ORF4 which overlaps ORF1 at the N terminus, albeit with a late start codon. This additional ORF is conserved among most viruses of the species Rocahepevirus ratti (Fig. 1a). However, the function of ORF4 remains unclear [15]. The sequence identities between Rr-KY2016 and selected genotype C1 strains of Rocahepevirus ratti range from 76.7 to 85.9% and 79.7 to 89.2% at the nucleotide level for ORF1 and ORF2, respectively, while the amino acid identities increase to 88.0-95.0% and 92.4-96.4%. In contrast, its similarity to C2 and C3 genotypes and to other Rocahepevirus species is significantly lower. These results classified Rr-KY2016 as a novel variant within the species Rocahepevirus ratti.
## Phylogenetic analysis and taxonomic placement
We further investigated the phylogenetic relationship among the newly sequenced shrew and rat HEVs and previously known orthohepeviruses using their ORF1 and ORF2 sequences. The Rocahepevirus ratti species contains HEV variants from both rodents and carnivores. Strains of Rocahepevirus ratti can be divided into three branches, represented by three genotypes of this species, which are found in rats (Rattus spp.), ferrets and field mice (Apodemus spp.) and were formerly designated as HEV-C1, HEV-C2 and HEV-C3, respectively [16]. In both ORF1 and ORF2 trees, the rat HEV Rr-KY2016 falls into the HEV-C1 genotype. It is clustered with other HEV-C1 strains identified in Rattus rattus and Rattus norvegicus from different continents, as well as some strains causing human infections but probably originating in rats. It shows the closest phylogenetic relationship to the human-infecting HEV-C1 strain reported in France and another Rattus rattus hepevirus from Sierra Leone in western Africa (Fig. 2). The phylogenetic analysis revealed that four shrew HEVs identified by independent teams, including Co-KY2016, formed a novel clade of orthohepeviruses. Within this clade, the Kenyan shrew HEV is more closely related to the other two shrew HEV strains also from Crocidura, but with a different geographical origin. One was discovered in Ethiopia, while the other was from Germany. These three Crocidura HEVs form a sub-clade separate from the HEV variants found in Episoriculus caudatus (Fig. 2). In comparison to the genera Rocahepevirus, Avihepevirus and Chirohepevirus, these shrew HEVs share relatively close phylogenetic relationships to Palsahepevirus, which includes the prototype human HEV, though they remain divergent from either of the two Paslahepevirus species. The topology of the phylogenetic tree of full-length genome sequences was generally consistent with that of ORF1 (Fig. S1). Our findings expand knowledge about the worldwide prevalence and genetic diversity of orthohepeviruses in small mammals.
According to ICTV, the taxonomy of hepeviruses species is based on the phylogenetic analysis of partial amino acid sequence from the methyltransferase, RNA-directed RNA polymerase and the capsid proteins (https://ictv.global/report/chapter/hepeviridae/ hepeviridae). However, the species or genus demarcation criteria are not specified, and many newly identified HEV variants have not been taxonomically assigned. Due to their phylogenetic divergence and low similarity to other hepeviruses, the shrew HEV Co-KY2016, along with the other HEV variants from Crocidura spp. shrews, is likely to be assigned at least as an independent species within Palsahepevirus. However, whether they represent a new species or even a new genus needs to be determined through further analysis.
## DISCUSSION
Here, we identified the full-length genome of two novel rodent and shrew orthohepeviruses from Kenya, compared their genome sequences, and analysed their phylogenetic relationship with other hepeviruses. Previous studies revealed a strong correlation between the phylogeny of hepeviruses and that of corresponding host taxa, suggesting the long-term coevolutionary relationship between the viruses with their hosts [12]. In our analysis, clustering of hepeviruses from the same or similar Crocidura shrew species regardless of geographical distance [12,17,18], as well as the distinct separation between the shrew HEVs and other hepeviruses, may provide more evidence for the virus-host co-speciation in the evolutionary history of hepeviruses.
Rodents have been recognized as important reservoirs of hepeviruses, as they are the major hosts of the genus Rocahepevirus (HEV-C) and harbour a high diversity of HEV-C strains [19]. Previous studies reported that rodents may also be infected with human HEV, further suggesting the possible cross-species transmission of hepeviruses among humans and rodents [20].
Moreover, an evolution study demonstrated that rodents may act as major drivers of hepevirus genealogy, and the ancient origin of paslahepeviruses may be even associated with rodent hepeviruses [17]. Besides rodents, we and other groups have discovered diverse hepeviruses in different species of shrews [12,17,18]. These findings suggest that shrews may be a previously neglected reservoir for hepeviruses. Uncovering and sequencing more novel hepeviruses from shrews in the future will help elucidate the role of these small mammals in hepevirus evolution.
Rodents and shrews are two large groups of small mammals known as zoonotic sources of emerging viruses such as hantaviruses, parahenipaviruses and Borna disease virus [21,22]. Acute hepatitis in humans caused by the interspecies transmission of animal paslahepeviruses, especially swine HEV, has been documented for a long time. In recent years, individual teams from Spain, Canada, France and China reported events of human infection by rat-derived rocahepeviruses that are distinct from prototypical human and swine HEV, which has drawn increasing attention [23][24][25][26]. Importantly, previous infection with conventional human HEV fails to provide cross-protection against rat HEV infection due to their high divergence in the antigenic properties, enhancing the risk of rat HEV to public health [27]. In addition to humans, rat HEV may also jump to other animal hosts. Previous studies have provided experimental evidence affirming that rat HEV is capable of infecting non-human primates and pigs [28,29]. Epidemiological survey also revealed herd prevalence of rat HEV infection in pig farms in Spain [30]. These findings suggested that other animals may serve as the vectors or intermediate hosts for rat HEV amplification and transmission, further increasing the likelihood of rat HEV emergence in humans.
Though we did not have experimental or epidemiological evidence for the spillover of rat HEV Rr-KY2016, its close similarity to another rat HEV identified in a human patient, with their capsid proteins (ORF2) sharing 94.7% amino acid sequence identity, implies its possible zoonotic capacity (Table S1) [26]. Moreover, a similar hepevirus strain was detected in the same rat species from Sierra Leone, indicating a wide circulation of related rat HEV in Africa. These rat HEV may have the potential to cause sporadic disease outbreaks among local human populations in Africa. Regarding the shrew HEV, little information is available about the currently known strains beyond their genome sequences, no matter of Co-KY2016 or those from Ethiopia or Germany. Future studies that analyse their virological features and investigate their spillover potential are, therefore, needed. Considering a relatively high proportion of HIV-infecting immunocompromised populations plus the inadequate sanitation and poor hygiene level in some areas, which can facilitate the virus spread, the threat that zoonotic hepeviruses pose to human health in Africa should not be underestimated. Proactive surveillance of hepeviruses targeting rodents and shrews as wildlife reservoirs is thus warranted for early identification of novel hepevirus strains with cross-species transmission risk.
In conclusion, the current study is the first to report and analyse the full-genome sequences of orthohepeviruses from rats and shrews in Kenya. The newly sequenced rat HEV is closely related to other rat-derived HEV-C strains, including ones that are known to have caused zoonotic infection, while the novel shrew HEV may be putatively classified to a novel Paslahepevirus species or a novel genus in the subfamily Orthohepevirinae. This study suggests the spillover potential of hepeviruses carried by rats and shrews in Kenya. It provides a basis for further functional studies on these viruses, which can help improve understanding of their cross-species infection risk and enhance preparedness against future disease emergence.
## References
1. Wang, Meng (2021) "Hepatitis E virus: host tropism and zoonotic infection" *Curr Opin Microbiol*
2. Wang, Meng (2021) "Structural and molecular biology of hepatitis E virus" *Comput Struct Biotechnol J*
3. Rasche, Sander, Corman et al. (2019) "Evolutionary biology of human hepatitis viruses" *J Hepatol*
4. Khuroo, Khuroo, Khuroo (2016) "Transmission of hepatitis E virus in developing countries" *Viruses*
5. Kamar, Selves, Mansuy et al. (2008) "Hepatitis E virus and chronic hepatitis in organ-transplant recipients" *N Engl J Med*
6. Wu, Wu, Xia (2020) "Hepatitis E virus infection during pregnancy" *Virol J*
7. Purdy, Drexler, Meng et al. (2022) "ICTV virus taxonomy profile: hepeviridae 2022" *J Gen Virol*
8. Santos-Silva, Hemnani, Lopez-Lopez et al. (2023) "A systematic review of hepatitis E virus detection in camels" *Vet Sci*
9. Wu, Meng, Wei et al. (2022) "Prevalence of hepatitis E Virus (HEV) in feral and farmed wild boars in Xinjiang, Northwest China" *Viruses*
10. Slot, Zaaijer, Molier et al. (2017) "Meat consumption is a major risk factor for hepatitis E virus infection" *PLoS One*
11. Chen, Wang, Zhang et al. (2025) "Substantial spillover burden of rat hepatitis E virus in humans" *Nat Commun*
12. Li, Wang, Han et al. (2025) "Identification of novel rodent and shrew orthohepeviruses sheds light on hepatitis E virus evolution" *Zool Res*
13. Onyuok, Hu, Li et al. (2019) "Molecular detection and genetic characterization of novel rna viruses in wild and synanthropic rodents and shrews in Kenya" *Front Microbiol*
14. Sikes, Gannon, Mammalogists (2011) "Guidelines of the American society of mammalogists for the use of wild mammals in research" *J Mammal*
15. Kobayashi, Takahashi, Jirintai et al. (2018) "An analysis of two open reading frames (ORF3 and ORF4) of rat hepatitis E virus genome using its infectious cDNA clones with mutations in ORF3 or ORF4" *Virus Res*
16. Wang, Li, Zhou et al. (2018) "Chevrier's field mouse (Apodemus chevrieri) and père david's vole (Eothenomys melanogaster) in China carry orthohepeviruses that form two putative novel genotypes within the species orthohepevirus C" *Virol Sin*
17. Jo, Cassiano, De Oliveira-Filho et al. (2024) "Ancient evolutionary origins of hepatitis E virus in rodents" *Proc Natl Acad Sci*
18. Haring, Litz, Jacob et al. (2024) "Detection of novel orthoparamyxoviruses, orthonairoviruses and an orthohepevirus in European white-toothed shrews" *Microb Genom*
19. Wang, Harms, Yang et al. (2020) "Orthohepevirus C: an expanding species of emerging hepatitis E virus variants" *Pathogens*
20. He, Innis, Shrestha et al. (2002) "Evidence that rodents are a reservoir of hepatitis E virus for humans in Nepal" *J Clin Microbiol*
21. Ochola, Li, Obanda et al. (2022) "Discovery of novel DNA viruses in small mammals from Kenya" *Virol Sin*
22. Sah, Mohanty, Chakraborty et al. (2022) "Langya virus: a newly identified zoonotic henipavirus" *J Med Virol*
23. Sridhar, Yip, Lo et al. (2022) "Hepatitis E virus species c infection in humans, Hong Kong" *Clin Infect Dis*
24. Andonov, Robbins, Borlang et al. (2019) "Rat Hepatitis E virus linked to severe acute hepatitis in an immunocompetent patient" *J Infect Dis*
25. Rivero-Juarez, Frias, Perez et al. (2022) "Orthohepevirus C infection as an emerging cause of acute hepatitis in Spain: first report in Europe" *J Hepatol*
26. Rodriguez, Marchand, Sessa et al. (2023) "Orthohepevirus C hepatitis, an underdiagnosed disease?" *J Hepatol*
27. Sridhar, Situ, Cai et al. (2021) "Multimodal investigation of rat hepatitis E virus antigenicity: Implications for infection, diagnostics, and vaccine efficacy" *J Hepatol*
28. Yang, Li, Li et al. (2022) "Experimental cross-species transmission of rat hepatitis E virus to rhesus and cynomolgus monkeys" *Viruses*
29. Yadav, Boley, Lee et al. (2024) "Rat hepatitis E virus cross-species infection and transmission in pigs" *PNAS Nexus*
30. Caballero-Gómez, García-Bocanegra, Cano-Terriza et al. (2025) "Zoonotic rat hepatitis E virus infection in pigs: farm prevalence and public health relevance" *Porcine Health Manag* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737675&blobtype=pdf | # An Assessment of Regional Genetic Diversity of HIV-1
Anastasiia Antonova, Anna Kuznetsova, Anna Kuznetsova, Aleksei Mazus, Ekaterina Loifman, Liudmila Grigoreva, Denis Kleimenov, Evgeniia Bykonia, Dmitry Shcheblyakov, Irina Favorskaya, Andrei Pochtovyi, Elena Tsyganova, Inna Kulikova, Andrei Plutnitskii, Vladimir Gushchin, Aleksandr Gintsburg
## Abstract
This study aimed to assess the genetic diversity of HIV-1 in the Far Eastern Federal District (Russia) to implement effective anti-epidemic measures, including the development of an anti-HIV vaccine and the selection of optimal antigens. The first stage of the study included an analysis of HIV-1 nucleotide sequences obtained in Khabarovsk city from 2022 to 2024. The second stage of the study included an additional download of nucleotide sequences from the Los Alamos HIV Sequence Database for phylogenetic cluster analysis. Additionally, an analysis of drug resistance mutations was conducted. The results showed the following distribution of HIV-1 genetic variants: A6-72.15%, CRF63-10.13%, URFs-7.59%, C-5.06%, B-3.8%, and CRF157-1.27%. The phylogenetic cluster analysis revealed a statistically significant difference in the number of clusters depending on the genetic variant. Among drug resistance mutations (DRMs), those associated with nucleoside reverse transcriptase inhibitors (NRTIs) were the most frequently observed, accounting for 55.7% (95% CI: 44.75%-66.65%). The most commonly detected NRTI DRMs were A62V (43.04%) and M184V (13.92%). The results of this study highlight several important indicators for public health, particularly in the development of vaccines aimed at combating HIV infection.
## 1. Introduction
Currently, HIV infection remains a significant threat to public health worldwide, despite notable progress in the development and availability of antiretroviral therapy (ART). By the end of 2024, the global number of people living with HIV (PLWH) was 40.8 million, with 927689 individuals in the Russian Federation [1].
One of the key characteristics of HIV-1, which simultaneously poses a challenge to achieving success in combating the HIV pandemic globally, is the extremely high genetic and, consequently, antigenic variability [2]. The high frequency of mutations and the ability to recombine lead to the formation of many HIV-1 genovariants. To date, 10 "pure" HIV-1 subtypes and over 170 circulating recombinant forms (CRFs) have been identified [3]. Moreover, the differences between these HIV-1 genovariants can affect the effectiveness of HIV prevention and therapy, complicating the selection of optimal antigens and the development of a universal vaccine [2,4,5].
Currently, there is a steady increase in HIV-1 genetic diversity over time, both globally and regionally. Assessing this genetic diversity is critical to developing effective strategies to combat the pandemic [6]. This assessment helps identify predominant genetic variants in different geographic regions, enabling the selection of optimal antigens and evaluation of the effectiveness of existing treatment regimens within target populations [6,7].
The Russian Federation has several distinctive features that can influence the course of the epidemic and the effectiveness of anti-epidemic measures: Russia encompasses a vast territory where people of various age groups, nationalities, and cultural practices coexist, with differing lifestyles and mobility patterns. Additionally, the Russian Federation shares borders with 18 countries, contributing to high genetic diversity and the rapid spread of HIV-1 due to migration processes. The unique nature of the distribution of HIV-1 genetic variants in the territory of the Russian Federation is attributed to the "founder effect"-in the late 1990s, the virus of sub-subtype A6 entered the country through injecting drug users (IDUs) in the southern regions, and then quickly and widely spread throughout the country [8]. Presently, this variant accounts for 70-80% of the total diversity. Also, the following HIV-1 genovariants have been reported in Russia: A1, B, C, F1, G, CRF01, CRF02, CRF03, CRF11, CRF63, and unique forms. The distribution of HIV genetic variants is uneven across regions, and the Far Eastern Federal District (FEFD) exhibits high genetic diversity and a predominance of recombinant forms [9].
Thus, the Far Eastern Federal District is of particular interest for studying the HIV-1 genetic diversity due to several factors. Firstly, the FEFD borders areas that are epicentres of the spread of specific HIV-1 genetic variants, such as the CRF01 and CRF07 in the countries of Southeast Asia [10]. Migration and tourist activity of the population between the FEFD and neighbouring countries create conditions for the importation and spreading of new HIV-1 variants. Secondly, the FEDF has a high prevalence of drug addiction, which can also contribute to the active transmission of HIV-1 [11].
Also, prior research has indicated a significant and pronounced upward trend in HIV incidence in the FEFD, with an annual average growth rate (AAGR) of 8.14%. A similar upward trend was also observed in prevalence, escalating from 128.23 to 347.83 cases per 100,000 people during the 2011-2022 period. Despite a high ART coverage in the FEFD (92%), viral suppression rates among those infected were the lowest in Russia, reaching only 62% [12]. Consequently, such indicators may also influence the spread of drug-resistant variants of HIV-1.
The aim of this study is to assess the genetic diversity of HIV-1 in the Far Eastern Federal District using cluster analysis methods to identify predominant subtypes, recombinant forms, and potential transmission clusters. This is essential for the implementation of effective anti-epidemic measures, including the development of an anti-HIV vaccine and the selection of optimal antigens.
## 2. Materials and Methods
## 2.1. Study Design and Data Collection
In this study, we analysed the genetic diversity of HIV-1 in the Far Eastern Federal District of Russia, an area characterised by high genetic diversity, particularly due to its borders with countries that have different HIV-1 genetic compositions.
The first stage of the study included an analysis of HIV-1 nucleotide sequences obtained in the Federal District (Khabarovsk city) from 2022 to 2024 (n = 79).
The second stage of the study included an additional download of nucleotide sequences from the Los Alamos HIV Sequence Database for phylogenetic cluster analysis.
Additionally, an analysis of drug resistance mutations was conducted for the 79 patients mentioned earlier.
## 2.2. Sequence Analysis and Dataset
The material of the first stage of the study included sequences (n = 79), obtained from HIV-infected ART-treated patients at the "Centre for AIDS and Infectious Diseases Prevention and Control" of the Khabarovsk Region Ministry of Health (Russia) (hereinafter referred to as the AIDS Centre) in the period of 2022-2024. The total number of nucleotide sequences was 79. All sequences used in this study are available through GenBank via accession numbers: PP221634-PP221712.
Each nucleotide sequence was accompanied by patient demographic and epidemiological data collected through a standardised study according to national regulations, and all data were anonymised and coded to ensure confidentiality in accordance with the ethical standards of the Russian Federation.
Firstly, the HIV-1 pol gene region, encoding the PR-RT (2253-3369 bp according to the HXB2 strain, GenBank accession number: K03455), was sequenced from 79 patient samples using the commercial genotyping kits (the AmpliSens ® HIV-Resist-Seq (Central Research Institute of Epidemiology, Moscow, Russia)).
Then, multiple sequence alignments were generated using the ClustalW module within AliView v.1.27 [13]. Regions of ambiguity in the initial alignments were subjected to manual refinement. The length of the final alignment was 1101 bp, i.e., it covered the complete protease coding region and 268 amino acids of the reverse transcriptase (2253-3353 bp according to the HXB2 strain, GenBank accession number: K03455).
Determination of the HIV-1 genetic variant was carried out using specialised online programs: COMET HIV-1 [14], HIVdbProgram Sequence Analysis (version 9.8), presented on the website of Stanford University [15], and REGA HIV-1 Subtyping Tool (version 3.46) [16], according to the algorithm described in the previous study [9]. To investigate potential recombination events, sequences containing undetermined genetic variants were subjected to additional analysis using the jpHMM algorithm [17]. To confirm the genotyping results, phylogenetic analysis was performed. Well-defined, unique recombinant forms (based on jpHMM results) were preliminarily excluded from the phylogenetic analysis. HIV-1 subtype references were downloaded (https://www.hiv.lanl.gov/content/ sequence/NEWALIGN/align.html, accessed on 4 August 2024) and further added to the analysis. Phylogenetic analyses were performed via the maximum likelihood method using IQ-TREE (version 2.0.3) with the following command-line arguments: iqtree -s [sequence alignment file] -m MFP -bb 1000. The best-fit nucleotide substitution model was determined automatically: GTR + F + R4 (chosen according to BIC). Phylogenetic tree visualisation and annotation were performed using the iTOL (Interactive Tree Of Life) v7 software [18].
For the second part of the study, nucleotide sequences of HIV-1 variants different from A6, which is the most widespread genetic variant in Russia, were selected. Among the sequences studied, these were the genetic variants: B, C, CRF63, and CRF157. Additional sequences were downloaded from the Los Alamos HIV Sequence Database (https:// www.hiv.lanl.gov/content/index/, accessed on 4 August 2025) with following criteria: 1. nucleotide sequences were obtained in large cities (with a population of more than 100 thousand people, and the city of sample collection is exactly known) of the Far Eastern Federal District (included: Khabarovsk, Vladivostok, Blagoveshchensk, Yuzhno-Sakhalinsk, Nakhodka, Chita, Yakutsk, Ulan-Ude), 2. the pol gene coding region, 3. the genetic variant, which is exactly defined and different from A6.
To identify phylogenetically related sequences, a BLAST (version 2.2.30) search (https: //www.hiv.lanl.gov/content/sequence/BASIC_BLAST/basic_blast.html, accessed on 4 August 2024) was performed on both the studied and downloaded sequences. Additionally, reference nucleotide sequences of different genetic variants and geographic clades were also added. After removing all duplicate sequences by "Patient Code", "PAT id", "Name", and "Accession", phylogenetic analysis was repeated as described previously. Additionally, background information for sequences was downloaded, including gender, age, route of infection, and country of infection. The total number of sequences downloaded from GenBank was 692 nucleotide sequences, 176 of which were from the Far East. Collection dates of sequences from the Far East ranged from 2012 to 2024.
## 2.3. Cluster Analysis
ClusterPicker v.1.2.3 software was used to identify potential transmission clusters between Russian non-A6 HIV-1 genetic variants and sequences from other regions and countries [19]. A total of 708 HIV-1 pol sequences were analysed. Phylogenetic clusters were defined as groups of sequences with an intra-cluster genetic distance threshold of ≤1.5% and 0.5% (0.015 and 0.005 nucleotide substitutions per site) and a bootstrap support value ≥ 90% [20].
## 2.4. Analysis of Drug Resistance Mutations
Analysis of drug resistance mutations (determination of DRMs to protease inhibitors (PIs), nucleoside reverse transcriptase inhibitors (NRTIs), and non-nucleoside reverse transcriptase inhibitors (NNRTIs)) was performed using the Sierra algorithm implemented in the HIVdb Program: Mutations Analysis Tool version 9.7 (Stanford University HIV Drug Resistance Database; https://hivdb.stanford.edu/hivdb/by-sequences/, accessed on 4 August 2024).
## 2.5. Statistical Analysis
Statistical analysis was performed to assess differences in demographic, epidemiological, and clinical characteristics of HIV-1 genetic variants (identified 2 groups ("pure" subtypes and recombinant forms) and resistance mutations (to PIs, NRTIs, and NNRTIs).
Categorical data assessed in the study were presented as proportions and frequencies and compared using the chi-square (χ 2 ) test; in case of instability (for small sample sizes and/or for expected observation values < 5.0 in more than 1 cell-for four-field tables or in more than 20%-for arbitrary ones), Fisher's exact test was used. DRM prevalence estimates were calculated with 95% confidence intervals (CIs). A statistically significant result was determined by p < 0.05.
Data analysis was performed using the R programming language (RStudio v.1.3.1093, Inc. Software, Boston, MA, USA) and STATISTICA v. 6.0 (StatSoft, Tulsa, OK, USA).
## 3. Results
## 3.1. Profile of the Study Cohort
A total of 79 HIV-1 sequences were analysed, obtained from ART-treated patients in the 2022-2024 period. The study cohort's average duration time on ART treatment was 1101 days, equivalent to 3 years. The median age of participants was 40 years, ranging from 29 to 51 years. Of the cohort, 45 (56.96%) individuals were male and 34 (43.04%) were female. The main risk factor was sexual contact (67, 84.81%), followed by intravenous drug user (IDU) (10, 12.66%), and two cases of mother-to-child transmission (2.53%). Median (IQR) CD4 cell count was 435 (259.75-617.75) cells/mm and median (IQR) CD4 cell percentage was 20 (13-30.25 %). Median (IQR) HIV RNA was 11,020 (1770-44,670) copies/mL
## 3.2. Phylogenetic Analysis
The combined refined results of the primary identification of the HIV-1 genetic variants and phylogenetic analysis showed the following ratio: A6-57 (72.15%), CRF63-8 of sequences (10.13%), URFs-6 (7.59%), C-4 (5.06%), B-3 (3.8%), and CRF157-1 (1.27%) (Figure 1). Unique recombinant forms were represented by different compositions of genovariants, such as BC, BG, CRF63B, and A6B.
## 3.3. Phylogenetic Clusters
The phylogenetic tree revealed 51 distinct clusters (with an intra-cluster genetic distance threshold of ≤1.5%). Among these, 15 (29.41%) clusters contained 31 sequences (17.61%) from the Far Eastern Federal District (Figure 2). These 15 clusters were composed of 10 clusters of CRF63 (66.67%), 2 clusters of CRF02 (13.33%), and 1 cluster (6.67%) for each genetic variant: B, C, and G. Thus, the phylogenetic cluster analysis revealed a statistically significant difference in the number of clusters depending on the genetic variant: CRF63 vs. CRF02 (p-value = 0.003) and CRF63 vs. B/C/G (p-value < 0.001). And 31 sequences were composed of 21 sequences of CRF63 (67.74%), 4 sequences of CRF02 (12.90%), and 2 sequences (6.45%) for each genetic variant: B, C, and G.
A comparison of cluster sizes revealed that clusters of CRF63 were larger than those of other genetic variants. Thus, the maximum cluster size for CRF63 was seven sequences, compared to only two sequences for other genetic variants. In total, 10 CRF63 clusters were identified: one with seven sequences, one with five sequences, two with three sequences, and six with two sequences. Two CRF02 clusters with two sequences were also identified. And one cluster containing two sequences was found for each genetic variant: B, C, and G.
For most clusters, the modes of transmission were unknown. However, among people with known transmission modes, infection through sexual contact prevailed. The largest cluster, formed by seven HIV-1 CRF63 sequences, deserves special attention, as it includes multiple modes of transmission, including nosocomial infection and injection drug use.
As for the origin of the infection, 11 clusters (73.33%) included 27 people (87.09%) from the studied cities of the Far Eastern Federal District (Yakutsk, Khabarovsk, Yuzhno-Sakhalinsk, Vladivostok, Birobidzhan, Ulan-Ude). The remaining four clusters (0.27%) also included people from Siberian cities (Kemerovo, Novosibirsk, Novokuznetsk, Krasnoyarsk).
With an intra-cluster genetic distance threshold of ≤0.5%, the phylogenetic tree revealed 16 distinct clusters. Among these, five (31.25%) clusters contained 10 sequences (5.68%) from the Far Eastern Federal District (Figure 2). These five clusters were composed of CRF63 and included 10 people from the studied cities of the Far Eastern Federal District (Yakutsk, Birobidzhan, and Ulan-Ude).
Moreover, on the phylogenetic tree, the studied sequences of the genetic variant CRF63 formed multiple reliable clusters among themselves, which indicates multiple cases of the transmission of the virus of this variant within the country. The sequences of the recombinant form CRF02 formed several reliable phylogenetic clusters: (1) with sequences of African origin and (2) with sequences identified in the territory of the former Soviet Union (Kyrgyzstan, Tajikistan, Uzbekistan). Also, the sequences of CRF03 formed a reliable cluster with sequences mainly from Russia, and the cluster also included one sequence each from the UK and Spain. The only sequence CRF157 detected in the study, identified in Russia, was included in a reliable cluster with the reference sequences.
The sequences of the least characteristic for Russia recombinant forms CRF01 were included in a reliable cluster with sequences from Vietnam and China; CRF07-with sequences from China, and CRF11 and "pure" subtype A7-with sequences of African origin.
The studied sequences of subtype B formed several reliable phylogenetic clusters both among themselves and with sequences from the Commonwealth of Independent States (CIS) countries, as well as Europe (Great Britain, Germany) and Asia (Thailand). The studied sequences of subtype C also formed multiple reliable ones among themselves, and also entered a reliable cluster with sequences of African origin. The studied sequences of subtype G formed several reliable phylogenetic clusters: (1) with sequences from countries mainly in Eastern Europe, and (2) with sequences of Asian origin.
Additionally, an analysis was conducted for the most common genetic variant in Russia, A6. The phylogenetic tree revealed 25 distinct clusters. Among these, eight (32%) clusters contained 19 sequences from the Far Eastern Federal District and Moscow (Central Federal District) (Figure S1). The remaining clusters were found in other cities and regions of the Russian Federation, or the data (about cities in Russia) were unclear. Sub-subtype A6 sequences were also identified, forming several significant phylogenetic clusters with sequences from Ukraine and Estonia.
## 3.4. Analysis of Drug Resistance Mutations
To assess the prevalence of key drug resistance mutations (DRMs), we analysed PR-RT sequences (n = 79) targeting resistance to protease inhibitors (PIs), nucleoside reverse transcriptase inhibitors (NRTIs), and non-nucleoside reverse transcriptase inhibitors (NNRTIs). DRMs associated with NRTIs were the most frequently observed-55.7% (95% CI: 44.75-66.65%) (44/79), followed by those related to NNRTIs-27.85% (95% CI: 17.97-37.73%) (22/79) and PIs-6.33% (95% CI: 0.96-11.70%) (5/79).
The most commonly detected NRTI DRMs were A62V (43.04%, 34/79) and M184V (13.92%, 11/79); NNRTI DRMs were E138A (11.39%, 9/79), V106I (7.59%, 6/79), and H221Y (7.59%, 6/79); PI DRMs was M46I (3.8%, 3/79). It is worth noting that A62V and E138A are polymorphic mutations for HIV-1 subtype A6.
## 4. Discussion
This study is focused on the molecular epidemiology of HIV-1 genetic variants in Russia, using the Far Eastern Federal District as a case study. The study includes an analysis of HIV-1 transmission clusters in the region, as well as an examination of mutations leading to drug resistance. The results provide practical knowledge for improving HIV-1 prevention and treatment strategies in Russia and hold particular significance for vaccine development.
The demographic and epidemiological characteristics of the study population demonstrated a predominance of males and heterosexual transmission; the median age was 40 years, which generally coincides with such indicators both in individual regions of Russia and across the country as a whole [12,21,22].
Our analysis revealed the following distribution of HIV-1 genetic variants: the subsubtype A6 virus was dominant in the Far Eastern Federal District, accounting for 72.15%, which is typical for HIV-1 in the Russian Federation and may be attributed to the "founder effect", as this genetic variant was responsible for the rapid and widespread dissemination of HIV across the country in the late 1990s [8]. Despite the significant presence of subsubtype A6, the region also exhibited high genetic diversity, including circulating and unique recombinant forms. The second most common variant was CRF63, accounting for 10.13%. This genetic variant of HIV-1 was identified in Siberia in the early 2010s, and its widespread distribution in the Far Eastern Federal District may be linked to the proximity of federal districts along with the "founder effect" [23]. The subtype B virus accounted for 3.8%, subtype C for 5.06% (which is rare in Russia), and one case of infection with a new recombinant form of HIV-1, identified in 2023 in the territory of Russia, in the Far Eastern Federal District, CRF157, was also detected. The share of recombinants was 7.59%, which corresponds to global trends towards an increase in recombinant form viruses on a global scale [2,5]. An analysis of additional sequences from the international Los Alamos National Laboratory database also identified the presence of CRF02, which entered Russia from Cameroon via Central Asian countries [24], CRF03, identified in the late 1990s in Kaliningrad (Russia) [25], subtype G, as well as sporadic detections of rare genetic variants of HIV-1 in Russia, such as CRF01, CRF07, CRF11, and A7, which are characteristic of Asian and African countries, respectively [9,26,27].
The phylogenetic cluster analysis identified 15 (or 5) clusters among the studied HIV-1 sequences. Among these, CRF63 formed a significantly larger proportion of the total clusters compared to any other non-A6 variant, suggesting more dynamic transmission patterns. It is noteworthy that these CRF63 clusters were formed by sequences obtained from the Siberian and the Far Eastern Federal Districts, indicating local transmission of the virus and underscoring the importance of national monitoring initiatives and studies of this nature.
Local transmission was also detected among the viruses of the HIV-1 genetic variants: CRF02, B, C, and G. At the same time, on the phylogenetic tree, the sequences of these genetic variants formed reliable clusters with sequences of diverse geographic origins: CRF02-from countries of the former USSR and Africa, B-from the CIS countries, Europe and Asia, C-from African countries, G-from Eastern European and Asian countries. This indicates a varied geographic origin and further spread of these genetic variants within our country. Sporadic cases of detection, without transmission clusters, were noted for the circulating recombinant forms CRF03 and CRF157, which were detected directly in Russia, as well as for CRF01, CRF07, and CRF11. For CRF03, limited introductions and reduced transmission efficiency may be attributed to the remoteness of the federal districts (Northwestern and Far Eastern), while for CRF157, a short circulation time may explain its sporadic presence. The sequences of recombinant forms CRF01 and CRF07 formed reliable phylogenetic clusters with sequences from Asian countries, which may be related to the recent popularisation of these tourist destinations.
Among the viruses of "pure" subtypes (predominantly HIV-1 sub-subtype A6), mutations associated with drug resistance to the NRTI class were prevalent. The primary mutations observed were A62V (43.04%) and M184V (13.92%). The former is a polymorphic mutation characteristic of HIV-1 sub-subtype A6, and its prevalence is linked to the dominance of this genetic variant. The M184V/I mutations confer resistance to 3TC/FTC, reducing vulnerability to these medications by more than 200 times [28][29][30], which are included in the first-line antiretroviral therapy regimen in Russia. This highlights the necessity for continuous monitoring of resistance trends to ensure the effectiveness of ART.
To date, several studies have been conducted on the development of anti-HIV vaccines, such as those in Thailand based on the CRF01 genetic variant and in South Africa focusing on clade C viruses, since these genetic variants are the most widespread in these regions [31].
Thus, when developing an HIV-1 vaccine for patients from Russia, it is essential to consider specific regional features of HIV-1, particularly the prevalence of sub-subtype A6. In 2019, an approach was proposed to create a stabilised HIV-1 envelope glycoprotein for vaccination strategies aimed at inducing broadly neutralising antibodies (bNAbs), based on the consensus sequence of all HIV-1 group M isolates [32]. This approach should probably also be considered in the context of HIV-1 sub-subtype A6 and other genetic variants that are most common in Russia. The results of this study indicate that viruses of genetic variants such as B and C, which are prevalent in Europe and America, and Africa, respectively, are also found in Russia [2].
Also, subtype-specific phenotypic differences, including coreceptor usage, replication fitness, disease progression, transmission biology, antigenicity, and mutational patterns, have been observed [33]. Even within subtypes, variations exist; for example, analysis of Pol and Env sequences has revealed mutations distinguishing A6 and A1 sequences, some in antibody-binding regions, potentially affecting immune response efficacy [34]. Subtype-specific signature amino acid residues, such as those in HIV-1 Tat protein (Indian HIV-1C), highlight the need for tailored vaccine design [35].
Thus, when assessing the effectiveness of vaccines under development, including those aimed at producing bNAbs, it is also important to use a panel of pseudoviruses that encompasses these genetic variants. Currently, no widely used panel includes the virus of sub-subtype A6 [36,37], despite its widespread distribution not only in Russia but also in European countries [9,27].
Additionally, based on the results of this study, it is important to emphasise the need for reliable epidemiological surveillance of CRF63 in the context of regional migration.
## References
1. (2025) "Documents"
2. Nair, Gettins, Fuller et al. (2024) "Global and regional genetic diversity of HIV-1 in 2010-21: Systematic review and analysis of prevalence" *Lancet Microbe*
3. Vidal, Koyalta, Richard et al. (2003) "High genetic diversity of HIV-1 strains in Chad, West Central Africa" *J. Acquir. Immune Defic. Syndr*
4. Korber, Gaschen, Yusim et al. (2001) "Evolutionary and immunological implications of contemporary HIV-1 variation" *Br. Med. Bull*
5. Hemelaar, Elangovan, Yun et al. (2019) "WHO-UNAIDS Network for HIV Isolation Characterisation. Global and regional molecular epidemiology of HIV-1, 1990-2015: A systematic review, global survey, and trend analysis" *Lancet Infect. Dis*
6. Mccutchan (2006) "Global epidemiology of HIV" *J. Med. Virol*
7. Khan, Hu, Miotto et al. (2017) "Analysis of viral diversity for vaccine target discovery" *BMC Med. Genom*
8. Bobkova (2013) "Current Status of HIV-1 Diversity and Drug Resistance Monitoring in the Former USSR" *AIDS Rev*
9. Antonova, Kazennova, Lebedev et al. (2023) "Recombinant Forms of HIV-1 in the Last Decade of the Epidemic in the Russian Federation" *Viruses*
10. Carr, Salminen, Koch et al. (1996) "Full-length sequence and mosaic structure of a human immunodeficiency virus type 1 isolate from Thailand" *J. Virol*
11. (2025) "Report on the Drug Situation in the Russian Federation in 2021"
12. Mazus, Antonova, Adgamov et al. (1716) "Current Trends in HIV Infection in the Republic of Crimea"
13. Larsson (2014) "AliView: A Fast and Lightweight Alignment Viewer and Editor for Large Datasets" *Bioinformatics*
14. Struck, Lawyer, Ternes et al. (2014) "COMET: Adaptive Context-Based Modeling for Ultrafast HIV-1 Subtype Identification" *Nucleic Acids Res*
15. Shafer (2006) "Rationale and Uses of a Public HIV Drug-Resistance Database" *J. Infect. Dis*
16. Pineda-Peña, Faria, Imbrechts et al. (2013) "Automated Subtyping of HIV-1 Genetic Sequences for Clinical and Surveillance Purposes: Performance Evaluation of the New REGA Version 3 and Seven Other Tools" *Infect. Genet. Evol*
17. Schultz, Bulla, Abdou-Chekaraou et al. (2012) "Recombination Analysis in Viruses with Circular Genomes Such as the Hepatitis B Virus" *Nucleic Acids Res*
18. Letunic, Bork (2021) "Interactive Tree Of Life (ITOL) v5: An Online Tool for Phylogenetic Tree Display and Annotation" *Nucleic Acids Res*
19. Ragonnet-Cronin, Hodcroft, Hué et al. (2013) "UK HIV Drug Resistance Database. Automated analysis of phylogenetic clusters"
20. Hassan, Pybus, Sanders et al. "Defining HIV-1 transmission clusters based on sequence data" *AIDS*
21. Nechaeva, Hiv (2021) "Situation in Russia at the Beginning of Implementation of State Strategy Aimed to Combat the Spread of HIV Infection" *Tuberc. Lung Dis*
22. Ogarkova, Antonova, Kuznetsova et al. (2023) "Current Trends of HIV Infection in the Russian Federation" *Viruses*
23. Baryshev, Bogachev, Gashnikova (2014) "HIV-1 genetic diversity in Russia: CRF63_02A1, a new HIV type 1 genetic variant spreading in Siberia" *AIDS Res. Hum. Retroviruses*
24. Moskaleychik, Laga, Delgado et al.
25. Kornilaeva, Pronin, Zhernov et al. (2015) "Rapid spread of the HIV-1 circular recombinant CRF02-AG in Russia and neighboring countries" *Vopr. Virusol*
26. Liitsola, Tashkinova, Laukkanen et al. (1998) "HIV-1 genetic subtype A/B recombinant strain causing an explosive epidemic in injecting drug users in Kaliningrad" *AIDS*
27. Montavon, Vergne, Bourgeois et al. (2002) "Identification of a new circulating recombinant form of HIV type 1, CRF11-cpx, involving subtypes A, G, J, and CRF01-AE"
28. Partsuneva, Gancheva, Dimitrova et al. "Distinct Molecular Epidemiology, Transmission Patterns, and Resistance Mutations of HIV-1 Subtypes A1, A6, and A7"
29. (2025) "HIV Drug Resistance Database. Major Nucleoside RT Inhibitor (NRTI) Resistance Mutations. Available online"
30. Boucher, Cammack, Schipper et al. (1993) "High-level resistance to (-) enantiomeric 2 ′ -deoxy-3 ′ -thiacytidine in vitro is due to one amino acid substitution in the catalytic site of human immunodeficiency virus type 1 reverse transcriptase" *Antimicrob Agents Chemother*
31. Tisdale, Kemp, Parry et al. (1993) "Rapid in vitro selection of human immunodeficiency virus type 1 resistant to 3 ′ -thiacytidine inhibitors due to a mutation in the YMDD region of reverse transcriptase" *Proc. Natl. Acad. Sci*
32. Karaseva, Sesorova (2025) "Development of a Vaccine Against HIV Infection Caused by the Human Immunodeficiency Virus: Domestic and Foreign Practice"
33. Sliepen, Han, Bontjer et al. (2019) "Structure and immunogenicity of a stabilized HIV-1 envelope trimer based on a group-M consensus sequence" *Nat. Commun*
34. Lynch, Shen, Gnanakaran et al. (2009) "Appreciating HIV type 1 diversity: Subtype differences in Env" *AIDS Res. Hum. Retroviruses*
35. Abidi, Aibekova, Davlidova et al. "Origin and evolution of HIV-1 subtype A6" *PLoS ONE*
36. Neogi, Gupta, Sahoo et al. (2012) "Genetic characterization of HIV type 1 Tat exon 1 from a southern Indian clinical cohort: Identification of unique epidemiological signature residues" *AIDS Res. Hum. Retroviruses*
37. (2025) "CATNAP: Compile, Analyze and Talle Nab Panels"
38. Decamp, Hraber, Bailer et al. (2014) "Global panel of HIV-1 Env reference strains for standardized assessments of vaccine-elicited neutralizing antibodies" *J. Virol*
39. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12754555&blobtype=pdf | # Cytomegalovirus reactivation risk after autologous hematopoietic stem cell transplantation: Results of a Tunisian study
Mouna Louiza, Ben Moussa, Yasmine Chelbi, Rachid Kharrat, Rabeb Berred, Raihane Lakhal, Manel Hamdoun, Olfa Bahri
## Abstract
BACKGROUNDCytomegalovirus (CMV) reactivation is a potentially severe complication in immunocompromised patients, yet its incidence and impact in recipients of autologous hematopoietic stem cell transplantation (AHSCT) remain insufficiently documented.
AIMTo assess the frequency, timing, and outcomes of CMV reactivation in patients undergoing AHSCT at Aziza Othmana Hospital.
METHODSWe conducted a retrospective descriptive study of all patients who underwent AHSCT between January 2022 and December 2024 and had at least one posttransplant plasma viral load (VL) assessment. CMV VL was quantified by realtime polymerase chain reaction using TaqMan probes (GeneProof ® ) with a sensitivity threshold of 150 IU/mL.
RESULTSAmong 277 AHSCT recipients, 17 (6.1%) experienced CMV reactivation. Their median age was 43 years, with a sex ratio of 0.46 (male/female). Underlying diseases included large B-cell lymphoma (n = 5), multiple myeloma (n = 3), and Hodgkin's lymphoma (n = 4). The median time to reactivation was 26 days post-Ben Moussa ML et al. CMV reactivation after autologous HSCT WJV https://www.wjgnet.com 2 December 25, 2025 Volume 14 Issue 4transplant (11 days after neutrophil recovery). Median peak VL was 1325 IU/mL (range: 150-641000 IU/mL). Six patients required antiviral therapy (median peak VL: 30150 IU/mL), while 11 had spontaneous resolution (median peak VL: 1320 IU/mL). Two patients died in the context of CMV reactivation.
CONCLUSIONCMV reactivation occurs in a noteworthy proportion of AHSCT recipients and may lead to severe outcomes. Routine VL monitoring in the early post-transplant period is crucial, and preemptive therapy should be initiated once clinically relevant VL thresholds are reached to prevent progression to CMV disease and associated mortality.
## INTRODUCTION
Cytomegalovirus (CMV) reactivation is a well-known and potentially severe complication in immunocompromised hosts, particularly recipients of allogeneic hematopoietic stem cell transplantation (allo-HSCT), where it is a major cause of morbidity and non-relapse mortality [1]. In contrast, its clinical significance in recipients of autologous hematopoietic stem cell transplantation (AHSCT) has historically been minimized and remains insufficiently documented. This perception stems from the absence of graft-versus-host disease and a typically faster immune reconstitution, which theoretically should lower the risk [2].
However, the myeloablative conditioning regimens used prior to AHSCT induce a period of profound cytopenia and cellular immunodeficiency, creating a credible risk for viral reactivation. Emerging evidence suggests that CMV reactivation in AHSCT recipients is not as rare as once thought and can lead to serious end-organ disease, treatment delays, and even fatal outcomes, especially in patients with certain underlying malignancies like lymphomas [3].
Despite these risks, there are no standardized guidelines for CMV monitoring and preemptive therapy in the autologous setting. Practices vary widely between institutions, and the viral load (VL) thresholds that should trigger intervention are not well-defined [4,5].
Therefore, the objective of this retrospective cohort study was to evaluate the epidemiology and clinical course of CMV reactivation in a large, contemporary cohort of AHSCT recipients. We sought to determine its incidence, identify timing patterns, characterize the VL dynamics, and assess associated outcomes to build an evidence base for optimal management strategies.
## MATERIALS AND METHODS
We conducted a retrospective descriptive study of all patients who underwent AHSCT between January 2022 and December 2024 at Aziza Othmana Hospital (Tunis) and had at least one post-transplant plasma VL assessment. As this was an exhaustive retrospective cohort study, all consecutive patients meeting the inclusion criteria during the study period were included, and no sample size calculation was performed a priori. Data were retrospectively collected from patients' medical records. We collected information on patients' demographics, underlying diseases, CMV serostatus, conditioning regimens, CMV reactivation patterns, VL dynamics, treatment modalities, clinical outcomes, and overall survival.
## Definitions
CMV reactivation was defined as a single plasma CMV DNA level ≥ 150 IU/mL detected by real-time polymerase chain reaction (PCR). Patients with detectable but sub-threshold VLs (< 150 IU/mL) were not considered to have reactivation.
Neutrophil engraftment was defined as achieving an absolute neutrophil count > 0.5 × 10 9 /L for 3 consecutive days.
Probable CMV disease: Compatible clinical signs and symptoms along with CMV detection (such as viremia) but without histopathological confirmation from tissue biopsy. Proven CMV end-organ disease was defined according to established criteria: (1) CMV pneumoniae required the presence of pulmonary infiltrates on imaging with histopathological or bronchoalveolar lavage confirmation of CMV; and (2) CMV colitis was defined by gastrointestinal symptoms with endoscopic and histopathological evidence of CMV infection.
## CMV monitoring protocol
CMV VL was quantified using real-time PCR with TaqMan probes (GeneProof ® , Czech Republic) with a lower limit of quantification of 150 IU/mL.
## Antiviral prophylaxis and treatment strategy
All patients received antiviral prophylaxis with acyclovir 400 mg orally twice daily from day-7 until neutrophil engraftment, followed by 400 mg twice daily until day +30 post-transplant in the absence of reactivation. Preemptive therapy was implemented at the clinician discretion.
First-line preemptive therapy consisted of (1) Ganciclovir: 5 mg/kg intravenously every 12 hours for induction (7-14 days) followed by 5 mg/kg daily for maintenance, provided adequate blood counts (absolute neutrophil count > 1.0 × 10 9 /L, platelets > 50 × 10 9 /L); (2) Valganciclovir: 900 mg orally twice daily for induction followed by 900 mg daily for maintenance (alternative to intravenous ganciclovir when oral route feasible); and (3) Foscarnet: 90 mg/kg intravenously every 12 hours for induction followed by 90 mg/kg daily for maintenance (reserved for patients with cytopenias or ganciclovir resistance).
Treatment duration was individualized based on VL response, with induction therapy continued until two consecutive negative PCR results, followed by maintenance therapy for 2-4 weeks.
All transfused blood products used in the transplant unit are systematically leukoreduced to minimize the risk of transfusion-transmitted CMV infection.
## Statistical analysis
Descriptive statistics were performed on demographic and clinical data, presenting counts and percentages for categorical variables and median (interquartile ranges) for continuous variables. χ 2 test or Fisher's exact test was used to compare categorical variables, while Mann-Whitney U test was used for continuous variables, depending on data distribution and sample sizes. Time to reactivation was calculated from the date of stem cell infusion to the first positive CMV PCR result. P < 0.05 was considered statistically significant. All analyses were performed using Statistical Package for the Social Sciences version 25.0 (IBM SPSS Statistics 25).
## RESULTS
## Cohort characteristics
A total of 277 autologous stem cell transplant (ASCT) recipients were included in the study. Among them, 17 patients (6.1%) experienced CMV reactivation, defined as a VL ≥ 150 IU/mL. The median age of affected patients was 43 years (interquartile ranges: 32-58), with a notable female predominance (sex ratio male/female: 0.46). The distribution of underlying diseases in this group was as follows: (1) Large B-cell lymphoma (n = 5, 26.3%); (2) Multiple myeloma (n = 3, 21.1%); and (3) Hodgkin's lymphoma (n = 4, 21.1%). The remaining six patients had other lymphoma subtypes.
## Patients without antiviral treatment
Of the reactivated patients, 11 (65%) did not require specific antiviral therapy. In this group, the median age was 39 years. The median initial VL was 690 IU/mL and the median peak VL was 1320 IU/mL. The median time to reactivation was 25 days post-transplant, with neutrophil engraftment occurring at a median of 12 days. Ten patients had a favorable course with spontaneous clearance of CMV viremia. One patient developed probable CMV gastrointestinal disease and died.
## Patients with antiviral treatment
The remaining six patients (35%) required antiviral therapy. Their median age was 45 years. The median initial VL was 1631 IU/mL, and the median peak VL reached 30150 IU/mL. The median time to reactivation was 24 days posttransplant, with a median time to neutrophil engraftment of 11 days. Antiviral regimens included foscarnet (n = 3), ganciclovir (n = 2), and valganciclovir (n = 1). Five treated patients achieved viral clearance, with a median time to negativity of 46 days and a median duration from treatment initiation to viral clearance of 33 days. Among them, three developed probable CMV disease, all with favorable outcomes.
## Two fatalities occurred in this cohort (11.8% mortality rate among reactivated patients)
Patient 1 was a 55-year-old female with diffuse large B-cell lymphoma, transplanted on D + 0. Neutrophil engraftment occurred on D + 12. CMV was first detected at D + 28 with a VL of 2450 copies/mL, peaking at 196000 copies/mL on D + 36. Foscarnet therapy was started on D + 40, but the patient died on D + 41 with gastrointestinal symptoms consistent with probable CMV gastrointestinal disease.
Patient 2 was a 59-year-old female with diffuse large B-cell lymphoma, transplanted on D + 0. Neutrophil engraftment occurred on D + 10. CMV reactivation was first detected on D + 30 with a VL of 915 copies/mL, peaking at 4360 copies/ mL. The patient developed probable CMV pneumoniae associated with candidemia on D + 33 and died on D + 35.
In both cases, the diagnosis of CMV disease was based on compatible clinical presentation and significant viremia in the absence of histopathological confirmation.
## Comparative analysis
When comparing patients requiring antiviral therapy with those who spontaneously cleared CMV, initial VL (P = 0.015) and peak VL (P = 0.007) were significantly higher in the treated group. No significant differences were observed for age (P = 0.309), sex (P = 0.205), time to reactivation (P = 0.66), or time to neutrophil engraftment (P = 0.884). Detailed comparisons are summarized in Table 1.
## DISCUSSION
This retrospective cohort study of 277 ASCT recipients reported a 6.1% incidence of CMV reactivation, defined by a VL ≥ 150 IU/mL. This is consistent with the clinical understanding that CMV reactivation is less frequent after autologous transplantation than after allogeneic transplantation, where rates commonly range between 20%-50%, depending on factors like conditioning regimens and patient risk profiles. Variability in CMV reactivation rates across studies may be attributed to differences in conditioning intensity, immune recovery, monitoring frequency, and transfusion policies. In our setting, systematic leukoreduction of transfused blood products and the absence of graft-versus-host disease likely contribute to the relatively low reactivation rate observed. The 6.1% numbers fit well within the 4%-20% range reported in previous literature for ASCT patients, highlighting the relative rarity of CMV reactivation in this setting compared to allo-HSCT cohorts, which show much higher rates (up to approximately 50%) [6,7].
The predominance of female patients and the underlying diseases (mainly B-cell lymphomas and multiple myeloma) reflect typical populations undergoing ASCT, supporting the external validity of this cohort [8].
Among the 17 reactivated patients, 6 (35.3%) required antiviral therapy, while the remaining 11 (64.7%) experienced spontaneous viral clearance. Our analysis demonstrated that both initial VL and peak VL were significantly higher in patients who required treatment (P = 0.015 and P = 0.007, respectively), whereas age, sex, time to reactivation, and time to neutrophil engraftment showed no significant association with therapeutic need (Table 1). This key finding aligns with current preemptive therapy protocols that rely heavily on VL as a trigger for intervention rather than patient demographics or other parameters, as demonstrated by Tan et al [9]; who showed that preemptive antiviral therapy relies primarily on VL measurements for initiation in hematopoietic cell transplant recipients. Their study found that starting treatment at lower VLs shortened the duration of viremia and reduced clinical risk, underscoring VL as the main determinant for therapy rather than patient demographics or timing variables. Similarly, other studies have confirmed that VL monitoring represents the most reliable predictor of clinically significant CMV infection in transplant recipients [10]. Furthermore, a recent systematic review by Sadowska-Klasa et al [11]; concluded that antiviral preemptive therapy initiated at CMV VL thresholds of approximately 2-3 Log10 IU/mL effectively prevents CMV disease progression, providing robust evidence supporting VL as the key trigger for treatment initiation across transplant populations [9,11].
Clinical outcomes were generally favorable across both treatment groups, reflecting the overall good prognosis associated with CMV reactivation in the autologous setting. Among patients with spontaneous viral clearance, the majority experienced uncomplicated courses, although one patient developed probable CMV gastrointestinal disease and subsequently died. This observation underscores that even patients with lower VLs can occasionally progress to severe disease, emphasizing that CMV can remain a serious threat even in the absence of antiviral therapy, consistent with findings by Fesler et al [12], who noted that while ASCT patients have a relatively low risk of CMV reactivation, outcomes can be serious when CMV disease develops, supporting the need for vigilant monitoring.
Among the six treated patients, four achieved successful viral clearance with favorable outcomes, while two fatalities occurred. Both deaths were associated with exceptionally high VLs and severe end-organ involvement, including gastrointestinal disease in one case and pneumopathy with concurrent candidemia in the other. These outcomes reinforce that despite preemptive antiviral treatment, CMV can remain a serious threat at high viral burdens, aligning with the findings of Lee et al [3], who analyzed CMV reactivation after autologous stem cell transplantation and reported variable outcomes, including some mortality associated with severe CMV disease despite antiviral treatment, highlighting the significant clinical impact of CMV in this setting. Three of the treated patients developed probable CMV disease but all responded favorably to antiviral therapy, demonstrating the efficacy of prompt therapeutic intervention in appropriately selected cases.
The overall CMV-related mortality rate of 17.6% (3/17 reactivated patients) in our cohort resonates with reports from other autologous transplant settings, underscoring the potential severity of CMV complications despite their relative infrequency in the ASCT context. These outcomes emphasize that while immune recovery post-AHSCT often suffices to control viral reactivation spontaneously, close monitoring remains essential to identify patients at risk for severe complications.
The study highlights ongoing clinical challenges: Although CMV reactivation post-ASCT is less common, it can lead to severe outcomes, demanding vigilant monitoring. There remain no standardized guidelines specific to CMV monitoring and treatment thresholds in ASCT recipients, resulting in varying clinical practices and potentially affecting patient outcomes. Future prospective studies with larger cohorts and standardized monitoring protocols are necessary to optimize management strategies [13].
## Study limitations
Our findings are limited by the lack of detailed demographic and clinical data for the full cohort, preventing the assessment of other potential risk factors such as comorbidities, disease status at transplant, or prior therapy. Additionally, our study has several limitations inherent to its retrospective and single-center design. The frequency of VL monitoring was not standardized and was performed at the discretion of the treating clinician, potentially leading to an underestimation of the true incidence of reactivation, particularly short-lived, low-level viremias. Furthermore, the small number of events limits the power of statistical analysis to identify independent risk factors for reactivation or for progression to high-level viremia.
## CONCLUSION
This comprehensive analysis of CMV reactivation dynamics and clinical outcomes in 277 AHSCT recipients provides valuable real-world evidence that enriches our understanding of viral complications in this population. Our results demonstrate that while CMV reactivation occurs in only 6.1% of patients, it represents a clinically significant event that can lead to severe complications and mortality, particularly in patients with high VLs. The strong correlation between VL parameters and treatment necessity, independent of demographic factors or clinical timing, reinforces VL quantification as both a predictive and therapeutic decision-making tool, consistent with current international guidelines. Based on our findings and their alignment with existing literature, we advocate for implementing a structured, riskadapted management approach for CMV in AHSCT recipients that includes: (1) Regular monitoring of CMV DNAemia via PCR during the first 2-3 months post-transplant for all at-risk patients, particularly CMV-seropositive recipients; (2) Preemptive therapy initiated upon detection of VLs exceeding predefined thresholds (approximately 1000-5000 IU/mL or 2-3 Log10 IU/mL) to prevent progression to end-organ disease; and (3) Sustained vigilance for late reactivation in patients with prolonged immunosuppression or additional risk factors. These findings also highlight that, although CMV reactivation is less frequent in AHSCT than in allo-HSCT, its clinical impact can be significant in cases of high VL or delayed diagnosis.
Despite the inherent limitations of our retrospective, single-center design, this study contributes important insights into the clinical course and outcomes of CMV reactivation in a real-world autograft population. The variability in current monitoring practices and the demonstrated potential for severe outcomes underscore the urgent need for standardized, evidence-based protocols. Future prospective, multicenter studies with larger cohorts and comprehensive risk factor assessment are essential to definitively establish optimal monitoring schedules, refine VL thresholds for treatment initiation, and develop risk stratification algorithms that will optimize CMV management and improve outcomes for AHSCT recipients.
## References
1. Jakharia, Howard, Riedel (2021) "CMV Infection in Hematopoietic Stem Cell Transplantation: Prevention and Treatment Strategies" *Curr Treat Options Infect Dis*
2. Degli-Esposti, Hill (2022) "Immune control of cytomegalovirus reactivation in stem cell transplantation" *Blood*
3. Lee, Jeon, Yu et al. (2023) "Prognostic outcomes of cytomegalovirus reactivation after autologous stem cell transplantation" *Int J Med Sci*
4. Alexander, Badoglio, Labopin et al. (2025) "Autoimmune Diseases Working Party (ADWP) of the EBMT. Monitoring and management of CMV and EBV after autologous haematopoietic stem cell transplantation for autoimmune diseases: a survey of the EBMT Autoimmune Diseases Working party (ADWP)" *Bone Marrow Transplant*
5. (2025) "Guidance for Prevention and Treatment of Cytomegalovirus (CMV) in Stem Cell Transplant Recipients at Atrium Health Wake Forest Baptist Updated: Winter 2025"
6. Lin, Wu, Liu (2025) "Epidemiology, clinical outcomes, and treatment patterns of cytomegalovirus infection after allogeneic hematopoietic stem cell transplantation in China: a scoping review and meta-analysis" *Front Microbiol*
7. Muraro, Mariottini, Greco et al. (2025) "Attendees of the ECTRIMS Focused Workshop on HSCT. Autologous haematopoietic stem cell transplantation for treatment of multiple sclerosis and neuromyelitis optica spectrum disorder -recommendations from ECTRIMS and the EBMT" *Nat Rev Neurol*
8. Georges, Bar, Onstad et al. (2020) "Survivorship after Autologous Hematopoietic Cell Transplantation for Lymphoma and Multiple Myeloma: Late Effects and Quality of Life" *Biol Blood Marrow Transplant*
9. Tan, Waggoner, Pinsky (2015) "Cytomegalovirus load at treatment initiation is predictive of time to resolution of viremia and duration of therapy in hematopoietic cell transplant recipients" *J Clin Virol*
10. Sadowska-Klasa, Xie, Zamora et al. (2025) "Cytomegalovirus Viral Load Continues to Predict Poor Outcomes in Adults and Children Despite Improved Hematopoietic Cell Transplantation Success" *Open Forum Infect Dis*
11. Sadowska-Klasa, Leisenring, Limaye et al. (2024) "Cytomegalovirus Viral Load Threshold to Guide Preemptive Therapy in Hematopoietic Cell Transplant Recipients: Correlation With Cytomegalovirus Disease" *J Infect Dis*
12. Fesler, Poole, Goldenberg et al. (2021) "Outcomes of Cytomegalovirus Monitoring in Autologous Transplantation: A Single Institution Experience" *Blood*
13. Nho, Lee, Cho et al. (2025) "How Should Cytomegalovirus Infection Be Managed in Allogeneic Hematopoietic Stem Cell Transplant Recipients? A Clinical Grand Round" *Infect Chemother* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12792359&blobtype=pdf | ## Abstract
Background. West Nile virus (WNV) has ubiquitous distribution in Africa. Over the years, the geographical range of WNV activity has increased and the virus has become established even in non-endemic areas where it has not been previously detected.
Methods. This serological-survey investigated the prevalence of anti-WNV IgM among patients with febrile illnesses at Gwagwalada metropolis, Abuja. Between the period of May and August 2016, a total of 171 patients attending the University of Abuja Teaching Hospital were recruited for the study. Serum samples were immediately harvested, stored and analyzed using the indirect ELISA for anti-WNV IgM antibodies using kits endorsed by the World Health Organization. Socio-demographic variables and clinical data was gotten using a self-administered interviewer-based questionnaires.
Results. Out of the 171 febrile participants, the overall prevalence of WNV IgM antibodies was 66.1%. With regards to participants preventive measures against WNV and associated risk factors, significant association was observed between WNV IgM seropositivity and the use of mosquito repellents (p =0.016).
Conclusion. Out of the 171 febrile participants, the overall prevalence of WNV IgM antibodies was 66.1%. With regards to participants preventive measures against WNV and associated risk factors, significant association was observed between WNV IgM seropositivity and the use of mosquito repellents (p =0.016). Findings from this study necessitate the need for routine surveillance of WNV. More so, infected patients should be closely monitored in order to detect possible associated sequelae.
Disclosures. All Authors: No reported disclosures |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584691&blobtype=pdf | # Molecular xenosurveillance of Aedes mosquitoes reveals dengue virus serotype-2 during an outbreak in Dhaka, Bangladesh
Prakash Ghosh, Anupama Hazarika, Md Arko, Ayon Chowdhury, Sabera Sultana, Nishad Mithila, Shariful Shahid, Md Haque, Md Islam, Dinesh Mondal, Rajib Chowdhury, Ayon Arko, Sabera Chowdhury, Rajib Sultana, Chowdhury
## Abstract
The increased risk of emerging infectious diseases is particularly pro nounced in resource-constrained regions with limited capacity to undertake surveillance, presenting a global challenge for the timely identification and mitigation of disease threats. Consequently, the development and implementation of feasible, cost-effective, and sustainable surveillance strategies are essential. Molecular xenosurveillance (MX) represents an innovative approach for arbovirus monitoring, including dengue surveillance, which utilizes field-collected mosquito specimens to detect circulating pathogens and provide early outbreak warnings. Bangladesh is highly endemic for dengue, and with limited resources, the implementation of MX will add value to ongoing efforts to scale up surveillance. This study aimed to test the suitability of MX in detecting dengue viruses in Aedes vectors and recommend its integration into the existing surveillance system. Adult Aedes mosquitoes were collected using BG-Senti nel traps during pre-monsoon dengue vector surveys in Dhaka city. Aedes mosquitoes were identified morphologically and screened in pools using real-time RT-PCR. Infec tion prevalence was estimated using the PoolTestR program. Out of 228 Aedes aegypti mosquitoes distributed across 26 pools, one pool with 10 mosquitoes tested positive for the dengue virus. The estimated infection prevalence was 0.450% (95% CI: 0.026%-1.960%) and 0.510% (95% CI: 0.053%-2.100%) using maximum likelihood and Bayesian estimates, respectively. The DENV-2 serotype was detected in the positive pool. This study presents the first successful implementation of the MX methodology to detect dengue virus presence in field-collected Ae. aegypti specimens from Dhaka, Bangladesh. These findings provide empirical support for enhanced vector surveillance protocols and offer valuable insights for strengthening comprehensive dengue control efforts.
IMPORTANCEThe timely detection of arboviral pathogens like dengue virus, through surveillance is critical for developing early warning systems and preventing outbreaks. However, limited surveillance capacity in resource-constrained settings creates gaps that can be addressed by implementing sustainable and cost-effective surveillance strategies. This study demonstrates the feasibility of molecular xenosurveillance (MX), which is a vector-based, non-invasive surveillance method for dengue virus surveillance in Aedes aegypti mosquitoes in Dhaka, Bangladesh. By detecting and identifying dengue virus serotype-2 (DENV-2) using real-time RT-PCR, this study also provides evidence of the method's potential for generating deeper epidemiological insights. These findings highlight the value of MX as a complementary tool to existing surveillance systems. Integrating MX into routine entomological surveillance could significantly enhance national preparedness and response to dengue outbreaks.
T he World Health Organization (WHO) estimated that 80% of the worldwide populace is vulnerable to one or more vector-borne diseases (VBDs), contributing nearly 17% to the overall burden of communicable diseases (1). Meanwhile, the Southeast Asian region has become the epicenter for emerging infectious diseases (EIDs), particularly VBDs having epidemic or pandemic potential (2). Vital factors, including population growth and high density, improper urbanization, mobility, and shifts in the environment, have a substantial impact on the emergence of infectious diseases (1).
Dengue virus infection is one of the VBDs primarily transmitted by infectious female Aedes mosquitoes. The WHO has flagged dengue as one of the top ten global health concerns (3). Over the past two decades, since 2000, the number of reported dengue cases increased 10-fold worldwide, rising from 500,000 to 5.2 million (4).
The earliest recorded dengue case in Bangladesh dates back to 1964. The illness was informally termed "Dacca fever" at that time. However, the disease did not draw attention as a major health problem until a major outbreak happened in 1999 at Dhaka, followed by an epidemic in 2000 (5,6). Since then, Dhaka and other metropolitan cities have been facing the scourge of dengue outbreaks periodically, with Dhaka being the epicenter. The Directorate General of Health Services (DGHS), Bangladesh, officially reported 101,354, 28,429, 61,763, 321,179, and 101,214 hospitalized dengue cases annually in 2019, 2021, 2022, 2023, and 2024, respectively (7)(8)(9)(10). Among the four dengue virus serotypes, DENV-1 and DENV-2 were prevalent between 2013 and 2016, whereas in the outbreak in 2019, DENV-3 serotype was the most prevalent one in Bangladesh (7). During the outbreak in 2023, the serotype DENV-2 reappeared as the predominant serotype (11).
In the absence of established therapeutic or preventive intervention, effective strategies for controlling and minimizing transmission revolve around implementing vector control measures and enhancing public awareness. WHO has denoted vector surveillance, monitoring, and evaluation of interventions as one of the four pillars to achieve effective and sustainable vector control (1). In recent years, the notion of establishing an early warning system has been proposed in various countries as an approach to contain dengue transmission (12)(13)(14)(15). Developing an effective early warning system to predict outbreaks of dengue is sophisticated as it involves several interacting factors including host, environment, vector, and dengue virus itself. Therefore, indicators spanning meteorological, epidemiological, population, and entomological (both vector abundance and infection) are essential for establishing a robust early warning system. Xenosurveillance, which involves the use of blood-fed hematophagous arthropods to monitor and identify vertebrate pathogens (16,17), offers a valuable approach in this context. Tracking dengue viruses in both immature and adult mosquitoes is a pivotal strategy for detecting the onset of the epidemic phase and guiding timely measures (18).
Previous entomological studies on dengue in Bangladesh have mostly concentrated on assessing vector density and distribution. However, estimating and mapping the prevalence of dengue virus in mosquitoes are also essential (19,20), as mentioned earlier. Successful implementation of a vector control program necessitates comprehen sive information about the vectors, encompassing their density, resistance to control methods, and infection rates (21). Despite the high dengue burden and risk of transmis sion in Bangladesh, especially in Dhaka, there is inadequate data on the vector infection rate. Low socio-economic regions of Dhaka were found to be abundant with Aedes mosquitoes, but their infection with dengue viruses has not been explored (20).
Molecular xenosurveillance or xenomonitoring (MX) is a modern tool to analyze genetic material (DNA or RNA) from field-derived vectors to detect the presence of pathogens for disease surveillance. It has emerged as a potential surveillance technique due to substantial advancements in laboratory techniques. MX offers a sensitive and tractable approach for tracking pathogen circulation in the vector population even before in the human population, making it a valuable tool for early outbreak detec tion and response. Periodical MX has successfully been applied as a post-mass drug administration surveillance tool for global programs to eliminate lymphatic filariasis (22,23). Several countries endemic to dengue are now putting efforts into establishing MX for assessing the prevalence of dengue infection in vectors, identifying transmission hotspots, predicting outbreaks through predictive modeling, and eventually, devising comprehensive early warning systems to implement timely control measures (18,(24)(25)(26). This innovative approach has the potential to revolutionize VBD/EID surveillance and is seemingly fit for programmatic use in Bangladesh (22). Thus, we have introduced xenosurveillance for the first time in Dhaka city for detecting and identifying dengue virus in Aedes aegypti vectors. The study aimed to establish MX for the dengue in settings like Bangladesh, thereby strengthening existing entomological surveillance and guiding large-scale implementation in areas prone to dengue outbreaks.
## RESULTS
## Vector distribution and physiological status
A total of 1,350 mosquitoes were collected (Table 1) from the 26 traps (6 days collection period) where the number of Ae. aegypti, Culex quinquefasciatus, Culex tritaeniorhynchus, and Armigeres subalbatus were 228, 992, 104, and 26, respectively. Cx. quinquefascia tus was the most abundant species, accounting for 73.48% of total captures, significantly exceeding the expected proportion under an equal-distribution assumption (χ² = 1,753.8, df =3, P < 0.001). The other three species Ae. aegypti (16.89%), Cx. tritaeniorhyn chus (7.70%), and Ar. subalbatus (1.93%) were significantly under-represented compared to Cx. quinquefasciatus, with Ar. subalbatus being particularly rare.
Among the collected Ae. aegypti, 148 and 80 were female and male, respectively. In female Ae. aegypti population, 48, 82, and 18 were fed, unfed, and gravid, respectively (Table 1). During the collection period, two traps had only Ae. aegypti, one trap had only Culex spp., and in seven traps, no mosquito was trapped. The captured Ae. aegypti (both female and male) were further distributed in 26 pools based on the trapping sites.
## Molecular screening of the pools and prevalence estimation
Out of these 26 pools, one pool from Basabo area containing 10 Ae. aegypti mosquitoes tested positive for the dengue virus in pan-dengue RT-qPCR screening. Upon further examination utilizing RT-PCR, the virus within the positive pool was confirmed as a DENV-2 serotype. Using the PoolPrev function of the PoolTestR tool, the PCR results were analyzed considering the entire data set from a single study region (Dhaka City). The maximum likelihood estimate indicated a dengue infection prevalence of 0.450% (95% CI: 0.026%-1.960%). On the other hand, the Bayesian estimation suggested a slightly higher prevalence rate of 0.51% (95% CI: 0.053%-2.100%). The Minimum Infection Rate (MIR) was determined to be 4.39, considering only one dengue-infected mosquito in the positive pool.
## DISCUSSION
MX has been well acknowledged by the scientific community for the purpose of VBDs/ EIDs surveillance and monitoring. This tool complements vector management strategies and has the potential to alert authorities in advance of outbreaks. However, this technique is yet to be integrated into the ongoing vector surveillance in Bangladesh. Therefore, the overarching goal of this study was to establish a MX tool for the detection and identification of dengue virus in vector mosquitoes.
This study provides the very first evidence of dengue infection in adult Aedes mosquitoes detected through MX in Bangladesh. The presence of the infection in the vectors denotes the ongoing transmission of dengue virus in the community. The occurrence of the outbreak during the study further bolsters the findings of the current study and the transmission.
The earlier entomological surveillance studies conducted in Bangladesh were confined to estimating house index (percentage of houses positive with larvae/pupae), container index (percentage of water holding containers positive with larvae/pupae), Breteau index (number of containers positive with larvae/pupae per 100 inspected houses), adult vector abundance, breeding sites, and insecticide resistance status (19,20,(27)(28)(29). Since the abundance does not proportionately represent the vector infection rate and some earlier studies failed to correlate the mentioned indices with the number of cases and the transmission of virus from host to vector, therefore, it is critical to estimate the vector infection rate (30,31).
In this study, we confirmed the existence of the dengue virus in the vector and distinguished its serotype as DENV-2, which is similar to the predominant serotype (68.1%) found in the patients during the outbreak of 2023 in Bangladesh (32). Serotyping the dengue virus in vectors offers valuable insights into the concurrent circulation of serotypes, thereby complementing the prediction of the magnitude of outbreaks and disease severity (33).
Several dengue-endemic countries have already introduced MX to aid vector-based surveillance. In India, in an endemic urban area, the minimum infection rate was found to be 4.5, which is similar to the minimum infection rate in this study (26). The same study reported one pool of male Ae. aegypti to be tested positive for DENV, indicat ing the importance of including male mosquitoes. Although male mosquitoes do not blood-feed and are not direct vectors, the presence of viral RNA in males may indicate vertical (transovarial) transmission from infected female mosquitoes to their offspring. This mechanism could play a role in maintaining the virus within mosquito populations a "-" indicates that data were not collected because it was outside the scope of interest.
during inter-epidemic periods and justifies our study to include male mosquitoes for testing. A study conducted in Brazil reported a minimum infection rate of 19.8, where adult Aedes vectors were collected for 3 years ( 18). An important constraint in this study is the relatively small sample size. Additionally, the sample collection period was very short, which limits the exploration of the seasonal variation of the infectivity rate. Thus, to employ the insights from xenosurveillance in the development of a successful early warning system, a substantial quantity of vectors should be collected year-round and tested for the disease-causing pathogen. Despite proper pool screening methods being used in this study, the observed infection rate does not truly represent Dhaka city. This can be overcome by increasing spatial and temporal coverage (34).
Although this pilot study has limitations, the findings of the study will help in tailoring a robust dengue surveillance strategy. Forecasting dengue outbreaks only through short-term infection surveillance is almost impossible. Thus, year-round surveillance programs are required to provide more precise and accurate insights. Moreover, the inclusion of infection surveillance in the immature mosquitoes (larvae and pupae) to the activity is crucially important for endemic settings to evaluate the conservation of the virus during an inter-epidemic period (35). This tool also has the potential to be used as a proxy to host infection estimation for vector-borne disease surveillance in the future. However, a comprehensive understanding of the correlation between pathogeninfection rates in both humans and vectors is indispensable (22). In addition to disease surveillance, data from MX can aid in formulating or determining the effectiveness of vector control interventions. In Indonesia, the vector infection data derived from MX promises to serve as baseline information for Wolbachia-mediated vector control strategies (31). In addition to the detection of a single pathogen, multiplex molecular assay can be integrated into MX for the detection of multiple pathogens transmitted by the same vector. It is worth noting that dengue, chikungunya, and zika viruses share the same vector; therefore, MX of Aedes might provide insights into the combined prevalence of these pathogens or the emergence of a pathogen in a particular area. The current MX tool demands cost-and expertise-intensive methods. Hence, to render MX more cost-effective and feasible, it is imperative to devise a rapid nucleic acid extraction method and simpler molecular assays (RT-RPA, RT-LAMP) alternative to RT-PCR without compromising sensitivity and specificity.
## Conclusion
This study demonstrates the potential of MX to detect dengue virus prevalence in vector mosquitoes and provide valuable epidemiological insights, both of which are critical components for developing robust early warning systems. Furthermore, large-scale prospective studies are essential to optimize the tool towards its integration into existing disease surveillance programs.
## MATERIALS AND METHODS
## Field collection
As part of a routine pre-monsoon dengue vector survey by the entomological team of the Communicable Disease Control Unit, DGHS, Government of Bangladesh, this cross-sectional study was conducted. Thus, adult mosquitoes were collected from different areas in Dhaka, Bangladesh, by trained entomologists and entomological technicians. Here, BG Sentinel-2 traps were used with 1 tablespoon of yeast with 15 mL of 10% sugar as bait. The areas encompassed Rampura, Basabo, Kamalapur, Mirpur, Baunia, Banani, and Dhanmondi within the Dhaka North City Corporation (DNCC) and Dhaka South City Corporation (DSCC) (Fig. 1). A total of 26 traps were set up in the randomly selected houses between 20-25 June 2023. Each trap was set up in a corner of the house for 24 h (09:00 AM-09:00 AM). The activities carried out to accomplish the study are shown in Fig. 2.
## Mosquito identification and processing
After collecting, the mosquitoes were placed in a -20°C freezer for at least 20 min, within 2 h to immobilize the mosquitoes. Afterward, the adult Ae. aegypti mosquitoes were separated from the blend and subjected to morphological speciation and physiological status determination. Mosquito specimens were then pooled based on traps with a number no greater than 10 and labeled accordingly. The pools were then stored at -80°C until the RNA was extracted.
## RNA extraction
RNA from the mosquito pools was extracted employing Qiagen RNeasy Mini Kit. Buffer RLT (600 µL) was added to each tube followed by homogenization using a handheld mini homogenizer with plastic micropistilles until finely homogenized. Ethanol (800 µL; 70%) was added and thoroughly mixed by pipetting. Subsequently, 700 µL of the blend was shifted to an RNeasy spin column and centrifuged at 8,000 × g for 15 seconds and the flow-through was disposed of. The remaining blend is put in the same column and the step is repeated. After introducing the samples into the RNeasy Mini spin columns, we adhered to the manufacturer's guidelines for the subsequent stages in the RNeasy Mini kit protocol. This included applying 700 µL of Buffer RW1, followed by 500 µL of Buffer RPE for each of the two consecutive wash steps. RNA was eluted using 40 µL RNase-free water.
## Real-time one-step RT-PCR
Dengue pan-serotype real-time one-step RT-PCR was performed on RNA extracted from every mosquito pool. The primers and probes used were derived from the work of Leparc-Goffarf et al. (36) and synthesized by Integrated DNA Technologies (IDT). The single-plex RT-PCR was performed on an Applied Biosystems 7500 real-time PCR instrument using 2 µL of extracted RNA, 10 µL of 2× QuantiNova Probe RT-PCR Master Mix, 0.2 µL of Probe RT Mix, 0.1 µL of ROX reference dye, 450 nM of forward and reverse primers, and 225 nM of probes (Taq), with the final volume adjusted to 20 µL using nuclease-free water. The thermal cycling conditions were as follows: 45°C for 12 min, 95°C for 5 min, followed by 40 cycles of 95°C for 5 seconds and 60°C for 40 seconds, while fluorescence data were collected at 60°C.
The RNA from dengue-positive pool was subsequently analyzed using the U.S. Food and Drug Administration (FDA) approved CDC DENV-1-4 Real-Time RT-PCR multi plex assay (package insert, catalog number KK0128; available at https://www.cdc.gov/ dengue) to determine the dengue virus serotype (11). The multiplex RT-PCR was performed on a Bio-Rad CFX96 Real-Time PCR Detection System using 5 µL of extracted RNA and along with oligonucleotide primers and fluorescent probes (TaqMan). The thermal cycling conditions were as follows: 50°C for 30 min, 95°C for 2 min, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 min. Fluorescence data were collected at 60°C. A sample was considered positive if the amplification curve had a cycle threshold (Ct) ≤37, while Ct values >37 were considered negative. In each of the RT-PCR run, nuclease-free water was used as the negative control and molecular standards were used as positive controls.
## Data analysis
Mosquito specimen collection information along, with their distribution in the pool, was entered into a single database. The corresponding molecular screening results from RT-PCR were analyzed using the 7500 Fast Software v2.3 and incorporated in the above-mentioned database. The prevalence of the infection was estimated using the PoolTestR program (with default parameters). It is an R language (R Core Team, 2020) based package for analyzing data from complex MX surveys involving testing of pooled or grouped samples (26). The minimum Infection Rate (MIR) was also determined by dividing the number of DENV-positive pools by the total number of mosquitoes tested and then multiplying by 1000 (18).
## References
1. (2017) "Global vector control response 2017-2030"
2. Coker, Hunter, Rudge et al. (2011) "Emerging infectious diseases in southeast Asia: regional challenges to control" *Lancet*
3. (2019) "Ten threats to global health in 2019"
4. (2023) "Dengue -global situation"
5. Yunus, Bangali, Mahmood et al. (2001) "Dengue outbreak 2000 in Bangladesh: from speculation to reality and exercises" *Dengue Bull*
6. Sharmin, Viennet, Glass et al. (2015) "The emergence of dengue in Bangladesh: epidemiology, challenges and future disease risk" *Trans R Soc Trop Med Hyg*
7. Hossain, Noman, Mamun et al. (2023) "Twenty-two years of dengue outbreaks in Bangladesh: epidemiology, clinical spectrum, serotypes, and future disease risks" *Trop Med Health*
8. (2022) "Dengue press release -31"
9. (2023) "Dengue press release -31"
10. (2024) "Dengue press release -31"
11. Hasan, Zamil, Trina et al. (2024) "Resurgence of dengue virus serotype 2: findings from the 2023 Bangladesh outbreak" *Am J Trop Med Hyg*
12. Baharom, Ahmad, Hod et al. (2018) "Early warning and response system (EWARS) for dengue outbreaks: Recent advancements towards widespread applications in critical settings" *Risk Manag Healthc Policy*
13. Chang, Tseng, Hsu et al. (2015) "Reassess vector indices threshold as an early warning tool for predicting dengue epidemic in a dengue non-endemic country" *PLoS Negl Trop Dis*
14. Racloz, Ramsey, Tong et al. (2012) "Surveillance of dengue fever virus: a review of epidemiological models and early warning systems" *PLoS Negl Trop Dis*
15. Fauver, Weger-Lucarelli, Fakoli et al. (2018) "Xenosurveillance reflects traditional sampling techniques for the identification of human pathogens: a comparative study in West Africa" *PLoS Negl Trop Dis*
16. Grubaugh, Sharma, Krajacich et al. (2015) "Xenosurveillance: a novel mosquito-based approach for examining the human-pathogen landscape" *PLoS Negl Trop Dis*
17. Medeiros, Costa, Branco et al. (2018) "Dengue virus in Aedes aegypti and Aedes albopictus in urban areas in the state of Rio Grande do Norte, Brazil: importance of virological and entomological surveillance" *PLoS One*
18. Priyanka, Bashar, Islam (2022) "Distribution and abundance of dengue vector(S) mosquito and their breeding preferences in five selected administrative wards of Dhaka North city corporation" *Bangladesh. PHOA*
19. Paul, Dhar-Chowdhury, Haque et al. (2018) "Risk factors for the presence of dengue vector mosquitoes, and determinants of their prevalence and larval site selection in Dhaka" *PLoS One*
20. Horstick, Runge-Ranzinger (2019) "Multisectoral approaches for the control of vector-borne diseases, with particular emphasis on dengue and housing" *Trans R Soc Trop Med Hyg*
21. Cameron (2021) "The use of molecular xenomonitoring for surveillance of mosquito-borne diseases" *Philos Trans R Soc Lond B Biol Sci*
22. Subramanian, Jambulingam, Krishnamoorthy et al. (2020) "Molecular xenomonitoring as a post-MDA surveillance tool for global programme to eliminate lymphatic filariasis: field validation in an evaluation unit in India" *PLoS Negl Trop Dis*
23. Ramesh, Jeffries, Castanha et al. (2015) "No evidence of Zika, dengue, or chikungunya virus infection in field-caught mosquitoes from the Recife Metropolitan Region" *Wellcome Open Res*
24. Gutiérrez-Bugallo, Rodríguez-Roche, Díaz et al. (2018) "Spatio-temporal distribution of vertically transmitted dengue viruses by Aedes aegypti (Diptera: Culicidae) from Arroyo Naranjo" *Tropical Med Int Health*
25. Chand, Godbole, Shivlata et al. (2021) "Molecular xenomonitoring of Dengue, Chikungunya and Zika infections: a year-round study from two Dengue endemic districts of central India" *J Vector Borne Dis*
26. Islam, Haque, Hossain et al. (2019) "Role of container type, behavioural, and ecological factors in Aedes pupal production in Dhaka, Bangladesh: an application of zero-inflated negative binomial model" *Acta Trop*
27. Haque, Dhar-Chowdhury, Hossain et al. (2023) "Spatial evaluation of dengue transmission and vector abundance in the city of Dhaka" *Bangladesh. Geographies*
28. Al-Amin, Gyawali, Graham et al. (2023) "Insecticide resistance compromises the control of Aedes aegypti in Bangladesh" *Pest Manag Sci*
29. Sharif, Khan, Hasan et al. (2022) "Spatial association of Aedes aegypti with dengue fever hotspots in an endemic region"
30. Rahayu, Saraswati, Supriyati et al. (2019) "Prevalence and distribution of dengue virus in aedes aegypti in yogyakarta city before deployment of wolbachia infected Aedes aegypti" *Int J Environ Res Public Health*
31. (2023) "Dengue situation report"
32. Soo, Khalid, Ching et al. (2016) "Meta-analysis of dengue severity during infection by different dengue virus serotypes in primary and secondary infections" *PLoS One*
33. Mclure, Neill, Mayfield et al. (2021) "PoolTestR: an R package for estimating prevalence and regression modelling for molecular xenomonitoring and other applications with pooled samples" *Environ Model Softw*
34. Lequime, Paul, Lambrechts (2016) "Determinants of arbovirus vertical transmission in mosquitoes" *PLoS Pathog*
35. Leparc-Goffart, Baragatti, Temmam et al. (2009) "Development and validation of realtime one-step reverse transcription-PCR for the detection and typing of dengue viruses" *J Clin Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645917&blobtype=pdf | # Engineering a TGEV S-trimer chimera with PEDV D0-NTD generates potent neutralizing antibodies against both viruses
Ding Zhang, Yang Yuan, Guangli Hu, Yunfei Xie, Xiumei Meng, Zhaotian Zhang, Yu Zhang, Qi Liao, Ashenafi Gebremariam, Hanqin Shen, Guiqing Peng, Yuejun Shi
## Abstract
Porcine epidemic diarrhea virus (PEDV) and transmissible gastroenteritis virus (TGEV), two economically devastating swine enteric coronaviruses, persistently threaten global pork production via severe clinical morbidity. Vaccination is the most effective means to prevent morbidity and mortality caused by porcine enteric coronavi rus infection. Hence, the development of an effective and safe vaccination strategy to protect against both PEDV and TGEV infections is urgently needed. Here, we designed and produced trimerized full-length PEDV and TGEV spike (S) proteins using an efficient mammalian expression vector system in HEK293F cells and reported that compared with the PEDV spike protein (2 7 ), the TGEV spike protein induces significantly more potent neutralizing antibodies (2 12 ). Moreover, transmission electron microscopy (TEM) with negative staining revealed that the TGEV spike protein exhibited excellent homogeneity. We subsequently used the TGEV S protein trimer as the backbone to display differ ent domains of the PEDV S protein and successfully constructed seven spike protein chimeras. We further screened two chimeric S proteins through expression and mouse experiments, which included TGEV S-PEDV(D0/NTD) and TGEV S-PEDV(D0/NTD/CTD), in which the S protein of TGEV was replaced by the PEDV S D0-NTD and D0-NTD-CTD, respectively. Evaluation of the immunogenicity and efficacy of the chimeric S proteins in piglets confirmed that piglets in the TGEV S-PEDV(D0/NTD)-immunized group generated high levels of neutralizing antibodies (2 13.6 against TGEV; 2 5.8 against PEDV) against both PEDV and TGEV. Our findings suggest that the TGEV S-PEDV(D0/NTD) is a promising candidate for combating both PEDV and TGEV infections.
IMPORTANCEThe design strategy of multivalent and multitarget single antigens facilitates the development of vaccines targeting PEDV and TGEV. Optimizing the presentation of PEDV core antigen epitopes on the basis of the S protein trimer structure will provide novel insights into the development of subunit vaccines. Here, we compared the immunogenicities of the TGEV S protein and PEDV S protein and then designed a bivalent chimeric S protein candidate, TGEV S-PEDV(D0/NTD), in which the correspond ing segments on the TGEV S protein were replaced with D0-NTD domains from the PEDV S protein. We demonstrated that the TGEV S protein exhibits greater immunoge nicity than the PEDV S protein and that TGEV S-PEDV(D0/NTD) induces broad-spectrum neutralization protection against PEDV and TGEV. Our results demonstrate the effective ness of the chimeric S protein and provide a feasible method for the development of efficient bivalent subunit vaccines against PEDV and TGEV. KEYWORDS porcine epidemic diarrhea virus (PEDV), transmissible gastroenteritis virus (TGEV), chimeric spike protein, subunit vaccine, cross-protection, immunogenicity P orcine epidemic diarrhea (PED) and transmissible gastroenteritis (TGE), caused by porcine epidemic diarrhea virus (PEDV) and transmissible gastroenteritis virus
(TGEV), respectively, are two highly contagious intestinal diseases affecting the global swine industry (1)(2)(3)(4)(5)(6). All ages of swine are highly susceptible to PEDV and TGEV infection, particularly piglets within 2 weeks of age, which show a notably high mortality rate (7,8). Before 2010, PEDV infections in China caused sporadic and endemic out breaks. Since October 2010, a severe epidemic caused by the highly pathogenic porcine epidemic diarrhea virus (PEDV) has been continuously spreading among pig populations in China, resulting in considerable economic losses (6,(9)(10)(11)(12). TGEV was first described in the United States in 1946 and was subsequently found in Europe, Asia, Africa, and South America, causing major losses to the global pig farm industry (13)(14)(15)(16). The clinical symptoms caused by TGEV are similar to those caused by PEDV, and they often present clinically mixed infections. However, cross-protection between the two viruses is limited (17). Therefore, there is an urgent need to develop a novel protective, safe, and affordable vaccine against both PEDV and TGEV.
Both PEDV and TGEV are members of the Alphacoronavirus genus (18,19), and the genomes of PEDV and TGEV are approximately 28 kb in length with 5′-capped and 3′-polyadenylated untranslated regions (UTRs) (20,21). The genomes of both PEDV and TGEV are conventionally organized around seven major open reading frames. These include the replicase genes ORF1a and ORF1b, encoding the pp1a/pp1ab polyproteins that are processed into 16 nsps (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16) for replication and transcription, and the genes for the structural proteins Spike (S), Envelope (E), Membrane (M), and Nucleocapsid (N) and the ORF3 gene encoding an accessory protein (20). Among these viral pro teins, the S protein on the surface of the virus is the main envelope glycoprotein. The trimeric structures of the PEDV S protein have been successively determined (22)(23)(24). The S protein comprises S1, a receptor-binding subunit that is responsible for host cell attachment and receptor binding, and S2, a membrane fusion subunit that is involved in triggering the fusion of the viral envelope and target cell membrane during infection (25,26). The S1 region (cell attachment domain) is considered a key target of neutralizing antibodies, and the S1-D0 and NTD regions induce the production of additional neutralizing antibodies (27). Therefore, the S1-D0 and S1-NTD regions may be key epitopes of the PEDV S protein. Currently, PEDV strains are classified into genotypes G1 and G2, with the G2 genotype being the clinically prevalent strain. The G2 genotype is further divided into the G2a, G2b, and G2c strains, among which the G2c strain has replaced G2a and G2b to become the dominant prevalent strain (10,28,29). Faced with the rapid evolution of PEDV genotypes and immune evasion, monitoring the variation and evolution of coronaviruses like PEDV and developing corresponding vaccines have become a critical priority in the clinical prevention and control of this virus (30). Vaccination is one of the most important measures for controlling PEDV and TGEV outbreaks (20,31). Recently, researchers have developed a PEDV-TGEV-PDCoV trivalent inactivated vaccine, and it has been shown to provide effective protection in pigs against all three enteric coronaviruses (32). Both inactivated and live-attenuated virus vaccines are commercially available and have been widely used to prevent PEDV and TGEV infection (33,34). However, the emergence of highly virulent strains and recurrent outbreaks, even on vaccinated farms, highlights the limitations of traditional vaccines and the need for effective vaccines (35). Compared with total viral vaccines, protein vaccines of subunits have several advantages that make them appealing for the development of new vaccines, such as enhanced safety features, elimination of infectious viral genetic material, antigen structural redesign, multi-antigen combinations, and strategic adjuvant integration; this approach significantly improves therapeutic efficacy (36,37). Therefore, our primary focus is to develop a chimeric S protein subunit vaccine against both PEDV and TGEV.
PEDV and TGEV are two major coronaviruses responsible for severe diarrhea and mortality in piglets. Because of the rapid mutation of PEDV, methods to quickly develop a vaccine for the effective prevention and control of both viruses are urgently needed. Here, we designed seven chimeric S proteins encoding different domains of the PEDV S1 region with the TGEV S protein trimer as the backbone. We investigated the expression levels, immunogenicity, and protective efficacy of these chimeric S proteins in vitro and in mice. Piglets were then immunized with two chimeric S proteins [TGEV S-PEDV(D0/ NTD) and TGEV S-PEDV(D0/NTD/CTD)] to evaluate their broad-spectrum immunoge nicity against PEDV and TGEV. Our results demonstrated that the TGEV S-PEDV(D0/ NTD) induced better immunity than the TGEV S-PEDV(D0/NTD/CTD) did, suggesting a promising strategy for preventing PEDV and TGEV infections.
## RESULTS AND DISCUSSION
## Expression and purification of recombinant proteins
To increase antigen expression, the TGEV S and PEDV S genes were optimized and synthesized, and we engineered TGEV and PEDV spike trimers by incorporating the GCN4 trimer domain, a Strep-tag II, and an 8× His tag. For each, we introduced the 2P mutation (E1139P/L1140P for TGEV S; S1076P/L1077P for PEDV S) and removed 59 or 67 residues from the C-terminus, respectively. As shown in Fig. 1A, the cDNA encoding the recombinant proteins was subcloned and inserted into the pCAGGS expression vector and successfully expressed in HEK293F cells. Proteins were purified by His-tag affinity chromatography, followed by size exclusion chromatography (SEC). The SEC chromato gram of the purified proteins revealed that TGEV S-Trimer eluted as a single peak with a molecular weight corresponding to a molecular weight of above 669 kDa and that PEDV S-Trimer eluted as a single peak with a molecular weight at or below approximately 669 kDa, suggesting that both TGEV S-Trimer and PEDV S-Trimer predominantly form a heavily glycosylated homotrimer (Fig. 1B andD). The expression and purification of the TGEV S protein and the PEDV S protein were confirmed through SDS-PAGE and western blot analyses. As shown in Fig. 1C andE, the bands corresponding to TGEV S-monomer and PEDV S-monomer appeared near the protein standard at 180 kDa, consistent with the molecular weight of a glycosylated monomer that showed no evidence of cleavage by endogenous proteases. These results indicated that purified trimeric TGEV S and PEDV S antigens were obtained.
## The ability of the TGEV S protein to induce neutralizing antibodies is superior to that of PEDV
We next compared the immunogenicity of TGEV S with that of PEDV S in a mouse model. The study design and experimental groups are summarized in Fig. 2A: group 1 included mice inoculated with TGEV S (n = 4); group 2 included mice inoculated with PEDV S (n = 4); and group 3 included mice inoculated with normal saline as a mock (n = 4). All the mice in groups 1 and 2 received two inoculations at 2-week intervals. Blood was collected from the mice on day 21. The mice experienced no adverse events after vaccination. Serum samples were collected to evaluate antibody responses using enzyme-linked immunosorbent assay (ELISA) or virus neutralization tests. As shown in Fig. 2B andC, the TGEV Sspecific IgG titer response to the TGEV S protein was similar to the PEDV Sspecific IgG titer response to the PEDV S protein. However, at the specific IgG antibody level, immunization with TGEV S resulted in a weaker cross-reaction than with PEDV S, and vice versa.
In addition to the IgG response, the neutralizing ability of antibody production is a pivotal factor influencing the quality of immunity. To further evaluate the humoral response induced by S proteins in experimental animals, neutralizing antibodies against TGEV and PEDV were also detected. The levels of TGEV-neutralizing antibodies in the TGEV S-immunized group were significantly higher, with titers of approximately 2 12 in the serum, than those in the mock group (Fig. 2D). The levels of PEDV-neutralizing antibodies in the PEDV S-immunized group were significantly higher, with titers of approximately 2 7 in the serum, than those in the mock group (Fig. 2E). The mean TGEV-neutralizing antibody titer in the TGEV S-immunized group was significantly higher than the mean PEDV-neutralizing antibody titer in the PEDV S-immunized group. Moreover, mice immunized with TGEV S and those immunized with PEDV S did not exhibit cross-protection in terms of serum neutralizing antibodies (Fig. 2D andE). The results from transmission electron microscopy (TEM) analysis of negatively stained samples (120 kV) and cryo-EM samples (300 kV) demonstrated high homogeneity of the TGEV spike protein (Fig. 2F andG), with the vast majority of particles exhibiting well defined triangular morphology, demonstrating exceptional structural homogeneity. This well-preserved native-like architecture likely contributes to the robust immunogenicity of the TGEV S protein. Our results indicate that the TGEV S elicits a stronger humoral immune response than the PEDV S in mice.
## Antigenic epitopes of TGEV and PEDV S proteins are primarily located in the S1-D0, S1-NTD, and S1-CTD regions
We performed B-cell epitope prediction to analyze the effective linear epitopes of the TGEV S protein and the PEDV S protein. A total of 16 B-cell linear epitopes were identified in TGEV S and PEDV S (Fig. 3A). The predicted positive residues of TGEV S and PEDV S (the corresponding linear epitopes) are displayed on the structural surface (Fig. 3B andC). Among these epitopes, 12 and 9 B-cell linear epitopes were located in the S1 domain of TGEV and PEDV, respectively (Fig. 4A). The results show that B-cell linear epitopes in both proteins were predominantly located in the S1 region.
To further determine which subdomains within the S1 subunit harbor the highest density of B-cell epitopes, structural and discontinuous epitopes in the TGEV S and PEDV S proteins were predicted with the DiscoTope 3.0 server. A total of 198 and 188 amino acid residues located on the S protein were predicted to be conformational epitopes for TGEV and PEDV, respectively. Among these residues, 133 and 128 residues were located in the S1 domain of TGEV and PEDV, respectively (Fig. 4A). The above findings demonstrate that the number of predicted B-cell epitopes on S1 is significantly higher than that on S2 for both TGEV and PEDV. Additionally, the epitopes of TGEV S and PEDV S are contained primarily within the D0, NTD, and CTD subdomains of the S1 subunit (Fig. 4B andC). These results suggest that retaining the different subdomains of S1 with the majority of B-cell epitopes that can induce protective Abs is vital for designing effective chimeric S protein subunit vaccines.
## Optimized expression strategy for the chimeric protein TGEV S-PEDV(D0/ NTD)
Our previous study suggested that in alpha-CoVs, subunit vaccines should prioritize the S-trimer rather than the S1-RBD (38). Moreover, the above results demonstrate that the immunogenic efficacy of the TGEV S protein surpasses that of the PEDV S protein, and the TGEV spike protein exhibited excellent homogeneity. Therefore, we engineered seven chimeric spike proteins with the TGEV S protein trimer as the backbone, which expressed different domains of the PEDV S1 subunit (Fig. 5A). To preliminarily screen suitable chimeric S proteins, the expression levels of these S protein chimeras on different days were determined by western blot analysis. As shown in Fig. 5B, among the seven chimeric spike proteins, the expression levels of three spike protein chimeras-TGEV S-PEDV(D0/NTD), TGEV S-PEDV(D0/NTD/CTD), and TGEV S-PEDV(S1)-were close to those of TGEV S and PEDV S, whereas the other four chimeras presented lower expres sion levels. In these constructs, the D0-NTD domain (aa 1-492) within the TGEV spike trimer was replaced with the corresponding PEDV D0-NTD region (aa 1-474) [designated TGEV S-PEDV(D0/NTD)], the D0-NTD-CTD domain (aa 1-675) within the TGEV spike trimer was replaced with the corresponding PEDV D0-NTD-CTD region (aa 1-640) [designated TGEV S-PEDV(D0/NTD/CTD)], and the S1 domain (aa 1-815) within the TGEV spike trimer was replaced with the corresponding PEDV S1 region (aa 1-764) [designated TGEV S-PEDV(S1)]. The results indicated that we successfully screened three chimeras with high expression levels as candidate chimeric S proteins (Fig. 5C).
To explore the immunogenicity of these three chimeras, a western blot analysis was conducted with serum from mice immunized with TGEV S or PEDV S as the primary antibody. As expected, all three spike protein chimeras demonstrated specific reactivity with both TGEV S-and PEDV S-positive sera (Fig. 6A andB). In addition, the stability of these three spike protein chimeras was assessed under the following conditions: (i) short-term stability (15 days of storage at 4°C) and (ii) freeze-thaw stability (six freezethaw cycles at -80°C). As depicted in Fig. 6C through E, SDS-PAGE analysis revealed that there was no significant degradation after six freeze-thaw cycles in all three spike protein chimeras. Moreover, all three chimeras could be stored at 4°C for at least 7 days without observable degradation (Fig. 6F through H). In detail, TGEV S-PEDV(D0/NTD) can be stored at 4°C for 7 days, and TGEV S-PEDV(D0/NTD/CTD) and TGEV S-PEDV(S1) can be stored at 4°C for at least 15 days. Together, these results indicate that TGEV S-PEDV(D0/NTD), TGEV S-PEDV(D0/NTD/CTD), and TGEV S-PEDV(S1) exhibit promising immunogenicity, short-term stability, and freeze-thaw stability.
## The chimeric protein [TGEV S-PEDV(D0/NTD)] can induce neutralizing antibodies against both TGEV and PEDV in mice
We next assessed the immunogenicity of chimeric S proteins in a mouse model. The study design and experimental groups are summarized in Fig. 7A: group 1 was inocula ted with TGEV S (n = 4), group 2 was inoculated with PEDV S (n = 4), group 3 was inoculated with TGEV S-PEDV(D0/NTD) (n = 4), group 4 was inoculated with TGEV S-PEDV(D0/NTD/CTD) (n = 4), group 5 was inoculated with TGEV S-PEDV(S1) (n = 4), protein-immunized mice. As shown in Fig. 7E and F, all three chimeric S proteins induced TGEV S-and PEDV Sspecific IgG antibodies in the serum. We subsequently measured the neutralizing antibodies against TGEV and PEDV in the sera of the three chimeric S protein-immunized mice. As shown in Fig. 7G, the mean neutralizing antibody titers against TGEV in the TGEV S-PEDV(D0/NTD) group were similar to those in the TGEV S group, whereas the mean neutralizing antibody titers against TGEV in the TGEV S-PEDV(D0/NTD/CTD) group and in the TGEV S-PEDV(S1) group were lower than those in the TGEV S group, but these titers were still higher than those in the PEDV S group, indicating a partial neutralizing protective capacity. As shown in Fig. 7H, although the level of neutralization against PEDV in the three chimeric S protein groups was lower than that in the PEDV S group, the mean neutralizing antibody titers against PEDV induced by all the three chimeric S protein groups were greater than those in the TGEV S group. These results suggest that the three chimeric S proteins could effectively induce varying levels of neutralizing antibodies against both TGEV and PEDV in mice.
## The chimeric protein [TGEV S-PEDV(D0/NTD)] can induce neutralizing antibodies against both TGEV and PEDV in piglets
Based on the results obtained from mouse experiments, we selected TGEV S-PEDV(D0/ NTD) and PEDV(D0/NTD/CTD) for subsequent immunogenicity evaluation in piglets. The study design is summarized in Fig. 8A. Briefly, five groups of 2-month-old piglets born from TGEV-and PEDV-negative sows were vaccinated intramuscularly with chimeric S proteins [TGEV S, PEDV S, TGEV S-PEDV(D0/NTD), TGEV S-PEDV(D0/NTD/CTD)] or normal saline as a mock and then administered a booster immunization 2 weeks later. Serum samples were collected at weekly intervals until 8 weeks (0, 7, 14, 21, 28, 35, 42, 49, and 56 days postvaccination, dpv) after which the neutralizing antibody titers against TGEV and PEDV were determined. Analysis of the neutralizing antibody titers revealed that piglets immunized with TGEV S exhibited a significantly high level of neutralizing antibodies against TGEV after booster immunization (14 dpv) and increased rapidly to the average peak titer of approximately 2 14.7 at 21 dpv (Fig. 8B), whereas the neutralizing antibodies against PEDV reached a peak titer of approximately 2 7.3 at 35 dpv in the PEDV S group (Fig. 8C). Furthermore, we determined the neutralizing antibody titers against TGEV and PEDV in the sera of these chimeric S protein-immunized piglets. As shown in Fig. 8B andC, the TGEV neutralizing antibody levels in the TGEV S-PEDV(D0/NTD) group were close to those in the TGEV S group, with an average peak titer of approximately 2 13.6 at 21 dpv. Moreover, the TGEV S-PEDV(D0/NTD) group presented sufficient neutralizing antibodies against PEDV, with an average peak titer of approximately 2 5.8 at 28 dpv.
In addition, the PEDV neutralizing antibody levels in the TGEV S-PEDV(D0/NTD/CTD) group were close to those in the PEDV S group, with an average peak titer of approx imately 2 7.2 at 28 dpv, whereas the TGEV S-PEDV(D0/NTD/CTD) group presented no neutralizing antibodies against TGEV (Fig. 8B andC). Together, these results showed that TGEV S-PEDV(D0/NTD) induced broad-spectrum neutralizing antibodies against TGEV and PEDV.
## Design strategy for a multivalent single antigen targeting the S proteins of PEDV and TGEV
In the swine industry, diarrhea caused by TGEV and PEDV causes serious economic losses worldwide. Thus, the development of protective vaccines against TGEV and PEDV remains a top priority. Recently, subunit vaccines based on the trimeric ectodomain of the S protein were developed and demonstrated favorable immune effects in mice and piglets (35,39,40). In this study, we first obtained trimeric full-length TGEV S and PEDV S and demonstrated that the immunogenic efficacy of TGEV S is superior to that of PEDV S, indicating that the TGEV S protein in its trimeric form can be used as a backbone to design a TGEV/PEDV S chimeric protein (Fig. 2). Furthermore, we predicted the potential B-cell epitopes for TGEV S and PEDV S, and the results revealed that the distribution of potential B-cell epitopes in TGEV S was similar to that in PEDV S. The epitopes of TGEV S and PEDV S were located mainly in S1-D0, S1-NTD, and S1-CTD (Fig. 3 and4). Therefore, the corresponding segments were replaced on TGEV S with different domains from those of PEDV S1; thus, seven distinct chimeric S proteins aimed at the formation of a multiepitope single antigen were designed. By comparing their expres sion levels, we subsequently screened three chimeric S proteins, TGEV S-PEDV(D0/NTD), TGEV S-PEDV(D0/NTD/CTD), and TGEV S-PEDV(S1). We speculate that the composition of these three chimeric S proteins may have a lesser impact on the stability of the S trimer protein structure and that these three chimeric S proteins exhibited high short-term stability and freeze-thaw stability (Fig. 6).
Neutralizing antibodies are critical indicators for evaluating their immune protective effect since they can directly reflect the protective capacity of a vaccine (41,42), and broad-spectrum activity is also a key factor in the design of an effective vaccine. In a mouse model, three chimeric S proteins elicited differential neutralizing antibody responses against both TGEV and PEDV. Notably, TGEV S-PEDV(D0/NTD) induced neutralizing antibody levels against TGEV comparable to those of the full-length TGEV S, whereas TGEV S-PEDV(D0/NTD/CTD) and TGEV S-PEDV(S1) showed significantly lower induction (Fig. 7). We further determined the neutralizing activity of the antibodies in the serum of immunized piglets. The results showed that TGEV S-PEDV(D0/NTD) could produce broad-spectrum neutralizing antibodies against both TGEV and PEDV. Considering that TGEV S-PEDV(D0/NTD/CTD) did not produce neutralizing antibodies against TGEV, we surmised that TGEV RBD played a crucial role in inducing neutralizing antibodies, which was confirmed in a mouse model (Fig. 8). Accordingly, we speculate that replacing the corresponding segments on TGEV S with D0-NTD domains from PEDV S is an effective strategy for designing novel bivalent subunit vaccines (Fig. 9).
In summary, on the basis of our research results, we propose a research and development strategy for producing efficient broad-spectrum subunit vaccines against both TGEV and PEDV. Replacing the corresponding segments on TGEV S with D0-NTD domains from PEDV S does not affect the protein expression level. Furthermore, through immunogenicity evaluation in mice and weaned piglets, it was confirmed that TGEV S-PEDV(D0/NTD) can provide broad-spectrum cross-neutralization protection against TGEV and PEDV. Our findings provide a novel strategy for the development of safe and effective bivalent subunit vaccine candidates against TGEV and PEDV in the future.
## MATERIALS AND METHODS
## Cell lines and viruses
Vero cells (CCL-81) and PK-15 cells were obtained from the American Type Culture Collection (ATCC) and were cultured in Dulbecco's modified Eagle medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin at 37°C in a 5% CO 2 humidified atmosphere. HEK293F cells preserved in our laboratory were maintained in serum-free Union 293 medium (Union, Cat# UP1000) with shaking at 120 rpm and 37°C in a humidified atmosphere comprising 8% CO 2 . The PEDV G2 strain CT P10 (GenBank accession no. MN114121) was propagated in Vero cells supplemented with 5 µg/mL trypsin (Gibco). The TGEV strain WH-1 (GenBank accession no. HQ462571) was propagated in PK-15 cells in Dulbecco's modified Eagle's medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS) in an incubator.
## B-cell epitope prediction analysis
Homology modeling of the S protein of PEDV CT P10 and TGEV WH-1 was conduc ted using the SWISS-MODEL (http://swissmodel.expasy.org/) to perform B-cell epitope predictions of the S-trimers (1). The cryo-electron microscopy (cryo-EM) structure of the PEDV PT52 S-trimer (PDB ID: 7W6M) was used as a template. According to previous research, B-cell epitopes were predicted and analyzed (IEDB, https://www.iedb.org/). Briefly, structure-based B-cell epitopes were predicted using DiscoTope 3.0 (https://services.healthtech.dtu.dk/services/DiscoTope-3.0/) with a confidence threshold of 0.90 (recall up to ~50%, default). For linear B-cell epit ope prediction, ABCpred, the IEDB server, and BcePred were employed with default parameters to optimize the sensitivity, specificity, and positive predictive value (43)(44)(45). To improve the accuracy of epitope prediction, the results of these three prediction sites were synthesized, and the overlapping part of the epitopes predicted by at least two websites was selected as the linear epitope of the dominant B cells. All the predicted residues were labeled in the corresponding structures using PyMOL (Schrödinger LLC). No epitope was predicted within the signal peptide (SP), FP, or TM domains of the two S proteins.
## Plasmid construction, protein expression, and purification
The human-optimized codon gene encoding the PEDV spike protein (GenBank accession no. QHB92364.1) and the human-optimized codon gene encoding the TGEV spike protein (GenBank accession no. ADY39740.1) were synthesized (GenScript Biotech). The DNA sequences of PEDV S (residues 1-1,319) and TGEV S ectodomain (residues 1-1,388) were cloned and inserted into a pCAGGS vector with a C-terminal GCN4 trimerization motif followed by a Strep-tag and an 8 × HisTag sequence. The other eukaryotic expression plasmids were constructed as shown in Fig. 5A. All S proteins were stabilized with a dual-proline (2P) mutation ( 1076 SL 1077 → 1076 PP 1077 for PEDV S, 1139 EL 1140 → 1139 PP 1140 for TGEV S). The expression plasmid was transiently transfected into a suspension of HEK293F cells using polyetherimide (PEI), and the cells were cultured for another 6 days at 37°C and 8% CO 2 . The proteins were harvested from the supernatants of the cell culture medium and purified on a Ni-NTA column. After affinity purification, the proteins were concentrated and subsequently subjected to additional purification using a size exclusion chromatography (SEC) column (Superose 6 increase 10/300 Gl; GE Healthcare, U.S.A.), utilizing a running buffer consisting of 20 mM HEPES (pH 7.6) and 150 mM NaCl. All the proteins were exchanged in phosphatebuffered saline (PBS) and stored at -80°C.
## SDS-PAGE and western blotting
SDS-PAGE and western blotting were carried out as described by us earlier (46). The collected protein samples were mixed with 5 × reducing loading buffer, boiled for 10 min, and then loaded onto 10% SDS-PAGE gels. The proteins were electrophoresed for 2.5 h at 80 V in a Bio-Rad MINI-PROTEAN Tetra system (Bio-Rad Laboratories). The gel was stained with Coomassie Brilliant Blue R-250 (Bio-Rad) for 30 min at room tempera ture and then decolorized with eluent overnight. For western blotting, after the proteins were resolved by SDS-PAGE, they were transferred onto a polyvinylidene fluoride (PVDF) membrane. The PVDF membranes were blocked with 5% skim milk/TBST overnight at 4°C. The membrane was then incubated with diluted mouse serum (1:500 dilution) or an anti-His-tag monoclonal antibody (1:5,000 dilution) for 2 h at room temperature. The membranes were subsequently incubated with a goat anti-mouse secondary antibody (Abbkine, no. A21010) for 1 h (1:5.000) at room temperature. Finally, the membranes were visualized using an enhanced chemiluminescence system (Amersham Imager 600, GE Healthcare).
## Electron microscopy
For negative staining experiments, the spike protein samples were directly analyzed using a Talos L120C G2 transmission electron microscope (Thermo Fisher Scientific).
Copper grids with 300 mesh and a thin layer of continuous carbon floated on top (EM Sciences) were glow-discharged at 15 mA for 60 s. The spike protein samples were diluted to 0.005 mg/mL, and 5 µL was applied to the grids for 30 s. Then, three drops containing 10 µL of uranyl acetate (3%, vol/vol) staining solution were applied to the parafilm (Bemis). Samples on the grids were then stained with the third drop of uranyl acetate for 60 s, followed by allowing it to rapidly combine with the first two drops. The specimen was finally gently blotted from the side with filter paper, air-dried at 25 °C for 30 min, and stored until imaging.
For cryo-EM sample preparation, 3 µL of spike protein solution (1 mg/mL) was applied to a glow-discharged (40 s, 15 mA) holey carbon Cu Quantifoil grid (R1.2/1.3, 300 mesh). The grids were blotted for 3 s at 100% humidity and 280 K and then plunge-frozen in liquid ethane using a Vitrobot (ThermoScientific). Cryo-EM micrographs with a defocus from -1 to -2 µm were collected on a Gatan K3 direct electron detector in superresolu tion mode on a Krios G4 cryoTEM (Thermo Fisher). A series of micrographs was acquired at ×105,000 magnification with a pixel size of 1.648 Å pixel -1 and a dose of 50 e -Å -2 for sample quality assessment. Automated data collection was performed by EPU software (Thermo Fisher Scientific).
## Design of the mouse vaccination experiments
Female BALB/c mice (four per group) aged 6 weeks were immunized with different proteins at 0 and 2 weeks. Proteins (20 µg) diluted in normal saline were mixed 1:1 with QuickAntibody-Mouse3W adjuvant (BioDragon). The mice were intramuscularly inoculated with 100 µL of this solution (50 µL into each hind leg). One week after the final immunization, sera were collected for subsequent assays.
## Design of the pig vaccination experiment
To determine the efficacy of the chimeric S proteins, 25 2-month-old PEDV/TGEV-naive piglets were placed in separate rooms and randomly assigned to four experimental groups and one mock group: the PEDV S-immunized group (n = 5), the TGEV S-immu nized group (n = 5), the TGEV S-PEDV(D0/NTD)-immunized group (n = 5), the TGEV S-PEDV(D0/NTD/CTD)-immunized group (n = 5), and the PBS-immunized group (n = 5). The piglets in the immunization groups were injected intramuscularly with the S vaccines (100 µg per piglet) or with PBS. At 14 dpv, the piglets in the immunization groups received a booster dose of the vaccine. All vaccines were formulated in Montanide ISA 201 VG adjuvant. Blood samples were collected from the anterior vena cava weekly. Each serum sample was tested for SN titers against PEDV and TGEV.
## Enzyme-linked immunosorbent assay (ELISA)
Serum Sspecific antibodies in each group were detected at 21 DPI. Titers of Sspecific antibodies in the serum were determined via indirect ELISA with purified S protein as the antigen, as described previously (38). Briefly, ELISA plates were coated with purified PEDV or TGEV S protein at 0.1 µM/well in citratebuffered saline (CBS, pH 9.6) overnight at 4°C and subsequently blocked with 1% (wt/vol) bovine serum albumin (BSA) in phosphatebuffered saline (PBS) containing 0.05% Tween 20 (PBST) at 37°C. For antibody detection, the plates were incubated with 10-fold serially diluted sera for 1 h at 37°C. After standard washes, 100 µL of diluted horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG (Boster, Wuhan, China) was used as the secondary antibody, and 3,3′,5,5′-tetramethylbenzidine (TMB) (Beyotime) was used as the substrate for detection. The final absorbance [optical density (OD)] was measured at 450 nm and 630 nm using a Spark 10 M microplate reader (Tecan) after the reaction was stopped with 2 M H 2 SO 4 . Serum from mice immunized with normal saline was used as a control.
## Neutralization assays
To detect PEDVspecific neutralizing antibodies, the serum-neutralizing antibody titers were determined with a virus neutralization test in 96-well cell culture plates (46). Briefly, the serum samples were heat-inactivated at 56°C for 30 min and serially diluted twofold with DMEM. The diluted samples were mixed with an equal volume of 200 TCID 50 of PEDV and incubated at 37°C for 1 h. Subsequently, 100 µL of the virus-serum mixture was inoculated into Vero cells in 96-well plates at 37°C for 1 h, followed by the addition of 100 µL of DMEM with 5 µg/mL trypsin, and the mixture was maintained at 37°C in a 5% CO 2 incubator. After incubation for 24-48 h, the neutralizing antibody titers were expressed as the reciprocal of the highest serum dilution that inhibited PEDVspecific CPE.
To detect TGEVspecific neutralizing antibodies, the serum-neutralizing antibody titers were determined with a virus neutralization test in 96-well cell culture plates. Briefly, the serum samples were heat-inactivated at 56°C for 30 min and serially diluted 2-fold with DMEM. The diluted samples were mixed with an equal volume of TGEV (200 TCID 50 ) and incubated at 37°C for 1 h. Subsequently, 100 µL of the virus-serum mixture was inoculated into PK-15 cells in 96-well plates. After incubation at 37°C for 48-72 h, the neutralizing antibody titers were expressed as the reciprocal of the highest serum dilution that inhibited TGEVspecific CPE.
## Statistical analysis
Statistical analysis was carried out using GraphPad Prism 8.0. Statistical significance was determined using an unpaired two-tailed Student's t test. The Data are presented as the means ± SDs (95% confidence intervals). *, P < 0.05 was considered statisti cally significant, **, P < 0.01 was considered highly significant, and ****, P < 0.0001 was considered extremely significant. All experiments were further confirmed using biological replicates.
## References
1. Boniotti, Papetti, Lavazza et al. (2016) "Porcine epidemic diarrhea virus and discovery of a recombinant swine enteric coronavirus" *Italy. Emerg Infect Dis*
2. Gu, Yue, Xing et al. (2012) "Progress in genetically engineered vaccines for porcine transmissible gastroenteritis virus" *Rev Med Vet (Toulouse)*
3. Hanke, Jenckel, Petrov et al. (2014) "Comparison of porcine epidemic diarrhea viruses from Germany and the United States" *Emerg Infect Dis*
4. Lee (2015) "Porcine epidemic diarrhea virus: an emerging and reemerging epizootic swine virus" *Virol J*
5. Pearce, Schweer, Schwartz et al. (2016) "Pig jejunum protein profile changes in response to a porcine epidemic diarrhea virus challenge" *J Anim Sci*
6. Sun, Cai, Chen et al. (2012) "Outbreak of porcine epidemic diarrhea in suckling piglets" *China. Emerg Infect Dis*
7. Ahmad, Yeap, Ali et al. (2012) "scFv antibody: principles and clinical application" *Clin Dev Immunol*
8. Xia, Yang, Wang et al. (2018) "Impact of TGEV infection on the pig small intestine" *Virol J*
9. Li, Li, Liu et al. (2011) "New variants of porcine epidemic diarrhea virus" *Emerg Infect Dis*
10. Zhang, Zou, Peng et al. (2023) "Global dynamics of porcine enteric coronavirus PEDV epidemiology, evolution, and transmission" *Mol Biol Evol*
11. Lee, Lee (2013) "Outbreak-related porcine epidemic diarrhea virus strains similar to US strains, South Korea" *Emerg Infect Dis*
12. Stevenson, Hoang, Schwartz et al. (2013) "Emergence of Porcine epidemic diarrhea virus in the United States: clinical signs, lesions, and viral genomic sequences" *J Vet Diagn Invest*
13. Lee, Moro, Baker (1954) "Transmissible gastroenteritis in pigs" *J Am Vet Med Assoc*
14. Paton, Ibata, Sands et al. (1997) "Detection of transmissible gastroenteritis virus by RT-PCR and differentiation from porcine respiratory coronavirus" *J Virol Methods*
15. Cubero, León, Contreras et al. (1993) "Transmissible gastroenteritis in pigs in south east Spain: prevalence and factors associated with infection" *Vet Rec*
17. Hess, Bollwahn, Pospischil et al. (1980) "Current aspects in the etiology of viral diarrheas of swine: occurrence of infections with the epizootic viral diarrhea (EVD) virus" *Berl Munch Tierarztl Wochenschr*
18. Lin, Gao, Oka et al. (2015) "Antigenic relationships among porcine epidemic diarrhea virus and transmissible gastroenteritis virus strains" *J Virol*
19. Laude, Rasschaert, Delmas et al. (1990) "Molecular biology of transmissible gastroenteritis virus" *Vet Microbiol*
20. Brian, Baric (2005) "Coronavirus genome structure and replication" *Curr Top Microbiol Immunol*
21. Kong, Jia, Xiao et al. (2023) "Prevention and control of swine enteric coronaviruses in China: a review of vaccine development and application" *Vaccines (Basel)*
22. Chen, Knutson, Rossow et al. (2019) "Decline of transmissible gastroenteritis virus and its complex evolutionary relationship with porcine respiratory coronavirus in the United States" *Sci Rep*
23. Huang, Draczkowski, Wang et al. (2022) "In situ structure and dynamics of an alphacoronavirus spike protein by cryo-ET and cryo-EM" *Nat Commun*
24. Wrapp, Mclellan (2019) "The 3.1-Angstrom cryo-electron microscopy structure of the porcine epidemic diarrhea virus spike protein in the prefusion conformation" *J Virol*
25. Kirchdoerfer, Bhandari, Martini et al. (2021) "Structure and immune recognition of the porcine epidemic diarrhea virus spike protein" *Structure*
26. Li (2016) "Structure, function, and evolution of coronavirus spike proteins" *Annu Rev Virol*
27. Li, Van Kuppeveld, He et al. (2016) "Cellular entry of the porcine epidemic diarrhea virus" *Virus Res*
28. Li, Li, De Esesarte et al. (2017) "Cell attachment domains of the porcine epidemic diarrhea virus spike protein are key targets of neutralizing antibodies" *J Virol*
29. Ma, Li, Li et al. (2025) "Changes in the motifs in the D0 and SD2 domains of the S protein drive the evolution of virulence in enteric coronavirus porcine epidemic diarrhea virus" *J Virol*
30. (2025) *Full-Length Text Journal of Virology*
31. Zhang, Liu, Liu et al. (2024) "The emergence of novel variants of the porcine epidemic diarrhea virus spike gene from 2011 to 2023"
32. Li, Yang, Qian et al. (2025) "Molecular characteristics of the immune escape of coronavirus PEDV under the pressure of vaccine immunity" *J Virol*
33. Du, Yan, Zhang et al. (2023) "Virus-like particle vaccines with epitopes from porcine epidemic virus and transmissible gastroenteritis virus incorporated into self-assembling ADDomer platform provide clinical immune responses in piglets" *Front Immunol*
34. Fan, Yi, Zhong et al. (2025) "A trivalent enteric coronaviruses inactivated vaccine provides effective protection against PEDV, TGEV, and PDCoV" *Vet Microbiol*
35. Song, Moon, Kang (2015) "Porcine epidemic diarrhea: a review of current epidemiology and available vaccines" *Clin Exp Vaccine Res*
36. Li, Li, Huang et al. (2022) "Isolation and oral immunogenicity assessment of porcine epidemic diarrhea virus NH-TA2020 strain: one of the predominant strains circulating in China from 2017 to 2021" *Virol Sin*
37. Xu, Yang, Yang et al. (2025) "Development and immunogenicity study of subunit vaccines based on spike proteins of porcine epidemic diarrhea virus and porcine transmissible gastroenteritis virus" *Vet Sci*
38. Du, Tai, Zhou et al. (2016) "Vaccines for the prevention against the threat of MERS-CoV" *Expert Rev Vaccines*
39. Meena, Singhvi, Srichandan et al. (2022) "RBD decorated PLA nanoparticle admixture with aluminum hydroxide elicit robust and long lasting immune response against SARS-CoV-2" *Eur J Pharm Biopharm*
40. Shi, Shi, Sun et al. (2021) "Insight into vaccine development for Alpha-coronaviruses based on structural and immunological analyses of spike proteins" *J Virol*
41. Song, Li, Wang et al. (2024) "Efficacy evaluation of a bivalent subunit vaccine against epidemic PEDV heterologous strains with low cross-protection" *J Virol*
42. Guo, Wang, Song et al. (2024) "Immunogenicity and protective efficacy of a trimeric full-length S protein subunit vaccine for porcine epidemic diarrhea virus" *Vaccine (Auckl)*
43. Yu, Wang, Ma et al. (2017) "Immunogenicity of eGFP-marked recombinant Lactobacil lus casei against transmissible gastroenteritis virus and porcine epidemic diarrhea virus" *Viruses*
44. Hu, Luo, Liao et al. (2025) "Neutralizing antibody levels as a key factor in determining the immunogenic efficacy of the novel PEDV alpha coronavirus vaccine" *Veterinary Quarterly*
45. Saha, Raghava (2006) "Prediction of continuous B-cell epitopes in an antigen using recurrent neural network" *Proteins*
46. Grifoni, Sidney, Zhang et al. (2020) "A sequence homology and bioinformatic approach can predict candidate targets for immune responses to SARS-CoV-2" *Cell Host Microbe*
47. Saha, Raghava (2004) "BcePred: prediction of continuous B-cell epitopes in antigenic sequences using physico-chemical properties. Artificial immune systems"
48. Zhang, Xie, Liao et al. (2024) "Development of a safe and broadspectrum attenuated PEDV vaccine candidate by S2 subunit replace ment" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548439&blobtype=pdf | # Comparison of the biological properties of bat-derived filovirus envelope glycoproteins
Francois Edidi-Atani, Munyeku Yannick, Bazitama, Hiroko Miyamoto, Akina Mori-Kajihara, Hayato Sugiura, Manabu Igarashi, Jean Jacques Muyembe-Tamfum, Steve Ahuka-Mundeke, Ayato Takada
## Abstract
Although some filoviruses, such as Ebola virus (EBOV) and Marburg virus (MARV), are highly pathogenic in humans, novel filoviruses, including Lloviu virus (LLOV), Bombali virus (BOMV), Mengla virus (MLAV), and Dehong virus (DEHV), whose biological properties are poorly understood, have been found in bats. In this study, we character ized the envelope glycoproteins (GPs) of these bat-derived filoviruses (BatFiloVs). We first confirmed that virus-like particles consisting of their GPs, nucleoproteins, and matrix proteins were filamentous. Interestingly, although BatFiloVs were serologically distinct, some previously established monoclonal antibodies (MAbs) (e.g., 6D6) successfully neutralized vesicular stomatitis Indiana viruses pseudotyped with LLOV, BOMV, or DEHV GPs. The pseudotyped viruses bearing BatFiloV GPs utilized human TIM-1 and C-type lectins for entry into cells, although the efficiency tended to be lower than for EBOV and/or MARV GP-pseudotyped viruses. These viruses broadly infected cultured cells derived from various animal species, including humans and bats. However, viruses pseudotyped with DEHV and MARV GPs failed to infect the Yaeyama flying fox cell line, whereas the other pseudotyped viruses infected this cell line. Interestingly, the virus bearing BOMV GP showed the greatest ability to infect cell lines derived from Angolan free-tailed bats, the only known host species of BOMV. We identified unique amino acid residues at the interface between GP and its receptor (i.e., Niemann-Pick C1), which might explain these differences. Our results suggest that the biological properties of filovirus GPs are generally consistent with their phylogenetic relationship and that BatFiloVs may have differential pathogenicity and host range restriction.IMPORTANCE Filoviruses, such as EBOV and MARV, are known to cause severe hemor rhagic fever in humans and nonhuman primates. With the recent advancements in next-generation sequencing, novel filoviruses have been detected in bats. However, their pathogenicity and host tropism remain largely unknown. Here, we focus on the filovirus spike protein GP, which plays a crucial role in the viral lifecycle, and discuss the biological properties of BatFiloVs. We studied the primary structures of GPs, virus particle morphol ogy, antigenic differences of GPs, neutralizing capacities of anti-EBOV and -MARV GP MAbs, usage of some attachment factors during the entry into cells, and GP-mediated cellular tropism. The present study provides fundamental information for understanding the BatFiloV ecology, host ranges, and potential risks as zoonotic pathogens for humans. This knowledge will guide public health interventions to prevent virus spillovers and the development of surveillance strategies and specific countermeasures.
Dianlovirus, and Loebevirus, comprising 17 species and 18 viruses (1,2). Of these, Ebola virus (EBOV), Sudan virus (SUDV), Tai Forest virus (TAFV), and Bundibugyo virus (BDBV) in the genus Orthoebolavirus, and Marburg virus (MARV) and Ravn virus (RAVV) in the genus Orthomarburgvirus, are known to be pathogenic in humans and cause severe hemorrhagic fever with high mortality rates. Reston virus (RESTV), which causes lethal infection in nonhuman primates, is believed to be nonpathogenic to humans, although its pathogenicity has not yet been fully clarified (3). Almost all filovirus disease outbreaks have occurred in Equatorial Africa, except for some imported cases reported in Europe and the United States, suggesting that diseases caused by filoviruses pathogenic to humans are endemic primarily in the African continent (4). The natural reservoir hosts for filoviruses are not fully elucidated, except for MARV, which has been frequently isolated from or detected in Egyptian fruit bats (Rousettus aegyptiacus) (5)(6)(7).
The recent advancement of new pathogen detection tools, such as next-generation sequencing, has led to the expansion of the Filoviridae family with the discovery of new filoviruses in bats: Lloviu virus (LLOV) in the genus Cuevavirus, Bombali virus (BOMV) in the genus Orthoebolavirus, Mengla virus (MLAV), and Dehong virus (DEHV) in the genus Dianlovirus (1,2). Among these bat-derived filoviruses (BatFiloVs), the Lloviu virus (LLOV) genome was first detected in insectivorous bats (Miniopterus schreibersii) in some European countries (8)(9)(10), and infectious LLOV was subsequently isolated in the same bat species in Hungary and Italy (11,12). The Bombali virus (BOMV) genome was detected in free-tailed bats (Mops condylurus and Chaerephon pumilus) in several African countries (13)(14)(15)(16). The Mengla virus (MLAV) genome was detected in fruit bats (Rousettus leschenaultii) in China (17). More recently, Dehong virus (DEHV) was isolated from fruit bats (Rousettus leschenaultii) in China (18).
Known filovirus particles are enveloped, variously shaped, but predominantly filamentous and contain a linear, negative-sense, and non-segmented RNA genome. The RNA genomes of the above-mentioned filoviruses encode at least seven struc tural proteins: envelope glycoprotein (GP), major matrix protein (VP40), nucleoprotein (NP), polymerase cofactor (VP35), replication/transcription protein (VP30), minor matrix protein (VP24), and RNA-dependent RNA polymerase (L) (2). In addition to these structural proteins, orthoebolaviruses express a nonstructural soluble GP (sGP and ssGP) through RNA editing (2,(19)(20)(21). Among these viral proteins, GP is the sole protein present on the virus surface and is responsible for receptor binding and fusion of the virus envelope with the host cell membrane (22,23). This protein is the only target of neutralizing antibodies to filoviruses (24). It exhibits genetic and antigenic variation, whereas other filovirus proteins (e.g., NP and VP40) are relatively conserved among filovirus species (25). Thus, GP is expected to induce more species-specific and less cross-reactive antibodies than other filovirus proteins (26). The GP monomer consists of 2 subunits, GP1 and GP2, which are linked by a disulfide bridge (27). The GP1 subunit contains a receptor-binding domain (RBD) responsible for attachment to the host cell receptor Niemann-Pick C1 (NPC1) and the mucin-like domain (MLD), which is heavily glycosylated with large amounts of N-and O-linked glycans (28)(29)(30). Its highly variable amino acid sequences and sugar chain structures suggest different GP properties among filovirus species (23). During viral entry into cells, GP1 is digested with host proteases to expose RBD, followed by the interaction with NPC1 on the endosomal membrane (31). The GP2 subunit contains an internal fusion loop, two heptad repeats, a membrane-prox imal external region, a transmembrane domain, and a cytoplasmic tail (32).
Currently, BatFiloVs are poorly characterized since infectious viruses have rarely been isolated. Their biological properties, such as pathogenicity in humans and nonhuman primates, transmission routes, and host range remain largely unknown, whereas it was reported that immunodeficient mice infected with recombinant LLOV or BOMV showed low pathogenicity or minimal signs of diseases (15,33) and that ferrets survived without signs of disease regardless of the dose and exposure routes after LLOV infection (34).
In the present study, we focused on filovirus envelope GP, which is thought to play a crucial role in viral pathogenicity and tropism, and compared biological properties among mammalian filoviruses (i.e., BatFiloVs and human/nonhuman primate-pathogenic filoviruses). We first analyzed the primary structures of the GPs, the morphology of virus-like particles (VLPs) consisting of GP, NP, and VP40, and the antigenic differences of GPs among filoviruses. Then, using replication-incompetent vesicular stomatitis Indiana virus (VSIV) pseudotyped with filovirus GPs, we investigated the neutralizing activities of several previously established monoclonal antibodies (MAbs) against EBOV and MARV GPs, the ability of GPs to use host attachment factors, and cellular tropism in cell lines derived from various animal species. Our findings provide valuable insights into the functional properties of BatFiloV GPs and contribute to a better understanding of their zoonotic risk.
## RESULTS
## Comparison of the primary structures among BatFiloV, EBOV, and MARV GPs
In the phylogenetic tree of mammalian filoviruses, GPs are divided into EBOV-like and MARV-like phylogroups, like other proteins such as NP and L, as described previously (18) (Fig. 1). We first compared the primary structures of BatFiloV, EBOV, and MARV GPs (Fig. S1 andS2). As expected, the N-terminal one-third regions and C-terminal one-third regions were relatively conserved among the viruses, and the middle regions, principally corresponding to their MLDs, were highly divergent. We identified 10 highly conserved cysteine residues in the relatively conserved regions. Additionally, two cysteine residues in GP1 (C121 and C147) were conserved only among EBOV, LLOV, and BOMV. We found that cysteine residues contributing to the disulfide bridge that links GP1 and GP2, as well as those involved in intramolecular subunit stabilization, were conserved (35). Two cysteine residues (C670 and C672), known to be required for GP acylation, were also conserved (36). The potential cleavage site motifs recognized by ubiquitous host proteases such as furin were found in all the GP sequences: RRRR for LLOV as previously described (37), RAKR for BOMV, RKRR for DEHV, and two motifs (RSKR and KKKR) for MLAV. N-glycosylation motifs were also found in all GPs: 17 sites for EBOV and 9 sites for BOMV as described in a previous study (38), 20 sites for LLOV, 23 sites for MARV, 15 sites for MLAV, and 19 sites for DEHV. We found two fully conserved N-glycosylation sites in GP2 (N563 and N618) among all orthoebolaviruses, MARV, and LLOV (38). The O-glycosylation site prediction revealed 80 sites for EBOV, 88 sites for LLOV, 61 sites for BOMV, 100 sites for MARV, 59 sites for MLAV, and 97 sites for DEHV, most of which were likely involved in forming their MLDs (Fig. S1). LLOV, MARV, MLAV, and DEHV MLDs were located over the cleavage sites. Consistent with the phylogenetic relationships, the overall characteristics of GP primary structures of LLOV and BOMV were closer to EBOV than to MARV, whereas MLAV and DEHV were closer to MARV than to EBOV.
## Morphology of VLPs consisting of BatFiloV GP, VP40, and NP
A unique characteristic of filovirus particles is their filamentous shape, and it has been shown that morphologically similar VLPs can be produced by the expression of GP, VP40, and NP (37,39). However, information on the morphology of BatFiloV particles remains limited. To address this, we generated VLPs by transient expression of BatFiloV GP, VP40, and NP in cultured cells and examined their morphology using electron microscopy (Fig. 2; Fig. S3). We confirmed that VLPs composed of the LLOV, BOMV, MLAV, or DEHV proteins exhibited a filamentous shape with densely arrayed spikes on their surfaces, similar to those of EBOV. Their diameters were uniform (approximately 80-90 nm), whereas their length varied. BOMV VLPs tended to be slightly thinner. These results suggest that BatFiloV GP, VP40, and NP also play crucial roles in producing virus particles with the characteristic filamentous shape.
## Antigenic comparison among filovirus GPs
Mammalian filoviruses are phylogenetically divided into four genera: Orthoebolavirus, Orthomarburgvirus, Cuevavirus, and Dianlovirus (2), three of which (Orthoebolavirus, Cuevavirus, and Dianlovirus) include BatFiloVs. However, information on the antigenic differences among BatFiloV GPs and other filovirus GPs remains limited (40). To investi gate antigenic relationships among GPs of mammalian filoviruses, we produced mouse antisera against VLPs of ten mammalian filoviruses selected from each genus and tested their IgG reactivities to the respective GP antigens (Fig. 3). We found that anti-LLOV, anti-BOMV, anti-MLAV, and anti-DEHV GP sera showed exclusive reactivity to their homolo gous GP antigens. Anti-EBOV sera showed slight cross-reactivity with BDBV and RESTV GPs. In general, all antisera demonstrated high specificity to homologous GP antigens with limited cross-reactivity to other viral antigens.
## Cross-reactivity of anti-EBOV and anti-MARV GP MAbs against BatFiloVs
The recent discovery of novel filoviruses highlights the urgent need to develop com pounds or drugs for pan-filovirus treatment. Therapeutic use of neutralizing MAbs is a potential option in the event of BatFiloV emergence in the human population (41). In this study, some previously established MAbs targeting EBOV and MARV GPs were tested for their ability to neutralize other filoviruses, including BatFiloVs, using VSIV pseudotyped (VSVΔG*) with filovirus GPs (Fig. 4). We found that 6D6, which has been shown to inhibit EBOV, SUDV, BDBV, TAFV, and RESTV infection (42), efficiently neutralized VSVΔG*LLOV-GP and VSVΔG*BOMV-GP but not VSVΔG*MLAV-GP, VSVΔG*DEHV-GP, or VSVΔG*MARV-GP. ADI-15946, which is known to inhibit EBOV, BDBV, and TAFV infection (43,44), efficiently neutralized VSVΔG*BDBV-GP and VSVΔG*TAFV-GP and showed slight neutralization of VSVΔG*BOMV-GP and VSVΔG*SUDV-GP but did not neutralize the other viruses. mAb114, a drug approved for EBOV treatment (45), showed strong inhibitory activity against VSVΔG*EBOV-GP and slightly neutralized VSVΔG*BDBV-GP. Previously reported EBOV-specific KZ52, 133/3.16, and 226/8.1 (24,46) showed no cross-neutraliz ing activity against the viruses with other filovirus GPs. Of the two anti-MARV GP neutralizing MAbs, MR78 and MR191 (47)(48)(49), MR191 neutralized VSVΔG*DEHV-GP to some extent, but not VSVΔG*MLAV-GP. These findings indicate that the epitope of 6D6 is widely shared among LLOV, BOMV, and other orthoebolaviruses and that MR191 recognizes a common epitope partially conserved between DEHV and MARV GPs.
## Roles of cellular attachment factors in BatFiloV entry
Human T-cell immunoglobulin and mucin domain-1 (hTIM-1) has been shown to facilitate the attachment of EBOV and MARV to cell surfaces (50). We investigated the potential of hTIM-1 to promote BatFiloV entry into cells using pseudotyped viruses (Fig. 5A). All tested viruses were found to infect hTIM-1-expressing cells more efficiently than control cells, although the extent of enhancement varied among the viruses. (51)(52)(53). To confirm the involvement of DC-SIGN and hMGL in BatFiloV entry into host cells and to compare their efficiency among filoviruses, we investigated the infectivity of the pseudotyped viruses in cells expressing these C-type lectins. As expected, all tested viruses infected DC-SIGN-expressing cells more efficiently than control cells, although the degree of enhancement varied among the viruses (Fig. 5B). There was no significant difference in infectivity between VSVΔG*EBOV-GP and VSVΔG*LLOV-GP (both showing more than 30-fold increases). Notably, VSVΔG*LLOV-GP exhibited significantly higher efficiency to infect DC-SIGN-expressing cells than three other BatFiloVs (i.e., VSVΔG*BOMV-GP, VSVΔG*MLAV-GP, and VSVΔG*DEHV-GP), which showed approxi mately 11-fold, 4-fold, and 4-fold higher infectivity in DC-SIGN-expressing cells, respec tively, than in control cells. VSVΔG*MLAV-GP and VSVΔG*DEHV-GP exhibited lower efficiency in infecting DC-SIGN-expressing cells than VSVΔG*MARV-GP and VSVΔG*BOMV-GP.
Similar to the findings in DC-SIGN-expressing cells, all the tested viruses showed higher infectivity in hMGL-expressing cells than in control cells (Fig. 5C). As expected, VSVΔG*EBOV-GP, VSVΔG*RESTV-GP, and VSVΔG*MARV-GP showed remarkable increases
## Cellular tropism of VSIVs pseudotyped with BatFiloV GPs
To investigate GP-mediated cellular tropism, 21 cell lines of various animal origins, including human and bat cell lines, were infected with the pseudotyped viruses, and infectious units (IUs) were determined for each cell line (Fig. 6A; Table 1). As expected, VSVΔG*VSIV-G uniformly infected these cells at high titers (approximately 10 5 -10 8 IUs/ mL). The viruses pseudotyped with BatFiloV GPs, as well as those with EBOV, RESTV, and MARV, also displayed broad cellular tropism, infecting human, monkey, hamster, pig, dog, bovine, mouse, and bat cell lines. The overall infectivity patterns of VSVΔG*LLOV-GP and VSVΔG*BOMV-GP in these cell lines were similar to those of VSVΔG*RESTV GP and VSVΔG*EBOV-GP, except in the cell line derived from a straw-colored fruit bat (ZFBK13-76E), which is known to be less susceptible to EBOV (54,55). In contrast, the infectivity patterns of VSVΔG*MLAV-GP and VSVΔG*DEHV-GP resembled those of VSVΔG*MARV-GP. Interestingly, VSVΔG*MARV-GP and VSVΔG*DEHV-GP failed to infect the cell line derived from a Yaeyama flying fox (FBKT1), which is known to have reduced susceptibility to MARV (37,55), whereas VSVΔG*MLAV-GP successfully infected this cell line (Fig. 6A). When the relative infectivities compared to Vero E6 cells were determined for all cell lines (Fig. 6B), it was also found that VSVΔG*BOMV-GP exhibited increased ability to infect cell lines (MoKi3 C1, MoKi3-P, and MoLu6 Prim) derived from Angolan free-tailed bats, the only known host animal species for BOMV, with 100-fold to 1,000fold higher titers than the other viruses tested.
## Amino acid differences found in GPs among filoviruses and NPC1 loops among cell lines
It has been shown that the interaction between GP RBD and the NPC1 protein is one of the key determinants of filovirus host tropism (55)(56)(57). NPC1, located in late endosomes, acts as the fusion receptor for filovirus entry into cells. This protein contains two loop regions in its domain C (NPC1-C) that interact with GP RBD (Fig. 7A) (29,30). Thus, amino acid variations at the interface between the NPC1-C loops and GP RBD significantly influence cell susceptibility and host tropism (54)(55)(56)(58)(59)(60). First, we focused on the molecular mechanism underlying the differential infectivity of VSVΔG*MLAV-GP, VSVΔG*DEHV-GP, and VSVΔG*MARV-GP in FBKT1 cells by comparing amino acid residues in GP RBD. Although the overall similarity among MARV, DEHV, and MLAV was high, we identified two unique amino acid residues in one of the NPC1-C loop-interacting regions of MLAV GP (isoleucine and valine at positions 113 and 114 [EBOV numbering], respec tively), which differed from those in MARV and DEHV (Fig. 7B). Isoleucine at position 113 was also found in EBOV, RESTV, BOMV, and LLOV GPs, whereas valine at position 114 was unique to MLAV GP. Next, to understand the molecular basis of the preference of VSVΔG*BOMV-GP for MoKi3 C1, MoKi3-P, and MoLu6 Prim cells, the amino acid sequen ces of the NPC1-C loops among the human-, monkey-, and bat-derived cell lines used in this study, as well as GP sequences, were compared (Fig. 7C). We found one unique amino acid residue on BOMV GP (glutamic acid at position 148 [EBOV numbering]) and three unique amino acid residues (histidine in loop 1, and glutamine/valine in loop 2) in NPC1 of Mops condylurus bat cells.
## DISCUSSION
BatFiloVs are emerging viruses that remain poorly characterized despite their potential to cause disease in humans (40). Studies on their envelope GPs, which play a crucial role in the viral lifecycle, are important for understanding their pathogenicity, ecology, and host range, as well as for clarifying the potential risk as zoonotic pathogens for humans. This knowledge will guide public health interventions to prevent virus spillover and support the development of surveillance strategies and specific medical countermeasures such as vaccines and therapeutics. Consistent with the phylogenetic relationship among filoviruses based on NP, VP35, and L amino acid sequences (18), GP sequences of BatFiloVs were also divided into two phylogroups, EBOV-like and MARV-like, which include BOMV/LLOV and MLAV/DEHV, respectively. Amino acid sequence comparisons among GPs revealed that BatFiloV GPs generally shared common features with other known mammalian filoviruses, including N-terminal signal peptides, MLDs, furin cleavage sites, conserved cysteine residues, and C-terminal transmembrane/cytoplasmic regions. However, interestingly, two furin cleavage site motifs, RSKR and KKKR, were found in MLAV GP at positions 337-340 and 403-406, respectively. Considering the overall similarity among MLAV, DEHV, and MARV GPs and their multiple alignment data, KKKR at positions 403-406 was assumed to be the primary furin cleavage site of MLAV GP, although further studies are needed to deter mine the biological significance of the RSKR motif. The primary structures and glycan compositions of MLDs also varied among BatFiloVs. However, since the MLD regions were tentatively predicted in this study solely based on the predicted O-glycosylation sites, additional detailed structural and molecular analyses will be required for a more accurate comparison in the future.
Assuming that bats are the reservoir for all filoviruses, one limitation in comparing BatFiloVs with previously identified human-pathogenic filoviruses is that the latter may harbor mutations associated with host adaptation, potentially influencing pathogenic ity and transmissibility. It is currently not possible to address this issue for orthoebo laviruses, since the origins of human-pathogenic viruses (e.g., EBOV and SUDV) have never been isolated from bats. In contrast, sequences of MARVs are available from both bat and human isolates. We compared the amino acid sequences of GPs from MARVs isolated from humans during different outbreaks (Angola, Uganda, and DRC) with those from bats in Uganda and Sierra Leone and found that amino acid residues considered important for potential pathogenicity (e.g., MLD) and host range (e.g., NPC1 binding site) were generally conserved among all variants (data not shown), suggesting limited evidence for the adaptation of bat viruses to humans. Nonetheless, detailed studies of parallel isolates from both reservoirs and spillover hosts would provide critical insights into the mechanisms of host adaptation and virulence of filoviruses.
We found that VLPs consisting of BatFiloV GP, VP40, and NP were morphologically similar to those of EBOV and MARV (37), suggesting common functions of these viral proteins and cellular machinery required to form filamentous particles, whereas the reason for the slightly smaller diameters of BOMV VLPs still needs to be clarified in future studies. Most importantly, although infectious MLAV has not yet been isolated and its morphology remains unverified, our findings strongly suggest that infectious MLAV particles are also expected to be filamentous. It was also demonstrated that mouse antisera to VLPs of respective mammalian filoviruses exhibited limited cross-reactivity to heterologous GP antigens. These results suggest that each BatFiloV GP is serologically distinct from those of the other known mammalian filoviruses, and antigens such as VLPs and purified GPs can serve as target antigens for serological assays to detect specific antibodies against respective BatFiloVs. Since no epidemics or diseases caused by BatFiloVs have been reported in humans or domestic animals to date, they may not currently pose a major public health problem. However, it is possible that BatFiloV spillover occurs but goes unrecognized. Moreover, in general, the evolution of viruses through a few mutations can alter their host range and pathogenicity. Thus, it is important to know in advance whether previously established therapeutics for EBOV and MARV are also effective against BatFiloVs. Currently, the use of MAbs has gained importance in the treatment of filovirus infections, particularly following the Food and Drug Administration (FDA) approval of two MAb therapies for Ebola virus disease after a randomized controlled clinical trial in the Democratic Republic of the Congo (41), mAb114 (Ebanga, Ridgeback Biotherapeutics, Miami, FL, USA) and REGN-EB3 (Inmazeb, Regeneron Pharmaceuticals, Tarrytown, NY, USA). In our study, some previously established MAbs against EBOV or MARV exhibited the potential to neutralize BatFiloV GP-pseudotyped viruses. Notably, 6D6, a pan-orthoebolavirus MAb targeting the highly conserved internal fusion loop in GP2 (42), neutralized VSVΔG*BOMV-GP and VSVΔG*LLOV-GP due to their shared epitope in this region. This finding demonstrated the extension of its neutralizing capacity to both BOMV, a newly identified member of the Orthoebolavirus genus, and LLOV, a virus belonging to the Cuevavirus genus. VSVΔG*BOMV-GP was also neutralized by ADI-15946, another cross-reactive MAb that targets a distinct epitope in the base region of GP, crosslinking the GP1 and GP2 subunits, whereas this antibody lacks neutralizing activity against RESTV, likely due to amino acid sequence divergence within its epitope (44). The other tested anti-EBOV MAbs (mAb114, KZ52, 226/8.1, and 133/3.16) were primarily EBOV-specific and failed to neutralize VSVΔG*BOMV-GP. Further investigation for assessing the neutralizing potential of additional MAbs such as MBP047, MBP087, and MBP43, which recognize conserved epitopes on the GP base, heptad repeat, and the membrane-proximal external regions, respectively, as well as polyclonal sera against orthoebolavirus that neutralize EBOV and BOMV (62), will provide more information on conserved epitopes on GPs. It was also observed that anti-MARV GP MR191, which targets RBD, exhibited slight neutral izing activity against VSVΔG*-DEHV-GP, suggesting that the RBD epitope is partially shared between MARV and DEHV. Interestingly, MR191 has been reported to neutral ize human immunodeficiency virus-based pseudoviruses bearing MLAV GP (63). By contrast, neutralization against VSVΔG*-MLAV-GP was not observed in our study using the VSIV-based pseudotype system. This was most likely due to the differences in GP incorporation levels per virus particle, as previously suggested (64).
Filoviruses utilize multiple cellular proteins to infect a variety of cells (23). Among these, hTIM-1 and C-type lectins are known as attachment factors/receptors. In this study, we compared the efficiency of hTIM-1-and C-type lectin-mediated entry among BatFiloVs, EBOV, MARV, and RESTV, using pseudotyped viruses. In hTIM-1-expressing cells, VSVΔG*LLOV-GP infected more efficiently than other BatFiloV GP-pseudotyped viruses, confirming that the efficiency of hTIM-1-mediated viral entry varies depending on viral surface GPs, even under the same pseudotyping conditions (65). In DC-SIGN-expressing cells, VSVΔG*LLOV-GP, as well as VSVΔG*EBOV-GP, infected more efficiently than other viruses. We assume that the number of N-linked sugar chains (EBOV ≒ LLOV >BOMV) is critical for the difference among the EBOV-like phylogroups, whereas the structure of sugar chains (i.e., lower amount of high-mannose-type carbohydrate) may explain the reduced ability of the MARV-like phylogroups to infect DC-SIGN-expressing cells. On the other hand, viruses pseudotyped with BatFiloV GPs exhibited weaker infectivity enhancement than those with EBOV, MARV, and RESTV GPs in hMGL-expressing cells, which may be explained by the extent of O-glycosylation with terminal galactose as suggested previously (52). Taken together, these results suggest that the C-type lectin-mediated entry shows different specificities depending on the number and structure of target glycans. Previous studies have suggested that the ability to utilize the C-type lectins to promote cellular entry correlates with differences in pathogenicity among filoviruses (52,53,66). The relatively low efficiency of BatFiloVs to utilize C-type lectins may suggest limited potential to infect C-type lectin-expressing cells such as dendritic cells, macrophages, hepatocytes, and endothelial cells, all of which are known as preferred targets of pathogenic filoviruses (23). Accordingly, low pathogenic potential has been demonstrated in animal models for BOMV and LLOV infections (15,33).
The host range of BatFiloVs remains unknown, although several studies have shown that BatFiloVs have broad cell tropism in vitro (11-13, 17, 18, 37). However, these previous studies used only a limited number of bat cell lines and did not directly compare infectivity among BatFiloVs. Using pseudotyped viruses, we confirmed that all BatFiloVs had similar potential to infect a variety of cell lines, including human-derived cells. Interestingly, differences in susceptibility to filoviruses were observed among the bat-derived cell lines. In FBKT1 cells, VSV*G-DEHV-GP and VSV*G-MARV-GP displayed reduced infectivity compared with the other pseudotyped viruses. In contrast, the virus bearing MLAV GP, which also belongs to the MARV-like phylogroup, successfully infected this cell line, likely due to the presence of isoleucine and valine at positions 113 and 114 of MLAV GP (EBOV numbering). The inability of VSVΔG*MARV and VSVΔG*DEHV GP to infect FBKT1 cells could also be explained by unique amino acid residues (threonine, glutamic acid, and threonine at positions 425, 426, and 427, respectively) in NPC1-C loop 1 of FBKT1, as described previously (55). Notably, all Angolan free-tailed bat (Mops condylurus)-derived cell lines tested (MoKi3 C1, MoKi3-P, and MoLu6 Prim) exhibited higher susceptibility to VSVΔG*BOMV-GP than to other viruses. This may be explained by unique amino acid residues found in BOMV GP (glutamic acid in a loop 2-interacting region) and NPC1-C of this bat species (glutamine and valine in loop 2), as polymor phisms of NPC1-C loops are a key determinant of filovirus cell tropism (54,55,(58)(59)(60).
In the present study, we focused solely on the biological characterization of BatFiloV GPs; however, the data obtained are insufficient to draw conclusions about the pathogenic potential of these novel filoviruses in humans. Although one option might be to use a surrogate animal model, such as hamsters infected with recombinant VSIV carrying filovirus GP genes (67), it has not been proven that viral pathogenicity in the surrogate model parallels that of authentic viruses. Thus, it is essential to use nonhuman primates and either natural isolates or infectious filoviruses generated by a reverse genetics approach, particularly for MLAV, for which infectious forms are not yet available. The development of reverse genetics systems for this virus could be an effective approach to circumvent this difficulty. Another limitation of our experiments is the lack of mutagenesis studies on GP and NPC1 to confirm the molecular basis of the increased susceptibility of cells derived from the Angolan free-tailed bat to BOMV. Since our previous mutagenesis studies on EBOV, MARV, and LLOV GPs have clarified the molecular mechanisms underlying their host range restrictions in certain bat cell lines (54,55), we believe that a similar approach is applicable to BOMV GP and NPC1 of this bat species.
The present study provides valuable insights by comparing the biological properties of BatFiloV GPs with those of human-pathogenic filoviruses. Our findings indicate that BatFiloVs share some key characteristics with EBOV and MARV, while suggesting that they may not be as pathogenic as EBOV and MARV. To date, the pathogenicities and transmission routes of these novel filoviruses remain unknown. However, given their affinity for human cell lines, it is crucial to assess the potential risk of human exposure to these viruses. As no human infections have been reported so far, this assessment is especially important in regions where people live in close contact with bats and other wildlife.
## MATERIALS AND METHODS
## Cells
Expi293F (Gibco, Waltham, MA, USA) cells were grown in suspension using Expi293 Expression Medium (Thermo Fisher Scientific, Waltham, MA, USA) at 37°C in 8% CO 2 with rotation at 125 revolutions per minute (rpm). Human embryonic kidney 293 (HEK293), HEK293T, and HEK293T expressing hTIM-1 (HEK293T-hTIM-1) (65), African green monkey kidney Vero E6, human hepatocellular carcinoma (Huh-7), human lung adenocarcinoma epithelial (A549), baby hamster kidney (BHK), mouse embryo fibroblast (NIH3T3), and Madin-Darby bovine kidney (MDBK) cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich, St Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, St Louis, MO, USA), 100 units/mL penicillin, and 0.1 mg/mL streptomycin. Madin-Darby canine kidney (MDCK) cells were grown in DMEM with 10% calf serum (Gibco, Waltham, MA, USA), 100 units/mL penicillin, and 0.1 mg/mL streptomycin. Swine testis (ST) cells were grown in Eagle's minimum essential medium (Sigma-Aldrich, St Louis, MO, USA) supplemented with 10% FBS. Bat cells (BKT1, DemKT1, FBKT1, SuBK12, YubFKT1, ZFBK13-76E, ZFBK11-97, and ZFBK15-137RA), human chronic myelogenous leukemia K562 cells, and K562 clones expressing DC-SIGN (K562-DC-SIGN) or hMGL (K562-hMGL) (52,53) were grown in Rosewell Park Memorial Institute (RPMI) 1640 medium (Gibco, Waltham, MA, USA) supplemented with 10% FBS, 100 units/mL penicillin, and 0.1 mg/mL streptomycin. Other bat cells derived from Mops condylurus (MoKi3 C1, MoKi3-P, and MoLu6 Prim) were grown in DMEM/Ham's F12 (Sigma-Aldrich, St Louis, MO, USA) supplemented with 15% FBS, L-glutamine (Gibco, Waltham, MA, USA), and Anti-Anti (Gibco, Waltham, MA, USA). All these cells were grown at 37°C in a 5% CO 2 incubator.
## Construction of plasmids expressing GP, NP, and VP40
The nucleotide sequences of BOMV (ON871047), MLAV (NC_055510.1), and DEHV (OP924273.1) were retrieved from GenBank. The coding regions of GP, NP, and VP40 were synthesized in the pUCFa vector (FASMAC Co., Ltd., Kanagawa, Japan), amplified by KOD One PCR master mix (TOYOBO, Osaka, Japan) using specific primers, and cloned into a mammalian expression vector, pCAGGS, using the In-Fusion HD cloning Kit (TAKARA Bio, CA, USA). Plasmids expressing a soluble form of BOMV, MLAV, and DEHV GPs, with the C-terminal His-tagged truncated transmembrane region, were also amplified by KOD One PCR using specific primers containing hexa-His tag sequences and cloned as described above. A designed trimerization motif sequence (GCN3 motif ) (68) was inserted between the His tag and residue 640 of DEHV GP to promote trimeric folding. The sequences of all genes were confirmed by Sanger sequencing. The expression plasmids for MARV, EBOV, SUDV, RESTV, TAFV, BDBV, and LLOV were constructed as previously described (37,69,70).
## Production and purification of VLPs
Plasmids (pCAGGS) encoding GP, NP, and VP40 of each filovirus (BatFiloVs, TAFV, and RESTV) were used for the transfection of HEK293T cells using TransIT LT-1 (Mirus Bio LLC, WI, USA) according to the manufacturer's instructions. Forty-eight hours after transfec tion, the supernatant was collected, and VLPs were purified by ultracentrifugation at 28,000 rpm (SW32Ti rotor, Beckman Coulter) at 4°C for 2 h with a 25% sucrose cushion. VLP pellets were resuspended in phosphate-buffered saline (PBS).
## Pseudotyped viruses
VSIV containing the green fluorescent protein (GFP) gene instead of the receptor-binding VSV G protein gene (VSVΔG*VSIV-G) (22), complemented with GPs of EBOV, SUDV, BDBV, TAFV, RESTV, MARV, LLOV, BOMV, MLAV, and DEHV (VSVΔG*EBOV-GP, VSVΔG*SUDV-GP, VSVΔG*BDBV-GP, VSVΔG*TAFV-GP, VSVΔG*RESTV-GP, VSVΔG*MARV-GP, VSVΔG*LLOV-GP, VSVΔG*BOMV-GP, VSVΔG*MLAV-GP, and VSVΔG*DEHV-GP, respectively), were generated. Infectious units (IUs) of these pseudotyped viruses were determined as described previously (22).
## Virus titration
The background residual infectivity of parental VSVΔG*VSIV-G was abolished before determining the infectivities of pseudotyped viruses by pretreatment with the anti-VSV G MAb VSV-G(N)1-9 (71). K562, K562-DC-SIGN, K562-hMGL, HEK293T, and HEK293T-hTIM-1 grown on 96-well plates were infected with viruses (50-150 IUs/well determined in K562 and HEK293T, respectively). At 20 h postinoculation, GFP-positive cells were counted using an IN-Cell Analyzer 2500 HS (GE Healthcare, Waukesha, WI, USA). The relative percentages of infectivity in hTIM-1-, DC-SIGN-, and hMGL-expressing cells were calculated by setting IUs in control cells (control K562 or HEK293T) to 100%. To determine the infectivities in adherent cell lines of different animal origins, cell mono layers grown on 96-well plates were infected with a serial dilution of VSVs pseudotyped filovirus GPs. At 20 h postinoculation, GFP-positive cells were counted using an IN-Cell Analyzer 2500 HS, and IUs were determined for each cell line. The relative infectivities were calculated by setting IUs in Vero E6 cells to 1.
## Neutralization test
Pseudotyped viruses pretreated with VSV-G(N)1-9 were diluted with DMEM containing 2% FBS to obtain 500-1,500 IUs/50 µL and mixed with an equal volume of serial dilutions of MAbs (6D6, ADI15946, mAb114, KZ52, 133/3.16, 226/8.1, MR78, MR191), followed by incubation for 30 min at room temperature. MAbs 6D6, 133/3.16, and 226/8.1 were obtained from our repository, and ADI15946 (Creative Diagnostics, Shirley, NY, USA), mAb114 (Ridgeback Biotherapeutics, Miami, FL, USA), KZ52 (Absolute Antibody Ltd., Cleveland, UK), and MR78 (ProteoGenix Inc., Morrisville, NC, USA), MR191 (ProteoGenix Inc., Morrisville, NC, USA) were purchased. The mixture (100 µL) was added to a confluent monolayer of Vero E6 cells grown in 96-well plates. Twenty hours later, GFP-positive cells were counted as described above. Percentages of infectivity were calculated by setting IUs in cells infected with each virus alone to 100%.
## Mouse antisera and enzyme-linked immunosorbent assay (ELISA)
Five-week-old BALB/c mice were immunized twice intraperitoneally with purified VLPs (BatFiloVs, TAFV, and RESTV) (100 µg per head) at 3-week intervals. Antisera were collected 14 days after the second immunization. Previously produced antisera to EBOV, SUDV, BDBV, and MARV stored at -80°C were also used (37). Filovirus GP-based ELISA was performed as described previously (69). Soluble forms of EBOV, SUDV, RESTV, TAFV, BDBV, MARV, LLOV, BOMV, MLAV, and DEHV GPs employed as antigens were purified from the supernatant of Expi293F cells transfected with His-tagged GP-expressing plasmids using the Ni-NTA purification system (Invitrogen, CA, USA). Each antiserum was serially diluted with PBS containing 0.05% Tween 20 and 1% skim milk. Bound antibodies were visualized with horseradish peroxidase-conjugated goat anti-mouse IgG (H + L) (Jackson ImmunoResearch Laboratories Inc., USA) and 3,3' ,5,5'-tetramethylbenzidine (Sigma-Aldrich, St Louis, MO, USA). The reaction was stopped by adding 1 N phosphoric acid to the mixture, and the optical density at 450 nm (OD 450 ) was measured.
## Electron microscopy
Transmission electron microscopy was performed as previously described (37). Purified VLPs (10 µL) fixed with 0.25% glutaraldehyde overnight were placed on collodion-car bon-coated copper grids (Nisshin EM Co. Ltd., Tokyo, Japan) for 2 minutes at room temperature. Then, the grids were washed three times with 10 µL of PBS droplets, negatively stained with 10 µL of 2% phosphotungstic acid hydrate (pH 5.8) (Thermo Fisher Scientific, USA) for 45 seconds, and dried using filter paper. For immunogold staining, we used MAb LGP14-2 for LLOV, MAb ZGP42/3.7 for BOMV and EBOV (37,69), mouse antisera against MLAV and DEHV VLPs produced as described above, and a 10 nm gold-conjugated goat anti-mouse IgG (H + L) polyclonal antibody (Cytodiagnostics Inc., Burlington, Canada). The purified VLPs (10 µL) were placed onto collodion-carbon-coated copper grids for 10 minutes, followed by blocking with PBS containing 3% BSA for 10 minutes. The grids were then treated with 10 µL of the MAbs or mouse antisera for 30 minutes, washed five times with 10 µL PBS droplets, and incubated with the gold-conju gated goat anti-mouse IgG antibody for 30 minutes. Next, the grids were washed three times with 10 µL PBS droplets and once with 10 µL ultrapure water before being stained with 10 µL 2% phosphotungstic acid hydrate (pH 5.8) for 45 seconds and dried using filter paper. Samples were examined with a transmission electron microscope (HT-7800, Hitachi High-Tech Corporation, Tokyo, Japan) at 80 kV.
## Sequence analyses
The evolutionary history was inferred using the neighbor-joining method (72). Opti mal trees based on GP, NP, and L of 10 mammalian filoviruses were constructed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (10,000 replicates) was shown next to the branches (73). The tree was drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method (74) and displayed as the number of amino acid substitutions per site. This analysis involved 10 whole amino acid sequences of GP, NP, and L. All ambiguous positions were removed for each sequence pair (pairwise deletion option). Evolutionary analyses were conducted in MEGA11 (75). The amino acid sequences of BatFiloVs used in phylogenetic analysis were retrieved from GenBank as described above. The sequences of EBOV (Mayinga-76), SUDV (Boniface-76), BDBV (Bundibugyo/2007), TAFV (Cote d'Ivoire-95), RESTV (Reston-89), MARV (Angola/2005), and LLOV (Asturias/2003) (GenBank under accession numbers AF086833.2, FJ968794.1, FJ217161.1, U28006, U23152.1, DQ447660.1, and JF828358, respectively) were also used. The potential glycosylation sites were predicted using NetNGlyc-1.0 (https://serv ices.healthtech.dtu.dk/services/NetNGlyc-1.0/) and NetOGlyc-4.0 (https://services.health tech.dtu.dk/services/NetOGlyc-4.0/) (DTU Health Tech, Lyngby, Denmark), and amino acid residues showing a score above the threshold (0.5) were selected.
## Molecular modeling
Three-dimensional models of the NPC1-C and EBOV GP complex were prepared based on a previous study (61) (Protein Data Bank [PDB] code 5F1B). The three-dimensional structures shown in the figure of this study were prepared using PyMOL (Schrödinger LLC).
## References
1. Biedenkopf, Bukreyev, Chandran et al. (2023) "Renaming of genera Ebolavirus and Marburgvirus to Orthoebolavirus and Orthomarburgvi rus, respectively, and introduction of binomial species names within family Filoviridae" *Arch Virol*
2. Biedenkopf, Bukreyev, Chandran et al. (1955) "ICTV virus taxonomy profile: Filoviridae 2024" *J Gen Virol*
3. Olival, Hayman (2014) "Filoviruses in bats: current knowledge and future directions" *Viruses*
4. Languon, Quaye (2019) "Filovirus disease outbreaks: a chronological overview" *Virology (Auckl)*
5. Towner, Amman, Sealy et al. (2009) "Isolation of genetically diverse Marburg viruses from Egyptian fruit bats" *PLoS Pathog*
6. Amman, Bird, Bakarr et al. (2020) "Isolation of Angola-like Marburg virus from Egyptian rousette bats from West Africa" *Nat Commun*
7. Kajihara, Bm, Changula et al. (2019) "Marburgvirus in Egyptian fruit bats"
8. Negredo, Palacios, Vázquez-Morón et al. (2011) "Discovery of an ebolavirus-like filovirus in Europe" *PLoS Pathog*
9. Kemenesi, Kurucz, Dallos et al. (2016) "Re-emergence of Lloviu virus in Miniopterus schreibersii bats" *Emerg Microbes Infect*
10. Changula, Kajihara, Mori-Kajihara et al. (2018) "Seroprevalence of filovirus infection of Rousettus aegyptiacus bats in zambia" *J Infect Dis*
11. (2025) *Full-Length Text Journal of Virology*
12. Kemenesi, Tóth, Mayora-Neto et al. (2022) "Isolation of infectious Lloviu virus from Schreiber's bats in Hungary" *Nat Commun*
13. Tóth, Hume, Suder et al. (2023) "Isolation and genome characterization of Lloviu virus from Italian Schreibers's bats" *Sci Rep*
14. Goldstein, Anthony, Gbakima et al. (2018) "The discovery of Bombali virus adds further support for bats as hosts of ebolaviruses" *Nat Microbiol*
15. Forbes, Webala, Jääskeläinen et al. (2019) "Bombali virus in Mops condylurus Bat" *Kenya. Emerg Infect Dis*
16. Bodmer, Breithaupt, Heung et al. (2023) "In vivo characterization of the novel ebolavirus Bombali virus suggests a low pathogenic potential for humans" *Emerg Microbes Infect*
17. Düx, Lwitiho, Ayouba et al. (2024) "Detection of Bombali virus in a Mops condylurus Bat in Kyela" *Viruses*
18. Yang, Tan, Anderson et al. (2019) "Characterization of a filovirus (Měnglà virus) from Rousettus bats in China" *Nat Microbiol*
19. He, Hu, Yan et al. (2024) "Isolation, characterization, and circulation sphere of a filovirus in fruit bats" *Proc Natl Acad Sci*
20. Sanchez, Trappier, Mahy et al. (1996) "The virion glycoproteins of Ebola viruses are encoded in two reading frames and are expressed through transcriptional editing" *Proc Natl Acad Sci*
21. Volchkov, Becker, Volchkova et al. (1995) "GP mRNA of Ebola virus is edited by the Ebola virus polymerase and by T7 and vaccinia virus polymerases" *Virology (Auckl)*
22. Mehedi, Falzarano, Seebach et al. (2011) "A new Ebola virus nonstructural glycoprotein expressed through RNA editing" *J Virol*
23. Takada, Robison, Goto et al. (1997) "A system for functional analysis of Ebola virus glycoprotein" *Proc Natl Acad Sci*
24. Takada (2012) "Filovirus tropism: cellular molecules for viral entry" *Front Microbiol*
25. Takada, Feldmann, Stroeher et al. (2003) "Identification of protective epitopes on ebola virus glycoprotein at the single amino acid level by using recombinant vesicular stomatitis viruses" *J Virol*
26. Pappalardo, Juliá, Howard et al. (2016) "Conserved differences in protein sequence determine the human pathogenicity of Ebolaviruses" *Sci Rep*
27. Nidom, Nakayama, Nidom et al. (1998) "Characterization of Ebola virus entry by using pseudotyped viruses: identification of receptor-deficient cell lines" *PLoS One*
28. Feldmann, Nichol, Klenk et al. (1994) "Characteri zation of filoviruses based on differences in structure and antigenicity of the virion glycoprotein" *Virology (Auckl)*
29. Carette, Raaben, Wong et al. (2011) "Ebola virus entry requires the cholesterol transporter Niemann-Pick C1" *Nature*
30. Côté, Misasi, Ren et al. (2011) "Small molecule inhibitors reveal Niemann-Pick C1 is essential for Ebola virus infection" *Nature*
31. Misasi, Chandran, Yang et al. (2012) "Filoviruses require endosomal cysteine proteases for entry but exhibit distinct protease preferences" *J Virol*
32. Malashkevich, Schneider, Mcnally et al. (1999) "Core structure of the envelope glycoprotein GP2 from Ebola virus at 1.9-A resolution" *Proc Natl Acad Sci*
33. Fletcher, Feldmann, Takada et al. (2023) "Pathogenicity of Lloviu and Bombali viruses in type i interferon receptor knockout mice" *J Infect Dis*
34. Fletcher, Donnell, Rhoderick et al. (2024) "Lack of Lloviu virus disease development in ferret model" *Emerg Infect Dis*
35. Jeffers, Sanders, Sanchez (2002) "Covalent modifications of the Ebola virus glycoprotein" *J Virol*
36. Ito, Watanabe, Takada et al. (2001) "Ebola virus glycoprotein: proteolytic processing, acylation, cell tropism, and detection of neutralizing antibodies" *J Virol*
37. Maruyama, Miyamoto, Kajihara et al. (2014) "Characterization of the envelope glycopro tein of a novel filovirus, lloviu virus" *J Virol*
38. Peng, Rayaprolu, Parvate et al. (2022) "Glycan shield of the ebolavirus envelope glycoprotein GP" *Commun Biol*
39. Noda, Sagara, Suzuki et al. (2002) "Ebola virus VP40 drives the formation of virus-like filamentous particles along with GP" *J Virol*
40. Munyeku-Bazitama, Edidi-Atani, Takada (2024) "Non-Ebola filoviruses: potential threats to global health security" *Viruses*
41. Mulangu, Dodd, Davey et al. (2019) "A randomized, controlled trial of Ebola virus disease therapeutics" *N Engl J Med*
42. Furuyama, Marzi, Nanbo et al. (2016) "Discovery of an antibody for pan-Ebolavirus therapy" *Sci Rep*
43. Wec, Herbert, Murin et al. (2017) "Antibodies from a human survivor define sites of vulnerability for broad protection against Ebolaviruses" *Cell*
44. West, Wec, Moyer et al. (2019) "Structural basis of broad ebolavirus neutralization by a human survivor antibody" *Nat Struct Mol Biol*
45. (2025) *Full-Length Text Journal of Virology*
46. Corti, Misasi, Mulangu et al. (2016) "Protective monother apy against lethal Ebola virus infection by a potently neutralizing antibody" *Science*
47. Maruyama, Rodriguez, Jahrling et al. (1999) "Ebola virus can be effectively neutralized by antibody produced in natural human infection" *J Virol*
48. Flyak, Ilinykh, Murin et al. (2015) "Mechanism of human antibodymediated neutralization of Marburg virus" *Cell*
49. Hashiguchi, Fusco, Bornholdt et al. (2015) "Structural basis for Marburg virus neutralization by a cross-reactive human antibody" *Cell*
50. King, Fusco, Flyak et al. (2018) "The Marburgvirus-neutralizing human monoclonal antibody MR191 targets a conserved site to block virus receptor binding" *Cell Host Microbe*
51. Kondratowicz, Lennemann, Sinn et al. (2011) "T-cell immunoglobulin and mucin domain 1 (TIM-1) is a receptor for Zaire Ebolavirus and Lake Victoria Marburgvirus" *Proc Natl Acad Sci*
52. Alvarez, Lasala, Carrillo et al. (2002) "Ctype lectins DC-SIGN and L-SIGN mediate cellular entry by Ebola virus in cis and in trans" *J Virol*
53. Takada, Fujioka, Tsuiji et al. (2004) "Human macrophage C-type lectin specific for galactose and N-acetylgalactosamine promotes filovirus entry" *J Virol*
54. Matsuno, Kishida, Usami et al. (2010) "Different potential of C-type lectin-mediated entry between Marburg virus strains" *J Virol*
55. Takadate, Manzoor, Saito et al. (2020) "Receptormediated host cell preference of a bat-derived filovirus" *Lloviu Virus. Microorganisms*
56. Takadate, Kondoh, Igarashi et al. (2020) "Niemann-pick C1 heterogene ity of bat cells controls filovirus tropism" *Cell Rep*
57. Ng, Ndungo, Kaczmarek et al. (2015) "Filovirus receptor NPC1 contributes to species-specific patterns of ebolavirus susceptibility in bats"
58. Hoffmann, Hernández, Berger et al. (2016) "The glycoproteins of all filovirus species use the same host factors for entry into bat and human cells but entry efficiency is species dependent" *PLoS One*
59. Martinez, Ndungo, Tantral et al. (2013) "A mutation in the Ebola virus envelope glycoprotein restricts viral entry in a host species-and cell-type-specific manner" *J Virol*
60. Ndungo, Herbert, Raaben et al. (2016) "A single residue in Ebola virus receptor NPC1 influences cellular host range in reptiles" *mSphere*
61. Lasso, Grodus, Valencia et al. (2025) "Decoding the blueprint of receptor binding by filoviruses through large-scale binding assays and machine learning" *Cell Host Microbe*
62. Wang, Shi, Song et al. (2016) "Ebola viral glycoprotein bound to its endosomal receptor Niemann-pick C1" *Cell*
63. Mcmullan, Flint, Jenks et al. (2022) "Characterization of Bombali virus, a new bat filovirus" *SSRN J. htt*
64. Cooper, Achi, Rong (2022) "Comparative analyses of small molecule and antibody inhibition on glycoprotein-mediated entry of Měnglà virus with other filoviruses" *J Med Virol*
65. Steeds, Hall, Slack et al. (2020) "Pseudotyping of VSV with Ebola virus glycoprotein is superior to HIV-1 for the assessment of neutralising antibodies" *Sci Rep*
66. Hattori, Saito, Miyamoto et al. (2022) "Single nucleotide variants of the human TIM-1 IgV domain with reduced ability to promote viral entry into cells" *Viruses*
67. Marzi, Akhavan, Simmons et al. (2006) "The signal peptide of the ebolavirus glycoprotein influences interaction with the cellular lectins DC-SIGN and DC-SIGNR" *J Virol*
68. Saito, Maruyama, Nagata et al. (2020) "A surrogate animal model for screening of Ebola and marburg glycoprotein-targeting drugs using pseudotyped vesicular stomatitis viruses" *Viruses*
69. Boré, Timothy, Tipton et al. (2024) "Serological evidence of zoonotic filovirus exposure among bushmeat hunters in Guinea" *Nat Commun*
70. Nakayama, Yokoyama, Miyamoto et al. (2010) "Enzyme-linked immunosorbent assay for detection of filovirus species-specific antibodies" *Clin Vaccine Immunol*
71. Changula, Yoshida, Noyori et al. (2013) "Mapping of conserved and species-specific antibody epitopes on the Ebola virus nucleoprotein" *Virus Res*
72. Nakayama, Tomabechi, Matsuno et al. (2011) "Antibody-dependent enhancement of Marburg virus infection" *J Infect Dis*
73. Saitou, Nei (1987) "The neighbor-joining method: a new method for reconstructing phylogenetic trees" *Mol Biol Evol*
74. Felsenstein (1985) "Confidence limits on phylogenies: an approach using the bootstrap" *Evolution*
75. Zuckerkandl, Pauling (1965) "Evolutionary divergence and conver gence in proteins"
76. Tamura, Stecher, Kumar (2021) "MEGA11: molecular evolutionary genetics analysis version 11" *Mol Biol Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724203&blobtype=pdf | # Coexistence of virome-encoded health-associated genes and pathogenic genes in global habitats
Min Qian, Dong Zhu, Ke-Yu Yao, Shu-Yue Liu, Meng-Ke Li, Mao Ye, Yong-Guan Zhu
## Abstract
Viral remnants constitute approximately 8% of the human genome, reflecting extensive historical gene exchange between viruses and their hosts. Some viral genomes harbor genes acquired through horizontal gene transfer that are associated with potential benefits to human health, alongside genes associated with pathogenicity. However, their global distribution, functional characteristics, and coexistence patterns remain poorly understood. Here, using the Integrated Microbial Genomes and Virome (IMG/VR v4) database, we identified 4,556 viruses carrying gene segments associated with human health across eight habitat types spanning 13 regions and 76 countries worldwide. Among viruses with identifiable hosts, those distributed in humans ( 478) accounted for the highest proportion. The viral genes associated with human health included BCO1 (beta-carotene oxygenase 1), bioB (biotin synthase), COQ2 (4-hydroxy benzoate polyprenyltransferase), GPX1 (glutathione peroxidase 1), GSTs (glutathione transferases), GSTT1 (glutathione S-transferase theta 1), GULO (L-gulonolactone oxidase), and menA (1,4-dihydroxy-2-naphthoate polyprenyltransferase). These genes not only associate with human health but also function as auxiliary metabolic genes in viral genomes. Notably, four pathogenic genes were found in viral sequences carrying health-associated genes, with potential for transcription and expression, indicating functional interactions. Experimental transduction of the viral bioB gene into Escheri chia coli altered the expression of host pathogenic genes GCH1 (GTP cyclohydrolase IA) and UGDH (UDP-glucose 6-dehydrogenase), supporting potential cross-regulatory interactions. Overall, this study incorporates health-associated genes into viral genomics, highlighting their coexistence with pathogenic genes, and provides new insights into virus-host coevolution and potential biotechnological applications. IMPORTANCE Viruses are the most abundant biological entities on Earth and key drivers of microbial evolution through horizontal gene transfer. While often studied for their pathogenic effects, viruses can also carry genes that influence host metabolism and health. Genes associated with human health have been identified in viral genomes, yet their global distribution, functions, and coexistence with pathogenic genes remain largely unexplored. This study integrates datasets of health-associated genes into viral genomic analyses, revealing for the first time the coexistence of viral health-associated genes with those linked to pathogenicity. This dual genetic potential is observed across diverse habitats, highlighting viruses as multifaceted reservoirs of both beneficial and harmful genes. The study findings advance understanding of viral functional diversity and open new avenues for exploring viral roles in microbial ecology, biotechnology, and human health.
V iral remnants make up approximately 8% of the human genome, including sequences that encode genes potentially beneficial to human health as well as those associated with pathogenicity, underscoring the complex and long-standing relationship between viruses and their human hosts (1)(2)(3)(4)(5). Viruses have interacted with humans for over 300,000 years, influencing host adaptation, evolution, and disease propagation (2,6,7). Throughout history, virus-induced diseases have caused major public health crises, ranging from the 20th-century Spanish influenza (8) to HIV/AIDS in the 21st century (9) and, most recently, the COVID-19 epidemic (10). These events underscore the pathogenicity of viruses, primarily driven by virulence genes that disrupt host cell structures and functions, leading to various diseases (11)(12)(13)(14).
Despite their pathogenic potential, viruses have also evolved genes that benefit human hosts through millennia of host-virus interactions (15,16). Some viruses carry genes associated with potential health benefits, acquired through horizontal gene transfer (HGT) from host cells, forming auxiliary metabolic genes (AMGs) (16)(17)(18)(19)(20)(21). These AMGs enhance viral adaptation to diverse environments and can positively influence host metabolism during infection (21)(22)(23)(24). For example, the gene psbA (photosystem II P680 reaction center D1 protein) serves as a biomarker for classifying oceanic and freshwater environments and has potential applications in source tracking (24,25). Similarly, genes arsC (arsenate reductase) and arsM (arsenite methyltransferase) in soil lysogenic viruses encode arsenic metabolism, influencing bacterial responses to environmental arsenic exposure (26). In marine environments, viral AMGs such as glnK (nitrogen regulatory protein P-II 2), norB (nitric oxide reductase subunit B), nirK (nitrite reductase), and nirA (ferredoxin-nitrite reductase) contribute to the nitrogen cycle (23). In nutrient-scarce environments like the South Pole, the viral gene cl enhances host survival by suppressing the expression of pckA (phosphoenolpyruvate carboxykinase) in the host, conferring selective advantages (27,28). These findings highlight the role of AMGs in supporting host metabolism and environmental adaptation.
While the ecological roles of viral AMGs are increasingly recognized, significant knowledge gaps remain regarding virus-encoded proteins that are associated with human health or pathogenicity. The occurrence, functions, and geographic distribution of these health-associated genes in viruses remain incompletely characterized (29,30). Health-associated genes are viral genes that are directly or indirectly linked to processes beneficial to human health, such as anticancer activity and vitamin synthesis, in contrast to pathogenic genes, which contribute to disease. Moreover, although viral genes that contribute to both human health and pathogenicity have been identified (13,14,27,28), the mechanisms that allow these seemingly opposing functions to coexist remain poorly understood. Viruses may balance the expression of these genes through sophisticated regulatory networks, including transcriptional control and epigenetic modifications, to optimize survival and transmission across diverse hosts and environmental conditions (31)(32)(33).
To address these gaps, this study analyzed four representative types of health-associ ated genes, including anticancer, vitamin synthesis, antioxidant, and longevity genes, identified through a literature search (Table S1). Using the global public database Integrated Microbial Genomes and Virome (IMG/VR v4), the present study investigated the geological distribution and habitat characteristics of viruses carrying these gene sequences worldwide (Tables S2 andS3). Additionally, this research examined host-virus relationships and phylogenetic patterns (Tables S3 andS4) and explored the mechanisms balancing human health and pathogenic functions in viruses across different environ ments. By introducing health-associated genes into Escherichia coli to simulate phagemediated transduction of the AMG bioB (biotin synthase), this study demonstrated that these genes significantly influence the differential expression of pathogenic genes GCH1 (GTP cyclohydrolase IA) and UGDH (UDP-glucose 6-dehydrogenase) in the host (Table S6). These findings confirm that viruses can compensate for health-associated gene functions in hosts during HGT (Fig. 1). This study provides novel insights into the distribution, functions, and potential applications of viral genes involved in human health, advancing the understanding of host-virus dynamics in applied and environmental microbiology.
## RESULTS
## Extensive global distribution of viruses carrying health-associated genes
Fifty-five representative health-associated genes were identified, encompassing functions related to disease prevention and physiological resilience, including anticancer activity, vitamin biosynthesis, antioxidant production, and longevity support (34)(35)(36)(37)(38)(39)(40)(41). These genes were screened across 7,159 uncultivated viral genomes (UVIGs), of which 6,583 contained geolocation metadata and 6,829 were linked to specific habitats. Viruses harboring health-associated genes were detected in 13 global regions spanning 76 countries. The highest numbers were observed in the USA (3,866), followed by Canada (671) and the UK (258) (Fig. 2A; Table S3). These viruses were classified into eight major viral taxa: Alsuviricetes, Caudoviricetes, Faserviricetes, Megaviricetes, Naldaviricetes, Papovaviricetes, Revtraviricetes, and Tectiliviricetes. Tailed bacteriophages (Caudoviricetes) encoded the largest number of health-associated genes (5,531), followed by giant viruses (Megaviricetes, 1,262) (Fig. 2B; Table S3). Among the 40 distinct health-associated genes identified, HSP70 (heat shock 70 kDa protein) was the most frequently encoded (472 occurrences), followed by GPX1 (glutathione peroxidase 1; 468) and SVCT2 (solute carrier family 23 member 2; 448) (Fig. 2C; Table S3). The viruses encoding these genes were associated with diverse environments, grouped into eight categories: humans, animals, freshwater, marine, other aquatic, soil, plants, and others. The highest numbers were found in freshwater (2,758), followed by ocean water (1,511) and human-associated samples (769).
Functionally, these genes were categorized into seven groups: anticancer, coenzyme Q10, glutathione synthesis, polyphenol, longevity, fat-soluble vitamin, and water-solu ble vitamin. Among these, anticancer-related genes were the most prevalent (1,540 occurrences), followed by genes associated with longevity (1,373) and water-soluble vitamin synthesis (1,006) (Fig. 2D; Table S3). Collectively, these findings reveal a broad global distribution of viruses encoding health-associated genes and provide theoretical support for understanding viral survival strategies and their potential roles in mediating beneficial host interactions.
## High host prediction rate in viruses carrying health-associated genes in human habitats
Host prediction analysis was performed on 4,556 viral sequences, of which 904 contained assignable host information, including 904 kingdom-and phylum-level annotations and 896 class-level assignments (Fig. 3; Table S2). Across these 904 viral sequences, 26 host types were predicted, with 22 classified at the class level and four unclassified beyond the phylum level. The dominant predicted hosts included c_Gammaproteobacteria (389 sequences), c_Clostridia (238 sequences), and c_Bacilli (128 sequences). Viral sequences were globally distributed across 11 regions: Asia, Africa, Europe, North America, South America, Oceania, Antarctica, Pacific Ocean, Indian Ocean, Arctic Ocean, and Unsure. The majority of sequences originated from North America (401 sequences), likely due to higher sampling density in that region.
Habitat analysis revealed that most host-associated viral sequences were linked to humans (478 sequences), followed by plants (64 sequences). Among 514 phage sequences, 280 were classified as virulent and 233 as temperate. Viruses carrying human health-associated AMGs were detected in both groups, with temperate phages harboring 22.3% more such genes per viral genome on average than virulent phages. Overall, phages carrying health-associated AMGs accounted for 4% of all phage sequences, whereas non-phage viruses carrying similar genes represented only 1% of total non-phage sequences.
## Virus-encoded health-associated genes exhibit the potential for transcription and translation
Through DRAM-V and VIBRANT prediction tools (42,43), BCO1 (beta-carotene oxy genase 1), bioB, COQ2 (4-hydroxybenzoate polyprenyltransferase), GPX1, GSTs (gluta thione transferases), GSTT1 (glutathione S-transferase theta 1), GULO (L-gulonolactone oxidase), and menA (1,4-dihydroxy-2-naphthoate polyprenyltransferase) were identified as virus-encoded health-associated AMGs (Table S5). In total, 133 viral sequences with high confidence carried both health-associated AMGs and virus marker genes. Phyloge netic and functional analyses revealed evolutionary relationships among GPX1, GSTT1, and bioB across different viral sequences (Fig. 4A; Table S4). Within the same geo graphic regions, such as North America, viruses carrying the GPX1 gene from marine and freshwater habitats exhibited pairing distances of 0.3-0.4, indicating relatively close phylogenetic relationships. In contrast, viruses carrying bioB from human-asso ciated samples in Europe displayed higher pairing distances (>0.9), reflecting stron ger consanguinity ties (44)(45)(46). Notably, viruses carrying GPX1 in humans from North America and Africa exhibited pairing distances near 0, indicative of exceptionally close evolutionary relationships (45)(46)(47). Promoter and terminator regions located immedi ately upstream and downstream of health-associated AMGs (BCO1, bioB, COQ2, GPX1, GSTs, GSTT1, GULO, and menA) were predicted, suggesting potential for transcription and translation into functional proteins (23,48)(Fig. 4B). Structural models of these eight genes, generated using SWISS-MODEL, exhibited Global Model Quality Estimation (GMQE) indices of 0.56, 0.95, 0.95, 0.80, 0.61, 0.97, 0.94, and 0.92, indicating the capacity of viral sequences to encode proteins with complete structures and functional roles (44, 49, 50) (GMQE >0.5) (Fig. 4C).
## Pathogenic genes in viral genomes carrying health-associated genes
Four types of pathogenic genes (GCH1, NAMPT [nicotinamide phosphoribosyltransfer ase], UGDH, and P4HA [prolyl 4-hydroxylase]) were recognized in viral sequences carrying health-associated genes. These pathogenic genes were found to be transcriptionally active (Fig. 5A). Specifically, GCH1 mutations can lead to tetrahydrobiopterin deficiency, resulting in phenylketonuria (37,48,51); elevated NAMPT expression may enhance cancer cell growth and survival, contributing to tumor development (52,53); UGDH dysfunction can disrupt extracellular matrix composition, affecting tissue structure and function, with links to cancer and fibrosis (54); and increased P4HA expression may facilitate tumor progression and metastasis (55). Despite evidence of transcriptional activity, the GMQE values of the four types of pathogenic genes were 0.63, 0.53, 0.59, and 0.67 (Fig. 5B), which were generally lower than those of health-associated genes, suggesting a reduced likelihood of translation compared with health-associated genes (44,49,50). Notably, pathogenic and health-associated genes frequently co-occurred within viral genomes, typically separated by distances of ~1-2 kb (Fig. 5C), indicating a close genomic association.
## Balancing effects of environmentally mediated regulation of exogenous health-associated genes on the expression of pathogenic genes
The health-associated bioB gene was introduced via plasmid transduction into E. coli B21, which was previously confirmed through sequencing to naturally harbor the prophage-derived pathogenic genes GCH1 and UGDH (Fig. 6A; Table S6). To assess protein expression under physiologically relevant conditions, recombinant cultures were subjected to various environmental regimes, including temperatures of 20°C, simulating ambient conditions, and 37°C, mimicking the human intestinal environment (56). The pH was varied across 5.5, reflective of urinary conditions (57), and 8.5, corresponding to a pathological colonic setting (58). Successful heterologous expression of bioB was confirmed by SDS-PAGE and Western blot analyses, with the most prominent bands observed in the supernatant and inclusion body fractions under 20°C/pH 5.5, 20°C/pH 8.5, and 37°C/pH 8.5 conditions (Fig. 6B; Fig. S1).
The transcriptional impact of bioB expression on the pathogenic genes was subse quently evaluated using real-time quantitative PCR with gapA as an internal reference. A control group consisting of the same E. coli strain transformed with an empty vector was included in all experimental conditions for normalization. Under acidic and ambient temperature conditions of 20°C and pH 5.5, the expression levels of GCH1 and UGDH were found to be up-regulated to 2.3-fold and 6.7-fold, respectively, relative to the empty vector control. When the pH was elevated to 8.5 while maintaining the temperature at 20°C, a moderate up-regulation to 1.5-fold for GCH1 and 2.6-fold for UGDH was still observed. In contrast, under the human gut-mimicking condition of 37°C and pH 8.5, a notable down-regulation was recorded, with expression levels reduced to 0.8-fold for GCH1 and 0.4-fold for UGDH (Fig. 6C; Table S6). These findings collectively indicate that the functional expression of the health-associated bioB gene is associated with significant and environmentally modulated changes in the expression of pathogenic genes, with a particularly suppressive effect observed under conditions simulating the human intestinal environment.
## DISCUSSION
## Potential impact of database completeness and sequence quality on ecological inferences
Large-scale metagenomic analyses often face a trade-off between maximizing data completeness and maintaining sequence quality. The IMG/VR database, by including fragmented and low-quality (LQ) viral genomes, provides a more comprehensive census of viral sequence space, which is valuable for assessing macroecological trends. In the analysis of habitat associations and geographic distribution, these LQ genomes were retained to minimize sampling bias against underrepresented viral groups (Table S2). It is established that LQ genomes may contain host-derived contamination; however, for analyzing relative distribution patterns across environments, this introduces a conservative, non-systematic error. The consistent and statistically significant patterns observed across diverse, independently sampled habitats support that the inferred habitat and geographic distributions are robust and unlikely to be generated solely by host contamination. Nevertheless, it is emphasized that all subsequent in-depth analyses of gene function and evolution were rigorously restricted to high-quality viral genomes to ensure the validity of the predictions regarding transcription and translation.
## Global habitat distribution and phylogenetic development of health-associ ated genes carried in viruses
Globally, viruses carrying health-associated gene segments were distributed across diverse habitats, with freshwater environments showing the highest account (38.5%)(Fig. 2; Table S3). This predominance likely reflects the rich organic matter, high microbial diversity, and dense host populations in freshwater systems, which provide favorable ecological niches for viral replication and persistence (59,60). Geographically, the majority of these viruses were detected in the Americas (54.0%), indicating poten tial biases linked to sampling density, as well as possible biogeographic preferences associated with habitat type and host range diversity. Interestingly, many of these viral genomes contained genes associated with anticancer functions, suggesting a possible evolutionary strategy favoring host survival and adaptation within dynamic ecosystems (61,62). Alternatively, anthropogenic pressures, such as pollution and elevated cancer incidence associated with environmental stress, may have indirectly selected for viruses harboring such health-associated genes as a response to altered ecological conditions (33,63). Although the Americas exhibited the highest number of viral sequences carrying health-associated genes, the average number of such genes per viral genome was greatest in Honduras (3.33), followed by South Africa (2.62), and lower in the broader Americas (1.50). This pattern may reflect the high biodiversity and ecological complexity of Honduras, which could enhance opportunities for viral gene acquisition and retention (64). Similarly, dense human populations and limited public health infrastructure in parts of South Africa may promote viral adaptation through the selection and mainte nance of health-associated genes (65,66). These findings underscore the importance of future research in ecologically diverse but underexplored regions to establish more comprehensive and objective global patterns in the distribution and evolution of viral health-associated genes.
Phylogenetic analyses revealed that viruses carrying identical health-associated genes within the same habitat displayed high sequence similarity, consistent with descent from common ancestral genes. These genes likely play pivotal roles in viral adaptation and survival, explaining their strong conservation across similar ecologi cal contexts (67). Habitat type and geographic position jointly influenced the evolu tionary relatedness of these viruses. For instance, in North America, viruses carrying GPX1 exhibited close genetic relatedness between marine and freshwater environ ments (pairwise distance: 0.3-0.4), suggesting limited habitat separation and consis tent selective pressures. Conversely, in Europe, viruses carrying bioB displayed marked genetic divergence between human-associated and environmental habitats (pairwise distance >0.9), reflecting strong habitat isolation and niche-specific evolution. Notably, viruses carrying GPX1 in human-associated samples from North America and Africa exhibited nearly identical genomic features, implying historical human movement, convergent selective pressures, or similar ecological conditions across continents (68). Collectively, these findings highlight the profound influence of both geography and habitat on the evolutionary trajectories of viral genomes carrying health-associated genes.
## Mechanisms underlying trade-offs between phage-transduced health-associ ated gene and pathogenic gene expression in the host
In the test experiment, the virus-carried bioB gene was successfully transduced into E. coli BL21, demonstrating the integration of a health-associated AMG and its regula tory influence on host pathogenic genes (69,70). Under host-simulated physiological conditions (37°C, pH 8.5), bioB activity exhibited an antagonistic relationship with the pathogenic genes GCH1 and UGDH, indicating that up-regulation of health-associated genes can suppress the expression of bacterial pathogenic genes. This suppression may reduce the transmissibility of bacterial virulence. Conversely, under environmental stress conditions (20°C, pH 5.5), transient activation of pathogenic genes may enhance the transmission potential of virulence determinants.
The coexistence of pathogenic and health-associated genes may arise from cascade regulation (71) or co-acquisition via horizontal transfer (72). Over evolutionary time scales, viruses appear to finely regulate both pathogenic and health-associated genes to balance virulence and host advantage. Viruses carrying pathogenic genes infect hosts more efficiently, replicate within them, and thereby ensure transmission and survival. These genes enhance pathogenicity and help overcome host defenses, improving transmission efficiency (73). In contrast, carrying health-associated genes strengthens host adaptation and survival, which indirectly benefits viral persistence. Some phages integrate health-associated genes into host genomes after infection, further enhancing host environmental adaptation (74,75). By carrying and regulating both pathogenic and health-associated genes, viruses achieve a dynamic equilibrium between exploiting host resources and supporting host survival, ensuring long-term persistence. This balance reflects the complex interplay of host-virus interactions, simultaneously promoting viral adaptation and modulating disease transmission (76)(77)(78)(79). Understanding these mechanisms may offer novel strategies to suppress the expression of pathogenic genes while enhancing health-associated genes, potentially reducing risk and supporting human health (80,81).
Genomic analyses reveal that eukaryote-to-virus gene transfer occurs over twice as frequently as the reverse (82), highlighting the potential for health-associated genes to enter viral genomes. Viruses thus act as vectors for both pathogenic and health-associ ated genes, facilitating their dissemination. However, whether these virally transported eukaryotic genes can integrate into the human genome and exert functional effects remains unresolved. Future studies should systematically screen human genomes for virus-derived sequences and assess their functional impacts in model systems to clarify the physiological and health implications of virus-mediated gene flow.
## MATERIALS AND METHODS
## Literature review and data review
To define the concept of "health-associated genes, " a comprehensive literature search was conducted to identify genes directly or indirectly linked to processes that benefit human health. Peer-reviewed articles published over the past 5 decades were retrieved from X-MOL (https://www.x-mol.com) and Web of Science (https:// www.webofscience.com/wos) using keywords including "anticancer gene, " "vitamin synthesis gene, " "antioxidant gene, " and "longevity gene. " From the screening of 58 relevant studies, a non-redundant set of 55 genes was compiled. These genes, which collectively define the health-associated gene category in this study, were classified into four functional groups: 18 involved in vitamin synthesis (e.g., bioB [34] and menA [35]), 14 with antioxidant functions (e.g., COQ2 [36,37] and GPX1 [38]), 14 linked to anticancer proteins (e.g., GSTT1 [39]), and nine associated with longevity (e.g., SIRT1 [NAD-depend ent protein deacetylase sirtuin 1]) (40) and IGF1R (insulin-like growth factor 1 receptor) [41]) (Table S1).
## Acquisition and comparison of sequences
Amino acid and nucleotide sequences of relevant genes were downloaded from the KEGG database (83). Gene names, KO numbers, and function information were col lected. The target amino acid sequences were compared through the IMG/VR v4 database (84), obtaining the global viral sequences encoding health-associated genes (E-value <0.0001). A total of 8,999 sequences that met the criteria (bit score >60, identity >20%) were screened, resulting in 4,556 non-redundant viral sequences and 548 high-quality sequences (85,86). Global spatial visualization was carried out through ArcGIS 10.7.
## Viral genome scanning and homology search
Viral genomes were scanned by HMMER (v.3.1b2) (87). All viral protein sequences were compared against the HMM profiles of the ViPhOG database, which contains curated orthologous groups of viral proteins (86,87). Matches with an E-value ≤1e -5 were considered significant, and only the highest-scoring hit for each sequence was retained to minimize redundancy and false positives.
## Gene annotation and cluster visualization
Functional annotation was performed using DRAM-V (v.1.2.0), which identifies viral marker genes and AMGs (42). In instances where the auxiliary score falls below 4, the M flag is assigned, while the A, V, or T flags are not assigned. This designation of a gene as a potential AMG is based on the specific numerical values assigned to each flag (42). Predictions were cross-validated with the vitality pipeline (v.1.2.0) (43). Annotation results were further compared against the KEGG database to confirm functional assignments. Genes with ambiguous functions were retained only if multiple databases gave consistent annotations.
## Promoter and terminator prediction
Regulatory elements were predicted using BPROM and FindTerm on the Softberry platform (http://www.softberry.com/) (88,89). BPROM-predicted promoters were filtered to retain only those with linear discriminant function >2.75, while FindTerm predic tions of terminators were ranked by confidence, with only the highest-confidence sites retained. Predicted promoters and terminators were manually inspected to ensure genomic context consistency with viral genome architecture (88,89).
## Protein structure prediction
Protein tertiary and quaternary structures were predicted through the SWISS-MODEL server (https://swissmodel.expasy.org/) (90). Models were generated with default parameters, and only those with GMQE >0.5 were retained to ensure structural reliability. GMQE is a quality score provided by SWISS-MODEL, ranging from 0 to 1, which pre dicts the expected accuracy of the resulting protein structure based on target-template alignment and template quality; higher GMQE values indicate more reliable models (44,49,50). Gene annotation information, including viral marker genes and AMGs, was visualized and refined using Chiplot (https://www.chiplot.online/), enabling clear display of gene types and functional categories (91,92). Gene clusters carrying health-associated AMGs were only selected for visualization if they also contained confirmed viral marker genes, ensuring that the displayed clusters represent bona fide viral genomic contexts.
## Phylogenetic analysis
High-quality amino acid sequences were aligned using ClustalW implemented in MEGA11 (93,94). Phylogenetic trees were constructed using the neighbor-joining method, with bootstrap analysis based on 1,000 replicates to assess the robustness of the branching topology. Bootstrap values were calculated for all nodes and displayed on the final tree (45). The resulting trees were exported and refined using the Chiplot online platform, where branch labels and functional annotations of genes were added for visualization.
## Gene positional analysis
The genomic positions of pathogenic genes and health-associated genes within viral genomes were visualized using the "Gene Location Visualize" function in TBtools (95), allowing assessment of their spatial relationships.
## Bacterial strain and culture conditions
The Escherichia coli BL21 strain used in this study endogenously harbors the pathogenic genes GCH1 and UGDH. This strain was transformed with the plasmid pET30a-bioB, which carries the health-associated gene bioB, to generate the experimental strain. Transform ants were selected on Luria-Bertani agar plates containing 50 µg/mL kanamycin. For protein expression, cultures were grown under varying temperature conditions (20°C and 37°C) and pH gradients (5.5, 7.0, 8.5, and 9.0). Protein expression was induced with isopropyl β-d-1-thiogalactopyranoside at a final concentration of 0.2 mM (for 20°C cultures) or 0.5 mM (for 37°C cultures) when the OD 600 reached 0.6-0.8.
## Protein expression analysis by SDS-PAGE and Western blot
Cells were harvested and disrupted by ultrasonication. The soluble (supernatant) and insoluble (inclusion body) fractions were collected and analyzed. Proteins were separated on 12% SDS-polyacrylamide gels and visualized using Coomassie Brilliant Blue stain ing. For Western blotting, proteins were transferred onto a polyvinylidene fluoride membrane, which was subsequently blocked and incubated with primary antibodies, followed by horseradish peroxidase-conjugated secondary antibodies. Protein bands were detected using an enhanced chemiluminescence detection system.
## Genetic confirmation and gene expression analysis by quantitative real-time PCR (qRT-PCR)
The presence of the endogenous GCH1 and UGDH genes in the bacterial genome was confirmed by sequencing of PCR-amplified products (Table S6). Total RNA was extracted from bacterial cultures, reverse-transcribed into cDNA, and analyzed by qRT-PCR using a SYBR Green-based system. The gapA gene was used as an endogenous control for normalization. Relative gene expression levels were calculated using the 2 -ΔΔCt method (96). All experiments were performed in six independent replicates and included an empty vector control to ensure reliability.
## References
1. Madsen, Dai, Mcmahon et al. (2020) "Human antibodies targeting influenza B virus neuraminidase active site are broadly protective" *Immunity*
2. Sudhan, Sharma (2020) "Chapter 4 -Human viruses: emergence and evolution"
3. Furuse, Oshitani (2020) "Viruses that can and cannot coexist with humans and the future of SARS-CoV-2" *Front Microbiol*
4. Lander, Linton, Birren et al. (2001) "Initial sequencing and analysis of the human genome" *Nature*
5. Hoyt, Storer, Hartley et al. (2022) "From telomere to telomere: the transcriptional and epigenetic state of human repeat elements" *Science*
6. Long, Giotis, Moncorgé et al. (2016) "Species difference in ANP32A underlies influenza A virus polymerase host restriction" *Nature*
7. Delamater, Street, Leslie et al. (2019) "Complexity of the basic reproduction number (R0)" *Emerg Infect Dis*
8. Barry, Viboud, Simonsen (2008) "Cross-protection between successive waves of the 1918-1919 influenza pandemic: epidemiological evidence from US Army camps and from Britain" *J Infect Dis*
9. Sharp, Hahn (2011) "Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med 1:a006841"
10. Lu, Zhao, Li et al. (2020) "Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding" *Lancet*
11. Meena, Swapnil, Zehra et al. (2019) "Chapter 11 -Virulence factors and their associated genes in microbes"
12. Keen (2012) "Paradigms of pathogenesis: targeting the mobile genetic elements of disease" *Front Cell Infect Microbiol*
13. Kumar, Prasoon, Kumari et al. (2021) "SARS-CoV-2-specific virulence factors in COVID-19" *J Med Virol*
14. Khan, Geiger (2021) "Role of viral protein U (Vpu) in HIV-1 infection and pathogenesis" *Viruses*
15. Laanto, Hoikkala, Ravantti et al. (2017) "Long-term genomic coevolution of host-parasite interaction in the natural environment" *Nat Commun*
16. Jiang, Guo, Wang et al. (2020) "Isolation and complete genome sequence of a novel cyanophage, S-B05, infecting an estuarine Synechococcus strain: insights into environmental adaptation" *Arch Virol*
17. Gasper, Schwach, Hartmann et al. (2017) "Distinct features of cyano phage-encoded T-type phycobiliprotein lyase ΦCpeT: the role of auxiliary metabolic genes" *J Biol Chem*
19. Thompson, Zeng, Kelly et al. (2011) "Phage auxiliary metabolic genes and the redirection of cyanobacterial host carbon metabolism" *Proc Natl Acad Sci*
20. Crummett, Puxty, Weihe et al. (2016) "The genomic content and context of auxiliary metabolic genes in marine cyanomyoviruses" *Virology (Auckl)*
21. Breitbart, Thompson, Suttle et al. (2007) "Exploring the vast diversity of marine viruses" *Oceanog*
22. Focardi, Ostrowski, Goossen et al. (2020) "Investigating the diversity of marine bacteriophage in contrasting water masses associated with the east Australian current (EAC) system" *Viruses*
23. Lindell, Jaffe, Johnson et al. (2005) "Photosynthesis genes in marine viruses yield proteins during host infection" *Nature*
24. Gazitúa, Vik, Roux et al. (2021) "Potential virus-mediated nitrogen cycling in oxygen-depleted oceanic waters" *ISME J*
25. Mann, Cook, Millard et al. (2003) "Marine ecosystems: bacterial photosynthesis genes in a virus" *Nature*
26. Clokie, Shan, Bailey et al. (2006) "Transcription of a "photosynthetic" T4-type phage during infection of a marine cyanobacterium" *Full-Length Text Applied and Environmental Microbiology*
27. Tang, Yu, Tang et al. (2019) "Bacteriophages from arsenic-resistant bacteria transduced resistance genes, which changed arsenic speciation and increased soil toxicity" *Environ Sci Technol Lett*
28. Brum, Hurwitz, Schofield et al. (2016) "Seasonal time bombs: dominant temperate viruses affect Southern Ocean microbial dynamics" *ISME J*
29. Chen, Golding, Sawai et al. (2005) "Population fitness and the regulation of Escherichia coli genes by bacterial viruses" *PLoS Biol*
30. Chakraborty, Sharma, Bhattacharya et al. (2021) "Evolution, mode of transmission, and mutational landscape of newly emerging SARS-CoV-2 variants"
31. Chakraborty, Bhattacharya, Sharma et al. (2022) "Evolution, epidemiology, geographical distribution, and mutational landscape of newly emerging monkeypox virus" *Geroscience*
32. Baquero, Medvedeva, Martin-Gallausiaux et al. (2024) "Stable coexistence between an archaeal virus and the dominant methanogen of the human gut" *Nat Commun*
33. Sun, You, Li et al. (2023) "Variants of a putative baseplate wedge protein extend the host range of Pseudomonas phage K8" *Microbiome*
34. Wu, Eckhardt, Baccarelli (2023) "Molecular mechanisms of environmental exposures and human disease" *Nat Rev Genet*
35. Cleary, Campbell, Chang (1972) "Location of promoter and operator sites in the biotin gene cluster of Escherichia coli" *Proc Natl Acad Sci*
36. Xv, Chen, Duan et al. (2018) "Research progress on the anticancer effects of vitamin K2" *Oncol Lett*
37. Zhu, Fu, Richman et al. (2017) "A personalized model of COQ2 nephropathy rescued by the wild-type COQ2 allele or dietary coenzyme Q10 supplementation" *JASN*
38. Mollet, Giurgea, Schlemmer et al. (2007) "Prenyldiphosphate synthase, subunit 1 (PDSS1) and OH-benzoate polyprenyltransferase (COQ2) mutations in ubiquinone deficiency and oxidative phosphoryla tion disorders" *J Clin Invest*
39. Schwarz, Löser, Cheng et al. (2023) "Side-by-side comparison of recombi nant human glutathione peroxidases identifies overlapping substrate specificities for soluble hydroperoxides" *Redox Biol*
40. Zhuo, Cai, Xiang et al. (2009) "GSTMI and GSTTI polymor phisms and nasopharyngeal cancer risk: an evidence-based metaanalysis" *J Exp Clin Cancer Res*
41. Xu, Wang, Fozouni et al. (2020) "SIRT1 is downregulated by autophagy in senescence and ageing" *Nat Cell Biol*
42. Abdellatif, Trummer-Herbst, Heberle et al. (2022) "Fine-tuning cardiac insulin-like growth factor 1 receptor signaling to promote health and longevity" *Circulation*
43. Shaffer, Borton, Mcgivern et al. (2020) "DRAM for distilling microbial metabolism to automate the curation of microbiome function" *Nucleic Acids Res*
44. Kieft, Zhou, Anantharaman (2020) "VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences" *Microbiome*
45. Patil, Akki, Raghu et al. (2023) "Sugarcane polyphenol oxidase: structural elucidation using molecular modeling and docking analyses" *Process Biochem*
46. Saitou, Nei (1987) "The neighbor-joining method: a new method for reconstructing phylogenetic trees" *Mol Biol Evol*
47. Liu, Wu, Yu (2015) "Coalescent methods for estimating species trees from phylogenomic data" *J Syst Evol*
48. Sourdis, Nei (1988) "Relative efficiencies of the maximum parsimony and distance-matrix methods in obtaining the correct phylogenetic tree" *Mol Biol Evol*
49. Niederwieser, Blau, Wang et al. (1984) "GTP cyclohydrolase I deficiency, a new enzyme defect causing hyperphenylalaninemia with neopterin, biopterin, dopamine, and serotonin deficiencies and muscular hypotonia" *Eur J Pediatr*
50. Kiefer, Arnold, Kunzli et al. (2009) "The SWISS-MODEL Repository and associated resources" *Nucleic Acids Res*
51. Liang, Jiang, Yuan et al. (2022) "Differential performance of RoseTTAFold in antibody modeling" *Brief Bioinformatics*
52. Ichinose, Ohye, Matsuda et al. (1995) "Characterization of mouse and human GTP cyclohydrolase I genes: mutations in patients with GTP cyclohydrolase I deficiency ( * )" *J Biol Chem*
53. Kozako, Ohsugi, Uchida et al. (0190) "High NAMPT expression and anti-tumour activity of NAMPT inhibitor in adult T-cell leukemia/lymphoma" *Ann Oncol*
54. Tang, Sanz, Smith et al. (2023) "Chemistry-led investigations into the mode of action of NAMPT activators, resulting in the discovery of non-pyridyl class NAMPT activators" *Acta Pharm Sin B*
55. Harrington, Kamdar, Ning et al. "Australian Ovarian Cancer Study. 2023. UGDH promotes tumor-initiating cells and a fibroinflammatory tumor microenvironment in ovarian cancer" *J Exp Clin Cancer Res*
56. Elia, Rossi, Stegen et al. (2019) "Breast cancer cells rely on environmental pyruvate to shape the metastatic niche" *Nature*
57. Winfield, Groisman (2003) "Role of nonhost environments in the lifestyles of Salmonella and Escherichia coli" *Appl Environ Microbiol*
58. Grases, Costa-Bauzá, Gomila et al. (2012) "Urinary pH and renal lithiasis" *Urol Res*
59. Mccallum, Tropini (2024) "The gut microbiota and its biogeography" *Nat Rev Microbiol*
61. Antiqueira, Petchey, Santos et al. (2018) "Environmental change and predator diversity drive alpha and beta diversity in freshwater macro and microorganisms" *Glob Chang Biol*
62. Tseng, Chiang, Shiah et al. (2013) "Microbial and viral metagenomes of a subtropical freshwater reservoir subject to Full-Length Text Applied and Environmental Microbiology December"
63. *ISME J*
64. Rao, Schieber, Connor et al. (2017) "Pathogen-mediated inhibition of anorexia promotes host survival and transmission" *Cell*
65. Correa, Howard-Varona, Coy et al. (2021) "Revisiting the rules of life for viruses of microorganisms" *Nat Rev Microbiol*
66. Giraudeau, Sepp, Ujvari et al. (2018) "Human activities might influence oncogenic processes in wild animal populations" *Nat Ecol Evol*
67. Howard (1936) "The forests and flora of British Honduras" *Nature*
68. Achoki, Sartorius, Watkins et al. (2022) "Health trends, inequalities and opportunities in South Africa's provinces, 1990-2019: findings from the Global Burden of Disease 2019 Study" *J Epidemiol Community Health*
69. De Villiers (2021) "Bridging the health inequality gap: an examination of South Africa's social innovation in health landscape" *Infect Dis Poverty*
70. Lebeer, Vanderleyden, Keersmaecker (2008) "Genes and molecules of lactobacilli supporting probiotic action" *Microbiol Mol Biol Rev*
71. Massarat, Gymrek, Mcstay et al. (2023) "Human pangenome supports analysis of complex genomic regions" *Nature*
72. Fortier, Sekulovic (2013) "Importance of prophages to evolution and virulence of bacterial pathogens" *Virulence*
73. Srikumar, Kröger, Hébrard et al. (2015) "RNA-seq brings new insights to the intra-macrophage transcriptome of Salmonella Typhimurium" *PLoS Pathog*
74. Fotsing, Margoliash, Wang et al. (2019) "The impact of short tandem repeat variation on gene expression" *Nat Genet*
75. Anderson, Roth (1978) "Tandem chromosomal duplications in Salmonella typhimurium: fusion of histidine genes to novel promoters" *J Mol Biol*
76. Geoghegan, Holmes (2018) "The phylogenomics of evolving virus virulence" *Nat Rev Genet*
77. Chevallereau, Pons, Van Houte et al. (2022) "Interactions between bacterial and phage communities in natural environments" *Nat Rev Microbiol*
78. Feiner, Argov, Rabinovich et al. (2015) "A new perspective on lysogeny: prophages as active regulatory switches of bacteria" *Nat Rev Microbiol*
79. Duggal, Emerman (2012) "Evolutionary conflicts between viruses and restriction factors shape immunity" *Nat Rev Immunol*
80. Roossinck (2011) "The good viruses: viral mutualistic symbioses" *Nat Rev Microbiol*
81. Ghazal, García-Ramírez, González-Armas et al. (2000) "Principles of homeostasis in governing virus activation and latency" *Immunol Res*
82. Jarett, Džunková, Schulz et al. (2020) "Insights into the dynamics between viruses and their hosts in a hot spring microbial mat" *ISME J*
83. Gan, Xu, Xiong et al. (2021) "Probiotics: their action against pathogens can be turned around" *Sci Rep*
84. Singh, Rao (2021) "Probiotics: a potential immunomodulator in COVID-19 infection management" *Nutr Res*
85. Irwin, Pittis, Richards et al. (2022) "Systematic evaluation of horizontal gene transfer between eukaryotes and viruses" *Nat Microbiol*
86. Kanehisa, Goto (2000) "KEGG: kyoto encyclopedia of genes and genomes" *Nucleic Acids Res*
87. Camargo, Nayfach, Chen et al. (2023) "IMG/VR v4: an expanded database of uncultivated virus genomes within a framework of extensive functional, taxonomic, and ecological metadata" *Nucleic Acids Res*
88. Roux, Adriaenssens, Dutilh et al. (2019) "Minimum information about an uncultivated virus genome (MIUViG)" *Nat Biotechnol*
89. Roux, Paul, Bagby et al. "Eloe-Fadrosh EA. 2021. Ecology and molecular targets of hypermutation in the global microbiome" *Nat Commun*
90. Eddy (2011) "Accelerated profile HMM searches" *PLoS Comput Biol*
91. Deltcheva, Chylinski, Sharma et al. (2011) "CRISPR RNA maturation by trans-encoded small RNA and host factor RNase III" *Nature*
92. Rinaldi, Lund, Blanco et al. (2016) "The Shine-Dalgarno sequence of riboswitch-regulated single mRNAs shows liganddependent accessibility bursts" *Nat Commun*
93. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-MODEL: homology modelling of protein structures and complexes" *Nucleic Acids Res*
94. Ji, Tang, Zhang (2022) "Effects of salt stress on the morphology, growth and physiological parameters of" *Juglansmicrocarpa L. seedlings. Plants (Basel)*
95. Li, Li, Zhao et al. (2023) "Physiological, biochemical, and genomic elucidation of the Ensifer adhaerens M8 strain with simultaneous arsenic oxidation and chromium reduction" *J Hazard Mater*
96. Thompson, Gibson, Higgins (2003) "Multiple sequence alignment using ClustalW and ClustalX" *Curr Protoc Bioinforma Chapter*
97. Tamura, Stecher, Kumar (2021) "MEGA11: molecular evolutionary genetics analysis version 11" *Mol Biol Evol*
98. Chen, Chen, Zhang et al. (2020) "TBtools: an integrative toolkit developed for interactive analyses of big biological data" *Mol Plant*
99. Livak, Schmittgen (2001) "Analysis of relative gene expression data using real-time quantitative PCR and the 2 -ΔΔCT method" *Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12809542&blobtype=pdf | # ViMOP: a user-friendly and field-applicable pipeline for untargeted viral genome nanopore sequencing
Nils Petersen, Mia Le, Annick Renevey, Ehizojie Emua, Sarah Ryter, Giuditta Annibaldis, Jacob Camara, Sanaba Boumbaly, Cyril Erameh, Tanja Laske, Jan Baumbach, Philippe Lemey, Stephan G€ Unther, Sophie Duraffour, Liana Kafetzopoulou
## Abstract
Motivation: Untargeted, also known as metagenomic, nanopore sequencing is a powerful tool for virus genomic surveillance, particularly in resource-limited settings and when paired with the portability of the MinION device (Oxford Nanopore Technologies, ONT). However, a major bottleneck for global access is the absence of a user-friendly software capable of efficiently analyzing untargeted nanopore sequencing data to generate high-quality consensus genomes.
Results:We share ViMOP, a pipeline built on our long-term experience in nanopore field sequencing. The pipeline emphasizes field userfriendliness, flexibility and versatility to analyze reads generated directly from human clinical samples. The software assembles de novo contigs, matches contigs to known viral references and uses them to assemble consensus genomes. Executed with a single Nextflow command or via the EPI2ME Desktop interface (ONT), results are summarized in an HTML report. ViMOP, through its user-centered design, lowers the barrier to high-quality virus genome reconstruction and advances capacity for genomic surveillance.
## 1 Introduction
Genomic surveillance has proven essential for identifying and responding to emerging viral disease outbreaks such as with mpox and COVID-19 (Attwood et al. 2022, Kinganda-Lusamaki et al. 2025). Rapid characterization of viral agents, alongside dense coverage of consensus genomes to monitor virus evolution, has substantially informed outbreak control efforts. This approach has facilitated the reconstruction of chains of transmissions, provided insights into the origin(s) of virus (re)-emergence, and enabled the monitoring of potential vaccine escape variants (Quick et al. 2016, Sissoko et al. 2017, Harvey et al. 2021, Keita et al. 2021, Kinganda-Lusamaki et al. 2021). Today, real-time genomic data sharing serves as a cornerstone for public health surveillance and evidence-based decision making.
Outbreaks often occur in endemic regions with limited resources (Kinganda-Lusamaki et al. 2021, Koundouno et al. 2022, Kinganda-Lusamaki et al. 2025). Global disparities persist in access to genomic surveillance infrastructure, and solutions to facilitate broad access and field applicability must be prioritized (Brito et al. 2022, Carter et al. 2022). Nanopore-based sequencing using the MinION device (Oxford nanopore technologies, ONT) offers a portable approach suited for such settings (Quick et al. 2016, Quick et al. 2017, Kafetzopoulou et al. 2019, Keita et al. 2021, de Vries et al. 2022). We previously demonstrated the feasibility of untargeted (metagenomic) nanopore sequencing and onsite analysis during outbreak emergencies in low-and middleincome countries (LMICs) (Kafetzopoulou et al. 2019, Keita et al. 2021, Koundouno et al. 2022).
Recurring challenges in on-site field data analysis include the highly demanding technical setup in environments with unreliable power supply and internet access, as well as the need of advanced bioinformatics expertise. Operating bioinformatics tools from the command line requires expertise that may not be available when trained personnel are scarce. Also, factors such as field-sample quality (e.g. lack of cold chain) and sample type (e. g. blood, swab or seminal fluid) may lead to large fractions of host-reads, short reads or uneven read coverage, challenges that analysis pipelines should be equipped to handle. Ultimately, generating high-quality consensus genomes is a requirement for further downstream molecular analysis, and sharing with public health authorities (Carter et al. 2022).
Few tools support automated analysis of untargeted viral nanopore data (Table 1, available as supplementary data at Bioinformatics online). None fully satisfied our requirements for offline use, simple setup and operation by non-experts. Vir-MinION and VirPipe support viral detection but do not automate the transition to reference-based assembly, instead relying on de novo reconstruction that is often fragmented for untargeted sequencing data (Mastriani et al. 2022, Kim et al. 2023). VirDetector is limited when host reads, segmented viral genomes, or co-infections are present (Kaiser et al. 2025). INSaFLU-TELEVIR is the only option with a graphical interface, yet still requires command-line installation, separates detection and assembly, and depends on a large local database (Santos et al. 2024).
We thus developed ViMOP (Virus Metagenomics for Outbreaks Pipeline), a user-friendly, automated pipeline for virus detection and reference-guided genome assembly of untargeted sequencing reads from human clinical or animal samples. Integration into ONT's EPI2ME framework enables command-line-free operation, setup and updates. Alternatively, it runs with a single Nextflow command. Docker containers grant independence from the host operating system. ViMOP runs offline on a laptop (≥16 CPU cores, 30 GB RAM recommended) enabling flexible deployment.
## 2 Materials, methods, and results
## 2.1 Broad overview of the nanopore virus surveillance pipeline
ViMOP detects known viral species and generates consensus genomes from basecalled nanopore reads (FASTQ) (Fig. 1; see Table 2, available as supplementary data at Bioinformatics online for tool choice rationales). Following separation of viral from non-viral reads, viral reads are assembled de novo into contigs. These are used to identify closely related reference genomes from a local database. Viral consensus sequences are subsequently generated via reference-guided assembly. All results, including consensus sequences and alignments, are output along with an HTML report.
## 2.2 Installation
ViMOP can be installed via EPI2ME Desktop by entering the GitHub repository URL. After installation, the backend database can be downloaded using the setup functionality. Alternatively, these steps can be done in a single Nextflow command.
## 2.3 Database
ViMOP uses a structured backend database and provides functionalities for automated download and updates. Users can create their own database. Our database is tailored to hosts and viral species relevant to our research. It contains (i) sequence files for background read removal, (ii) a virus genome set, and (iii) a Centrifuge index for read and contig classification. For background read removal, our database comprises the human genome and transcriptome, two rodent genomes and one mosquito genome, as well as a set of reagent-associated sequences including common bacterial laboratory strains and expression vectors (Tables 3 and4, available as supplementary data at Bioinformatics online). The virus genome set is built based on all GenBank virus genomes (Benson et al. 2013). A curated subset of species relevant to our projects is maintained to ensure quality and consistency (Table 5, available as supplementary data at Bioinformatics online). Genomes with (a) >1% of unknown bases (N), (b) >98% sequence identity, or (c) <80% of a reference genome size, or not aligning to any reference genome (considered misassigned) were excluded. For SARS-CoV-2, a pre-selection from RVDB (Goodacre et al. 2018) was used. All remaining GenBank viral genomes are included without curation, and may require additional review upon detection. Groups comprising all genomes in our database for selected species or families were defined to filter for virus reads (Tables 5 and6, available as supplementary data at Bioinformatics online). Our Centrifuge index combines our virus genome set with sequences (RefSeq) of human, mouse, bacterial, and archaeal genomes (O' Leary et al. 2016). Database updates and extensions are implemented as needed or upon user request. In low-connectivity settings, users may restrict an update to individual components.
## 2.4 Read trimming
seqtk (Li, https://github.com/lh3/seqtk) trims 3' and 5' ends to remove primer sequences. The number of trimmed-bases is user-configurable.
## 2.5 Taxonomic classification and read cleaning
Non-viral reads are removed in three steps. First, the reads are taxonomically classified using Centrifuge (Kim et al. 2016). Reads classified as non-viral are removed, and an overview of the microbial content of the sample is displayed in the ViMOP report. Second, for thorough removal of nonviral reads, the remaining reads are mapped with minimap2 (Li 2018) against one or more sets of host genome sequences and reagent-associated sequences. The resulting output is called the cleaned read set. Third, ViMOP filters for viral reads. This is optional. The user can apply one or multiple filters in parallel to specifically map the reads against one or more virus species or family. All reads are mapped against target viral genomes using minimap2. The reads mapping to a viral reference are called target filtered reads and one read set is created for each filter. For both the set of cleaned reads, and each set of target-filtered reads, de novo assembly and reference genome searches are performed.
## 2.6 De novo assembly
Reads are assembled into contigs using Canu (Koren et al. 2017). An iterative approach was implemented to account for uneven genome abundance (e.g. in a co-infection, one viral species might be less abundant than the other), increasing the likelihood for broad viral species detection. At each iteration, reads mapping against any generated contigs are removed. The remaining reads are further used for another Canu run. This is done until a pre-defined maximum number of iterations is reached, or until no contig is created or no reads are left. Assembled contigs are then used for reference identification. A special routine is integrated in ViMOP for Canu runs of target filtered read sets that did not yield any contigs. In this case, reads are clustered using CD-HIT (Fu et al. 2012) and then used as queries, analogous to contigs.
## 2.7 Reference identification
Contigs are matched to reference genomes from the backend database using BLAST (Altschul et al. 1990). For each contig, the top hit is selected using the default BLAST ranking with ties being resolved by the database's internal ordering. Each unique hit is then used for reference-guided assembly. Users can manually add additional reference sequences in a FASTA file.
## 2.8 Reference-guided assembly
ViMOP assembles a consensus sequence for each reference identified. All reads are mapped to the reference sequence and large insertions and deletions are detected with Sniffles2 (Smolka et al. 2024). These are integrated into a first draft genome using BCFtools (Danecek et al. 2021). All reads are mapped to this draft, and the final consensus is built using either Medaka (ONT, https://github.com/nanoporetech/me daka) or Samtools (Danecek et al. 2021) to call single-nucleotide variants and small insertions and deletions. Per default, ViMOP uses Medaka if a model that fits the respective basecaller is available, and Samtools if not. In both cases, a minimum depth (default 20) is required to call bases. Samtools additionally requires a minimum proportion (70%) for a nucleotide to be called. Positions below the thresholds are masked with Ns. Samtools may alter the length of Nmasked areas due to insertion or deletion artifacts from sparse reads. Therefore, a custom script replaces deletions that are not covered by the minimum depth of reads with Ns and removes insertions of Ns.
## 2.9 Contig classification
Contigs are classified using Centrifuge. This can help to identify contigs that were not assigned to a reference or are only partially aligned to the reference hit, e.g. if these are from bacterial origin. The Centrifuge classification is only used for reporting and not for the assembly of the reference target selection.
## 2.10 Output
ViMOP produces consensus sequences (FASTA), contigs (FASTA), alignment files (BAM), tables as tab-delimited files, and a separate HTML report for each sample. The userfriendly report, organized in three sections, provides information on the outcome of all the pipeline steps. The first section includes read distributions, read filtering statistics, and an overview of the Centrifuge read classification. The second section lists the viruses identified and statistics on genome recovery. The third section lists all de novo assembled contigs with the corresponding BLAST hits, and Centrifuge classification.
## 2.11 Benchmark
We compared ViMOP against INSaFLU-TELEVIR and VirDetector using simulated viral reads, mixed host-virus datasets, and real samples from the Sequence Read Archive (Supplementary Methods, Results, and Tables 789101112, available as supplementary data at Bioinformatics online). While VirDetector failed to automatically assemble many of the genomes, ViMOP and INSaFLU-TELEVIR show comparable results in most cases. Comparison on simulated data shows, that ViMOP is more robust when it comes to high mutation levels. Different to INSaFLU, ViMOP also recovers partial genomes below 50% completeness. On the downside, ViMOP has longer runtimes.
## 3 Conclusions and outlook
ViMOP simplifies untargeted nanopore sequencing data analysis. By integrating field-applicability, accessibility, maintainability and versatility, ViMOP enables broad usage for virus detection and genome assembly. The graphical interface, offline capability, and HTML reports make it particularly attractive for small teams focused on laboratory implementation. ViMOP integrates 8-years of experience of nanopore sequencing development in laboratories across Sub-Saharan Africa, and within the European Mobile Laboratory (https:// www.emlab.eu). Integrated into our laboratory network, ViMOP will be continuously improved. ViMOP aims to strengthen global outbreak preparedness by supporting rapid, decentralized virus genomic surveillance where it is most urgently needed.
## References
1. Altschul, Gish, Miller (1990) "Basic local alignment search tool" *J Mol Biol*
2. Attwood, Hill, Aanensen (2022) "Phylogenetic and phylodynamic approaches to understanding and combating the early SARS-CoV-2 pandemic" *Nat Rev Genet*
3. Benson, Cavanaugh, Clark (2013) *Nucleic Acids Res*
4. Brito, Semenova, Dudas (2022) "Swiss SARS-CoV-2 Sequencing Consortium. Global disparities in SARS-CoV-2 genomic surveillance" *Nat Commun*
5. Carter, Yu, Sacks (2022) "Global genomic surveillance strategy for pathogens with pandemic and epidemic potential 2022-2032" *Bull World Health Organ*
6. Danecek, Bonfield, Liddle (2021) "Twelve years of samtools and bcftools" *Gigascience*
7. De Vries, Cogan, Gubala (2022) "Rapid, in-field deployable, avian influenza virus haemagglutinin characterisation tool using MinION technology" *Sci Rep*
8. Fu, Niu, Zhu (2012) "CD-HIT: accelerated for clustering the nextgeneration sequencing data" *Bioinformatics*
9. Goodacre, Aljanahi, Nandakumar (2018) "A reference viral database (RVDB) TO enhance bioinformatics analysis of high-throughput sequencing for novel virus detection" *mSphere*
10. Petersen
11. Harvey, Carabelli, Jackson (2021) "COVID-19 Genomics UK (COG-UK) Consortium. SARS-CoV-2 variants, spike mutations and immune escape" *Nat Rev Microbiol*
12. Kafetzopoulou, Pullan, Lemey (2019) "Metagenomic sequencing at the epicenter of the Nigeria 2018 Lassa fever outbreak" *Science*
13. Kaiser, Groschup, Sadeghi (2025) "VirDetector: A bioinformatic pipeline for virus surveillance using nanopore sequencing" *Bioinformatics*
14. Keita, Koundouno (2021) "Resurgence of Ebola virus in 2021 in Guinea suggests a new paradigm for outbreaks" *Nature*
15. Kim, Song, Breitwieser (2016) "Centrifuge: rapid and sensitive classification of metagenomic sequences" *Genome Res*
16. Kim, Park, Lee (2023) "VirPipe: An easy-TO-use and customizable pipeline for detecting viral genomes from nanopore sequencing" *Bioinformatics*
17. Kinganda-Lusamaki, Amuri-Aziza, Fernandez-Nuñez (2025) "Clade I mpox virus genomic diversity in the democratic republic of the Congo, 2018-2024: predominance of zoonotic transmission" *Cell*
18. Kinganda-Lusamaki, Black, Mukadi (2021) "Integration of genomic sequencing into the response to the Ebola virus outbreak in Nord Kivu, Democratic Republic of the Congo" *Nat Med*
19. Koren, Walenz, Berlin (2017) "Canu: scalable and accurate longread assembly via adaptive k-mer weighting and repeat separation" *Genome Res*
20. Koundouno, Kafetzopoulou (2022) "Detection of marburg virus disease in Guinea" *N Engl J Med*
21. Li (2018) "Minimap2: pairwise alignment for nucleotide sequences" *Bioinformatics*
22. Mastriani, Bienes, Wong (2022) "PIMGAVir and vir-minion: Two viral metagenomic pipelines for complete baseline analysis of 2nd and 3rd generation data" *Viruses*
23. O'leary, Wright, Brister (2016) "Reference sequence (RefSeq) database at NCBI: current status, taxonomic expansion, and functional annotation" *Nucleic Acids Res*
24. Quick, Grubaugh, Pullan (2017) "Multiplex PCR method for MinION and Illumina sequencing of Zika and other virus genomes directly from clinical samples" *Nat Protoc*
25. Quick, Loman, Duraffour (2016) "Real-time, portable genome sequencing for Ebola surveillance" *Nature*
26. Santos, Sobral, Pinheiro (2024) "INSaFLU-TELEVIR: an open web-based bioinformatics suite for viral metagenomic detection and routine genomic surveillance" *Genome Med*
27. Sissoko, Duraffour, Kerber (2017) "Persistence and clearance of Ebola virus RNA from seminal fluid of Ebola virus disease survivors: a longitudinal analysis and modelling study" *Lancet Glob Health*
28. Smolka, Paulin, Grochowski (2024) "Detection of mosaic and population-level structural variants with Sniffles2" *Nat Biotechnol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724381&blobtype=pdf | # A novel partitivirus confers dual contradictory effects to its host fungus: growth attenuation and virulence enhancement
Zhen Liu, Maoqiu Chen, Ioly Kotta-Loizou, Robert Coutts, Linghong Kong, Hamdy Aboushedida, Risky Sari, Hiromitsu Moriyama, Wenxing Xu
## Abstract
Mycoviruses possess a potential role for biological control due to their ability to reduce both virulence and vegetative growth in some phytopathogenic fungi. However, mycoviruses that enhance fungal pathogenicity have been poorly studied and characterized. In this study, a novel double-stranded RNA (dsRNA) fungal virus, tentatively named Sinodiscula camellicola partitivirus 1 (ScPV1), was identified in the phytopathogenic fungus Sinodiscula camellicola, isolated from tea leaves. ScPV1 possesses two genomic components of 1,835 bp and 1,697 bp in length, each containing an open reading frame (ORF) encoding a putative RNA-dependent RNA polymerase (RdRP) and coat protein (CP), respectively, as confirmed by mass spectrometry. Phylo genetic analysis of the amino acid sequences of the RdRPs from ScPV1 and related mycoviruses placed ScPV1 within a newly proposed genus, Epsilonpartitivirus, in the family Partitiviridae. The virus was purified using ultracentrifugation, and transmission electron microscopy revealed that ScPV1 dsRNA genomes are encapsidated in virus particles ca. 31 nm in size, ranging from 24.9 to 36.8 nm, together with the RdRP protein, which was of an unexpected size. Transfection with purified virions generated transfec tants with significantly reduced growth rates but with increased virulence, indicating that ScPV1 confers unusual effects on its host fungus. This finding represents a significant advancement in understanding the complex interactions between mycoviruses and their host fungi. IMPORTANCE Here, we identified a novel partitivirus, tentatively named Sinodiscula camellicola partitivirus 1 (ScPV1), marking the first report of a partitivirus from a phytopathogenic fungus infecting tea plants. ScPV1 is characterized by possession of two dsRNA genomic components encapsidated in particles of varying sizes, along with an RNA-dependent RNA polymerase protein of an expected size, which contained some unique amino acids, indicating its distinct molecular and morphological traits. Biological tests on transfectants generated following protoplast infection with purified virions demonstrated that ScPV1 impairs vegetative growth while enhancing virulence in its fungal host. This finding represents the first instance of a mycovirus responsible for hypervirulence on a phytopathogenic fungus through virion transfection, as well as the first case of a partitivirus conferring hypervirulence while reducing vegetative growth in a phytopathogenic fungus. We anticipate that these findings will significantly advance our understanding of the complex interactions between mycoviruses and their host fungi.KEYWORDS mycovirus, Camellia sinensis, two-sided effects, Sinodiscula camellicola partitivirus 1, hypervirulence, Partitiviridae M ycoviruses (or fungal viruses) either affect the biological traits of their host fungi or are latent (1). Some viruses can impair the growth of and reduce the
virulence in their host fungi, making them potential biocontrol agents. For example, Cryphonectria hypovirus 1 (CHV1) has been used to control chestnut blight in Europe, and Sclerotinia sclerotiorum hypovirulence-associated DNA virus 1 (SsHADV-1) has been applied to manage oilseed rape Sclerotinia stem rot in China (2,3). Conversely, certain mycoviruses can enhance specific biological traits in their host fungi, granting them an adaptive advantage in particular environments. This is illustrated by the cases of Beauveria bassiana victorivirus 1 (BbVV-1) and Beauveria bassiana polymycovirus 1 (BbPmV-1), which, when infecting the entomopathogenic fungus Beauveria bassiana, lead to an increase in virulence against insects (4). This hypervirulent trait is attractive since it enhances the potential biocontrol effect for host fungi against agricultural pests. Conversely, two mycoviruses have been demonstrated to be related to hypervir ulence in phytopathogenic fungi: a 6.0-kbp dsRNA, which is phylogenetically related to plant cryptic viruses, in Nectria radicicola, the causal fungus of ginseng root rot (5); and Leptosphaeria biglobosa quadrivirus-1 (LbQV-1) in Leptosphaeria biglobosa, causing Phoma stem canker (blackleg) of oilseed rape (6). The mycoviruses associated with hypervirulence in phytopathogenic fungi have implications that run counter to their potential use as biocontrol agents. Consequently, careful management and utilization of these mycoviruses is essential to prevent them from inadvertently increasing the virulence of their fungal hosts.
The family Partitiviridae contains a large number of viruses that infect fungi, plants, and protozoa and have been divided into the five genera Alphapartitivirus, Betapartivirus, Deltapartivirus, Gammapartitivirus, and Cryspovirus (7,8) and the two newly proposed genera Epsilonpartitivirus (9) and Zetapartitivirus (10,11). Members of this family possess genomes comprising two linear double-stranded RNA (dsRNA) molecules, sized between 1.4 and 2.4 kilobases (kbp), with the larger molecule encoding the RNA-dependent RNA polymerase (RdRP) and the smaller one encoding the capsid protein (CP), both of which are individually packaged into viral particles measuring 25-40 nanometers in size (8). Partitiviruses are generally known for their persistent and asymptomatic infections in their hosts (7); however, there are notable exceptions. For example, Colletotrichum alienum partitivirus 1 (CaPV1) has been shown to significantly reduce host virulence, mycelial growth, appressorial development, and appressorium turgor, while increasing conidial production with abnormal morphology (12). Similarly, Aspergillus fumigatus partitivirus 1 (13) and Heterobasidion partitivirus 3 (14) have been found to attenuate the growth rates of their host fungi, and Aspergillus flavus partitivirus 1 (10) and Sclerotinia sclerotiorum partitivirus 1 (15) have been observed to attenuate virulence. Additionally, Rosellinia necatrix partitivirus 2 (RnPV2) caused no apparent morphological changes to its host fungus Rosellinia necatrix, but induced obvious morphological changes in a Dicer-like 2 knockout mutant (dcl-2) of a non-natural host, Cryphonectria parasitica, which causes chestnut blight (16). Moreover, infection with a partitivirus has been shown to induce hypervirulence in Talaromyces marneffei, a thermal dimorphic clinical fungus (17). However, the extent to which partitivirus-mediated hypervirulence occurs in phytopathogenic fungi in nature is not known.
Tea, represented by the species Camellia sinensis (L.) Kuntze, is of considerable economic significance in China and provides a multitude of health benefits to humans. Within the spectrum of diseases that afflict tea plants, anthracnose emerges as one of the most severe, presenting a substantial challenge to the cultivation of tea (18). The pathogens responsible for tea anthracnose exhibit regional diversity across the globe, with the disease commonly attributed to Discula theae-sinensis (formerly known as Gloeosporium theae-sinensis) and various species of Colletotrichum (19). Recently, two new fungal species, Sinodiscula theae-sinensis and Sinodiscula camellicola, which belong to a newly proposed genus Sinodiscula within the family Melanconiellaceae, have been identified and characterized as causative agents of tea anthracnose (20). The disease caused by these fungi exhibits similar symptoms: initially, dark green or yellowishbrown watery spots appear, which then expand along the leaf veins, forming irregu lar-shaped spots. These spots gradually turn brown or reddish-brown and eventually become grayish-white (20). The visual similarity of the disease symptoms caused by these pathogens presents a challenge in differentiating them through unaided visual inspection. Consequently, a specific disease name, "tea leaf blight, " has been proposed to describe the symptoms induced by Sinodiscula species (20). Given that the majority of Chinese tea cultivars are susceptible to this fungal disease, the implementation of effective control measures is essential to protect the yield and quality of tea crops. This imperative necessitates the investigation of mycoviruses that infect related fungi as an understanding of these viral interactions could potentially lead to novel strategies for disease management.
In this study, a novel partitivirus was identified from S. camellicola isolated from tea plants, and its molecular, morphological, and biological traits were characterized. This partitivirus exhibits some unique molecular and morphological traits unreported in other partitiviruses and dual effects on the fungal host by simultaneously attenuating growth while enhancing virulence, marking the first instance of a partitivirus contributing to hypervirulence in a phytopathogenic fungus.
## RESULTS
## Partitivirus dsRNAs in S. camellicola strain WJT-1-1
S. camellicola strain WJT-1-1 isolated from tea leaves exhibited enhanced virulence but lower growth rates as compared to other S. camellicola strains, as exemplified by strain SST-5 (Fig. 1A andB). To investigate whether these biological changes were due to the presence of a virus, dsRNAs were extracted from the mycelia of strains WJT-1-1 and SST-5 and subjected to electrophoresis on a 1.0% agarose gel. No significant differences were observed in their RNA patterns (Fig. 1C). Further analysis involved subjecting the nucleic acids to agarose gel electrophoresis after digestion with DNase I and S1 nuclease, revealing two dsRNA bands termed dsRNA1 and 2 based on their sizes exclusively in strain WJT-1-1 (Fig. 1D). Random primers were employed for reverse transcription and polymerase chain reaction (RT-PCR) amplification and rapid amplification of cDNA ends (RACE). The sequences of full-length complementary DNA (cDNA) clones were deter mined for dsRNA1 (GenBank no. PQ201539), which was 1,835 bp in length, and dsRNA2 (PQ201540), which was 1,697 bp in length (Fig. 1E). The 5′ and 3′-untranslated regions (UTRs) were 78 bp and 47 bp in dsRNA1 and 100 bp and 91 bp in dsRNA2, respectively. The 5′-UTRs exhibited conservation, featuring the conserved stretch "CCCAUUAAA. " In contrast, the 3′-UTRs of dsRNA1 and dsRNA2 showed lower conservation (Fig. 1F). Genomic organization analysis of both dsRNAs revealed that dsRNA1 contains one open reading frame (ORF) on the positive strand. This ORF starts at nucleotide (nt) 79 and terminates at nt 1,788, encoding a putative protein of 569 amino acids (aa) with an estimated molecular weight of 65.5 kDa. Similarly, ORF2 in dsRNA2 spans from nt 101 and terminates at nt 1606 and putatively encodes a protein of 501 aa with a molecular weight of 55.7 kDa.
Blastx searches revealed that the proteins encoded by dsRNA1 and dsRNA2 share the highest identity with the RdRP gene (NCBI GenBank No. UOK20169, 92% coverage, 62.70% identity, E value = 0.0) and CP (88% coverage, 52.27% identity, E value = 5e-160) of Diplodia seriata partitivirus 1 (DsPV1) and Colletotrichum liriopes partitivirus 1 (ClPV1) (Table S1), supporting the notion that dsRNAs 1 and 2 harbor putative ORFs encoding RdRP and CP, respectively. Additionally, the putative RdRP shares serially high identities (92%-93% coverage, 61.71%-62.46% identity, E value = 0.0) with the RdRP of Colletotrichum eremochloae partitivirus 1, Erysiphe necator-associated partiti-like virus 1, and Metarhizium brunneum partitivirus 1 in the family Partitiviridae (Table S1). Based on their molecular characteristics and similarity to other viruses, these dsRNAs are proposed as a novel partitivirus, tentatively named Sinodiscula camellicola partitivirus 1 (ScPV1).
An additional 31 S. camellicola samples were collected from two distinct counties in Hubei Province: 23 from Xuan'en County (Enshi Tujia and Miao Autonomous Prefec ture) and nine from Zigui County (Yichang City), in China, and subjected to ScPV1 identification by RT-PCR using the extracted nucleic acids and the primer pair nomina ted RNA1-F2/RNA1-R2, resulting in amplicons in size of 411 bp (Fig. S1 and Table S2). It revealed an incidence rate of 39.1% in Xuan'en County samples, while none were detected in Zigui County samples, supporting its regional prevalence in specific areas.
## Multiple alignment and phylogenetic analysis of the RdRP reveals that ScPV1 belongs to a new viral genus
Multiple alignment analysis was performed for the putative RdRP sequences of ScPV1 and some closely related members in the family Partitiviridae. Six conserved motifs (Motifs III to VIII), similar to those observed in other related members, were detected, whereas ScPV1 RdRP contains several unique amino acids, including a phenylalanine (F) instead of leucine (L) in Motif IV, F instead of valine (V) or isoleucine (I) or L in Motif V, and serine (S) instead of aspartic acid (D) in Motif VII (Fig. 2A). A phylogenetic tree was constructed using the sequence of the ScPV1 RdRP gene and representative members from various genera within the family Partitiviridae. The analysis revealed that the ScPV1 RdRP gene clusters with those of members belonging to the newly proposed genus Epsilonpartitivirus, while also forming an independent branch within the tree (Fig. 2B). Based on the established threshold criteria for species distinction, which requires a 90% identity in RdRP sequences (21), ScPV1 is confirmed as a novel member of the genus Epsilonpartitivirus within the family Partitiviridae.
## ScPV1 is encapsidated in isometric particles
The encapsidation of ScPV1 dsRNA in particles was investigated using virions purified from strains WJT-1-1 and SST-5, which were subjected to stepwise sucrose gradient centrifugation (10% to 40% in 10% sucrose increments). Nucleic acids were extracted from each gradient fraction and subjected to electrophoretic analysis on 6% polyacryla mide gel electrophoresis (PAGE) gels, the results of which showed that the target dsRNA bands, which matched the sizes of those extracted from strain WJT-1-1 mycelia exclusively detected in nucleic acid preparations of strain WJT-1-1, sharply concentrated in the 30% sucrose fraction (Fig. 3A).
To identify presumed virion proteins, these were extracted from 10% to 40% sucrose gradient fractions and individually subjected to SDS-PAGE. The results indicated that two protein bands, designated as P55 and P100 according to their estimated sizes, were predominantly found in the 30% sucrose fraction. Additionally, a protein band matching the size of P100 was specifically concentrated in the 40% sucrose fraction. Both protein bands were excised and analyzed using peptide mass fingerprinting (PMF). The results revealed that 228 peptide segments from P55 matched the protein sequence encoded by dsRNA2, representing approximately 74% of the entire sequence (Table S3), confirming that P55 is the CP encoded by dsRNA2. Additionally, seven peptide segments from P100 matched the protein sequence encoded by dsRNA1, accounting for about 20% of the entire sequence (Table S4), indicating that P100 is likely the RdRP protein encoded by dsRNA1. No proteins were detected in the ScPV1-free strain SST-5 (Fig. 3B). Transmission electron microscopy (TEM) examination of these fractions revealed the presence of isometric VLPs, having a diameter that spanned from 24.9 to 36.8 nm, with Full-Length Text an average of 31 nm (Fig. 3C). In contrast, no viral particles were discerned in the control strain SST-5.
## ScPV1 infection reduces growth rate but increases host fungus virulence
To assess viral transmission, the ScPV1-infected strain WJT-1-1, acting as the donor of the virus, was cultured in direct contact with the uninfected strain SST-5, which served as the recipient. Following a 7-day contact-culture period, 24 mycelial disks (marked by asterisks in Fig. S2A) were excised from the colonies of strain SST-5 at six independent peripheral positions on the colonies and subjected to dsRNA extraction. After treatment with the S1 enzyme, the extracted dsRNAs were visualized on agarose gels, and the results showed the absence of any dsRNAs in the strain SST-5 sub-isolates, suggesting that ScPV1 is most likely difficult to be horizontally transmitted from strain WJT-1-1 to other strains of the fungus (Fig. S2B).
To further assess the biological consequences of ScPV1 infection, purified virus particles were transfected into the protoplasts of strain SST-5 using PEG mediation. A total of 110 protoplast-generated colonies were randomly picked and transferred on fresh plates incubated as above and subjected to nucleic acid extraction. The extracted nucleic acids were treated with S1 enzyme and analyzed by agarose gel electrophoresis, which revealed that six transfectants (representing a frequency of 5%) were infected with ScPV1 (Fig. 4A). This finding was further supported by RT-PCR identification using the extracted nucleic acids and the primer pair RNA1-F2/RNA1-R2 (Fig. 4B; Table S2).
To determine the ratios of dsRNA1 to dsRNA2 in the transfectants, strains SST-5-T1 to SST-5-T6 together with WJT-1-1 were subjected to dsRNA extraction, followed by DNase I and S1 nuclease treatment, and non-denaturing PAGE analysis. The results revealed that the dsRNA2/dsRNA1 ratios varied among transfectants, ranging from 1.32 in SST-5-T3 to 2.68 in SST-5-T6, but were consistently higher than the ratio (1.19) observed in the original strain WJT-1-1 (Fig. 4C andD).
Following culture on PDA, all the transfectants exhibited sparse hyphal patterns with petal-shaped edges, similar to the morphologies of strain WJT-1-1 (Fig. 5A). Further more, all the transfectants demonstrated reduced growth rates, ranging from 5.9 to 6.3 mm/day, as compared to their parent strain SST-5, which exhibited a growth rate of 6.5 mm/day (Fig. 5C). By contrast, most of the transfectants exhibited significantly increased virulence, resulting in lesions with areas of 102.0 to 130.4 mm 2 except for two transfectants (SST-5-T2 and -T4), which caused similar lesions (79.7 to 83.0 mm 2 ), as compared to the lesions (67.7 mm 2 ) elicited by strain SST-5 at 5 days post inoculation (dpi) on C. sinensis cultivar "Echa" tea leaves (Fig. 5B andD).
The observed biological effects might be influenced by intra-isolate variability, so the transfectant SST-5-T3, displaying relatively pronounced phenotypic alterations, was selected as the donor strain for back-introduction of ScPV1 into the virus-free recipient strain SST-5 through contact culture in four replicates (Fig. S3A). After three rounds of subculturing, RT-PCR analysis of twelve SST-5 subisolates confirmed ScPV1 infection in all cases (Fig. S3B). Six transfectants derived from SST-5-T3 (T3-1 to T3-6) were subjected to biological characterization. These transfectants exhibited reduced growth rates ranging from 5.9 to 6.3 mm/day compared to the parent strain SST-5 (6.5 mm/day; Fig. 6A andC). Furthermore, they demonstrated significantly enhanced virulence, causing lesions measuring 102.5 to 130.0 mm² on C. sinensis "Echa" tea leaves, whereas the parent strain SST-5 produced lesions averaging only 69.7 mm² (Fig. 6B andD).
## DISCUSSION
In this study, a new mycovirus, tentatively named ScPV1, was identified and character ized from S. camellicola isolated from tea plants. According to the species demarcation criteria within the family Partitiviridae (less than 90% aa sequence identity in the RdRP (21), as well as its genomic organization and phylogenetic analysis, ScPV1 is proposed as a new member belonging to the newly proposed genus Epsilonpartitivirus in the family Partitiviridae. To date, five mycoviruses, namely, Colletotrichum camelliae filamentous virus 1 (CcFV1, belonging to the family Polymycoviridae), Pestalotiopsis theae chrysovirus 1 (PtCV1, Chrysoviridae), Pestalotiopsis fici hypovirus 1 (PfHV1, Hypoviridae), Melanco niella theae mitovirus 1 (MtMV1, Mitoviridae), and Didymella theifolia botybirnavirus 1 (DtBRV1, Botybirnaviridae) have been identified from phytopathogenic fungi infecting tea plants, including Colletotrichum camelliae, Pestalotiopsis theae, Pestalotiopsis fici, S. camellicola, and Didymella theifolia, respectively, while no partitiviruses have been reported in fungi infecting tea plants (22)(23)(24)(25)(26). To our knowledge, this is the first report of a partitivirus from a phytopathogenic fungus infecting tea plants. The genomic sequences of two dsRNA fragments in ScPV1, nominated dsRNAs1 and 2, were determined with sizes of 1,835 bp and 1,697 bp, respectively. These genomes conform to the genomic size range of the family Partitiviridae, whose members typically have 1 or 2 segments of 1.4-2.3 kbp per segment (27). Each ScPV1 dsRNA segment contains a single ORF, encoding RdRP and CP in dsRNA1 and dsRNA2, respectively, a common feature within the Partitiviridae family (28). Following purification of ScPV1 from strain WJT-1-1 and application of the same procedure for strain SST-5, icosahedral virus-like particles ca. 31 nm in diameter were visualized via TEM in the mycelia of strain WJT-1-1 exclusively (Fig. 3C). Additionally, dsRNAs 1 and 2 were concentrated in the 30% fraction of sucrose gradients where both virus-like particles and structural proteins were detected (Fig. 3C; Tables S3 andS4). Additionally, P55, which corresponds in size to the CP encoded by ScPV1 dsRNA2, was identified through SDS-PAGE and PMF analysis (Fig. 3B; Table S3). These confirm that ScVP1 dsRNAs 1 and 2 are encapsidated in 31 nm viral particles encapsidated by P55. The major structural protein encoded by ScPV1 dsRNA 2 is an icosahedral T = 1 CP consisting of 60 polypeptide subunits. These observations match previous results of investigations on partitivirus particle structure including Fusarium poae virus 1 and Penicillium stoloniferum viruses S and F, characterized by cryo-electron microscopy (cryo-EM), three-dimensional image reconstruction, and X-ray crystallogra phy (29)(30)(31). TEM observation demonstrated that ScPV1 consists of isometric particles 24.9-36.8 nm in diameter, fitting the size range (25-40 nm) for known members of the Partitiviridae that are variable (12). ScPV1 particles also contain a P100 protein that was detected following SDS-PAGE and PMF analysis. These studies revealed that P100 generated six peptide fragments that matched the deduced ScPV1 RdRP sequence at amino acid positions 64-217, accounting for 20% of the entire coverage. These results suggest that P100 is most likely the viral RdRP, which was significantly larger than the predicted molecular weight ScPV1 RdRP concentrated in sucrose gradients in the 30% fraction with a molecular weight of 65.5 kDa, possibly due to some processing modifications. Notably, it was observed at a significantly higher level in the 40% fraction as compared with the concentration of CP bands on SDS-PAGE gels (Fig. 3B). Typically, one or two RdRP molecules are packaged in each partitivirus particle (8), so the peptide concentration ratio should be ca. 60:1 or 30:1, respectively, for CP versus RdRp in partitiviruses. Multiple alignment of the ScPV1 RdRP sequence with the RdRP sequences of related dsRNA viruses indicates six conserved motifs, namely, motifs III to VIII, and the triplet "GDD" in Motif-VI, which is conserved in the RdRP sequences of +ssRNA and most dsRNA viruses (Fig. 2A) (32); however, some unique amino acids within Motifs IV, V, and VII were observed in the ScPV1 RdRP. Fungal partitiviruses are transmitted intracellularly during cell division, hyphal anastomosis, and sporogenesis, and alpha-and betapartitiviruses infecting fungal hosts from the genus Heterobasidion are transmitted via hyphal contacts between somatically incompatible host strains (7). Here, horizontal transmission was assessed using two different strains, WJT-1-1 and SST-5, but was unsuccessful. However, it occurred at a high frequency (over 90%) between exogenous strains SST-5 and its transfectant SST-5-T3. This suggests that horizontal transmission of ScPV1 is strain-dependent. To date, virus transfection of protoplasts with partitiviruses has been reported only in a few fungal species, including R. necatrix, Aspergillus fumigatus, and S. sclerotiorum (13,15,16,33). In this study, virus transfection was achieved using purified particles, and ScPV1 was shown to be transmitted at a low frequency of 5%, which facilitates a critical assess ment of the biological traits of ScPV1. Analysis of the transfectants revealed significant variation in the relative abundance ratios of the two ScPV1 genomic segments among different clones. However, all transfectants maintained consistently higher segment ratios compared to the original WJT-1-1 strain, where the two segments showed nearly equivalent band intensities. Similarly, some transfectable partitviruses changed their dsRNA1-to-dsRNA2 ratios before and after transfection (16,34). Multipartite (+)ssRNA viruses typically maintain a defined ratio of encapsidated genomic segments in their hosts (35), a pattern also observed in many partitiviruses (14,36). The genomic segment ratios may affect the biological properties of ScPV1.
Most of the viruses in the family Partitiviridae are latent in their host fungi, but some can attenuate fungal virulence and have potential as biological control agents. For example, ClPV1 reduces virulence and sporulation of Colletotrichum liriopes (37); CaPV1 significantly decreases host virulence, mycelial growth, appressorial development, and appressorium turgor but increases conidial production with an abnormal morphology (12). Additionally, a gammapartitivirus reduced C. acutatum sporulation (38). Nearly all investigations have concentrated on partitivirus-elicited hypovirulence rather than hypervirulence, which has only been reported rarely in any phytopathogenic fungi, but has been found to enhance the virulence of T. marneffei, which is a dimorphic clinical fungus causing systemic mycosis in Southeast Asia (17). To our knowledge, there are very limited reported cases linking mycoviruses to the hypervirulence of fungi. In one case, a 6.0 kbp dsRNA has been shown to enhance the virulence of N. radicicola when horizontally transmitted into fungal strains, while the fungal virulence decreased following elimination of the dsRNA (5). In the other case, LbQV-1 causes hypervirulence together with significant alterations in pigmentation and rapid growth, as demonstrated by comparisons between virus-infected and virus-cured isogenic lines of L. biglobosa (6). Besides, there are several hypervirul cases reported in entomopathogenic fungi, including a mycovirus infecting B. bassiana (39), and one gammapartitivirus named Metarhizium flavoviride partitivirus 1 in Metarhizium flavoviride (40). In this study, ScPV1 was transmitted to another fungal strain using protoplast transfection with purified virions. Most of the transfectants showed lower growth rates and hypervirulent traits as compared with those of the parent strain, indicating that ScPV1 confers two contra dictory effects: growth attenuation and virulence enhancement in S. camellicola. We also observed that two transfectants (SST-5-T2/T4) caused shorter lesions as compared with other transfectants, similar to those of the parental strain, which is a likely phe nomenon observed in fungi as intra-isolate variability in different SST-5 derivatives. For example, 34 subisolates of Botrytis porri strain Bc-72 infected by Botrytis porri RNA virus varied greatly both in mycelial growth rates (1.0 to 6.4 mm/day) on PDA and in leaf lesion lengths (0 to 3.6 mm) on garlic leaves (41). To rigorously evaluate ScPV1's biological effects, the virus was reintroduced into the ScPV1-free SST-5 strain through anastomosis using transfectant SST-5-T3 (which exhibited the most pronounced phenotypic alterations). All resulting subisolates consistently displayed reduced growth rates coupled with enhanced virulence compared to the parental strain (Fig. 6A through D). Thus, we provide compelling evidence to demonstrate the first instance of a mycovirus, specifically a partitivirus, conferring hypervirulence to a phytopathogenic fungus through protoplast transfection with mycoviral virions. Most mycoviruses usually confer synergistic effects by attenuating the virulence and reducing the growth to their fungal host, and only Alternaria alternata chrysovirus 1 (AaCV1) was reported to cause dual effects in its host fungus (Alternaria alternata), which was evaluated by comparison of several virus-high-titer and virus-low-titer isolates of AaCV1-bearing A. alternata, in accordance with a 13-fold increase in the AK-toxin level (42,43). Besides, it was demonstrated that Phytophthora infestans RNA virus 2 (PiRV-2) could stimu late sporangia production but reduced vegetative growth in oomycete Phytophthora infestans between the isogenic PiRV-2-infected and -free isolates that were generated via hyphal anastomosis (44). To our knowledge, this is the first instance of a mycovirus responsible for hypervirulence on a phytopathogenic fungus through virion transfection and the first case of a partitivirus conferring dual contradictory effects (hypervirulence but reduced growth) in a phytopathogenic fungus.
In summary, a novel mycovirus nominated ScPV1, which belongs to the newly proposed genus Epsilonpartitivirus within the family Partitiviridae, has been identified and characterized from S. camellicola isolated from tea plants. This marks the first report of a partitivirus from a phytopathogenic fungus infecting tea. ScPV1 contains two dsRNA segments that encode an RdRP and a CP, respectively, and these seg ments are encapsidated into isometric virions of different sizes. Additionally, the RdRP exhibits an unexpected size and contains some unique amino acids, indicating that the virus possesses unique molecular and morphological characteristics. Biological tests conducted with transfectants generated through protoplast transfection with purified virions rigorously assessed the effects of ScPV1 on its host and demonstrated that ScPV1 impairs S. camellicola growth while simultaneously enhancing its virulence. This represents the first report of a mycovirus causing hypervirulence and conferring dual effects in a phytopathogenic fungus as demonstrated through virion transfection. This study is expected to expand our understanding of viral diversity, evolution, and biological traits of the family Partitiviridae.
## MATERIALS AND METHODS
## Fungal strains and cultures
S. camellicola strains WJT-1 to WJT-27 (including WJT-1-1) were isolated from tea gardens across a village in Xuan'en County, Enshi Tujia and Miao Autonomous Prefecture, Hubei Province, China. Strains SST, SS-1 to SS-3, JYC-1-1 to JYC-1-4, and LLX were collected from various villages in Zigui County, Yichang City, Hubei Province. All strains were isolated from diseased tea leaves showing the typical leaf blight symptoms, with the exception of strain SST, which originated from an asymptomatic tea leaf. All fungal isolates were cultured on potato dextrose agar (PDA) comprising 20% [wt/vol] diced potatoes, 2% glucose, and 1.5% agar) medium at room temperature (-25°C) unless otherwise stated.
## Extraction and validation of strain dsRNA
For dsRNA extraction, mycelial plugs of the isolates were placed on sterilized cellophane disks on PDA plates and allowed to grow for 3-5 days. Frozen mycelial powder (0.5 g) was then suspended in SDS buffer. The column method was carried out to extract total RNA of the strain as previously described (45), which was eluted with 30 µL of RNase-free water. Residual DNA and ssRNA were eliminated from the extracted nucleic acids using 2U of DNase I (Simgen, Hangzhou, China) and 10 U of S1 nuclease (TaKaRa, Dalian, China) at 37°C for 1 h. Purified dsRNA (1 µL) was subjected to electrophoretic separation on 1.2% (wt/vol) agarose gels and Tris-acetate-EDTA (TAE) buffer. The dsRNA in individual fractions was stained with ethidium bromide. ScPV1 genomic dsRNAs were then excised from the gel, purified using the DNA Gel Extraction Kit (Simgen, Hangzhou, China), dissolved in RNase-free water, and stored at -70°C until use. dsRNA extraction to ascertain the presence of viral dsRNAs. The purified virion solution was adsorbed onto a copper mesh for 3 min, and then the mesh was placed on absorbent filter paper and allowed to dry for 2 min. Finally, the mesh was placed in a negative staining solution for 3 min and dried for 2 min for electron microscopic observation (51). The length of the observed viruses was measured using ImageJ software (52).
## SDS-polyacrylamide gel electrophoresis
Proteins isolated from distinct sucrose gradient fractions underwent comprehensive fractionation procedures utilizing 12% SDS-PAGE. This electrophoretic analysis was performed in a buffer system composed of 25 mM Tris/glycine and 0.1% SDS, which provided optimal conditions for protein separation. The resulting gels were subjected to staining with Coomassie brilliant blue R-250 (Bio-Safe CBB; Bio-Rad, USA), and individual protein bands were excised for PMF analysis by Sangon Biotech, Co., Ltd, in Shanghai, China (41).
## Analysis of dsRNAs on horizontal transmission
Dual cultures were used to determine the potential horizontal transfection of fungi with dsRNA viruses (53). Mycelial disks (5 mm diameter) from ScPV1-infected donor strains and virus-free recipient strains were co-cultured in paired combinations on 9 cm Petri dishes, maintained at 25°C with four or six biological replicates per treatment. After 7 days incubation, mycelial clusters were collected from the peripheral hyphal growth area of the recipient strain and sub-cultured onto PDA plates covered with sterilized cellophane. Following an additional 7 days of incubation at 25°C in a constant tempera ture incubator, the cultured mycelia were harvested and subjected to dsRNA extraction.
## Biological testing
Fungal morphologies and growth rates were evaluated as previously described (54). Virulence tests were conducted on fresh detached tea leaves (C. sinensis cv. "Echa 1") in six replicates. Briefly, tea leaves were subjected to a three-step washing process with sterile water, followed by air-drying, prior to being inoculated. The leaves were pricked three times with a sterilized needle (0.5 mm in diameter) and inoculated with a mycelial disk. The inoculated leaves were incubated at 25°C, and disease progress was recorded every 24 h. The lesion areas were measured using ImageJ software, and the data were recorded.
## Preparation of protoplasts and transfection of virus particles
Protoplasts were prepared from the fresh mycelium of the ScPV1-free S. camellicola strain SST-5 that were growing vigorously as described previously (55). Briefly, ca. 0.5 g mycelia were mixed with 10 mL NaCl buffer (700 mM NaCl, 0.1 g lysing enzymes from Tricho derma harzianum and 0.01 g snailase, followed by incubation for 3 h at 30°C with shaking at 100 rpm/min to obtain protoplasts. Subsequently, protoplasts were filtered using a Millipore filter and counted with a hemocytometer slide, and 2.0 × 10 6 protoplasts were used for each transfection. ScPV1 virions (ca. 70.0-80.0 µg) were transfected using PEG 6000 as previously described (56). Following transfection, protoplast suspensions were evenly spread on Bottom Agar plates (200 g of sucrose, 3 g of yeast extract per liter, 3 g casein acid hydrolysate, and 10 g agar) and incubated at 25°C.
## Statistical analysis
The biological data were statistically analyzed using SPSS Statistics Software 27.0 with one-way analysis of variance (ANOVA) and comparison of means. The mean values for the biological replicates are presented as column charts with error bars representing SEM. The graphs were generated using both MS Excel and GraphPad Prism 7 (a software program provided by GraphPad). P-values < 0.05 were considered to indicate statistical significance.
## References
1. Myers, James (2022) *Mycoviruses. Curr Biol*
2. Yu, Li, Fu et al. (2010) "A geminivirus-related DNA mycovirus that confers hypoviru lence to a plant pathogenic fungus" *Proc Natl Acad Sci*
3. Heiniger, Rigling (1994) "Biological control of chestnut blight in Europe" *Annu Rev Phytopathol*
4. Filippou, Garrido-Jurado, Meyling et al. (2018) "Mycoviral population dynamics in Spanish isolates of the entomopathogenic fungus Beauveria bassiana" *Viruses*
5. Ahn, Lee (2001) "A viral double-stranded rna up regulates the fungal virulence of Nectria radicicola" *MPMI*
6. Shah, Kotta-Loizou, Fitt et al. (2019) "Identification, molecular characterization, and biology of a novel quadrivirus infecting the phytopathogenic fungus Leptosphaeria biglobosa" *Viruses*
7. Vainio, Chiba, Ghabrial et al. (2018) "ICTV virus taxonomy profile: Partitiviridae" *J Gen Virol*
8. Nibert, Ghabrial, Maiss et al. (2014) "Taxonomic reorganization of family Partitiviridae and other recent progress in partitivirus research" *Virus Res*
10. Nerva, Silvestri, Ciuffo et al. (2017) "Transmission of Penicillium aurantiogriseum partiti-like virus 1 to a new fungal host (Cryphonectria parasitica) confers higher resistance to salinity and reveals adaptive genomic changes" *Environ Microbiol*
11. Jiang, Wang, Yang et al. (2019) "Molecular characterization of a debilitation-associated partitivirus infecting the pathogenic fungus Aspergillus flavus" *Front Microbiol*
12. Gilbert, Holcomb, Allscheid et al. (2019) "Hiding in plain sight: new virus genomes discovered via a systematic analysis of fungal public transcriptomes" *PLoS One*
13. Zhu, Qiu, Gao et al. (2024) "A novel partitivirus conferring hypovirulence by affecting vesicle transport in the fungus Colletotrichum"
14. Bhatti, Jamal, Petrou et al. (2011) "The effects of dsRNA mycoviruses on growth and murine virulence of Aspergillus fumigatus" *Fungal Genet Biol*
15. Vainio, Korhonen, Tuomivirta et al. (2010) "A novel putative partitivirus of the saprotrophic fungus Heterobasidion ecrustosum infects pathogenic species of the Heterobasidion annosum complex" *Fungal Biol*
16. Xiao, Cheng, Tang et al. (2014) "A novel partitivirus that confers hypovirulence on plant pathogenic fungi" *J Virol*
17. Chiba, Lin, Kondo et al. (2013) "A novel victorivirus from a phytopathogenic fungus, Rosellinia necatrix, is infectious as particles and targeted by RNA silencing" *J Virol*
18. Lau, Lo, Chow et al. (2018) "Novel partitivirus enhances virulence of and causes aberrant gene expression in Talaromyces marneffei" *mBio*
19. Wang, Hao, Lbinxiaowang et al. (2016) "Diverse Colletotrichum species cause anthracnose of tea plants"
20. Kuntze) In *Sci Rep*
21. Li, Zhu, Ren et al. (2023) "Characteristics and Pathogenicity of Discula theae-sinensis isolated from tea plant (Camellia sinensis) and interaction with Colletotrichum spp" *Plants (Basel)*
22. Guo, Zhao, Gao et al. (2024) "A phylogenetic and taxonomic revision of Discula theae-sinensis, the causal agents of anthracnose on Camellia sinensis" *JoF*
23. Arjona-Lopez, Telengech, Jamal et al. (2018) "Novel, diverse RNA viruses from Mediterranean isolates of the phytopathogenic fungus, Rosellinia necatrix: insights into evolutionary biology of fungal viruses" *Environ Microbiol*
24. Shafik, Umer, You et al. (2021) "Characterization of a novel mitovirus infecting Melanconiella theae isolated from tea plants" *Front Microbiol*
25. Zhou, Li, Kotta-Loizou et al. (2021) "A mycovirus modulates the endophytic and pathogenic traits of a plant associated fungus" *ISME J*
26. Jia, Dong, Zhou et al. (2017) "A dsRNA virus with filamentous viral particles" *Nat Commun*
27. Ye, Shi, He et al. (2023) "A novel botybirnavirus with a unique satellite dsRNA causes latent infection in Didymella theifolia isolated from tea plants" *Microbiol Spectr*
28. Han, Liu, Kong et al. (2023) "A special satellite-like RNA of a novel hypovirus from Pestalotiopsis fici broadens the definition of fungal satellite" *PLoS Pathog*
29. Peyambari, Habibi, Fotouhifar et al. (2014) "Molecular characterization of a novel putative partitivirus infecting Cytospora sacchari, a plant pathogenic fungus" *Plant Pathol J*
30. Nibert, Tang, Xie et al. (2013) "structures of fungal partitiviruses" *Adv Virus Res*
31. Pan, Dong, Lin et al. (2009) "Atomic structure reveals the unique capsid organization of a dsRNA virus" *Proc Natl Acad Sci*
32. Tang, Ochoa, Li et al. (2010) "Structure of Fusarium poae virus 1 shows conserved and variable elements of partitivirus capsids and evolutionary relationships to picobirnavirus" *J Struct Biol*
33. Ochoa, Havens, Sinkovits et al. (2008) "Partitivirus structure reveals a 120-subunit, helix-rich capsid with distinctive surface arches formed by quasisymmetric coat-protein dimers" *Structure*
34. Mahillon, Decroës, Caulier et al. (2021) "Genomic and biological characterization of a novel partitivirus infecting Fusarium equiseti" *Virus Res*
35. (2025) *Full-Length Text Journal of Virology*
36. Sasaki, Kanematsu, Onoue et al. (2006) "Infection of Rosellinia necatrix with purified viral particles of a member of Partitiviri dae (RnPV1-W8)" *Arch Virol*
37. Chiba, Salaipeth, Lin et al. (2009) "A novel bipartite double-stranded RNA mycovirus from the white root rot fungus Rosellinia necatrix: molecular and biological characterization, taxonomic considerations, and potential for biological control" *J Virol*
38. Marsh, Huntley, Pogue et al. (1991) "Regulation of (+):(-)-strand asymmetry in replication of brome mosaic virus RNA" *Virology (Auckl)*
39. Jurvansuu, Kashif, Vaario et al. (2014) "Partitiviruses of a fungal forest pathogen have species-specific quantities of genome segments and transcripts" *Virology (Auckl)*
40. Zhu, Guo, Hu et al. (2021) "A novel partitivirus that confer hypovirulence to the plant pathogenic fungus Colletotrichum liriopes" *Front Microbiol*
41. Zhong, Chen, Lei et al. (2014) "Detection and characterization of a novel gammapartitivirus in the phytopathogenic fungus Colletotrichum acutatum strain HNZJ001" *Virus Res*
42. Kotta-Loizou, Coutts (2017) "Studies on the virome of the entomopathogenic fungus Beauveria bassiana reveal novel dsRNA elements and mild hypervirulence" *PLoS Pathog*
43. Guo, Zhang, Wu et al. (2024) "Transfection of entomopathogenic Metarhizium species with a mycovirus confers hypervirulence against two lepidopteran pests" *Proc Natl Acad Sci*
44. Wu, Zhang, Yang et al. (2012) "Characterization of a novel bipartite double-stranded RNA mycovirus conferring hypoviru lence in the phytopathogenic fungus Botrytis porri" *J Virol*
45. Okada, Ichinose, Takeshita et al. (2018) "Molecular characterization of a novel mycovirus in Alternaria alternata manifesting two-sided effects: Down-regulation of host growth and up-regulation of host plant pathogenicity" *Virology (Auckl)*
46. Fuke, Takeshita, Aoki et al. (2011) "The presence of double-stranded RNAs in Alternaria alternata Japanese pear pathotype is associated with morphological changes" *J Gen Plant Pathol*
47. Cai, Fry, Hillman (2019) "PiRV-2 stimulates sporulation in Phytophthora infestans" *Virus Res*
48. Fu, Zhang, Yin et al. (2022) "A novel heptasegmented positivesense single-stranded RNA virus from the phytopathogenic fungus Colletotrichum fructicola" *J Virol*
49. Wang, Zhao, Hong et al. (2017) "Characterization of causal agents of a novel disease inducing brownblack spots on tender tea leaves in China" *Plant Dis*
50. Liu, Fu, Jiang et al. (2009) "A novel mycovirus that is related to the human pathogen hepatitis E virus and rubi-like viruses" *J Virol*
51. Kumar, Stecher, Tamura (2016) "MEGA7: molecular evolutionary genetics analysis version 7.0 for bigger datasets" *Mol Biol Evol*
52. Xiang, Fu, Hong et al. (2017) "Characterization of a novel botybirnavirus isolated from a phytopathogenic Alternaria fungus" *Arch Virol*
53. Wang, Jiang, Wang et al. (2014) "Hypovirulence of the phytopathogenic fungus Botryosphaeria dothidea: association with a coinfecting chrysovirus and a partitivirus" *J Virol*
54. Miller (2001) "Microwave processing techniques for biological samples in a service laboratory"
55. Schneider, Rasband, Eliceiri (2012) "NIH Image to ImageJ: 25 years of image analysis" *Nat Methods*
56. Zhang, Nuss (2008) "A host dicer is required for defective viral RNA production and recombinant virus vector RNA instability for a positive sense RNA virus" *Proc Natl Acad Sci*
57. Fu, Crous, Bai et al. (2019) "Colletotrichum species associated with anthracnose of Pyrus spp" *Persoonia*
58. Kanematsu, Arakawa, Oikawa et al. (2004) "A reovirus causes hypovirulence of Rosellinia necatrix" *Phytopathology*
59. Yoo, Cho, Sheen (2007) "Arabidopsis mesophyll protoplasts: a versatile cell system for transient gene expression analysis" *Nat Protoc*
60. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12614646&blobtype=pdf | # Spike mutations that affect the function and antigenicity of recent KP.3.1.1-like SARS-CoV-2 variants
Bernadeta Dadonaite, Sheri Harari, Brendan Larsen, Lucas Kampman, Alex Harteloo, Anna Elias-Warren, Helen Chu, Jesse Bloom
## Abstract
SARS-CoV-2 is under strong evolutionary selection to acquire mutations in its spike protein that reduce neutralization by human polyclonal antibodies. Here, we use pseudovirus-based deep mutational scanning to measure how mutations to the spike from the recent KP.3.1.1 SARS-CoV-2 strain affect cell entry, binding to the ACE2 receptor, RBD up/down motion, and neutralization by human sera and clinically relevant antibodies. The spike mutations that most affect serum antibody neutraliza tion sometimes differ between sera collected before versus after recent vaccination or infection, indicating that these exposures shift the neutralization immunodominance hierarchy. The sites where mutations cause the greatest reduction in neutralization by post-vaccination or infection sera include receptor-binding domain (RBD) sites 475, 478, and 487, all of which have mutated in recent SARS-CoV-2 variants. Multiple mutations outside the RBD affect sera neutralization as strongly as any RBD mutations by modulat ing the RBD up/down movement. Some sites that affect RBD up/down movement have mutated in recent SARS-CoV-2 variants. Finally, we measure how spike mutations affect neutralization by three clinically relevant SARS-CoV-2 antibodies: VYD222, BD55-1205, and SA55. Overall, these results illuminate the current constraints and pressures shaping SARS-CoV-2 evolution and can help with efforts to forecast possible future antigenic changes that may impact vaccines or clinical antibodies.
IMPORTANCEThis study measures how mutations to the spike of a SARS-CoV-2 variant that circulated in early 2025 affect its function and recognition by both the polyclonal antibodies produced by the human immune system and monoclonal antibodies used as prophylactics. These measurements are made with a pseudovirus system that enables safe study of viral protein mutations using virions that can only infect cells once. The study identifies mutations that decrease recognition by current human antibody immunity; many of these mutations are increasingly being observed in new viral variants. It also shows the importance of mutations that move the spike's receptor-binding domain up or down. Overall, these results are useful for forecasting viral evolution and assessing which newly emerging variants have reduced recognition by immunity and antibody prophylactics.
predict which of these lineages have mutations that will enable them to be evolutionarily successful.
Deep mutational scanning is a powerful approach to measure how spike mutations affect key functional and antigenic properties of spike (2,9,(12)(13)(14), but the fact that both spike (8,15,16) and human population immunity (17)(18)(19)(20) are constantly evolving limits the utility of measurements made using older strains and human antibodies for understanding newer variants. Here, we use pseudovirus-based deep mutational scanning (2,21) to measure how thousands of mutations to the spike of the recentKP.3.1.1 variant affect cell entry, receptor binding, RBD up/down motion, and neutralization by human sera and therapeutic antibodies. Overall, our work provides detailed maps of the functional and antigenic effects of spike mutations that can help rationalize recent trends in SARS-CoV-2 evolution and identify mutations that affect key protein properties.
## RESULTS
## Pseudovirus-based deep mutational scanning of KP.3.1.1 spike
To measure how mutations in the SARS-CoV-2 spike affect cell entry, receptor binding, and escape polyclonal sera or therapeutic antibodies, we used pseudovirus-based deep mutational scanning (Fig. 1A) (2,21). This method produces genotype-phenotype-linked lentiviral particles that encode uniquely barcoded spike variants and can be used to measure the effects of mutations on different spike phenotypes (21) (Fig. S1A). Because these pseudoviruses are restricted to a single round of infection and require helper plasmids to produce viral particles, they cannot cause disease or transmit in humans, making them a safe tool for characterizing mutations in viral proteins at biosafety level 2.
We designed pseudovirus-based deep mutational scanning libraries for the spike protein from the recently circulating KP.3.1.1 strain. KP.3.1.1 is a descendant of the JN.1 lineage and was one of the major variants circulating from the second half of 2024 to early 2025 (22). Its spike shares many important antigenic mutations with the other current JN.1 descendant strains and is closely related to the spikes currently recommen ded as options for inclusion in SARS-CoV-2 vaccines (JN.1, KP.2, and LP.8.1) (Fig. 1B) (23).
We designed the deep mutational scanning libraries to contain all evolutionarily accessible and antigenically important mutations in the spike protein. Specifically, we included all mutations that have occurred at appreciable frequency during the SARS-CoV-2 evolution in humans, as well as every possible amino acid change at sites that have mutated frequently in recent variants and all sites within the RBD. We produced two independent pseudovirus libraries (Lib-1 and Lib-2), which contained 42,783 and 45,513 barcoded variants, respectively, and covered 95% of the 9,809 targeted amino-acid mutations with an average of 1.3 mutations per spike (Fig. S1B andC).
## Mutation effects on spike-mediated cell entry
We measured how mutations to KP.3.1.1 spike affect entry into 293T cells that were engineered to express medium levels of the ACE2 receptor (24) (Fig. 2A and interactive heat map at https://dms-vep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/cell_entry.html). The measured effects of mutations on cell entry were highly correlated between the two independent libraries (Fig. S1D). As expected, the stop codons were highly deleterious for cell entry, whereas amino acid mutations had varied effects (Fig. 2A). Single-residue deletions were well tolerated at many sites in the N-terminus domain (NTD), consistent with frequent NTD deletions in many circulating SARS-CoV-2 variants (25) (Fig. 2A). Amino acid mutations in the RBD had a range of effects, with some sites intolerant of mutations but others tolerant of many changes. Note that the measurements of mutation effects on cell entry here were made using only a single cell line, and some mutations may have different effects on cell entry in other cell lines due to differences in, for example, receptor or protease expression (26,27).
Our measurements suggest a possible reason why certain mutations have begun to recurrently evolve in recent JN.1-descended strains related to KP.3.1.1 after being rare in earlier variants. A number of these mutations-specifically T22N, K182R, G184S, F186L, R190S, A435S, and N487D-are better tolerated for cell entry in the KP.3.1.1 spike compared with the earlier pre-JN.1 XBB.1.5 lineage (Fig. 2B), as assessed by comparing our current deep mutational scanning to prior measurements for the XBB.1.5 spike (2). Therefore, shifts in mutational tolerance for specific mutations may be a contributor to the recent recurrent selection for these mutations.
## Mutation effects on ACE2 binding
To determine how spike mutations affect receptor binding, we measured how well each spike mutant pseudovirus was neutralized by soluble monomeric ACE2 protein (Fig. S2A). We and others have previously shown that ACE2 binding affinity to spike is proportional to neutralization of SARS-CoV-2 pseudovirus by soluble ACE2 protein (2,28,29). Namely, mutations that increase spike's binding to ACE2 also increase pseudovirus neutralization by soluble ACE2 protein, and mutations that decrease spike's ACE2 binding decrease pseudovirus neutralization by soluble ACE2. Therefore, incubating deep mutational scanning libraries with increasing amounts of monomeric ACE2 protein allows us to measure how mutations affect ACE2 binding. Note that this approach only works for spike mutants that retain at least some moderate ability to mediate pseudovirus entry in ACE2-expressing cells. Among the spike mutations that retain sufficient cell entry function, effects on cell entry and ACE2 binding show no correlation (Fig. S2B), demon strating that cell entry and ACE2 binding are distinct phenotypes, and ACE2 binding is often not the limiting factor for cell entry in our assays.
A variety of mutations both in the RBD and other regions of spike affect ACE2 binding, as measured by soluble ACE2 neutralization (Fig. 3A and B and inter active heatmap at https://dms-vep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/receptor_bind ing.html). The substantial effect of some mutations outside the RBD on ACE2 binding is because the interaction of the full spike with ACE2 is impacted by several distinct mechanisms: direct interaction of the RBD with ACE2, changes in RBD up (open) or down (closed) conformation, and changes to S1 shedding (30)(31)(32)(33). Interestingly, we measure mutations at sites distant to the RBD's ACE2 binding motif to have as large effects on ACE2 binding as mutations at sites in close proximity to ACE2, emphasizing the importance of conformational changes to spike in affecting ACE2 binding (Fig. 3B). Many ACE2 distal RBD mutations with the strongest binding effects are at sites near the base of the RBD in spike, suggesting their likely involvement in positioning the RBD in the up or down conformation (e.g., sites 332, 358, 390, 393, 395, 517, and 527; Fig. 3A andB; Fig. S2C). Among the sites in proximity to ACE2, certain mutations at site E493 cause the largest increase in receptor binding (Fig. 3A andB). Notably, site 493 interacts with ACE2 directly, recently substituted from Q to E in parents of KP.3.1.1 and several other current lineages, and has been previously shown to epistatically interact with two other recent mutations also present in KP.3.1.1 (L455S and F456L) (7,15).
There is a good correlation between the effects of RBD mutations on ACE2 binding in our KP.3.1.1 deep mutational scanning and similar data previously published for the XBB.1.5 spike (2) (Fig. 3C). However, there are some mutations with different effects on ACE2 binding in KP.3.1.1 and XBB.1.5, including A435S, which increases binding to ACE2 in KP.3.1.1 but decreases binding for XBB.1.5 (note this mutation also had contrasting effects on cell entry in the two spikes as described above) (Fig. 3D). The A435S mutation has been rare for most of SARS-CoV-2′s evolution but has recently occurred independently in multiple lineages including the JN. BA.3.2. In addition, E493D and E493N increase ACE2 binding by the KP.3.1.1 spike, but in XBB.1.5, mutating site 493 from its initial identity of Q to any of E, D, or N impairs ACE2 binding (Fig. 3D) (2,34).
## Mutation effects on serum neutralization
We measured how spike mutations affect neutralization by sera collected from seven human individuals pre-and post-exposure by vaccination or infection with JN.1descendant variants (Table S1). All seven individuals were adults who had originally been imprinted by vaccination with the early COVID-19 vaccine in 2021, followed by various further booster vaccinations and infections. For most (although not all) of these individuals, exposure to a JN.1-descendant spike via vaccination increased neutralizing serum titers against KP.3.1.1 (Fig. S3A andB). We used the pseudovirus libraries to measure how the KP.3.1.1 spike mutations affected neutralization by the sera from each individual, both pre-and post-vacci nation or infection with a JN-1 descendant spike. For the most part, mutations had similar effects on neutralization by sera from each individual collected preversus post-vaccination or infection (Fig. 4; Fig. S3C). Across all sera, the sites where mutations caused the most escape from serum neutralization were primarily in the RBD (Fig. 4 and interactive plot at https://dms-vep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/ polyclonal_sera_escape.html). RBD mutations at sites 332, 344, 357, 393, 428, 458, 470, and 518 caused the greatest serum escape both pre-and post-vaccination or infection (Fig. 4A). Some sites outside the RBD also reduced serum neutralization, including sites 50, 132, 200, and 222 in NTD, 572 in SD1, and 852 in S2. Notably, most of the sites where mutations caused the greatest escape in the RBD and all the strongest sites of escape outside the RBD are ones where mutations affect ACE2 binding (Fig. 3B, and next section), suggesting that mutations at these sites impact serum neutralization largely changing the RBD's up/down conformation, thereby indirectly affecting binding by antibodies targeting potent neutralizing epitopes on the RBD (2,(35)(36)(37). However, there are also some sites of appreciable escape where mutations do not affect RBD up/ down binding (e.g., 456, 458, 475, 478, and 487); these mutations likely directly escape binding by neutralizing antibodies rather than affecting RBD up/down conformation.
Although many mutations that reduce serum neutralization pre-and post-vaccina tion or infection were shared among the different sera, in a subset of individuals, exposure to a JN.1-descendant spike clearly shifts neutralization immunodominance. In adults 1, 3, 4, and 5, several RBD sites where mutations had little or no effect on serum neutralization before JN.1-descendant spike exposure became the dominant escape sites after vaccination or infection (Fig. 4B). These new escape sites include 403, 405, 475, 478, 487, 490, and 505. Notably, in circulating SARS-CoV-2 variants, many of these sites have recently acquired mutations that reduce serum neutralization. For example, the XFJ, JN. We validated the deep mutational scanning measurements of how mutations affect serum neutralization using standard SARS-CoV-2 pseudovirus neutralization assays (Fig. S4) (38). The deep mutational scanning measurements correlated well with changes in IC50 values measured in the standard neutralization assays (Fig. S4A). We also confirmed via standard neutralization assays that mutations A475V, H505E, K478I, and N487D cause a larger reduction in the neutralization by the serum from some individuals after versus before exposure to a JN.1-descendant spike (Fig. S4B), consistent with the deep mutational scanning.
## Sites where mutations affect the RBD up/down conformation
To identify sites in spike that affect RBD up/down conformation, we leveraged the previously noted fact that mutations at these sites have opposing effects on ACE2 binding and serum antibody neutralization escape: namely, mutations that put the RBD more in the up conformation increase ACE2 binding and also enhance neutralization (2,(35)(36)(37). Our measurements for the KP.3.1.1 spike show this relationship clearly: there is a strong inverse correlation between serum neutralization escape and ACE2 binding for mutations that affect both these phenotypes but are distal from the RBD's ACE2-binding motif (Fig. 5A). This inverse correlation is due to the fact that positioning RBD in the up conformation reveals the receptor-binding motif, which mediates binding to ACE2 but is also targeted by many potent neutralizing antibodies. Therefore, mutations that put the RBD more in the up conformation sensitize the spike to serum neutralization (negative escape values in our measurements), whereas mutations that put the RBD more in the down conformation tend to cause serum neutralization escape. By contrast, ACE2-proximal sites show no correlation between ACE2 binding and serum neutraliza tion (Fig. 5A) because they often both interact with the receptor directly and are directly targeted by neutralizing serum antibodies. Note that some ACE2 proximal sites may still modulate the RBD up/down conformation, but this modulation does not lead to the aforementioned consistent pattern on ACE2 binding and neutralization because the direct effects of mutations at these sites, both ACE2 binding and neutralizing antibody binding, can overwhelm the effect of the RBD up/down conformation modulation.
To estimate how much each site affects RBD up/down conformation, we cal culated the correlation between serum neutralization escape and ACE2 binding at each site, weighting it by the root mean square effect of mutations at each site on both phenotypes (Fig. 5B and interactive plot at https://dms-vep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/RBD_movement.html). Among the sites that stand out as strongly affecting RBD up/down conformation are many clade-defining mutations as well as some of the most frequently mutated sites through various periods of SARS-CoV-2 evolution in humans. Site 222 was one of the most frequently mutated sites just before Omicron emerged (39), sites 371 and 373 fixed mutations in all Omicron lineages (40), and sites 332, 356, and 570 fixed mutations in the BA.2.86 lineage, which is the ancestor of most currently circulating strains (41). The prevalence of mutations at sites that modulate RBD up/down conformation in major SARS-CoV-2 lineages suggests a strong selective pressure to balance receptor binding with resistance to neutralization by RBD-directed antibodies; indeed, evidence suggests that multiple recent SARS-CoV-2 variants have acquired mutations that position the RBD in a more closed conformation (36,42).
## Effects of mutations on neutralization by clinically relevant monoclonal antibodies
We next determined how mutations to spike affect neutralization by three clinically relevant monoclonal antibodies: BD55-1205 (12), SA55 (43), and VYD222 (44, 45) (Fig. 6). BD55-1205 and SA55 have maintained high neutralizing potency against currently circulating variants (12). SA55 is in clinical trials in China (41,46), BD55-1205 is licensed to Moderna Inc. (12), and VYD222 is currently the only SARS-CoV-2 antibody authorized for use in the USA for pre-exposure prophylaxis in immunocompromised individuals (it is the antibody in Pemivibart) (47). Knowledge of which mutations reduce neutralization by these antibodies is important for ongoing surveillance, as all other clinically approved SARS-CoV-2 antibodies have now been escaped by viral mutations (43,48). All three antibodies bind to sites around the RBD's receptor binding motif, with SA55 and VYD222 sharing especially similar structural epitopes (43,49). Our deep mutational scanning shows that all three antibodies are strongly affected by mutations at several sites in the range from 500 to 505, although the exact sites in this range where mutations have the most impact vary among the antibodies (Fig. 6). BD55-1205 neutralization is also affected by mutations at sites 456, 475, and 493, all of which interact with ACE2 (Fig. 6A). Most changes to 456 and 475 sites are deleterious for ACE2 binding (see letter colors in logoplots in Fig. 6), although A475V, which measurably escapes BD55-1205, is only mildly deleterious for ACE2 binding and has recently occurred in several JN.1-descendant lineages. SA55 is affected by mutations to sites 440 and 445, and to a lesser degree by mutations at 493 (Fig. 6B). VYD222 is also affected by mutations at site 440 in addition to mutations at sites 405 and 403 (Fig. 6C). Because sites 500-505 are primarily accessible in the RBD's up position, all three antibodies are affected by mutations that modulate RBD up/down movement, as has been noted previously (12,50,51). In particular, some mutations at sites 332, 357, and 427 affect neutralization by all three antibodies to various degrees, despite the fact that none of these sites are in the direct structural epitopes, presumably by putting the RBD more in the down conformation and so partially shielding the antibody epitopes. Interestingly, in our pseudovirus deep mutational scanning, mutations at site 505 cause significantly more escape from all three of the antibodies than previously published yeast-based RBD-only deep mutational scanning data suggest (Fig. S5) (6,43,45). We hypothesize that this difference is because RBD-only assays measure just the direct effects of mutations on antibody-RBD binding, whereas the pseudovirus deep mutational scanning also measures the impacts of mutations on RBD up/down movement that affect RBD epitope accessibility in the context of the full spike. Indeed, mutations at RBD motion-regulating sites 332, 357, and 427 affect neutralization by all three antibodies in full-spike but not in yeast-based RBD-only deep mutational scanning (Fig. S5). Similarly, mutations at site 505 both directly affect antibody-RBD binding and the up/down motion of the RBD due to this site's location in the inter-protomer interface in the down RBD spike conformation. Although site 505 is likely under significant evolutionary constraint because most mutations at that site reduce ACE2 binding, our serum-escape measurements described above suggest that this site may be starting to come under appreciable pressure for mutations from population immunity.
## DISCUSSION
Here, we have measured how mutations to the KP.3.1.1 spike affect several distinct phenotypes: cell entry, ACE2 binding, serum neutralization, RBD up/down motion, and neutralization by key monoclonal antibodies. These measurements provide several important insights into the selection pressures and molecular constraints currently shaping SARS-CoV-2 evolution.
First, our measurements underscore the substantial impact of mutations that affect RBD up/down motion on receptor binding and antibody neutralization. In the context of the full spike, mutations that affect RBD up/down motion impact ACE2 binding as much as mutations at RBD sites that interact with ACE2 directly. Mutations that affect RBD up/ down motion have a consistent signature: they have opposite effects on ACE2 binding and serum neutralization, since putting the RBD more up increases the accessibility of the receptor-binding motif to bind ACE2 but also makes it more susceptible to RBD-tar geting neutralizing antibodies (33,52). Many sites that affect RBD up/down motion have mutated in major lineages during the course of SARS-CoV-2 evolution in humans, emphasizing the importance of the balancing effects of RBD up/down movement on viral fitness via impacts on ACE2 binding and serum neutralization. Note that mutations that put the RBD in a more up conformation may promote the cross-species transfer of coronaviruses by increasing binding to receptors from new species (32,53,54); it appears that the spike of the first SARS-CoV-2 strains identified in humans had the RBD in a relatively more up conformation, and subsequent evolution has selected for mutations that position the RBD more down (36,42).
Second, our measurements identify sites where mutations cause the largest reductions in neutralization by human serum antibodies; there are already newly emerging viral lineages that carry some of these mutations. Most of the sites where mutations have the most impact on serum neutralization are in the RBD, as expected from prior work showing that RBD-directed antibodies are usually responsible for most serum neutralizing activity (55)(56)(57), although mutations at some NTD sites also have a substantial effect. Some of the top RBD sites of serum antibody escape are likely directly in the epitopes of neutralizing antibodies that sterically block receptor binding (e.g., sites 456, 458, 475, 478, and 487); mutations at some of these sites have recently been observed in new SARS-CoV-2 lineages. However, mutations at NTD and RBD sites that affect RBD up/down motion and so affect serum neutralization indirectly by conformational masking often have as much impact on serum neutralization as direct escape mutations in key RBD epitopes. As mentioned above, some of these up/down affecting sites have mutated in major lineages; however, such conformational escape is constrained by the fact that mutations that reduce serum neutralization by putting the RBD in a more down conformation also reduce ACE2 binding, and hence, they need to be buffered by other ACE2 affinity-increasing mutations.
Third, we find that exposure to a JN.1-descendant spike (via vaccination or infection) often shifts the neutralization immunodominance hierarchy to new epitopes. Specifically, for some individuals, vaccination or infection with a JN.1-descendant variant leads to mutations at new sites causing large reductions in neutralization; these new sites include several (e.g., 475, 478, and 487) that have acquired mutations in very recent SARS-CoV-2 lineages. Our data cannot determine the underlying mechanism responsible for this shift in serum neutralizing specificity. Once individuals have been imprinted by SARS-CoV-2 infection or vaccination, most of the neutralizing response to subsequent vaccinations and infections is driven by activation of pre-existing cross-reactive B cells (28,(58)(59)(60)(61). However, the affinity maturation of these pre-existing B cells can shift the balance of epitope targeting in polyclonal sera (59). In addition, sufficient exposures to new variants can activate naive B cells (61,62). The shifts in serum neutralizing specificity we observe after exposure to a JN.1-descendant variation could be due to some combination of boosting of pre-existing cross-reactive B-cells that were previ ously subdominant, affinity maturation of pre-existing B-cells to better target recently mutated epitopes, or activation of naive B-cells targeting new epitopes. Regardless of the underlying mechanism, the fact that exposure to recent variants changes the neutralization immunodominance hierarchy supports the idea that updating vaccines to more recently circulating variants (23) can shift the specificity of neutralizing antibodies to target newer SARS-CoV-2 variants.
The fact that exposure to recent JN.1-descendant variants can shift which spike mutations affect neutralization by the serum antibodies of imprinted adults highlights the increasing heterogeneity in antibody immunity across the human population. We recently showed that the epitopes targeted by the neutralizing antibodies of young children who had experienced just a single infection with a recent variant differ dramatically from those targeted by adults imprinted by infection or vaccination early in the SARS-CoV-2 pandemic (20). The current study only examined serum from imprinted adults, but it found heterogeneity even among such adults depending on whether they have been exposed to a JN.1-descendant variant. This increasing immune heterogeneity across the population may favor more co-circulation of multiple SARS-CoV-2 lineages rather than repeated rapid evolutionary sweeps by a single variant (63,64).
We also mapped how mutations affect neutralization by three clinically relevant monoclonal antibodies (BD55-1205, SA55, and VYD222) that have so far retained neutralizing activity against nearly all SARS-CoV-2 lineages (65,66). A major epitope targeted by all these antibodies is the 500-505 loop in the RBD, which has not mutated in any major lineage since the emergence of Omicron in 2021. Notably, these antibodies target functionally constrained RBD epitopes that overlap with the ACE2 binding motif and are only fully accessible in the up RBD conformation, and our data show that neutralization by all three antibodies is reduced by mutations that put the RBD in a more down conformation. In particular, mutations to site 505, which both affect RBD motion and form part of the epitope for all three antibodies, have a greater impact on pseudovi rus neutralization than was apparent in prior RBD-only yeast-display deep mutational scanning (6). Site 505 remains under substantial constraint, since most mutations at that site both reduce direct RBD-ACE2 binding affinity (15) and put the RBD in a more up conformation that increases its susceptibility to RBD-directed serum neutralizing antibodies. However, our results show that site 505 is now a serum neutralization escape mutation for some individuals who have been exposed to a JN.1-descendant variant, suggesting that such individuals now produce appreciable neutralizing antibodies directly targeting site 505. Therefore, site 505 might be under increasing pressure to mutate in circulating SARS-CoV-2 lineages, although additional changes to spike would likely be needed to overcome the pleiotropic effects such a mutation would have on ACE2 binding and RBD up/down conformation.
## MATERIALS AND METHODS
## Deep mutational scanning library design
Deep mutational scanning libraries were designed to cover all possible mutations in the RBD, and all tolerated and frequently mutated changes outside the RBD. To identify the tolerated and frequently mutated sites, we included mutations that occur more than 50 times among SARS-CoV-2 genomes deposited on GISAID (67), mutations that occur at least 10 times on UShER (68) spike phylogenetic tree, any mutation present in a recent SARS-CoV-2 lineage (at the time of library design these lineages were BA.2.86, JN.1, JN.1.11.1, and KP.3), and any mutations that occurred at least once in a Pango designated lineage (69). In addition, we introduced all possible amino-acid mutations at sites that fit any of the following criteria: mutated at least 50 times in a recent SARS-CoV-2 lineage, mutated along UShER spike phylogenetic tree at least 2,500 times, mutated repeatedly at least three times among any Pango-designated lineages, or had mutated in the KP.3.1.1 variant relative to the Wuhan-Hu-1 sequence. The above criteria were also applied for deletions, but deletions were only included if they were present at any site in the NTD or positions 331-354 or 434-508 in the RBD. Several mutations and sites to saturate were also included manually in library design, regardless of their frequency counts, based on reports of these mutations occurring in circulating lineages at the time of library design. The list of manually included mutations, as well as parameters for all other selection criteria, is at https://github.com/dms-vep/ SARS-CoV-2_KP.3.1.1_spike_DMS/blob/main/library_design/config.yaml. The full list of all mutations included in the library design is at https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS/blob/main/library_design/results/mutations_to_make.csv.
## Overview of library construction using Golden Gate assembly
Golden Gate assembly was used to create KP.3.1.1 spike coding plasmid libraries containing all the designed mutations (70-76) (Fig. S6). Due to the length of the spike sequence and the number of mutations we wanted to include in the library, it was cost-prohibitive to synthesize the spike gene as a single fragment for all spike variants we wanted to include. We therefore subdivided spike into 17 overlapping tiles between 250 and 290 nt in length (close to the maximum length that can be synthesized by Twist Bioscience as a single-stranded DNA [ssDNA] oligo pool) (Fig. S6A), computationally designed a pool of oligos, where each oligo is one of the tiles with a mutation we wanted to include in the library (Fig. S6B), and ordered all the oligos pooled together as ssDNA fragments from Twist Bioscience. From that ssDNA pool, we performed 17 individual PCR reactions to amplify oligos belonging to each tile using primers with flanking sequences containing BsmBI restriction sites (Fig. S6C). Golden Gate assembly was then used to assemble each tile pool and flanking spike sequences unique to each tile into a shuttle vector (Fig. S6D). The assembled shuttle vector pool was electroporated into bacteria, and the next day, plasmids were recovered for all 17 pools. The full spike sequence was amplified from each pool using primers with flanking sequences that match the lentiviral backbone as well as a barcode sequence in the reverse primer (Fig. S6E). All 17 barcoded spike pools were then pooled equimolarly, and HiFi assembly was used to clone the library (Fig. S6F), which, after pooling, had all designed mutations throughout the spike into a lentivirus backbone.
The sequence of the codon-optimized KP. data/KP311_GAA_assembly_fragments.csv. Tiles were designed manually, making sure that the overhangs for the fragments that will be assembled during the Golden Gate assembly step are unique for each fragment and have a sequence compatible with high fidelity assembly (77). The 1st and the 17th tile overlapped with a pGGAselect DNA shuttle vector that is provided in NEBridge Golden Gate Assembly Kit (BsmBI-v2) (E1602L). The oligo pool was designed using a script available at https://github.com/jbloomlab/gga_codon_muts_oligo_design. The script reads in tile sequences and the desired mutation spreadsheet and generates a fasta file with oligo sequences that can be uploaded directly for ordering an oligo pool from Twist Biosciences. We set the oligo design script to intentionally include 0.005 fraction of unmutated sequences for each tile in order to have some wildtype KP.3.1.1 spike in the final pseudovirus library, as well as avoid any mutation design that would introduce BsmBI cut sites. Sequences for designed oligos cover ing all 17 tiles are at https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS/blob/ main/library_design/results/mutagenesis_oligos.fa. A GitHub repository that selects the mutations to be included in the library and designs mutated oligos for each tile is at https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS/tree/main/library_design.
## Deep mutational scanning plasmid library cloning using Golden Gate assembly
To amplify individual tile pools from one ssDNA oligo pool, we performed 17 PCR reactions. For each reaction, we used KOD Hot Start Master Mix (Sigma-Aldrich, Cat. No. 71842), 0.3 µM of forward and reverse primer, and 2 ng of ssDNA oligo pool. Each reaction was started at 95°C for 2 min and then underwent 23 cycles of 95°C for 20 s, 62°C for 10 s, and 68°C for 25 s. To amplify flanking spike sequences for each tile, we used KOD Hot Start Master Mix, 0.3 µM of forward and reverse primer, and 1 ng of KP.3.1.1 spike coding lentiviral backbone (see above section for plasmid map). The full list for forward and reverse primers used in both reactions is at https://github.com/dms-vep/ SARS-CoV-2_KP.3.1.1_spike_DMS/blob/main/library_design/data/primers.csv. Expected size products were gel and Ampure XP bead purified (1:3 DNA to bead).
We then performed Golden Gate assembly using NEBridge Golden Gate Assembly Kit (BsmBI-V2). For the assembly, we used 100 fmol of amplified tile pool and flanking spike sequence fragments each and 50 fmol of pGGAselect shuttle plasmid (provided in NEBridge Golden Gate Assembly Kit). The assembly reactions were incubated at 42°C for 1 min, followed by 16°C for 1 min for 30 cycles, followed by 60°C for 5 min. The reactions were then purified using Ampure XP beads and eluted in 20 µL of water. In addition, 1 µL of purified assembly was then used to electroporate NEB 10-beta Electrocompetent E. coli cells (C3020K). Electroporated cells were then suspended in 1 mL of recovery media and shaken at 37°C for 1 h. After recovery, the cells were spun down, the recovery media were removed, and the cells were resuspended in chloramphenicol-con taining LB media for incubation at 37°C with shaking overnight. High transformation efficiency (~1 million colonies per tile library) was confirmed by diluting a small amount of recovered cells, plating on chloramphenicol-containing agar plates overnight, and counting colony-forming units the next day. High transformation efficiency at this and later steps is important to avoid any barcode duplication at later virus production steps due to lentivirus recombination. Note that here and in later electroporation steps, we used liquid cultures to amplify our plasmid libraries as opposed to high-density spread on bacterial culture plates, we used in the past, as this has been shown to be sufficient for a uniform plasmid amplification (78). After overnight growth, shuttle plasmid libraries for each tile were recovered using QIAprep Spin Miniprep Kit (Cat. No. 27106 × 4).
Next, the spike libraries for each tile were amplified and barcoded. We per formed PCR on each tile plasmid library using KOD Hot Start Master Mix, 10 ng of plasmid library, and 0.3 µM of forward (5′-gcacgcgCAGCCGAGCCACATCGCTCA-3′) and reverse (5′-gcggaactccactaggaacatttctctctcgaaTCTAGANNNNNNNNNNNNNNNNA GATCGGAAGAGCGTCGTGTAGGGAAAGAG-3′) primers; the latter primer contained a 16 nt barcode. After amplification, each spike tile library was purified by gel and Ampure XP beads. Note that gel purification at this step is important because we found cloning of some tiles produces a minor amount of truncated spike, and gel purification allowed us to recover only the full-length products. All barcoded spike libraries were then pooled equimolarly. We made two equimolar pools of barcoded spike libraries to make library-1 and library-2 biological replicates. All subsequent steps in library production were done in parallel for library-1 and library-2. NEBuilder HiFi DNA Assembly Master Mix (E2621S) was then used to assemble barcoded spikes into a lentivirus backbone, as described previously (21). See lentivirus backbone structure in Fig. S6F; plasmid for the backbone is available at Addgene #204579. Assembled backbones were electroporated into electrocompetent bacteria, and plasmids were amplified using liquid culture, as described above. As before, we confirmed high electroporation efficiency at this step and cultured at least 10 million colony-forming units per library replicate.
## Production of cell-stored deep mutational scanning libraries
To produce the cell-stored deep mutational scanning libraries, we used a method described previously (Fig. S1A) (21). In brief, we first used lentivirus backbones that carried barcoded spike libraries to produce VSV-G pseudotyped viruses. To do so, we transfected two 6-well plates of 293T cells with lentivirus helper plasmids (BEI: NR-52517, NR-52519, NR-52518) and VSV-G expression plasmid (Addgene #204156); 48 h after transfection, we collected VSV-G pseudotyped viruses from cell supernatant and used them to infect 293T-rtTA cells at low multiplicity of infection (<0.01), so that most infected cells were infected with only one viral variant. We then used puromycin to select for successfully transduced cells. The transduced cell library pool was then expanded and frozen at >15 M cells per aliquot in liquid nitrogen until further use.
## Long-read sequencing for variant-barcode linkage
To build a variant to the barcode lookup table for the deep mutational scanning libraries, we rescued VSV-G pseudotyped viruses from the cell-stored libraries. We use VSV-G pseudotyping at this stage to rescue all virus variants from the cells, regardless of how deleterious a mutation in spike may be. To do so, we transfected library cells with lentivirus helper plasmids and VSV-G expression plasmid, and 48 h after transfection, we recovered VSV-G pseudoviruses from cell supernatant, purified them from cell debris using a 0.45 µm SFCA Nalgene 500 mL Rapid-Flow filter unit (Cat. No. 09-740-44B), and concentrated using Pierce Protein Concentrator (ThermoFisher, 88537). We then used ~ 10 million transcription units of VSV-G pseudotyped viruses to infect 293T cells, and 15 h after infection, we recovered non-integrated viral genomes using the QIAprep Spin Miniprep Kit. We then performed two rounds of PCR to amplify the barcoded spikes in the recovered lentivirus genomes, minimizing the number of PCR cycles to avoid strand-switching. Long-read circular consensus sequencing was performed on amplified virus genomes using the PacBio Sequel IIe machine. Consensus sequence for each variant was determined using at least 2 CCSs per barcode. Variant-barcode lookup table for both biological KP.3.1.1 library replicates is at https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS/blob/main/results/variants/codon_variants.csv. On average, each variant had 1.25 and 1.27 mutations per spike for library-1 and library-2, respec tively.
## Measurement of mutation effects on cell entry effect
KP.3.1.1 spike pseudotyped viruses were produced from cell-stored libraries as described previously (2); 150 million library cells were plated into 5-layer flasks (Corning Falcon 875 cm² Rectangular Straight Neck Cell Culture Multi-Flask, Cat. No. 353144) in the presence of 1 µg/mL of doxycycline to induce spike expression from the TRE3G promoter in the lentivirus backbone. The next day, the cells were transfected with 50 µg of each lentiviral helper plasmid, and during transfection, the cell media were replaced with fresh serum-free media (Opti-MEM supplemented with 0.1% heat-inactivated FBS, 0.3% bovine serum albumin, 100 µg/mL of calcium chloride, 100 U/ml penicillin, and 100 µg/mL streptomycin). Serum-free media were used because they allowed better virus concentration in protein columns, as FBS tends to clog column filters; 48 h after transfection, the cell supernatant was collected, purified from cell debris, and concentra ted using protein columns. Protein column concentrated virus titers varied between 12 and 25 million transcription units per milliliter. VSV-G-pseudotyped viruses were also produced in parallel to spike pseudotyped libraries, using the protocol described in the section above. For cell entry effect measurements, both 3 million transcription units of spike pseudotyped libraries and 10 million transcription units of VSV-G-pseudotyped libraries were used to infect medium-ACE2 (24) cells and 293T cells, respectively. For spike-pseudotyped library infections, the cells were plated in the presence of 2.5 µg/mL of amphotericin B (Sigma, Cat. No. A2942), which we have shown in the past increa ses virus titers (21); 15 h after infection, non-integrated viral genomes were recovered using the QIAprep Spin Miniprep Kit, and amplicon libraries were prepared for Illumina sequencing as described previously using dual indexing for each sample to avoid index hopping on certain sequencing platforms (21). Sequencing was performed on NovaSeq X Plus and NextSeq 2000 platforms.
Mutation effects on cell entry were calculated using log enrichment ratio: log 2 n v post / n wt post / n v pre / n wt pre , where n v post is variant count post-infection (spike pseudotyped virus infection), n v pre is variant count pre-infection (VSV-G pseudotyped virus infection), and n wt post and n wt pre are unmutated variant counts postand pre-infection. The multidms (79) package was used to fit global epistasis models (80) on variant effect data to estimate the effects of individual mutations from the full libraries of both singly and multiply mutated spike variants. The values reported here are the median across the measurements with all replicates of both libraries.
## Measurement of mutation effects on receptor binding
To measure how mutations to spike affect ACE2 binding, we used soluble monomeric ACE2. Monomeric ACE2 was produced as described previously (2). First, we mixed 1.5 million transcription units of spike pseudotyped library virus per sample with RDPro pseudotyped virus at 1%-2% of total transcription units used. Use and produc tion of RDPro pseudotyped virus were described previously (2). RDPro is used in our experiments as a non-neutralizable standard to convert sequencing counts to fractional neutralization of each variant at each ACE2 concentration as described previously (2). The library virus was then mixed with increasing concentrations of soluble monomeric ACE2 and incubated at 37°C for 30 min. The ACE2 concentrations were selected such that they would cover most of the KP.3.1.1 spike pseudotyped virus neutralization range in order to identify mutations that both increase (spike variants that are neutralized well at low ACE2 concentrations) and decrease (spike variants that are neutralized at high ACE2 concentrations) ACE2 binding; specific concentrations used were 6, 13, 27, 54, and 115 µg/mL. After incubation, the libraries were used to infect medium-ACE2 cells in the presence of 2.5 µg/mL of amphotericin B, and 15 h post-infection, non-inte grated viral genomes were recovered and prepared for Illumina sequencing as described previously (21). After converting the sequencing counts to the fractional neutralization using the non-neutralized RDPro standard (2), we analyzed the data using a biophysical model implemented in the polyclonal software package (https://github.com/jbloomlab/ polyclonal) (81) to determine the effect of each mutation on ACE2 neutralization, reporting the values such that positive effects indicate improved ACE2 binding (higher neutralization by soluble ACE2). We performed ACE2 binding experiments with both library-1 and library-2 biological replicates. The values reported here are the median across both replicates. Mutations' effects on ACE2 binding are shown at https://dmsvep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/receptor_binding.html.
## Measurement of mutation effects on serum and antibody neutralization
Before performing sera and antibody selection experiments with deep mutational scanning libraries, we determined their potency by performing pseudovirus neutraliza tion assays on viruses pseudotyped with KP.3.1.1 spike. Pseudovirus neutralization assays were performed as described previously (38) and in "Standard pseudovirus neutralization assays, " below. Before use, all sera were inactivated for 1 h at 56°C.
For each sample, 1.5 million transcription units of spike pseudotyped library virus were mixed with RDPro pseudotyped virus at 1%-2% of total transcription units used. For each serum, we performed selection at three concentrations aiming to neutralize more than 60% of library variants in at least two of these concentrations. Our start ing serum dilution was twice the IC99 value as determined by standard pseudovirus neutralization, which typically significantly underestimates neutralization achieved for deep mutational scanning (perhaps due to differing amounts of spike on the surface of pseudoviruses used in standard neutralization assay versus library virus, or depletion of antibody molecules by the higher virion concentration in the library experiments). An example of neutralization achieved by different serum concentrations can be seen here https://dms-vep.org/SARS-CoV-2_KP.3.1.1_spike_DMS/notebooks/avg_escape_anti body_escape_adult-1_pre_vaccination.html in the probability escape plots. Generally, serum escape probabilities > 0.4 allow identification of mutations that affect serum neutralization. Antibodies were used in the following concentrations: BD55-1205, the concentrations were 0.73, 2.18, and 6.55 µg/mL; for SA55, 0.32, 0.95, and 2.84 µg/mL; and for VYD222, 100, 300, and 900 µg/mL. In standard pseudovirus neutralization assays, all these concentrations were above the IC99 value, but in deep mutational scanning data, these ranged between IC50 and IC99 for BD55-1205, IC5 and IC75 for SA55, and IC94 and IC99 for VYD222. After incubation, virus mixtures were used to infect medium-ACE2 cells in the presence of 2.5 µg/mL of amphotericin B, and 15 h post-infection, non-inte grated viral genomes were recovered and prepared for Illumina sequencing as described previously (21).
To determine mutations that affect serum or antibody neutralization, we used a biophysical model from the polyclonal (v6.16) package (81), which is implemented in dms-vep-pipeline-3 (v3.27.0) https://github.com/dms-vep/dms-vep-pipeline-3/tree/main.
Mean and individual sera escape plots and links to raw numeric escape val ues for each sera are at https://dms-vep.org/SARS-CoV-2_KP.
## Estimate of mutation effects on RBD up/down motion
To quantify a site's effect on RBD up/down motion, we used the following formula:
where R is Pearson correlation between mutation effects on serum escape (averaged across all sera) and ACE2 binding for site s. Positive R values were set to zero and then
$$Site effect on RBD motion = R s × -1 × 1 n s i = 1 n escape s, i 2 × 1 n s i = 1 n binding s, i2$$
## Standard pseudovirus neutralization assays
Desired mutations were cloned into KP.3.1.1 spike expres sion plasmid https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS/blob/main/ KP311_validation_notebooks/plasmid_maps/HDM_KP.3.1.1.gb, and the sequence was confirmed using whole plasmid sequencing. Spike pseudotyped lentiviruses were rescued by transfecting 293T cells with spike expression plasmids, Gag/Pol (BEI: NR-52517) helper plasmid, and pHAGE6_Luciferase_IRES_ZsGreen backbone; 48 h post-transfection, virus-containing cell supernatants were collected and titrated. Neutralization assays were performed as described in Crawford et al. (38) using medium-ACE2 cells (24) in the presence of 2.5 µg/mL of amphotericin B. For all neutralization assays, the starting dilution was 0.05, and we performed eight 3-fold serial dilutions. Fraction infectivity at each dilution was determined relative to serum-free controls, and the neutcurve (V2.1.0) package (82) was used to fit Hill curves to fraction infectivity data.
## Antibody production
Antibodies were ordered from Genscript Biotech using published variable sequences (12,43,44,83). Variable sequences and completely expressed polypeptide sequences are specified in Table S2. These sequences were codon-optimized, cloned into expression vectors, and expressed in Chinese hamster ovary-derived cells. Heavy chain variable sequences were cloned into a human IgG1 backbone. The light chain variable sequences for BD55-1205 and SA55 were cloned into a human kappa light chain backbone; VYD222 was cloned into a human lambda light chain backbone.
## Cells
293T, 293T-rtTA, medium-ACE2, and cell-stored library cells were all grown in D10 medium (Dulbecco's modified Eagle medium with 10% heat-inactivated fetal bovine serum, 2 mM l-glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin). For the deep mutational scanning library and 293T-rtTA cells, tetracycline-free FBS was used. Medium-ACE2 cells were grown in the presence of 2 µg/mL doxycycline, which induced ACE2 expression in these cells.
The computational analysis pipeline used to analyze deep mutational scanning data and make all associated figures is on GitHub at https://github.com/dms-vep/SARS-CoV-2_KP.3.1.1_spike_DMS. The GitHub repository is archived on Zenodo DOI: https:// doi.org/10.5281/zenodo.17282308. Sequencing data associated with this manuscript have been deposited to the SRA under BioProject PRJNA1305008.
## References
1. Bloom, Neher (2023) "Fitness effects of mutations to SARS-CoV-2 proteins" *Virus Evol*
2. Dadonaite, Brown, Mcmahon et al. (2024) "Spike deep mutational scanning helps predict success of SARS-CoV-2 clades" *Nature*
3. Lamb, Luka, Saathoff et al. (2024) "Mutational signature dynamics indicate SARS-CoV-2's evolutionary capacity is driven by host antiviral molecules" *PLoS Comput Biol*
4. Greaney, Starr, Bloom (2022) "An antibody-escape estimator for mutations to the SARS-CoV-2 receptor-binding domain" *Virus Evol*
5. Greaney, Loes, Gentles et al. (2021) "Antibodies elicited by mRNA-1273 vaccination bind more broadly to the receptor binding domain than do those from SARS-CoV-2 infection" *Sci Transl Med*
6. Jian, Wang, Yisimayi et al. (2025) "Evolving antibody response to SARS-CoV-2 antigenic shift from XBB to JN.1" *Nature*
7. Feng, Sun, Zhang et al. (2024) "Structural and molecular basis of the epistasis effect in enhanced affinity between SARS-CoV-2 KP.3 and ACE2" *Cell Discov*
8. Starr, Greaney, Hannon et al. (2022) "Shifting mutational constraints in the SARS-CoV-2 receptor-binding domain during viral evolution" *Science*
9. Starr, Greaney, Hilton et al. (2020) "Deep mutational scanning of SARS-CoV-2 receptor binding domain reveals constraints on folding and ACE2 binding" *Cell*
10. Ozono, Zhang, Ode et al. (2021) "SARS-CoV-2 D614G spike mutation increases entry efficiency with enhanced ACE2-binding affinity" *Nat Commun*
11. Khatri, Siddqui, Sadhu et al. (2023) "Intrinsic D614G and P681R/H mutations in SARS-CoV-2 VoCs Alpha, Delta, Omicron and viruses with D614G plus key signature mutations in spike protein alters fusogenicity and infectivity" *Med Microbiol Immunol*
12. (2025) *Full-Length Text Journal of Virology*
13. Jian, Wec, Feng et al. (2025) "Viral evolution prediction identifies broadly neutralizing antibodies to existing and prospective SARS-CoV-2 variants" *Nat Microbiol*
14. Lei, Qing, Odle et al. (2024) "Functional and antigenic characterization of SARS-CoV-2 spike fusion peptide by deep mutational scanning" *Nat Commun*
15. Tan, Mou, Lei et al. (2003) "High-throughput identification of prefusion-stabilizing mutations in SARS-CoV-2 spike" *Nat Commun*
16. Taylor, Starr (2024) "Deep mutational scanning of SARS-CoV-2 Omicron BA.2.86 and epistatic emergence of the KP.3 variant" *Virus Evol*
17. Jian, Feng, Yang et al. (2023) "Convergent evolution of SARS-CoV-2 XBB lineages on receptor-binding domain 455-456 synergistically enhances antibody evasion and ACE2 binding" *PLoS Pathog*
18. Meijers, Ruchnewitz, Eberhardt et al. (2023) "Population immunity predicts evolutionary trajectories of SARS-CoV-2" *Cell*
19. Suntronwong, Assawakosri, Klinfueng et al. (2025) "Age associated SARS-CoV-2 immune responses provide insights into population immunity over four years since the COVID-19 pandemic" *Sci Rep*
20. Raharinirina, Gubela, Börnigen et al. (2025) "SARS-CoV-2 evolution on a dynamic immune landscape" *Nature*
21. Dadonaite, Burrell, Logue et al. (2025) "SARS-CoV-2 neutralizing antibody specificities differ dramatically between recently infected infants and immuneimprinted individuals" *J Virol*
22. Dadonaite, Crawford, Radford et al. (2023) "A pseudovirus system enables deep mutational scanning of the full SARS-CoV-2 spike" *Cell*
23. Roemer, Neher (2024) "SARS-CoV-2_variant-reports_2024-12-20"
24. (2025) "Statement on the antigen composition of COVID-19 vaccines"
25. Farrell, Dadonaite, Greaney et al. (2022) "Receptorbinding domain (RBD) antibodies contribute more to SARS-CoV-2 neutralization when target cells express high levels of ACE2" *Viruses*
26. Cantoni, Murray, Kalemera et al. (2022) "Evolutionary remodelling of N-terminal domain loops fine-tunes SARS-CoV-2 spike" *EMBO Rep*
27. Bayati, Kumar, Francis et al. (2021) "SARS-CoV-2 infects cells after viral entry via clathrin-mediated endocytosis" *J Biol Chem*
28. Jackson, Farzan, Chen et al. (2022) "Mechanisms of SARS-CoV-2 entry into cells" *Nat Rev Mol Cell Biol*
29. Cao, Jian, Wang et al. (2023) "Imprinted SARS-CoV-2 humoral immunity induces convergent Omicron RBD evolution" *Nature*
30. Wang, Iketani, Li et al. (2022) "Antigenic characterization of the SARS-CoV-2 Omicron subvariant BA.2.75" *Cell Host Microbe*
31. Wang, Nair, Anang et al. (2021) "Functional differences among the spike glycoproteins of multiple emerging severe acute respiratory syndrome coronavirus 2 variants of concern"
32. Walls, Xiong, Park et al. (2019) "Unexpected receptor functional mimicry elucidates activation of coronavirus fusion" *Cell*
33. Tse, Acreman, Lax et al. (2024) "Distinct pathways for evolution of enhanced receptor binding and cell entry in SARS-like bat coronaviruses" *PLoS Pathog*
34. Uğurlu, Salminen, Yuan et al. (2024) "Altered spike dynamics drives immune evasion of SARS-CoV-2"
35. Taylor, Starr (2023) "Deep mutational scans of XBB.1.5 and BQ.1.1 reveal ongoing epistatic drift during SARS-CoV-2 evolution" *PLoS Pathog*
36. Wang, Mellis, Ho et al. (2024) "Recurrent SARS-CoV-2 spike mutations confer growth advantages to select JN.1 sublineages" *Emerg Microbes Infect*
37. Yajima, Anraku, Kaku et al. "Genotype to Phenotype Japan"
38. Consortium, Ito, Maenaka et al. (2024) "Structural basis for receptor-binding domain mobility of the spike in SARS-CoV-2 BA.2.86 and JN.1" *Nat Commun*
39. Wang, Guo, Mellis et al. (2025) "Antibody evasiveness of SARS-CoV-2 subvariants KP.3.1.1 and XEC" *Cell Rep*
40. Crawford, Eguia, Dingens et al. (2020) "Protocol and reagents for pseudotyping lentiviral particles with SARS-CoV-2 spike protein for neutralization assays" *Viruses*
41. Hodcroft, Zuber, Nadeau et al. (2021) "Spread of a SARS-CoV-2 variant through Europe in the summer of 2020" *Nature*
42. Zhao, Zhou, Tian et al. (2022) "Omicron SARS-CoV-2 mutations stabilize spike up-RBD conformation and lead to a non-RBM-binding monoclonal antibody escape" *Nat Commun*
43. Yang, Guo, Wang et al. (2024) "Structural basis for the evolution and antibody evasion of SARS-CoV-2 BA.2.86 and JN.1 subvariants" *Nat Commun*
44. Zhang, Lindenberger, Parsons et al. (2024) "SARS-CoV-2 Omicron XBB lineage spike structures, conformations, antigenicity, and receptor recognition" *Mol Cell*
45. (2025) *Full-Length Text Journal of Virology*
46. Cao, Jian, Zhang et al. (2022) "Rational identification of potent and broad sarbecovirusneutralizing antibody cocktails from SARS convalescents" *Cell Rep*
47. Ison, Popejoy, Evgeniev et al. (2023) "Efficacy and safety of adintrevimab (ADG20) for the treatment of high-risk ambulatory patients with mild or moderate coronavirus disease 2019: results from a phase 2/3, randomized, placebo-controlled trial (STAMP) conducted during delta predominance and early emergence of Omicron" *Open Forum Infect Dis*
48. Yuan, West, Foreman et al. (2025) "Structural and functional analysis of VYD222: a broadly neutralizing antibody against SARS-CoV-2 variants" *bioRxiv*
49. "SINOVAC reports on the clinical trial progress of its anti-COVID-19 antibody SA55 injection-SINOVAC: supply vaccines to eliminate human diseases"
50. Invivyd (2025) "Invivyd announces FDA authorization for emergency use of PEMGARDA (formerly VYD222) for pre-exposure prophylaxis (PrEP) of COVID-19 -Invivyd"
51. Cao, Wang, Jian et al. (2022) "Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies" *Nature*
52. Rappazzo, Tse, Kaku et al. (2021) "Broad and potent activity against SARS-like viruses by an engineered human monoclonal antibody" *Science*
53. Powers, Williams, Kreher et al. (2024) "Neutralization of recent SARS-CoV-2 variants by genetically and structurally related mAbs of the pemivibart lineage" *bioRxiv*
54. Wang, Guo, Ho et al. (2024) "Activity of research-grade pemivibart against recent SARS-CoV-2 JN.1 sublineages" *N Engl J Med*
55. Liu, Iketani, Guo et al. (2022) "Striking antibody evasion manifested by the Omicron variant of SARS-CoV-2" *Nature*
56. Tse, Lasso, Mclellan et al. (2025) "Bat sarbecovi rus WIV1-CoV bears an adaptive mutation that alters spike dynamics and enhances ACE2 binding" *bioRxiv*
57. Li, Cai, Li et al. (2025) "Structural and functional constraints on spike activation and host protease utilization limit cell entry of SARS-CoV-2-related bat coronaviruses" *J Virol*
58. Greaney, Loes, Crawford et al. (2021) "Comprehensive mapping of mutations in the SARS-CoV-2 receptor-binding domain that affect recognition by polyclonal human plasma antibodies" *Cell Host Microbe*
59. Piccoli, Park, Tortorici et al. (2020) "Mapping neutralizing and immunodominant sites on the SARS-CoV-2 spike receptor-binding domain by structure-guided high-resolution serology" *Cell*
60. Premkumar, Segovia-Chumbez, Jadi et al. (2020) "The receptor binding domain of the viral spike protein is an immunodo minant and highly specific target of antibodies in SARS-CoV-2 patients" *Sci Immunol*
61. Liang, Raju, Liu et al. (2024) "Imprinting of serum neutralizing antibodies by Wuhan-1 mRNA vaccines" *Nature*
62. Kaku, Starr, Zhou et al. (2023) "Evolution of antibody immunity following Omicron BA.1 breakthrough infection" *Nat Commun*
63. Addetia, Piccoli, Case et al. (2023) "Neutralization, effector function and immune imprinting of Omicron variants" *Nature*
64. Yisimayi, Song, Wang et al. (2024) "Repeated Omicron exposures override ancestral SARS-CoV-2 immune imprinting" *Nature*
65. Schiepers, Van 't Wout, Greaney et al. (2023) "Molecular fate-mapping of serum antibody responses to repeat immunization" *Nature*
66. Roemer, Neherra (2025) "SARS-CoV-2 variant reports"
67. Barrat-Charlaix, Neher (2024) "Eco-evolutionary dynamics of adapting pathogens and host immunity. Elife 13:RP97350"
68. Planas, Staropoli, Planchais et al. (2024) "Escape of SARS-CoV-2 variants KP.1.1, LB.1, and KP3.3 from approved monoclonal antibodies" *Pathog Immun*
69. Guo, Yu, Liu et al. (2025) "Antigenic and virological characteristics of SARS-CoV-2 variants BA.3.2, XFG, and NB.1.8.1" *Lancet Infect Dis*
70. Khare, Gurry, Freitas et al. (2021) "GISAID's role in pandemic response" *China CDC Wkly*
71. Turakhia, Thornlow, Hinrichs et al. (2021) "Ultrafast Sample placement on Existing tRees (UShER) enables real-time phylogenetics for the SARS-CoV-2 pandemic" *Nat Genet*
72. O'toole, Pybus, Abram et al. (2022) "Pango lineage designation and assignment using SARS-CoV-2 spike gene nucleotide sequences" *BMC Genomics*
73. Hanning, Walker, Beijerling et al. (2025) "Simple high-throughput encoding of deep mutational scanning libraries by oligo-based Golden Gate assembly" *bioRxiv*
74. Jann, Gagnon-Arsenault, Pageau et al. (2025) "Making deep mutational scanning accessible: a costefficient approach to construct barcoded libraries for genes of any length" *bioRxiv*
75. Álvarez-Rodríguez, Velandia-Álvarez, Toft et al. (2024) "Mapping the mutational landscape of a full viral proteome reveals distinct profiles of mutation tolerability" *bioRxiv*
76. Bakhache, Orr, Mccormick et al. (2024) "Uncovering structural plasticity of enterovirus A through deep insertional and deletional scanning"
77. Macdonald, Nedrud, Grimes et al. (2023) "DIMPLE: deep insertion, deletion, and missense mutation libraries for exploring protein variation in evolution, disease, and biology" *Genome Biol*
78. Coyote-Maestas, Nedrud, Okorafor et al. (2020) "Targeted insertional mutagenesis libraries for deep domain insertion profiling" *Nucleic Acids Res*
79. Nedrud, Coyote-Maestas, Schmidt (2021) "A large-scale survey of pairwise epistasis reveals a mechanism for evolutionary expansion and specialization of PDZ domains" *Proteins*
80. Potapov, Ong, Kucera et al. (2018) "Comprehensive profiling of four base overhang ligation fidelity by T4 DNA ligase and application to DNA assembly" *ACS Synth Biol*
81. Mateyko, De Boer (2024) "Culture wars: empirically determining the best approach for plasmid library amplification" *ACS Synth Biol*
82. Haddox, Galloway, Dadonaite et al. (2023) "Jointly modeling deep mutational scans identifies shifted mutational effects among SARS-CoV-2 spike homologs"
83. Otwinowski, Mccandlish, Plotkin (2018) "Inferring the shape of global epistasis" *Proc Natl Acad Sci*
84. Yu, Thornton, Hannon et al. (2022) "A biophysical model of viral escape from polyclonal antibodies" *Virus Evol*
85. Loes, Tarabi, Huddleston et al. (2024) "High-throughput sequencing-based neutralization assay reveals how repeated vaccinations impact titers to recent human H1N1 influenza strains" *J Virol*
86. (2024) "Emergency use authorization (EUA) for PEMGARDA center for drug evaluation and research (CDER) review memorandum"
87. (2025) "Length Text Journal of Virology" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12845124&blobtype=pdf | # No Evidence of Direct Transmission of Emerging Bluetongue Virus Strains Between Israel and Europe Based on Genomic Analyses (2013-2023)
Natalia Golender, Eyal Klement, Bernd Hoffmann
## Abstract
Bluetongue (BT) is an arthropod-borne viral disease primarily affecting domestic and wild ruminants. In recent years, several BTV serotypes and genotypes have been detected in Israel almost annually, raising questions about their origin and routes of introduction. Some BTV serotypes closely related to those first identified in Israel, including BTV-3, BTV-8, and BTV-12, were subsequently reported in Europe after a delay of several years. In this study, we sequenced the complete genomes of one representative strain of all newly identified Israeli BTV genotypes/serotypes-BTV-1, -4, -5, -8, and -11-first detected between 2021 and 2023. Additionally, complete sequences of enzootic Israeli BTV ( 2015) and eleven BTV-3 strains (2019-2023), with two representative strains for every year of isolation, except 2021 (three strains), were analyzed using phylogenetic, BLAST, and pairwise identity approaches. Genetic analyses revealed that recently identified Israeli and European BTV strains share common African ancestors, with some genomic "incursions" from Mayotte Island or the Arabian Peninsula. These incursions appeared more frequently in Israeli than in European strains. Nevertheless, nucleotide sequence differences of at least 2-3% across all genes indicate several years of independent evolution. The observed divergence suggests that no direct transmission of BTV occurred between Israel and Europe during the past decade.
## 1. Introduction
Bluetongue (BT) is a widely distributed viral disease of domestic and wild ruminants caused by bluetongue virus (BTV) [1]. There is no public health risk associated with BT [2,3]. Since BT causes substantial economic and animal health impacts, the World Organization for Animal Health (WOAH) classifies BT as a notifiable disease of global concern [3].
Severe clinical manifestations, often resulting in acute death, are primarily observed in sheep and certain wildlife species, such as white-tailed deer. The course of BT in sheep can vary from peracute to chronic, with a mortality rate ranging from 2% to 90%. Peracutely affected animals may die within 7-9 days of infection, most often due to severe pulmonary edema leading to dyspnea, frothing from the nostrils, and death by asphyxiation [2].
Infection in cattle is uncommon but may resemble that observed in sheep [4]. The clinical disease is usually characterized by fever, increased respiratory rate, lacrimation, hypersalivation resulting from oral vesicles and ulcers, stiffness or lameness with inflammation at the junction of the skin and coronary band, hyperesthesia, vesicular and ulcerative dermatitis, ocular and nasal discharge, swelling of the mouth, head, and neck, respiratory distress, and abortion [5]. Infection in goats is generally subclinical or causes only mild clinical signs [6].
Economic losses caused by BT are associated with direct and indirect costs. Direct costs are included in production losses, which are mostly linked with mortality, morbidity, reduced milk yield, abortions, and reduced fertility rate. Indirect losses are estimated at the national level, which are associated with control and surveillance costs, trade restrictions, vaccination, diagnosis, and vector monitoring [7].
BTV is a member of the Orbivirus genus within the family Sedoreoviridae. It is characterized by a double-stranded RNA genome divided into ten segments of different sizes [8,9]. The BTV genome consists of 10 linear double-stranded RNA segments (Seg-1 to Seg-10) encoding seven structural (VP1-VP7) and five nonstructural (NS1-NS5) proteins [10]. Like other segmented viruses, BTV undergoes genetic reassortment during co-infections, allowing the exchange of genome segments between strains [11]. The BTV outer-capsid proteins VP2 and VP5, encoded by genome segments 2 and 6 (Seg-2 and Seg-6), respectively, play key roles in cell attachment and entry during the early stages of infection. These proteins-particularly VP2-also contain epitopes that bind neutralizing antibodies generated during infection of the mammalian host. Differentiation within serotypes and between different serotypes is now based on both classical methods (serum neutralization tests) and molecular approaches (sequence analysis) [12]. Currently, thirty-six distinct BTV serotypes have been officially recognized based on sequence differences in the Seg-2 gene and virus neutralization tests [13,14].
Transmission of most BTV serotypes between mammalian hosts relies on competent, blood-feeding midges of the Culicoides species [1]. Vertical and horizontal transmissions have been described and/or hypothesized, but they are considered to be of less epidemiological importance [15][16][17][18][19][20].
Officially, the history of BT in Europe begins with the identification of BTV-3 in Cyprus in 1943, although it was possibly introduced into the country earlier and clinically diagnosed as "stomatitis" as early as 1924 [21,22]. The extent and severity of BTV outbreaks in Cyprus apparently varied from year to year, with the most virulent occurring in 1924, 1939, 1943, 1951, 1965, and 1977 [23]. Before 1998, BTV caused only two outbreaks elsewhere in Europe: a major outbreak of BTV-10 on the Iberian Peninsula between 1956 and 1960 [24] and a smaller outbreak of BTV-4 on several Greek islands in the Aegean in 1979-1980 [25]. BTV-4, closely related to these Greek BTV-4 strains but differing in several internal genes, circulated in the Mediterranean Basin (Cyprus and Turkey) between 1979 and 2004 [26]. This situation changed dramatically from 1998, when BTV-9 was detected on several Greek islands and spread rapidly throughout the Mediterranean Basin (Turkey, Bulgaria, Serbia, Montenegro, Kosovo, and Macedonia) [27], followed by incursions of BTV-1, BTV-2, BTV-4, and BTV-16 during 2000-2005 in Italy, Turkey, Greece, Balkan countries, and France (Corsica) [23]. Since 2006, a BTV-1 strain of Algerian origin has expanded northwards across the Iberian Peninsula and France [28]. In August 2006, BTV crossed latitude 50 • N for the first time, and BT outbreaks caused by BTV-8 occurred in the Netherlands, Belgium, Germany, France, and Luxembourg. In 2008, BTV-6 and BTV-11 were detected in Germany, the Netherlands, and Belgium [29,30]. On this occasion, mass vaccination campaigns implemented in Europe in the spring of 2008 quickly limited the spread of disease caused by BTV-8, and it was eradicated by 2011. However, after a three-year hiatus, in September 2015, BTV-8 re-emerged in central France and subsequently spread throughout the entire country. The renewed outbreak of this BTV-8 strain was most likely caused by the use of cryopreserved virus-positive semen or embryos. In the following years, BTV-8 outbreaks were reported in Switzerland, Germany, Belgium, and Spain [30][31][32].
In 2023, a new genetically distinct BTV-8 strain was also discovered in France [9]. In addition to BTV-8, serotypes 1, 2, 4, 9, and 16 have recently circulated in Europe [31]. The history of the current incursions of BTV-3 in Europe begins with its identification in Italy [33]. However, the spread of BTV-3 through continental Europe was registered in September 2023, when clinical disease was observed on four sheep farms in the Netherlands [34]. Cases of BTV-3 have since spread across 13 European countries (Netherlands, Belgium, Germany, UK, Denmark, France, Luxembourg, Switzerland, the Czech Republic, Portugal, Sweden, Austria, and Norway) [35].
Most recently, in October 2024, BTV-12 was detected in cattle and sheep in the Netherlands [36] and later in England (United Kingdom) [37].
A schematic map showing the first detection of each BTV serotype in Europe over the past decade is presented in Figure 1. The information was summarized based on recent publications and annual WOAH reports [31,33,[38][39][40][41][42]. In Israel, BT was first observed in 1944. From 1964 to 2004, five serotypes were found to be circulating: BTV-2, -4, -6, -10, and -16 [43]. Between 2004 and 2020, thirteen BTV serotypes were detected in Israel-BTV-1, -2, -3, -4, -5, -6, -8, -9, -10, -12, -15, -16, and -24-with BTV-3 and BTV-4 remaining enzootic in recent years [44].
The increasing frequency of incursions by new BTV strains, along with the temporal similarity in the emergence of novel serotypes in Israel and Europe, raises questions about the origin and possible transmission routes of these viruses.
In the present study, recently identified Israeli BTV strains were sequenced and compared with publicly available BTV sequences. The objectives were to (i) describe the BTV strains detected in Israel between 2020 and 2023 and (ii) compare Israeli and European BTV strains using BLASTn (NCBI) [45], pairwise, and phylogenetic analyses, in order to identify shared and distinct patterns in the emergence of previously undetected BTV strains in both regions.
## 2. Materials and Methods
## 2.1. Field Samples
Samples from weak, dead, or aborted domestic and wild/zoo ruminants submitted between 2021 and 2023 for routine arboviral infection testing to the Virology Department of the Kimron Veterinary Institute (KVI), Israel, were included in this study. Clinical specimens included placenta, brain, and internal organs from aborted fetuses; whole blood from symptomatic animals; and spleen or lung tissue from dead ruminants.
## 2.2. Virus Isolation (VI)
Due to coronavirus restrictions and the increased number of field samples received during 2021-2023-associated with simultaneous outbreaks of several arboviral diseases in livestock caused by multiple serotypes of BTV, bovine ephemeral fever virus (BEFV) [46], and equine encephalosis virus serotype 6 (EEV-6; arboviral season of 2023, unpublished data)-only selected BTV RT-qPCR-positive samples collected between 2021 and 2023 were inoculated into embryonated chicken eggs (ECE; VALO BioMedia North America LLC) according to the method described by Komarov and Goldsmit [47]. In brief, infectious material was inoculated intravenously into 9-11-day-old ECE and incubated in an egg incubator at 35 • C. Infected ECE were monitored for seven days post-inoculation. Whole embryos, in which embryonic death occurred, were homogenized and tested using pan-BTV RT-qPCR to confirm VI. Some of the viruses isolated in ECE were subsequently adapted to Vero (African green monkey kidney epithelial cells; source-ATCC), BHK-21 (source-ATCC), or BHK-BSR cells (baby hamster kidney cells or their clone, BSR), kindly provided by Prof. Eran Bachrach, Tel Aviv University, Israel. The procedures for sample preparation for VI and the entire VI process are described in Golender et al.'s study [46].
## 2.3. Nucleic Acid Extraction and Pan-BTV Real-Time Polymerase Chain Reaction (RT-PCR)
Ribonucleic acid (RNA) was extracted from tissue culture supernatants, chicken embryo homogenates, and field samples (whole blood, lung, brain, spleen) using the Mag-MAX™ CORE Nucleic Acid Purification Kit (Thermo Fisher Scientific, Austin, TX, USA) or the IndiMag Pathogen Kit (Indical Bioscience, Leipzig, Germany), following the manufacturers' instructions. Viral RNA detection was performed using the VetMAX™ BTV NS3 All Genotypes Kit (Applied Biosystems™, Thermo Fisher Scientific, Lissieu, France) or according to Wernike et al. [48], with amplification carried out using the AgPath-ID™ One-Step RT-PCR Kit (Life Technologies, Austin, TX, USA). Samples of cattle origin were additionally tested for epizootic hemorrhagic disease virus (EHDV) as described by Wernike et al. [48], and for BEFV following the protocols of Erster et al. [49] and Golender et al. [50]. According to the authors' and manufacturers' instructions for each RT-qPCR system, the cut-off value for all assays was a cycle threshold (Ct) of 40.
## 2.4. Type-Specific Real Time (RT-qPCR) and Conventional RT-PCRs
The method described by Lorusso et al. [51] was used for the detection of BTV-3. For the detection of BTV-1, BTV-4, and BTV-8 genomes, assays described by Maan et al. [52] were performed. During 2021-2023, samples that tested positive for BTV by RT-qPCR with Ct values ≤34 were further analyzed using BTV-3-and BTV-4-specific RT-qPCR assays, as these serotypes are currently enzootic in Israel.
Screening of pan-BTV positive samples by RT-qPCR for additional serotypes was performed retrospectively, based on the results of virus isolation and serotype determination by in-house RT-PCR, followed by confirmation using Sanger sequencing, as previously described [53]. Primers for the BTV-1-specific conventional RT-PCR are listed in the Supplementary Material (Table S1).
Due to the identification of a critical mutation in BTV-12 Israeli strains from 2020-2021 that affected the applicability of the assay described by Maan et al. [52], the oligonucleotide (nt) sequence of the probe was modified, tested for specificity, and used in a limited manner (Table S1). For this reason, only a limited number of BTV-positive samples collected during the early phase of the outbreak (June-July 2021) were tested using conventional BTV-12specific RT-PCR and subsequently confirmed by Sanger sequencing [53]. Screening of BTV-positive samples with the modified BTV-12 RT-qPCR assay was conducted on field samples received in early 2022 (January-June 2022).
All RT-qPCR assays were performed using the AgPath-ID™ One-Step RT-PCR Kit (Life Technologies, Austin, TX, USA), and conventional RT-PCR assays were carried out using the OneStep RT-PCR Kit (Qiagen, Hilden, Germany).
## 2.5. Whole Genome Sequencing (WGS) and Phylogenetic Analyses
Every successfully isolated BTV strain was sequenced by Sanger sequencing using serotype-specific primers for final confirmation of the serotype [53]. Based on these results, each virus was assigned to a specific serotype/genotype.
We sequenced the complete genomes of one representative strain for each newly identified Israeli BTV genotype/serotype-BTV-1, -4, -5, -8, and -11-first detected between 2021 and 2023. Only one BTV-8 isolate was available, and the single BTV-4 strain had already been successfully adapted to tissue culture, which determined the choice of these strains. The choice of a specific strain among several belonging to the same serotype (BTV-1, -5, and -11) was based on the lower Ct values in RT-qPCR obtained for the tissue-cultureadapted strains.
Additionally, we sequenced complete genomes of enzootic Israeli BTV-4 (2015) and eleven BTV-3 strains isolated during 2019-2023. To better understand the dynamics of BTV evolution, and as a continuation of our previously published study on Israeli BTV-3 [53], two representative strains from each year were included, except for 2021 (three strains). Two different sub-genotypes were detected simultaneously in 2020, and one strain from each sub-genotype was sequenced. Sample preparation for WGS of six Israeli strains was performed at the Friedrich-Loeffler-Institut, Germany. High-throughput sequencing (HTS) was carried out using the sequence-independent single-primer amplification (SISPA) approach [54], following the procedure described by Ries et al. [55] for double-stranded cDNA preparation. The resulting cDNA was submitted to Eurofins Genomics (Ebersberg, Germany) for genome sequencing on the Illumina platform.
RNA extraction for 11 additional Israeli strains was performed at the KVI, Israel, using either the Invisorb Spin Virus RNA Mini Kit (STRATEC Molecular GmbH, Berlin, Germany) or the GeneAll RiboSpin™ vRD II Kit (GeneAll Biotechnology, Songpa-gu, Seoul, Korea). Extracted RNA was transferred to the Technion Genomic Center (Technion-Israel Institute of Technology, Haifa, Israel) for high-throughput sequencing using the method previously described by Golender et al. [56].
The resulting nucleotide (nt) sequences were assembled and aligned pairwise using Geneious version 9.0.5 (Biomatters, Auckland, New Zealand) and/or BioEdit (https:// bioedit.software.informer.com/7.2/). Phylogenetic trees were constructed using MEGA X software [57]. For all phylogenetic analyses, the maximum likelihood (ML) method and the Tamura-Nei model were applied.
## 3. Results
## 3.1. Field Samples
As described in our previous investigations, the most suitable materials for molecular detection and virus isolation are whole blood from sick animals, and spleen from dead animals, as well as spleen, brain, and placenta from aborted fetuses. Overall, the proportion of positive samples is highest in blood samples [44,53]. A detailed analysis of the sample types used in the current study is not provided.
A total of 4640 field samples were tested between 2021 and 2023. The results are summarized in Table 1. Field samples were systematically tested for serotypes 3 and 4, while for other serotypes, samples were tested partially. During 2021-2023, BTV-3 and -4 were constantly identified. In general, during 2021 and 2023, an increased number of samples were submitted to the Department of Virology, KVI, likely associated with simultaneous outbreaks of BTV and BEFV in those years. The total number of tested samples was similar in 2021 (n = 1899) and 2023 (n = 1868) (Table 1), while the number of BEFV-positive samples was also comparable: 526 in 2021 (according to the annual reports of the Veterinary Services and the KVI) and 506 in 2023 [46]. The proportion of BTV-positive samples in 2021 was significantly higher (n = 629; 33.12%) than in 2023 (n = 360; 19.26%). The increased number of BTV-positive samples in 2021 reflected more severe outbreaks, which caused prominent clinical signs in affected animals and led veterinary clinicians and animal owners to submit more samples for arboviral diagnostics. Thirteen of the eighty-eight samples tested positive, indicating the continued circulation of BTV-12 at the end of 2021 (Table 1).
## 3.2. Virus Isolation (VI)
Data on VI during 2021-2023 are presented in Table 2. During this period, eight serotypes were isolated from field samples (Table 2). Throughout the entire study period, BTV-3 and BTV-4 were consistently isolated. Since BTV-12 generally caused a low viral load in the bloodstream of infected cattle (most BTV-positive cattle blood samples had pan-BTV RT-qPCR Ct values above 30), virus isolation was often unsuccessful. Detailed information on BTV identification during the past decade was described previously [44,53,58]. During 2021, five BTV serotypes were detected during the arboviral season: two enzootic strains (BTV-3 and BTV-4), BTV-11 (previously unreported in the region), a new genotype of BTV-12 that was first identified in Israel at the end of 2020, and a new genotype of BTV-1 (detailed genetic analyses of all Israeli BTV strains are presented in Section 3.5).
The BTV-1 strain identified at the end of 2021 continued to circulate in 2022 and became the dominant serotype that year. Additionally, BTV-3 and BTV-4 were identified and isolated (Tables 1 and2). In 2023, five serotypes were detected: BTV-3, BTV-4, BTV-5, BTV-6, and BTV-8. Three new genotypes were isolated that year, BTV-4, BTV-5, and BTV-8, although the current BTV-8 genotype was first molecularly detected in Israel in 2019 (Tables 1 and2).
Summarized information on BTV identification in Israel during 2013-2023 is presented in Figure 2.
## 3.4. Sequencing
Information about selected strains for sequencing is presented in Table 3. The increased number of completely sequenced Israeli BTV-3 strains can be attributed to the detection of several genetically distinct, non-ancestral BTV-3 genotypes, as determined by Sanger sequencing, suggesting multiple independent introduction events into Israel. Information about selected strains for sequencing is presented in Table 3.
Phylogenetic analysis of Segment 2 (Seg-2) indicated that the Israeli BTV-1 strain ISR-3279/1/21, isolated in 2021, clustered with a BTV-1 strain from Oman identified in 2020, sharing 98.25% nt identity. These two strains also exhibited high similarity to BTV-1 strains circulating in North African and Southern European countries since 2006, with 97.66-98.25% nt identity. A new BTV-1 strain into Israel, which is markedly distinct from the BTV-1 strain previously identified in 2019, that clustered with a BTV-1 strain from Sudan (Figure 3a). Based on Seg-2 sequences, Israeli BTV-1 strains from 2019 and 2021 shared only 93.15% nt identity.
Phylogenetic analysis of Seg-6 showed that both Israeli BTV-1 strains from 2019 and 2021 clustered with a South African strain isolated in 2017 (Edinburgh VR49). The European BTV-1 strains formed a more distant Seg-6 clade relative to both genotypes of Israeli BTV-1 strains, as would be expected based on Seg-2 (Figure 3b). Since Seg-6 data are not available for the BTV-1 strain from Oman, its relationship to the Israeli viruses could not be assessed. Among the European strains, recently identified viruses clustered with strains from the Mediterranean Basin isolated approximately 1.5 decades ago, suggesting continuous circulation of these strains in southern Europe.
$$• BTV-1$$
## •
BTV-3
Phylogenetic analysis of Seg-2 demonstrated that the Tunisian BTV-3 strain TUN2016/ Zarsis is the closest relative of all Israeli BTV-3 strains. However, a more detailed analysis revealed the presence of several distinct subclusters among the Israeli strains. Two strains, ISR-2019/13 (isolated in 2013) and ISR-2262/2/16 (isolated in 2016), were more distantly related to the most recent Israeli BTV-3 strains. As previously reported [53], BTV-3 strains circulating between 2016 and 2020 were extremely closely related, sharing 99.59-99.62% nucleotide identity, and likely represent ancestral lineages. The BTV-3 strains circulating in Israel from 2020 to 2023 formed an additional subcluster, sharing 98.87-99.28% nt identity with the earlier strains (2016-2020) based on Seg-2 (Figure 4a). In comparison, both recently emerging European BTV-3 variants-Mediterranean and Central European types detected in 2023-were closely related to each other, sharing 97.51-97.55% nt identity, and clustered with the Zimbabwean strain ZIM2002/01. Pairwise comparison of the European (both variants) and Israeli BTV-3 strains revealed 93.92-93.96% nt identity. BLASTn and phylogenetic analyses of Seg-6 revealed that BTV-3 sequences were closely related to Seg-6 sequences of BTV-6, -9, -13, -14, and -16. According to these analyses, reassortment events most frequently occurred between BTV-3, -6, and -16. Phylogenetic analysis of Seg-6 of Israeli BTV-3 strains showed results consistent with those obtained for Seg-2. All Israeli BTV-3 isolates clustered together, sharing 99.08-99.87% nt identity. Similar to Seg-2, strains from 2016-2020 formed a distinct subcluster, while strains from 2020-2023 formed another separate subcluster (Figure 4b). Considering the European BTV-3 strains, the Italian isolates obtained between 2018 and 2024 clustered with Tunisian BTV-3 strains detected between 2016 and 2021, sharing 99.63-99.88% nt identity. These Italian strains also showed 97.18-97.49% nt identity with the Israeli BTV-6 strain ISR-2050/1/19 (2019) and 97.68-97.98% nt identity with the Tunisian BTV-3 strain TUN2016/Zarsis. According to phylogenetic analysis, the European BTV-3 strains exhibited 94.45-95.26% nt identity with the Israeli BTV-3 strains. Notably, the Seg-6 sequences of the BTV-3 strains recently detected in Central Europe (2023-2024) clustered with South African BTV-3 and BTV-16 strains, showing 96.88-97.19% nt identity (Figure 4b).
## •
BTV-4 and BTV-11
Phylogenetic analysis of Seg-2 showed that Israeli BTV-4 strains are highly similar to each other and, since 2001, have formed a distinct genetic cluster, sharing 98.97-99.45% nt identity. The strain that emerged in 2023 (ISR-1621/23) clustered with a Sudanese BTV-4 strain (96.21% nt identity), but shared only 93.04-93.18% identity with previously circulating Israeli BTV-4 strains.
At least three distinct BTV-4 genotypes are currently circulating in Europe. In 2023, a BTV-4 strain was identified in Israel from calves imported from Portugal [59]. Its partial Seg-2 sequence (ISR-1692/3/23) showed close relatedness (97.97% nt identity) to BTV-4 strains that circulated in North Africa and Spain between 2006 and 2010. Comparison of European BTV-4 strains circulating since 2014 with those from the Mediterranean region (Italy, Spain, France) revealed Seg-2 identities of 97.64-97.97%, and phylogenetic analysis grouped them within a single major cluster. However, when divergent BTV-4 and outgroup BTV-24 strains were excluded, European, Middle Eastern, and African strains separated into two closely related subclusters (Figure S1). In contrast, the Portuguese BTV-4 strain isolated in Israel (ISR-1692/3/23) and its related Spanish and Moroccan strains (2004-2010) exhibited lower identity (90.87-94.22%) with other European BTV-4 strains, forming a distinct lineage within the broader BTV-4 cluster (Figure 5a).
Phylogenetic analysis of Seg-2 demonstrated that the recently emerged Israeli BTV-11 strain clustered with South African strains, sharing 95.52-95.59% nt identity (Figure 5b).
Phylogenetic analysis of Seg-6, including the newly identified Israeli BTV-4 and BTV-11 strains, revealed that the 2023 BTV-4 strain (ISR-1621/23) formed a distinct monophyletic branch. BLASTn analysis showed its closest relatedness to South African BTV-17 (Prieska_VR07_2014) and Israeli BTV-24 (ISR2008/02), with 96.56% nt identity to both. Notably, Israeli BTV-4 strains circulating until 2006 clustered with a Turkish BTV-4 strain isolated in 1978 (Figure 5c), indicating a long-standing local lineage. Since the emergence of BTV-24 in Israel in 2008, local BTV-4 strains appear to have incorporated Seg-6 from BTV-24, as all subsequent local BTV-4 strains have consistently contained BTV-24-like Seg-6 sequences.
The emerging Israeli BTV-11 strain (ISR-3265/2/23) clustered with South African and Portuguese BTV-10 strains, sharing 95.98% nt identity (Figure 5c). Among European BTV-4 strains, two main clusters were identified: the first includes strains circulating in continental Europe since 2014 (notably Italian and French strains) with 99.08-99.63% nt identity, while the second includes Mediterranean strains from 2020-2021, sharing 99.74-99.87% identity. Unfortunately, the Seg-6 sequence of the "Portuguese" BTV-4 strain isolated in Israel was not available for analysis.
## •
BTV-5
According to phylogenetic, BLASTn, and pairwise analyses of Seg-2, the Israeli BTV-5 strain ISR-2089/7/23 showed the closest relationship with the Zambian strain ZAM KASAMA KS08, sharing 98.18% nt identity, followed by the Nigerian strain NIG1982/03, with 96.82% nt identity. In contrast, comparison with previously circulating Israeli BTV-5 https://doi.org/10.3390/pathogens15010038 strains (detected between 2006 and 2016; available sequences ISR2009/13 and ISR2011/04) revealed only 78.07-78.32% nt identity (Figure 6a). Phylogenetic, BLASTn, and pairwise analyses of Seg-6 of the newly emerged Israeli BTV-5 strain (ISR-2089/7/23) indicated the closest identity with the Nigerian BTV-5 strain NIG1982/03, sharing 98.11% nt identity (Seg-6 data for the Zambian strain ZAM KASAMA KS08 was not available). In contrast, the nt identity between ISR-2089/7/23 and previously circulating Israeli BTV-5 strains (2006-2016) was only 89.96-90.10% (Figure 6b).
## •
BTV-8
Phylogenetic analysis of Seg-2 illustrated the presence of two distinct genotypes of BTV-8 in Israel. The first genotype, detected between 2008 and 2019 (and responsible for outbreaks in 2010 and 2015-2016), was closely related to the European BTV-8 strains circulating since 2006 ("old strains") [58]. The second, "new" genotype was first identified in Israel in 2019 [58] and re-detected in 2023. This new genotype forms a distinct monophyletic cluster within BTV-8.
Among currently circulating European BTV-8 strains, two separate genotypes have been identified: one circulating in continental Europe and another spreading across Mediterranean Europe. The nt identity between these two European genotypes is 96.42-96.70%. The Israeli "new" BTV-8 genotype shares 94.89-94.99% nt identity with the "Mediterranean" European strains, and 95.34-95.61% nt identity with the "old" BTV-8 strains (both Israeli and European) (Figure 7a). Phylogenetic analysis of Seg-6 of recently circulating BTV-8 strains in Eurasia revealed results consistent with the Seg-2 phylogeny. The only exception was that the Israeli "new" strains clustered with a BTV-8 strain from Oman (2020), sharing 95.60% nt identity. Similar to the Seg-2 phylogenetic analysis, the closest phylogenetic and pairwise nt identity for Seg-6 was observed between the "old" and "new" European strains (97.49-97.62%). The new Israeli BTV-8 strains formed a distinct monophyletic cluster.
Comparison of the "new" Israeli strains with the European "Mediterranean" and "old" genotypes showed nt identities of 94.86-94.93% and 95.41-95.60%, respectively (Figure 5b). Interestingly, the new Israeli BTV-8 strains clustered with the Omani BTV-8 strain, while BLASTn analysis indicated their closest identity with the Nigerian strain NIG1982/07, sharing 95.32-95.91% nt identity. The "Mediterranean" European BTV-8 strains clustered with the BTV-8 strain from Mayotte (2016; strain 5191), which was confirmed by BLASTn analysis, sharing 98.84-98.96% nt identity (Figure 7b).
## •
BTV-12
Phylogenetic and pairwise analyses of the Israeli BTV-12 strain revealed its close relationship and clustering with the Zambian strain ZAM MONGU ZC10 (2018), sharing 96.73% nt identity. Pairwise comparison of the current Israeli BTV-12 strain with the earlier Israeli BTV-12 genotype detected in 2010-2011 showed 95.96-96.42% nt identity. In contrast, the recently detected BTV-12 strain in the Netherlands clustered with the strain from Mayotte Island, sharing 98.56% nt identity, followed by the South African strain isolated in 2017, which showed 96.56% nt identity with the strain from the Netherlands (Figure 8a). Interestingly, phylogenetic analysis of Seg-6 revealed a notably closer relationship between the Israeli BTV-12 strain ISR-2717/1/20 and the recently identified BTV-12 strains from the Netherlands (NET2024/24023518) and Mayotte Island (strain 24-01(3804), 2024). The Israeli strain shared 97.52% nt identity with the Mayotte strain and 96.96% with the strain from the Netherlands. Notably, the strain from the Netherlands and Mayotte strains exhibited a very high nt identity of 99.21% (Figure 8b). S2 and Figure S2. In brief, BLASTn, pairwise, and phylogenetic analyses indicated that the majority of internal genes of the currently circulating Israeli strains are of local origin (see the total number of probable reassortment events in Supplementary Table S2). Most genes displayed a very high nt identity (typically >99%) with previously circulating Israeli strains, followed by strains of African origin. Notably, sequence similarity with strains from the Arabian Peninsula was minimal.
Interestingly, the recently identified BTV-3, BTV-4, and BTV-12 strains appear to share a probable common ancestor with some older Israeli strains, suggesting a shared evolutionary origin. However, the relatively low nt identity values exclude the possibility of direct viral transfer between these geographic regions, with the exception of Seg-7 of BTV-12, which showed 99.35% nt identity.
Considering BTV-12 cases in Israel and Europe, the currently identified BTV-12 strain from Mayotte Island exhibited close identity with both European and Israeli strains. In contrast, the BTV-4 strain isolated in 2015 (the oldest isolate analyzed in this study) showed higher identity with locally circulating strains from the pre-2015 period than with strains currently circulating in Israel.
## 4. Discussion
Israel is situated in a unique geographic location at the junction of three continents. Its close proximity to Africa, Asia, and Europe facilitates the potential introduction and spread of arboviral infections from all these directions. Historically, Israel has been endemic for BTV since the beginning of monitoring efforts [43].
During the past decade, numerous arboviruses belonging to the families Sedoreoviridae, Peribunyaviridae, and Rhabdoviridae, which had not been previously detected in Israel, have been identified and isolated on an annual basis [60]. Most of the identified simbuviruses showed close genetic relationships with African strains, with the exception of the most recent Akabane virus (AKAV) strains detected in 2018, which exhibited high identity with Turkish AKAV [60]. Interestingly, in 2023, a "Mayotte-like" BEFV strain was identified in Israel, whereas only local strains had been detected for more than two decades, with the exception of a transient introduction of a "Turkish" BEFV strain in 2008 [46].
Regarding other non-BTV orbiviruses, during the past decade, EHDV-1, -6, and -7 have been detected in Israel, with EHDV-1 and -6 likely originating from Africa [59]. Notably, this virus was previously detected in North African countries in 2006 and in Turkey in 2007, before reaching Israel in 2016-indicating a delay of approximately ten years and suggesting a natural mode of spread [59]. Furthermore, in 2023, an EEV-6 strain was identified and isolated in Israel, causing a severe outbreak in horses (unpublished data), which most likely also originated from Africa.
The most recent orbiviral outbreaks, caused by EHDV-8, spread from the North African Mediterranean region into large parts of Europe. This virus first emerged in Tunisia in 2021, also pointing on African origin of the virus. Further EHDV-8 was subsequently identified in Italy and Spain the following year [61]. In 2023, EHDV-8 spread rapidly across Spain, Portugal, and France, and outbreaks were still being reported in 2024 in Spain, France, Portugal, and Andorra [62]. Historically, until 1998, the emergence and spread of previously unrecognized BTV strains were reported predominantly within the Mediterranean region. For example, the BTV-1 strain detected in Greece in 2001 was genetically similar to isolates originating from the Far East [63]. Another distinct BTV-1 strain emerged for the first time in Algeria in 2006 [64] and subsequently disseminated across the southern Mediterranean basin, spreading northward to France and even posing a potential threat to the United Kingdom. Following its incursions in 2006 and 2010, BTV-1 reappeared in Sardinia in the autumn of 2012 and persisted in subsequent years, invading Corsica, Sicily, and mainland Italy [65]. The BTV-4 strain identified in Greece in 1999 was closely related to isolates recorded in the 1960s and 1970s from Cyprus and Turkey [26]. Additionally, BTV-2 entered Tunisia in 1999, spread to Algeria and Morocco in 2000, and subsequently advanced into the western and central Mediterranean islands as well as mainland Italy. Moreover, a distinct BTV-4 strain, unrelated to those circulating in the eastern Mediterranean basin, appears to have spread from North Africa to Spain, Portugal, and Corsica between 2003 and 2005 [25]. During 2006-2009, incursions of BTV-6, -8, and -11 were reported in Europe. The BTV-8 strain was confirmed to have originated from sub-Saharan Africa, with climatic changes and altered wind patterns suggested as contributing factors influencing vector distribution [66]. At the same time, genetically similar strains such as BTV-6 and BTV-11 were detected in Germany, the Netherlands, and Belgium. These viruses showed close genetic identity to live attenuated vaccine strains, suggesting a possible link to their unauthorized or accidental use [25].
In Israel, a different pattern of BTV spread than that observed in Europe was evident. Phylogenetic, pairwise, and BLASTn analyses of Israeli BTV-1 strains (ISR-2050/19, identified in 2019, and ISR-3279/1/21, detected in 2021) showed that they do not cluster together, indicating distinct evolutionary origins. Notably, the Israeli BTV-3 strains are most closely related to the Tunisian BTV-3 strain TUN2016/Zarsis, whereas two different European BTV-3 strains show a closer relationship with other African BTV-3 isolates. Since 2013, several distinct incursions of different BTV-3 genotypes have been recorded in Israel: one in 2013 (ISR-2019/13) and two genetically distinct genotypes in 2016 (ISR-2153/16 and ISR-2262/2/16) [53]. The strain ISR-2153/16 was likely the ancestor of the local BTV-3 strains circulating between 2016 and 2020. In 2020, a new introduction of BTV-3 occurred, and two groups of strains were detected simultaneously. Comparison of strains circulating during 2016-2020 and 2020-2023 revealed that the outer genes (Seg-2 and Seg-5) and only three of the seven internal genes (Seg-7, -8, and -10) shared a common ancestor. Since 2021, only one genotype-the most recent-of BTV-3 has been detected in Israel. Considering reassortment events, during 2016-2023, only a single case of reassortment between a BTV-3 strain and other local strains was identified among fully sequenced isolates (strain ISR-1434/1/23) [53]. Nevertheless, the recently identified and analyzed European and Israeli BTV-3 strains showed a close relationship of their Seg-7 sequences, with a minimum nt identity of 96.37%. Analysis of European BTV-3 revealed that Mediterranean BTV-3 strains are more closely related to continental European BTV-3 strains and to the Israeli BTV-6 strain ISR-2095/3/17. Notably, the newly emerging Mediterranean BTV-3, -4, and -8 strains exhibit multiple reassortment events both among themselves and with different local BTV strains.
Analyses of European BTV-8 revealed the presence of two distinct genotypes: one circulating in continental Europe since 2006, which was also introduced into Israel, and a second genotype that emerged in Mediterranean European countries in 2023. The European BTV-8 strains are more closely related to each other than to the recently emerged Israeli BTV-8 strains (2019-2023). These findings suggest a common origin for the European strains and indicate separate routes of introduction for the new BTV-8 strains into Israel, the European Mediterranean region, and continental Europe.
Detailed analysis of recently identified BTV-12 strains in Israel and the Netherlands revealed close genetic identity with the BTV-12 strain recently detected on Mayotte Island in 2024 for the majority of genes (7/10 for Israeli strains and 8/10 for the Netherlands strain). Notably, phylogenetic analysis of Seg-7 showed that all "western" BTV-12 strains clustered together, further supporting a probable common origin.
For an extended period, local Israeli BTV-4 strains remained closely related to other regional BTV-4 strains, although frequent reassortment events with multiple serotypes and genotypes were continuously observed (GeneBank, NCBI [45]). In 2023, a new BTV-4 strain was introduced into Israel and rapidly spread throughout the country. Phylogenetic and BLASTn analyses of this strain (ISR-1621/23) suggest that it has likely undergone reassortment with local Israeli strains and shares common ancestry with recently identified Israeli BTV-1, -8, and -11 strains detected between 2019 and 2023.
Notably, the most recent European BTV-4 strains cluster with the newly emerging Israeli BTV-5 based on internal gene analyses, suggesting possible co-circulation and a likely African origin for these viruses. Regarding European BTV-4, at least three distinct genotypes are currently circulating, primarily in Portugal, Spain, and France. Reassortment events between European BTV-3, -4, and -8 have also been observed, facilitating the spread of some reassorted strains into new regions, including the Mediterranean basin (e.g., BTV-3 and BTV-8).
Repeated introductions of orbiviruses into new regions in recent years may reflect changing environmental or climatic conditions influencing arbovirus transmission dynamics. Potential routes of introduction include windborne or vehicle-assisted dispersal of infected vectors, natural expansion of arthropod populations into previously uncolonized areas, and legal or illegal movement of infected domestic or wild ruminants or biological materials. According to recent studies, the majority of BTV incursions into Italy over the past 20 years have originated from Northern Africa, likely due to wind-blown dissemination of infected midges [33]. The spread of BTV-8 in Europe in 2006-2007 and of BTV-3 in the Netherlands in 2023 was probably associated with the movement of infected animals and/or midges [67].
In summary, the available data suggest that the primary origin of newly introduced arboviruses in Israel is likely Africa, although some internal genes share ancestry with viruses isolated in the Arabian Peninsula, Europe, and the Comoros Archipelago in the southwestern Indian Ocean. Many knowledge gaps remain in the molecular investigation of both historical and contemporary BTV distribution. Whole-genome sequencing of new and archived BTV strains from diverse regions is needed to clarify and complete the understanding of BTV evolution and spread.
## 5. Conclusions
All genetic analyses indicate that the recently identified Israeli and European BTV strains share common ancestors, predominantly of African origin. Notably, some genomic segments derived from the Comoros/Mayotte region or the Arabian Peninsula are more frequently observed in Israeli strains than in European BTV strains. Differences of at least 2-3% in nucleotide sequences across all genes suggest that the Israeli and European BTV strains have evolved independently over several years. These genomic distinctions indicate that there has been no direct transfer of BTV strains between Israel and Europe in the past decade.
## References
1. The, Website (2025)
2. The, Veterinary, Website (2025)
3. The, Website (2025)
4. Elbers, Backx, Mintiens et al. (2008) "Field observations during the Bluetongue serotype 8 epidemic in 2006. II. Morbidity and mortality rate, case fatality and clinical recovery in sheep and cattle in the Netherlands" *Prev. Vet. Med*
5. Backx, Heutink, Van Rooij et al. (2007) "Clinical signs of bluetongue virus serotype 8 infection in sheep and goats" *Vet. Rec*
6. Salem, Ben Aicha, Kalthoum et al. "Estimation of the economic impact of a bluetongue serotype 4 outbreak in Tunisia" *Front. Vet Sci*
7. Mertens, Diprose, Maan et al. (2004) "Bluetongue virus replication, molecular and structural biology" *Vet. Ital*
8. Gondard, Postic, Garin et al. (2024) "Exceptional Bluetongue virus (BTV) and Epizootic hemorrhagic disease virus (EHDV) circulation in France in 2023" *Virus Res*
9. Stewart, Hardy, Barry et al. (2015) "Characterization of a second open reading frame in genome segment 10 of bluetongue virus" *J. Gen. Virol*
10. Vijaykrishna, Mukerji, Smith (2015) "RNA Virus Reassortment: An Evolutionary Mechanism for Host Jumps and Immune Evasion" *PLoS Pathog*
11. Maan, Maan, Belaganahalli et al. (2012) "Identification and differentiation of the twenty six bluetongue virus serotypes by RT-PCR amplification of the serotype-specific genome segment 2" *PLoS ONE*
12. Ries, Vögtlin, Hüssy et al. (2021) "Putative novel atypical BTV serotype '36' identified in small ruminants in Switzerland" *Viruses*
13. Ries, Sharav, Tseren-Ochir et al. (2020) "Putative novel serotypes '33' and '35' in clinically healthy small ruminants in Mongolia expand the group of atypical BTV" *Viruses*
14. Menzies, Mccullough, Mckeown et al. (2008) "Evidence for transplacental and contact transmission of bluetongue virus in cattle" *Vet. Rec*
15. Backx, Heutink, Van Rooij et al. (2009) "Transplacental and oral transmission of wild-type bluetongue virus serotype 8 in cattle after experimental infection" *Vet. Microbiol*
16. Darpel, Barber, Hope et al. (2016) "Using shared needles for subcutaneous inoculation can transmit bluetongue virus mechanically between ruminant hosts" *Sci. Rep*
17. Savini, Lorusso, Paladini et al. (2014) "Bluetongue serotype 2 and 9 modified live vaccine viruses as causative agents of abortion in livestock: A retrospective analysis in Italy" *Transbound. Emerg. Dis*
18. Batten, Darpel, Henstock et al. (2014) "Evidence for transmission of bluetongue virus serotype 26 through direct contact" *PLoS ONE*
19. Bréard, Schulz, Sailleau et al. (2018) "Bluetongue virus serotype 27: Experimental infection of goats, sheep and cattle with three BTV-27 variants reveal atypical characteristics and likely direct contact transmission BTV-27 between goats" *Transbound. Emerg. Dis*
20. Gambles (1949) "Bluetongue in sheen in Cyprus" *J. Comp. Pathol*
21. Sellers (1975) "Bluetongue in Cyprus" *Aust. Vet. J*
22. Mellor, Carpenter, Harrup et al. (2006) "Bluetongue in Europe and the Mediterranean Basin: History of occurrence prior to" *Prev. Vet. Med*
23. Manso-Ribeiro, Rosa-Azevedo, Noronha et al. (1957) *Bull. Off. Int. Epiz*
24. Wilson, Mellor (2009) "Bluetongue in Europe: Past, present and future" *Phil. Trans. R. Soc. Lond. B. Biol. Sci*
25. Breard, Sailleau, Nomikou et al. (1979) "Molecular epidemiology of bluetongue virus serotype 4 isolated in the Mediterranean Basin between" *Virus Res*
26. Osmani, Murati, Kabashi et al. (2001) "Evidence for the presence of bluetongue virus in Kosovo between" *Vet. Rec*
27. Hateley (2009) "Bluetongue in northern Europe: The story so far. Practice"
28. Eschbaumer, Hoffmann, Moss et al. (2010) "Emergence of bluetongue virus serotype 6 in Europe--German field data and experimental infection of cattle" *Vet. Microbiol*
29. De Clercq, Mertens, De Leeuw et al. (2009) "Emergence of bluetongue serotypes in Europe, part 2: The occurrence of a BTV-11 strain in Belgium" *Transbound. Emerg. Dis*
30. Niedbalski (2015) "Bluetongue in Europe and the role of wild life in the epidemiology of disease" *Pol. J. Vet. Sci*
31. Pascall, Nomikou, Bréard et al. (2020) "Frozen evolution" of an RNA virus suggests accidental release as a potential cause of arbovirus re-emergence" *PLoS Biol*
32. Cappai, Rolesu, Loi et al. (2019) "Western Bluetongue virus serotype 3 in Sardinia, diagnosis and characterization" *Transbound. Emerg. Dis*
33. Holwerda, Santman-Berends, Harders et al. (2023) *Emerg. Infect. Dis*
34. Agriculture, Website (2025)
35. Van Den Brom, Santman-Berends, Van Der Heijden et al. (2025) "Bluetongue virus serotype 12 in sheep and cattle in the Netherlands in 2024-A BTV serotype reported in Europe for the first time" *Vet. Microbiol*
36. (2025) "The Ruminant Health and Welfare Website"
37. Romero-Trancón, Valero-Lorenzo, Ruano et al. (2021) "Emerging Bluetongue Virus Serotype 4 in the Balearic Islands"
38. The, Website (2025)
39. The, Website (2025)
40. The, Uk Website (2025)
41. *EFSA Panel on Animal Health and Welfare*
42. More, Bicout, Bøtner et al. "Assessment of listing and categorisation of animal diseases within the framework of the Animal Health Law"
43. Shimshony (2004) "Bluetongue in Israel-A brief historical overview" *Vet. Ital*
44. Golender, Klement, Kovtunenko et al. (2023) "Comparative Molecular and Epidemiological Analyses of Israeli Bluetongue Viruses Serotype 1 and 9 Causing Outbreaks in 2018-2020"
45. Golender, Hoffmann, Kenigswald et al. "Ephemeral Fever Viruses in Israel 2014-2023: Genetic Characterization of Local and Emerging Strains. Pathogens 2024"
46. Komarov, Goldsmit (1951) "A disease similar to Blue Tongue in cattle and sheep in Israel" *Ref. Vet*
47. Wernike, Hoffmann, Beer (2015) "Simultaneous detection of five notifiable viral diseases of cattle by single-tube multiplex real-time RT-PCR" *J. Virol. Methods*
48. Erster, Stram, Menasherow et al. (2017) "High-resolution melting (HRM) for genotyping bovine ephemeral fever virus (BEFV)" *Virus Res*
49. Golender, Klement, Hoffmann (2025) "Development of New Probe-Based Real-Time RT-qPCR Assays for the Detection of All Known Strains of Bovine Ephemeral Fever Viruses" *Viruses*
50. Lorusso, Sghaier, Di Domenico et al. (2018) "Analysis of bluetongue serotype 3 spread in Tunisia and discovery of a novel strain related to the bluetongue virus isolated from a commercial sheep pox vaccine" *Infect. Genet. Evol*
51. Maan, Maan, Belaganahalli et al. (2016) "Development and Evaluation of Real Time RT-PCR Assays for Detection and Typing of Bluetongue Virus" *PLoS ONE*
52. Golender, Bumbarov, Elda et al. (2020) "Bluetongue Serotype 3 in Israel 2013-2018: Clinical Manifestations of the Disease and Molecular Characterization of Israeli Strains" *Front. Vet. Sci*
53. Chrzastek, Lee, Smith et al. (2017) "Use of Sequence-Independent, Single-Primer-Amplification (SISPA) for rapid detection, identification, and characterization of avian RNA viruses" *Virology*
54. Ries, Domes, Janowetz et al. (2020) "Isolation and Cultivation of a New Isolate of BTV-25 and Presumptive Evidence for a Potential Persistent Infection in Healthy Goats" *Viruses*
55. Golender, Klement, Ofer et al. (2023) "Hefer valley virus: A novel ephemerovirus detected in the blood of a cow with severe clinical signs in Israel in 2022" *Arch. Virol*
56. Kumar, Stecher, Li et al. (2018) "MEGA X: Molecular Evolutionary Genetics Analysis across Computing Platforms" *Mol. Biol. Evol*
57. Golender, Eldar, Ehrlich et al. (1955) "Genomic Analysis Illustrated a Single Introduction and Evolution of Israeli Bluetongue Serotype 8 Virus Population"
58. Golender, Hoffmann "The Molecular Epidemiology of Epizootic Hemorrhagic Disease Viruses Identified in Israel between 2015 and 2023" *Epidemiologia*
59. Golender, Bumbarov, Kovtunenko et al. (2015) "Identification and Genetic Characterization of Viral Pathogens in Ruminant Gestation Abnormalities"
60. Thabet, Sghaier, Ben Hassine et al. (1567) "Characterization of Epizootic Hemorrhagic Disease Virus Serotype 8 in Naturally Infected Barbary Deer (Cervus elaphus barbarus) and Culicoides (Diptera: Ceratopogonidae) in Tunisia" *Viruses*
61. Mertens, Maan, Prasad et al. (2007) "Design of primers and use of RT-PCR assays for typing European bluetongue virus isolates: Differentiation of field and vaccine strains" *J. Gen. Virol*
62. Cêtre-Sossah, Madani, Sailleau et al. (2011) "Molecular epidemiology of bluetongue virus serotype 1 isolated in 2006 from Algeria" *Res. Vet. Sci*
63. Lorusso, Sghaier, Ancora et al. (2014) "Molecular epidemiology of bluetongue virus serotype 1 circulating in Italy and its connection with northern Africa" *Infect. Genet. Evol*
64. Roy, Boyce, Noad (2009) "Prospects for improved bluetongue vaccines" *Nat. Rev. Microbiol*
65. Boender, Hagenaars, Holwerda et al. (2024) "Spatial Transmission Characteristics of the Bluetongue Virus Serotype 3 Epidemic in The Netherlands, 2023. Viruses"
66. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12569247&blobtype=pdf | # Correction: Evaluating the COVID-19 vaccination program in Japan, 2021 using the counterfactual reproduction number
Taishi Kayano, Yura Ko, Kanako Otani, Tetsuro Kobayashi, Motoi Suzuki, Hiroshi Nishiura
As a result of an error during figure preparation, the caption of Figure 2 In addition, in the Methods section, under the subheading 'Effective reproduction number' , Equation 9 was incorrectly given as
The correct Equation 9 appears below.
$$i total t = Rt 1-τ ∑ τ i total t-τ gτ .$$ |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772336&blobtype=pdf | # A novel targeted hybrid capture-NGS assay for sensitive detection of multiplex respiratory pathogens
Junyan Ma, Liping Yu, Kangchen Zhao, Qiao Qiao, Xiaojuan Zhu, Tao Wu, Heng Rong, Shuo Ning, Jinlei Guo, Yuhan Ding, Ying Chi, Lunbiao Cui, Yiyue Ge
## Abstract
Emerging respiratory infectious diseases represented by COVID-19, along with traditional respiratory infections, pose a serious threat to human health. Highthroughput sequencing (NGS), with its high sensitivity and ultra-high throughput, is particularly suitable for the detection of respiratory pathogens (RP) that are extremely diverse in types and frequently involved in mixed infections. In this study, by integrating a Micro-Targets Hybrid Capture (MT-Capture) system, we developed previously with NGS, we developed a novel assay (termed RP-MT-Capture NGS) for the detection of multi ple respiratory pathogens (more than 300 species/types). By optimizing probe design and hybridization capture procedures, RP-MT-Capture NGS achieved high detection sensitivity for different types of pathogens. For influenza viruses, this assay could acquire full-length sequences of hemagglutinin (HA) and neuraminidase (NA) genes for samples with CT values < 32, offering a robust tool for viral mutation surveillance and recombina tion analysis. The results of clinical sample detection showed that RP-MT-Capture NGS exhibited superior accuracy and sensitivity compared to TaqMan array and metagenomic NGS (mNGS) technologies for respiratory pathogen detection. Compared with tradi tional probe hybridization-based targeted NGS (tNGS), RP-MT-Capture NGS significantly shortens the wet lab experiment time to within 6 h. In summary, the RP-MT-Capture NGS assay developed in this study offers a novel tool for detecting multiple respiratory pathogens, with substantial clinical and public health relevance. IMPORTANCE Emerging and traditional respiratory infections pose threats to human health. These diseases are caused by a variety of pathogens, which often lead to co-infections and, thus, make detection difficult. This study combines a novel probe hybridization capture system with high-throughput sequencing to develop a new detection tool (RP-MT-Capture NGS), which can identify over 300 types of respiratory pathogens. For influenza viruses, it can reveal complete details of key viral genes, facilitating the tracking of viral mutations. Compared with existing detection methods, this new tool is more accurate, more sensitive, and has a higher throughput. It provides great value for clinical practice and public health in respiratory pathogen detection. KEYWORDS hybrid capture, MT-capture, tNGS, respiratory pathogens, multiplex detection R espiratory infectious diseases have long been a global public health focus due to their high incidence and mortality (1, 2). In recent years, globalization has driven the emergence of multiple novel respiratory infectious diseases with epidemic potential. Examples include severe acute respiratory syndrome (SARS), Middle East respiratory syndrome (MERS), human infections with highly pathogenic avian influenza (H5N1, H7N9), and Coronavirus Disease 2019 (COVID-19) (3,4). Key contributing factors include rapid population mobility, antibiotic misuse, environmental pollution, and pathogen
mutation and evolution. Besides emerging diseases, traditional respiratory pathogens also frequently cause infections and epidemics. Particularly in the post-COVID-19 era, the population's "immune debt (a phenomenon where the overall population's immune protection against certain pathogens decreases due to reduced exposure, potentially leading to increased infection rates when exposure resumes)" has led to a substantial increase in the incidence of common respiratory infectious diseases (5)(6)(7). Notably, the two most devastating infectious disease pandemics of the past century, the 1918 Spanish flu and the 21st-century COVID-19 outbreak, were both triggered by respiratory infections. The short incubation period, high contagiousness, and broad susceptibility of the population to respiratory pathogens pose significant challenges to the prevention and control of respiratory infectious diseases (8). Rapid, comprehen sive, and accurate pathogen diagnosis is the cornerstone of treating and preventing respiratory infectious diseases, playing a crucial role in the early selection of sensitive medications and timely implementation of effective prevention and control measures.
Conventional laboratory diagnostic approaches, including pathogen culture, biochemical assays, and immunological detection, are inherently time-consuming, labor-intensive, and limited in sensitivity (9)(10)(11). Furthermore, the "window period (the time interval between the initial infection with a pathogen and the point when the infection can be reliably detected by current diagnostic methods)" of infectious diseases represents a critical limitation for antibody detection technologies (12). In recent years, nucleic acid amplification diagnostics have emerged as a pivotal tool for pathogen detection, leveraging advantages such as rapidity, high sensitivity, and specificity (13). Notwithstanding, the most widely adopted technique, real-time PCR, can only detect a maximum of 4-6 pathogens per reaction tube. Given the vast diversity of respiratory pathogens (encompassing hundreds of species) and the frequent occurrence of mixed infections, the development of ultra-high-throughput detection methods for respiratory pathogens is urgently warranted.
High-throughput sequencing technologies, particularly metagenomic next-gener ation sequencing (mNGS) and targeted next-generation sequencing (tNGS), have emerged as powerful platforms for pathogen detection in clinical samples. mNGS holds the potential to identify all pathogens present in a clinical sample, yet the large amount of host genomic data often compromises its sensitivity for pathogen detec tion (14). By contrast, tNGS offers enhanced speed, sensitivity, and cost-effectiveness compared to mNGS, as it focuses on analyzing specific targets of clinical relevance (15,16). tNGS employs either multiplex PCR amplicon or probe capture to enrich target pathogens. While amplicon sequencing boasts advantages like fewer workflow steps, operational flexibility, and reduced turnaround time, it faces notable challenges (17,18). For instance, PCR amplification inherently introduces bias, leading to unequal target enrichment. Additionally, an increase in primer pairs heightens the risk of primer dimer/ multimer formation, which can diminish amplification efficiency or even cause amplification failure. Furthermore, Taq enzyme-mediated errors during amplification necessitate careful discrimination between genuine genetic variations and sequencing artifacts. Lastly, delayed primer updates may undermine the efficacy of amplicon sequencing in detecting emerging mutations (19). In comparison, capture sequencing offers distinct benefits. These include higher target number per panel, superior sequence uniformity, higher fault tolerance, and reduced risk of aerosol contamination.
However, the traditional liquid-phase hybrid capture sequencing process is highly time-consuming and cumbersome, requiring 2-3 days from nucleic acid extraction to capture library construction (20). To address the limitations of conventional liq uid-phase hybrid capture, our previous study developed a novel, rapid, and efficient Micro-Targets Hybrid Capture (MT-Capture) system, which has been successfully applied to SARS-CoV-2 whole-genome sequencing (21). Although respiratory multi-pathogen capture sequencing differs in purpose from whole-genome sequencing, their experi mental workflows, especially hybridization process, share consistency-both employ hybrid capture for library molecules of similar sizes. Thus, in this study, we developed a novel assay (RP-MT-Capture NGS) for detecting multiple respiratory pathogens (RP) by integrating MT-Capture technology with NGS. By optimizing probe design and hybridization capture procedures, RP-MT-Capture NGS achieved high detection sensitivity and remarkably short operation time (within 6 h for wet lab experiment). Compared with the MT-Capture whole-genome sequencing assay for SARS-CoV-2, the RP-MT-Capture NGS assay established in this study has two fundamental innovations. First, it enables the simultaneous detection of over 300 species/types of respiratory pathogens, whereas the former can only detect the coronavirus genome. Second, this assay can additionally obtain the full-length sequences of hemagglutinin (HA) and neuraminidase (NA) gene segments of influenza viruses, providing a reliable tool for influenza virus mutation surveillance and recombination analysis. Clinical evaluation of RP-MT-Capture NGS has demonstrated that the assay delivers exceptional detection performance and holds substantial promise for the identification of multiple respiratory pathogens.
## RESULTS
## The workflow and panel design principles of RP-MT-Capture NGS
A comprehensive overview of the RP-MT-Capture NGS workflow and bioinformatics analysis is illustrated in Fig. 1. MT-Capture can be completed within 2.5 h, resulting in the wet-lab operations of RP-MT-Capture NGS finishing within 6 h. Compared with mNGS, RP-MT-Capture NGS offers reduced costs and enhanced sensitivity. In contrast to the TaqMan array, this assay provides extensive coverage of major respiratory pathogens and enables high-throughput analysis of more samples in a single run. A detailed comparison of the performance characteristics of the three methods is presented in Table 1.
Given the diversity and complexity of respiratory pathogens, probe design should prioritize highly conserved genomic regions (Fig. 2A). This strategy not only cuts down probe costs but also boosts the efficiency of pathogen capture. RNA pathogens typically have fewer conserved regions. For these pathogens, it is recommended to select prevalent strains as reference genomes and moderately increase probe density to enhance capture efficiency. The 16S ribosomal RNA (rRNA), 18S rRNA, and the internal transcribed spacer (ITS) regions are frequently used as molecular markers for identifying bacterial or fungal species. However, the multicopy nature of these genes may render them less suitable as targets for RP-MT-Capture NGS. For example, designing probes for 16s region will capture more ribosomal targets, which can sequester the data of low-abundance pathogens and affect the detection rate of these pathogens. Our results confirmed that targeting 16s rRNA substantially compromised the detection sensitivity of RNA viruses (Fig. 2B andC). Therefore, we selected specific genes other than 16S, 18S, and ITS as targets for the design of bacterial or fungal probes (Fig. 2D; Table S1). Considering the detection sensitivity and cost-effectiveness, we used conserved regions of approximately 500 base pairs for probe design. Due to the epidemiological significance and unique characteristics of influenza A virus, its probe set was designed to target the full-length sequences of both the HA and NA genes.
## Analytical sensitivity of RP-MT-Capture NGS
The analytical sensitivity of RP-MT-Capture NGS was evaluated by testing serially diluted nucleic acid extracted from respiratory samples positive for different kinds of pathogens. The assay successfully detected influenza B virus, human coronavirus OC43, and human alphaherpesvirus in samples with CT values ≥ 33, as well as Klebsiella pneumoniae and Pseudomonas aeruginosa at CT values of 36 and 34, respectively (Fig. 3). These results demonstrate that the RP-MT-Capture NGS assay exhibits high analytical sensitivity.
## High-precision pathogen subtyping and full-length detection of HA/NA genes of influenza virus
The RP-MT-Capture NGS assay developed in this study can not only identify the species but also characterize the types of many common pathogens. As shown in Table 2, the assay accurately identified all the types of tested viruses. Additionally, it enabled multiplex detection of pathogens in a single reaction. For example, sample ML8 showed co-infection with human respiratory syncytial virus A, human orthopneumovirus, and human coronavirus OC43 (Table 2).
Furthermore, RP-MT-Capture NGS enables full-length detection of HA/NA genes of influenza virus, facilitating the monitoring of viral mutations and recombination events. Serial dilutions of H1N1, H3N2, and H7N9 positive samples were analyzed in parallel using RP-MT-Capture NGS and real-time PCR. Results confirmed that RP-MT-Capture NGS accurately identified all tested influenza subtypes. Notably, for samples with CT values < 32, the HA and NA gene coverage reached 100% (Fig. 4). The phylogenetic analysis revealed that the H1N1 HA gene belongs to clade 6B.1A and the H3N2 HA gene belongs to clade 3C.2a1b.2 (Fig. S1). The high sensitivity of HA/NA sequencing is valuable for both epidemic prevention and control and clinical treatment guidance.
## RP-MT-Capture NGS for clinical sample analysis
A total of 186 respiratory samples from acute respiratory infection (ARI) patients were analyzed using the RP-MT-Capture NGS assay. As shown in Fig. 5, 64 distinct pathogens were detected, comprising 14 kinds of Gram-positive bacteria, 13 kinds of Gram-neg ative bacteria, 8 kinds of fungi, 9 kinds of DNA viruses, 18 kinds of RNA viruses, 1 kind of mycoplasma, and 1 kind of chlamydia each. The most prevalent bacteria were Haemophilus parainfluenzae, Streptococcus anginosus, Streptococcus constellatus, Streptococcus intermedius, Haemophilus influenzae, Stenotrophomonas maltophilia, and Streptococcus pneumoniae. The top five viruses included Human betaherpesvirus 7 (Roseolovirus), Human gammaherpesvirus 4 (EB virus), Rhinovirus, Influenza A virus H3N2, and SARS-CoV-2.
Many detected pathogens (particularly bacteria) mentioned above are opportunis tic pathogens. To assess the carriage characteristics of these pathogens in healthy populations, RP-MT-Capture NGS was conducted on samples obtained from 50 healthy controls (HCs). Results showed that the detection rates of seven opportunistic patho gens were either approaching or exceeding 50% (Fig. 6A). We further compared the detection rates and relative abundances of the seven pathogens in ARI and HC samples. Results showed that the detection rates of these pathogens in ARI samples were not significantly different from, and in some cases even lower than, those in HC samples, suggesting that in some ARI patients, infection with other specific pathogens might suppress the colonization of these opportunistic pathogens (Fig. 6A). The results of pathogen loads confirmed this hypothesis; as shown in Fig. 6B, the relative abundances of these pathogens in some ARI samples were significantly lower than those in the HC samples. However, overall, the distribution ranges of the relative abundances of these pathogens in ARI samples were broader, and the average pathogen loads in positive ARI samples were higher (Fig. 6B). Streptococcus pneumoniae, Streptococcus anginosus, and Haemophilus influenzae showed significantly higher pathogen loads in ARI samples, suggesting these pathogens might proliferate secondarily due to decreased immune function, leading to disease. Taken together, these findings suggest that when inter preting positive results for opportunistic pathogens, it is essential to integrate clinical manifestations with the relative abundances of pathogens to determine their potential pathogenicity.
## RP-MT-Capture NGS exhibits superior accuracy and sensitivity compared to TaqMan array and mNGS for respiratory pathogen detection
One hundred and fifty-nine ARI samples were randomly selected from 186 samples that had undergone RP-MT-Capture NGS testing for parallel analysis using the TaqMan array. A comparative evaluation was conducted on the detection results of 41 common respiratory pathogens covered by both methods. Out of the 159 samples, no pathogen (among the 41 pathogens) was detected by either method in 10 samples, while at least 1 pathogen was identified in the remaining 149 samples (Fig. 7). The detection rate for RP-MT-Capture NGS was 95.97% (140/149), compared to 84.56% (126/149) for the TaqMan array. Among the 149 samples, we identified 22 pathogens, 17 of which were co-detected by both techniques. Except for rhinovirus, RP-MT-Capture NGS identified a significantly higher number of positive samples for most of the co-detected pathogens than the TaqMan array (Fig. 8).
In addition to the co-detected pathogens, among the 41 pathogens, 3 pathogens (SARS-CoV-2, Human betaherpesvirus 6, and Influenza B virus) were exclusively detected S2 andS3; Fig. S2). These data indicate that, for most respiratory pathogens, the RP-MT-Capture NGS offers superior performance compared to the TaqMan array. However, for samples of Human coronavirus HKU1 and Mycoplasma pneumoniae with low viral load, the TaqMan array shows better detection performance.
To further compare the performance of RP-MT-Capture NGS and mNGS in detecting respiratory pathogens, 38 specimens with relatively high pathogen loads as detected by RP-MT-Capture NGS were selected from the 186 ARI samples and subjected to mNGS. Results showed that mNGS achieved a detection rate of 73.1% for viruses, 100% for bacteria, and 0% for fungi, as compared to those of RP-MT-Capture NGS (Table 3). Further analysis showed that viral enrichment folds of RP-MT-Capture NGS ranged from 463 to 11,801 (Table S4).
## DISCUSSION
Respiratory infections are common diseases caused by diverse pathogens, including bacteria, viruses, fungi, mycoplasmas, and chlamydia. These pathogens exhibit high diversity in types and complex transmission patterns; notably, distinct pathogens can induce similar clinical manifestations, thereby increasing diagnostic and therapeutic challenges (22,23). Furthermore, respiratory pathogens frequently involve co-infections, and certain RNA viruses display high mutation rates, which further complicate diagnosis (24). Thus, rapid and accurate pathogen identification is critical for improving prognosis, reducing mortality, and controlling infectious disease spread. Nucleic acid amplification tests represented by real-time PCR have become important tools for pathogen detection due to their rapidity, sensitivity, and specificity. Various forms of multiplex real-time PCR and TaqMan array enable simultaneous detection of multiple pathogens. However, these methods either have a limited number of detection targets or a low sample throughput per test.
mNGS is an effective high-throughput molecular detection technology for multiple pathogens; however, this assay has limitations such as host interference, low sensitivity, high cost, and long turnaround time (25,26). Amplicon-based tNGS, despite high sensitivity, fails to cope with pathogen variations and cross-interference with excessive targets (27,28). In this study, we developed a novel RP-MT-Capture NGS assay, which enables simultaneous identification of over 300 respiratory pathogen species/types with high throughput and sensitivity. The high error tolerance of probes enables this assay to easily cope with pathogen variations. Compared with that of the time-consum ing traditional probe hybridization-based tNGS (20), RP-MT-Capture NGS significantly shortens wet-lab experiment time (within 6 h). While 16S rRNA is an ideal target for bacterial species identification (29), multicopy genes may not be suitable as detection targets for RP-MT-Capture NGS based on its detection principle. This is because when bacteria and RNA viruses coexist in a sample, 16S rRNA is likely to dominate most sequencing data, thereby impairing viral detection performance. Our experimental results have confirmed this hypothesis. Therefore, for all bacterial and fungal pathogens, we excluded multicopy genes such as 16S, 18S, and ITS as detection targets, and instead designed probes based on other pathogen-spe cific genes. This strategy helps enhance RNA virus detection while ensuring accurate identification of bacteria and fungi.
Among respiratory pathogens, influenza virus is highly variable, and its classification depends on two key proteins: HA and NA (30). Influenza viruses undergo major genetic variations such as antigenic shift every decade or so, which may give rise to new strains. These strains pose a significant potential threat to humans, making it necessary to continuously track and monitor the variations of influenza virus. Most reported respiratory multi-pathogen detection technologies based on NGS platforms can only identify pathogens but fail to monitor the variation of influenza viruses (31,32). The RP-MT-Capture NGS assay established in this study can detect full-length HA and NA genes, offering a powerful tool for monitoring viral mutations and recom bination analysis. Furthermore, amplicon-based tNGS has limited scalability. Due to cross-interference between multiplex PCR primer pairs, if the existing detection panel needs to be updated, for example, to include additional pathogens, the newly added primer pairs may affect the efficiency of previously optimized primers. Consequently, the entire detection system may require re-optimization and re-evaluation. In contrast, the RP-MT-Capture NGS assay established in this study offers significant advantages in updating detection panels. Simply adding new pathogen-specific probes to the original probe panel allows for easy updates, enabling the detection of newly emerging pathogens. In terms of cost-effectiveness, detecting >300 respiratory pathogens per sample costs ~70 US dollars, significantly lower than mNGS costs. For detecting opportunistic pathogens, the lack of well-defined virulence genes complicates distinguishing their infection from colonization (33)(34)(35). Following primary infection in ARI patients, opportunistic pathogens may exhibit two scenarios: first, the proliferation of primary pathogens inhibits the growth of opportunistic pathogens in some individuals, which may lead to a decreased detection rate of opportunistic pathogens (Fig. 6A); second, primary infection-induced disruption of the body's barriers and immune balance results in secondary proliferation of opportunistic pathogens in some individuals, which may cause an increased pathogen loads in positive individuals (Fig. 6B). Our results suggest that when interpreting positive results for opportunistic pathogens, it is essential to integrate clinical manifestations with the abundances of the pathogens. The higher the abundance, the greater the potential association with the current disease. Standardized thresholds for the abundance of opportunistic pathogens in healthy controls will aid clinicians in result interpretation but require future multicen ter large-scale studies and time to accumulate relevant data.
The RP-MT-Capture NGS assay was primarily developed to address the limitations of existing high-throughput detection technologies (e.g., high cost and low sensitivity of mNGS, as well as limited numbers of detection targets and samples for TaqMan array). Therefore, for performance evaluation, we mainly compared the detection perform ance of RP-MT-Capture NGS with that of mNGS and TaqMan array, rather than with traditional detection technologies. Although RP-MT-Capture NGS outperforms mNGS in detection sensitivity and TaqMan array in sample detection throughput, it still has certain limitations. First, the pathogen detection capability relies on the designed RP panel and cannot achieve full coverage of all pathogens. For instance, regarding rhinoviruses with over 120 serotypes, the RP panel cannot identify all the subtypes of rhinoviruses (36). This is related to the size of the target regions covered by the probes and the annotation of the reference genomes in the database. Some reference genomes are not clearly typed. In the future, the coverage of multi-type pathogens (such as rhinoviruses) can be improved by increasing probe density or optimizing database annotation. Second, the evaluation of the performance of RP-MT-Capture NGS in clinical sample detection was based on a relatively small sample size, which was also from a single research center. This may compromise the representativeness and generalizability of the study findings. Currently, we have begun to conduct multi-center validation of this assay to establish robustness. Finally, the library preparation and MT-Capture processes in this study were performed manually, involving numerous experimental steps. Automation of the workflow through an automated library preparation instrument would further reduce experimental time, minimize human-induced errors, and improve the result stability.
Collectively, the RP-MT-Capture NGS assay developed in this study offers a novel technical tool for detecting multiple respiratory pathogens. It allows simultaneous detection of most common respiratory pathogens, with substantial clinical and public health relevance.
## MATERIALS AND METHODS
## Clinical samples
A total of 236 nasopharyngeal swab or sputum samples collected by Jiangsu Pro vincial Center for Disease Control and Prevention were included in this study, com prising 186 samples from acute respiratory infection (ARI) patients and 50 samples from healthy control (HC) subjects. All samples were stored at -80°C until use. All procedures conducted in this study involving human materials were approved by the Ethics Committee of Jiangsu Provincial Center for Disease Control and Prevention, and informed consent was obtained from each participant involved. All the experiments were carried out in accordance with the Declaration of Helsinki.
## Analytical sensitivity and pathogen typing capability
To assess the analytical sensitivity of the RP-MT-Capture NGS assay, representative positive respiratory clinical samples were selected, including RNA viruses (influenza B virus, human coronavirus OC43), DNA viruses (human Alphaherpesvirus, human adenovirus), bacteria (Klebsiella pneumoniae, Pseudomonas aeruginosa), and Myco plasma pneumoniae. Nucleic acids were extracted from these samples, serially diluted, and analyzed using the RP-MT-Capture NGS assay, with parallel real-time PCR analysis for comparison. To evaluate the pathogen typing capability of the RP-MT-Capture NGS assay, clinical samples confirmed by real-time PCR to be positive for different subtypes of influenza viruses (H5N1 and H7N9), respiratory syncytial viruses (types A and B), human parainfluenza viruses (types 2, 3, and 4), and human adenoviruses (types 5 and 7) were selected for RP-MT-Capture NGS testing.
## Performance of full-length sequencing of influenza HA and NA genes
To evaluate the performance of the RP-MT-Capture NGS assay for full-length detection of influenza virus HA and NA genes, nucleic acids from clinical samples confirmed by real-time PCR to be positive for H1N1, H3N2, and H7N9 were selected. After serial dilution, the samples were analyzed using both RP-MT-Capture NGS and real-time PCR.
## Evaluation of RP-MT-Capture NGS with clinical specimens
To comprehensively assess the performance of RP-MT-Capture NGS in the identification of pathogens in clinical samples, we analyzed 236 samples, encompassing 186 acute respiratory infection (ARI) specimens and 50 healthy controls (HC). To further evaluate the detection accuracy of RP-MT-Capture NGS, 159 samples were randomly selected from the 186 ARI samples and subjected to parallel analysis by a TaqMan Respiratory Tract Microbiota Comprehensive Card (Thermo Fisher Scientific Inc., USA). The full panel of detectable pathogens for this card is detailed in Table S5. Discrepant results were resolved via real-time PCR reconfirmation. Additionally, among RP-MT-Capture NGS-pos itive samples, we chose 38 specimens with high pathogen loads for mNGS analysis. This enabled a direct comparison of pathogen identification efficacy between the two methods.
## Bioinformatics analysis
The sequencing data results were processed and analyzed via an automated analysis platform. The main workflow encompassed the following steps: tag identification and data splitting, matching data to corresponding samples; the data were aligned to the human reference (hg19) and classification reference database using Burrow-Wheeler Aligner (version 0.7.17-r1188), and human reads were filtered. For the establishment of the Reference Gene Database: first, based on the target genes of the designed probes, a BLAST search was conducted on NCBI, relevant reference genomes were downloaded, screened to remove sequences that did not meet the requirements, and then a reference gene database was constructed (Table S6). The sequencing data were compared with the constructed database to identify the matching reference genomes. Then, the sequence with the highest similarity was analyzed in combination with the pathogen list, so as to determine the type of detected pathogens.
In the sequencing data result report, Unique reads were defined as sequencing reads that uniquely and unambiguously align to a specific location in the reference sequences during the alignment process. Using this metric as the number of reads detected for a species provided a more reliable reflection of the true expression level or coverage of a specific gene or region, thereby avoiding noise introduced by multiple alignments. Subsequently, the number of detected Unique reads was normalized using RPM (Reads Per Million mapped reads), calculated as unique RPM = (unique reads × 10 6 )/ total mapped reads. This normalization step further eliminated the impact of varying sequencing depths across samples, facilitating more accurate comparisons between samples. Finally, the value of log10 (Unique RPM) was calculated by taking the base-10 logarithm of the Unique RPM. This transformation converted the skewed data distribu tion into one that is closer to a normal distribution, reducing the influence of extreme values on the data and providing a better representation of the data distribution.
## Statistical analysis
All statistical analyses were performed with SPSS software version 23.0. Categorical variables were expressed as percentages and compared using the chi-square test. The nonparametric Mann-Whitney U test was used to compare the levels of the relative abundances of opportunistic pathogens between HC and ARI groups. P < 0.05 was considered statistically significant.
## References
1. (2000) *Influenza A virus H*
2. "Human Respiratory Syncytial Virus ML8 Human Respiratory Syncytial Virus A Human Respiratory Syncytial Virus"
3. "ML11 Human Parainfluenza Virus 3 Human parainfluenza virus"
4. "Human gammaherpesvirus 4 5 3.1 ML12 Human Parainfluenza Virus 4 Human parainfluenza virus"
5. Troeger, Blacker, Khalil et al. (2016) "Estimates of the global, regional, and national morbidity, mortality, and aetiologies of lower respiratory infections in 195 countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study" *Lancet Infect Dis*
6. Calderaro, Buttrini, Farina et al. (2022) "Respiratory tract infections and laboratory diagnostic methods: a review with a focus on syndromic panel-based assays" *Microorganisms*
7. Gao, Cao, Hu et al. (2013) "Human infection with a novel avian-origin influenza A" *N Engl J Med*
8. Wang, Horby, Hayden et al. (2020) "A novel coronavirus outbreak of global health concern" *Lancet*
9. Cohen, Levy, Rybak et al. (2023) "Immune debt: recrudescence of disease and confirmation of a contested concept" *Infect Dis Now*
10. Tang, Lai, Chao (2022) "Changing epidemiology of respiratory tract infection during COVID-19 pandemic" *Antibiotics (Basel)*
11. Billard, Bont (2023) "Quantifying the RSV immunity debt following COVID-19: a public health matter" *Lancet Infect Dis*
12. Leung (2021) "Transmissibility and transmission of respiratory viruses" *Nat Rev Microbiol*
13. Liu (2025) "Isothermal nucleic acid amplification technologies and CRISPR-Cas based nucleic acid detection strategies for infectious disease diagnostics"
14. Zeng, Lloyd, Hamdy et al. (2025) "Identification and characterization of an invasive, hyper-aerotolerant Campylobacter jejuni strain causing bacteremia in a pediatric leukemia patient" *ASM Case Rep*
15. Lin, Zhai, Liu et al. (2025) "Metabolism-driven colorimetric "read-to-answer" sensor array for bacterial discrimination and antimicrobial susceptibility testing" *Anal Chem*
16. Zheng, Duan R Hua, Gong et al. (2022) "Accuracy of serological tests for COVID-19: a systematic review and meta-analysis" *Front Public Health*
17. Chiu, Tien, Lin et al. (2022) "Detection of respiratory pathogens by application of multiplex PCR panel during early period of COVID-19 pandemic in a tertiary hospital in Central Taiwan" *J Microbiol Immunol Infect*
18. Chen, Feng, Ye et al. (2021) "Application of metage nomic next-generation sequencing in the diagnosis of pulmonary infectious pathogens from bronchoalveolar lavage samples" *Front Cell Infect Microbiol*
19. Li, Wang, Zhang et al. (2021) "Etiology of severe pneumonia in children in alveolar lavage fluid using a high-throughput gene targeted amplicon sequencing assay" *Front Pediatr*
20. Ju, Zhou, Deng et al. (2024) "Low-cost high-throughput targeted sequencing for the accurate detection of respiratory tract pathogens" *J Infect Dev Ctries*
21. Samorodnitsky, Jewell, Hagopian et al. (2015) "Evaluation of hybridization capture versus amplicon-based methods for whole-exome sequencing" *Hum Mutat*
22. Qin, Wu, Zhang et al. (2023) "Effects of error, chimera, bias, and GC content on the accuracy of amplicon sequencing"
23. Daviña-Núñez, Pérez, Cabrera-Alvargonzález et al. (2024) "Performance of amplicon and capture based next-generation sequencing approaches for the epidemiological surveillance of Omicron SARS-CoV-2 and other variants of concern" *PLoS One*
24. Schultzhaus, Wang, Stenger (2021) "CRISPR-based enrichment strategies for targeted sequencing" *Biotechnol Adv*
25. Meng, Wang, Yu et al. (2024) "A novel fast hybrid capture sequencing method for high-efficiency common human coronavirus whole-genome acquisition"
26. Sun, Dai, Shan et al. (2024) "Pathogen characteristics of respiratory infections in the season after the COVID-19 pandemic between August and December 2023: evidence from direct-to-consumer testing-based surveillance in Guangzhou and Beijing" *Int J Infect Dis*
27. Baral, Saini, Kandpal et al. (2024) "The interplay of co-infections in shaping COVID-19 severity: expanding the scope beyond SARS-CoV-2" *J Infect Public Health*
28. Singh, Kulsum, Rufai et al. (2020) "Mutations in SARS-CoV-2 leading to antigenic variations in spike protein: a challenge in vaccine development" *J Lab Physicians*
29. Zhao, Zhang, Zhang (2024) "Application of metagenomic nextgeneration sequencing in the diagnosis of infectious diseases" *Front Cell Infect Microbiol*
30. Qian, Wang, Zhou et al. (2020) "Improving pulmonary infection diagnosis with metagenomic next generation sequencing" *Front Cell Infect Microbiol*
31. Li, Jiang, Liu et al. (2025) "Targeted next-generation sequencing for antimicrobial resistance detection in ventilator-associated pneumonia" *Front Cell Infect Microbiol*
32. Yang, Ding, Lin et al. (2023) "A tool to automatically design multiplex PCR primer pairs for specific targets using diverse templates" *Sci Rep*
33. Yang, Wang, Zhou et al. (2015) "High-throughput sequencing of 16S rDNA amplicons characterizes bacterial composition in bronchoalveolar lavage fluid in patients with ventilator-associated pneumonia" *Drug Des Devel Ther*
34. Haas, Mcgregor, Bouhaddou et al. (2023) "Proteomic and genetic analyses of influenza A viruses identify pan-viral host targets" *Nat Commun*
35. Yelagandula, Bykov, Vogt et al. (2021) "Multiplexed detection of SARS-CoV-2 and other respiratory infections in high throughput by SARSeq" *Nat Commun*
36. Gaston, Miller, Fissel et al. (2022) "Evaluation of metagenomic and targeted nextgeneration sequencing workflows for detection of respiratory pathogens from bronchoalveolar lavage fluid specimens" *J Clin Microbiol*
37. Doxey, Mazen, Homm et al. (2025) "Metatranscriptomic profiling reveals pathogen and host response signatures of pediatric acute sinusitis and upper respiratory infection" *Genome Med*
38. Mai, Chen, Zhang et al. (2025) "Construction of a pathogenic microorganism detection method based on thirdgeneration nanopore sequencing data" *BMC Infect Dis*
39. Nienhold, Mensah, Frank et al. (2022) "Unbiased screen for pathogens in human paraffin-embedded tissue samples by whole genome sequencing and metagenomics" *Front Cell Infect Microbiol*
40. Li, Yu, Wan et al. (2024) "Human respiratory organoids sustained reproducible propagation of human rhinovirus C and elucidation of virus-host interaction" *Nat Commun* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645981&blobtype=pdf | # The role of chikungunya virus capsid-viral RNA interactions in programmed ribosomal frameshifting
Jordan Farrington, Erin Rooney, Richard Hardy
## Abstract
Chikungunya virus (CHIKV) is a globally significant arthropod-borne virus that causes outbreaks in tropical and subtropical regions and is rapidly expanding in geographical range. Infection can lead to acute febrile illness with debilitating joint pain that may persist as chronic arthritis. Efficient CHIKV replication and virion assembly rely on coordinated interactions between viral RNA and the capsid protein, a structural component essential for particle formation. In addition to its well-established structural role, the alphavirus capsid protein has been shown in Sindbis virus to bind viral RNA and stabilize the genome, while in New World alphaviruses, it contributes to immune evasion through inhibition of host gene expression. However, its potential regulatory functions in viral RNA processes remain less understood. Here, we show that CHIKV capsid modu lates programmed ribosomal frameshifting (PRF) within the 6K/Transframe (TF) region, influencing the balance between structural protein synthesis and TF production. Using a previously engineered CHIKV mutant (9900), which contains silent mutations that reduce capsid binding within 6K/TF, we demonstrate that diminished capsid-vRNA interaction correlates with increased PRF efficiency. This is associated with enhanced viral replica tion in immune-competent cells, yet this enhanced replication is not further improved by JAK-STAT inhibition, indicating that capsid binding already modulates immune response pathways. These findings identify CHIKV capsid as a potential modulator of PRF, expanding our understanding of alphavirus gene expression and RNA-protein interac tions.
IMPORTANCEThe multifunctional roles of the chikungunya virus (CHIKV) capsid protein-particularly its RNA-binding properties and potential to influence translationrepresent important aspects of viral replication and pathogenesis. This study focuses on the CHIKV capsid's influence on PRF, a key step in viral protein synthesis. Simi lar capsid-vRNA binding regions have been described in other alphaviruses, such as Venezuelan equine encephalitis virus (B. D.
characterized by fever, rash, and debilitating joint pain that can persist for months to years (3,4). As global temperatures rise and Aedes mosquito habitation zones broaden, the risk of CHIKV outbreaks in naïve populations increases, underscoring the need for continued surveillance and research (5)(6)(7).
CHIKV has a single-stranded, positive-sense RNA genome with two large open reading frames (ORFs). The nonstructural ORF encodes replication proteins, while the structural ORF-expressed from a subgenomic RNA (sgRNA)-produces proteins essential for virion assembly (8)(9)(10). A key structural component is the capsid protein, derived from the sgRNA, which packages the viral genomic RNA into nucleocapsid cores (11)(12)(13)(14)(15). Beyond its structural role, the alphavirus capsid protein is a well-documented RNA-binding protein, as seen in Sindbis virus (SINV) and Venezuelan equine encephali tis virus (VEEV) (16,17). Its interaction with viral RNA (vRNA) contributes to genome stability and, in combination with host proteins, can modulate infection dynamics (16,17). With the increasing use of advanced RNA-binding analyses such as cross-linked immunoprecipitation and high-throughput sequencing (CLIP-Seq), cross-link-assisted mRNP purification, and RNA-interactome capture, the repertoire of host and viral RNAs interacting with alphavirus capsid proteins is expected to expand (18)(19)(20).
Building on these insights, our previous work mapped CHIKV capsid protein associations to the vRNA and explored the functional consequences of disrupting these interactions (21). Using CLIP-Seq, we identified a strong interaction between the CHIKV capsid protein and the 6K/Transframe (TF) coding region of the vRNA (21). However, the specific functional impact of capsid binding to this region remained unclear. Notably, this binding site overlaps a regulatory region involved in programmed ribosomal frameshift ing (PRF), a mechanism that finely regulates protein synthesis by enabling ribosomes to shift reading frames with temporal and stoichiometric precision (22,23).
In alphaviruses, the canonical polyprotein derived from the sgRNA includes 6K and the downstream glycoprotein E1 (8). However, when -1 PRF occurs at the 6K/TF junction, translation of 6K and E1 is reduced in favor of TF production (23,24). TF, a viroporin, enhances viral replication and virion maturation while also interfering with the host interferon (IFN) response through mechanisms that remain incompletely understood (25)(26)(27). While neither 6K nor TF is strictly required for viral proliferation, both contribute to efficient viral propagation (28,29).
Given the critical role of PRF in regulating the balance between 6K and TF synthesis, we hypothesized that capsid binding to the 6K/TF region may influence PRF efficiency and, in turn, impact CHIKV replication. While CHIKV capsid is well known for its struc tural and RNA-binding roles, its potential to modulate translational processes like PRF has not been explored. Here, we combine mutational analysis, dual-luciferase assays, and infection studies in immune-competent cells to investigate this interaction. These experiments reveal a previously unrecognized function for capsid in influencing PRF and offer insight into how alphaviruses may fine-tune gene expression in response to host immune pressure.
## RESULTS
## Capsid binding, not RNA sequence alone, modulates frameshifting
We previously mapped a prominent CHIKV capsid binding site to the 9900 region, which is located 12 bases upstream of a site containing essential elements for PRF in the 6K/TF coding sequence (Fig. 1A) (21). PRF is a mechanism that enables a nonrandom shift of the ribosome into alternative reading frames, mediated by two critical elements: a heptanucleotide sequence (slip site) and a downstream RNA secondary structure (Fig. 1A) (30).
Translation of alphavirus sgRNA produces a polyprotein that is processed into the capsid, E3, E2, 6K, and E1 structural proteins (Fig. 1B) (23). However, when a frameshifting event occurs, translation results in the production of a polyprotein processed into capsid, E3, E2, and TF proteins instead (Fig. 1B) (23). During frameshifting, the ribosome pauses on the slip site as it unwinds the downstream RNA structure. It then shifts back by one nucleotide, entering the -1 reading frame and accessing the TF coding sequence (30). Given the proximity of the 9900 region to the PRF signals in the 6K/TF coding region, we hypothesized that capsid binding might influence frameshifting events. To disrupt the capsid-vRNA interaction, we used the 9900 mutant sequence, which was engineered with silent mutations in the 6K/TF region to weaken this interaction, as previously shown by quantitative immunoprecipitation (qIP) following viral infection (Fig. 1B) (21).
To investigate frameshifting, we leveraged a dual-luciferase plasmid system to quantify frameshift efficiency across various synthetic DNA constructs (31,32). In this system, the CHIKV 6K/TF sequence (nucleotides 9785-10022 from CHIKV strain 181/clone 25) is inserted between the zero-frame renilla luciferase and -1-frame firefly luciferase coding sequences (Fig. 2A). The renilla luciferase serves as a control for translation efficiency, while the firefly luciferase acts as a reporter for TF synthesis, which occurs when the ribosome shifts into the -1 reading frame. This design allows us to measure the ratio of firefly to renilla luciferase activity to determine frameshift efficiency. To prevent interference with luciferase activity and ensure proper separation of the two reporter proteins, foot-and-mouth disease virus 2A peptide sequences were placed after Full-Length Text the renilla luciferase and test insert (Fig. 2A). This design minimizes potential effects of the test sequence on enzymatic function, ensuring consistent comparison between the in-frame control (explained below) and experimental constructs (33,34). Control constructs include the slip site knockout (SSKO), which prevents slippage at the heptanucleotide sequence; a 5′ stop codon to test for readthrough to the -1 frame encoding firefly luciferase; and a 3′ stop codon to evaluate for cryptic splice sites introduced by the test sequence (Fig. 2A) (35). In addition, an in-frame firefly luciferase control was generated for each test sequence. In this control, the firefly luciferase coding sequence was placed directly downstream of renilla luciferase and the inserted 6K/TF sequence, maintaining the 0-reading frame throughout (Fig. 2A) (35). This configuration allows uninterrupted translation of both luciferases and produces the maximal firefly signal expected in the absence of a frameshift event. Firefly activity from these in-frame constructs was used to normalize the firefly signal from experi mental constructs, enabling accurate calculation of frameshift efficiency. Frameshifting efficiency was calculated using the formula:
and then expressed relative to the corresponding in-frame firefly control (35). Together, these controls ensured that the observed frameshift is due to the 6K/TF sequence itself rather than artifacts of the assay system.
Given the sensitivity of PRF to RNA structure and potential interactions with host or viral factors, we first evaluated our wild type (WT), 9900, and related control con structs in a minimal, cell-free translation system using rabbit reticulocyte lysates (RRL). This approach not only allowed us to study PRF in the absence of viral proteins but also provided a controlled context, reducing the impact of host interactions. Since the 9900 mutations could conceivably influence RNA structure or host protein interactions, isolating their direct impact on translation in this simplified system was essential.
Renilla luciferase, serving as the translation control, consistently exhibited levels at least 10-fold higher than background, confirming translation of all constructs (Fig. 2B; Fig. S1A). Firefly luciferase, representing the frameshift product and serving as a proxy for TF synthesis, showed comparable levels between CHIKV and 9900 constructs (Fig. 2C; Fig. S1B). Analysis of frameshifting efficiency indicated that, in this system, frameshifting occurred in 7%-9.5% of translational events for both the WT and 9900 constructs with no statistically significant difference in frameshifting between the two sequences (Fig. 2D; Fig. S1C). In all cases, frameshifting frequency on control RNAs was significantly lower (0%-1.2%).
$$Firefly RLU-Mock Firefly RLU Renilla RLU-Mock Renilla RLU × 100 = % PRF,$$
## CHIKV capsid binds the 6K/TF region in reporter constructs.
While the RRL system provided a controlled environment to assess intrinsic frameshifting efficiency, it lacked the viral and cellular factors that could influence PRF modulation in a biological context. To explore capsid's role in modulating PRF in CHIKV, we incorporated it into the dual-luciferase system to evaluate its impact on frameshifting efficiency. CHIKV capsid expression was driven by an encephalomyocarditis virus (EMCV)-derived internal ribosome entry site (IRES) element positioned downstream of the frameshifting cassette to enable translation of capsid without dependence on upstream frame continuity. Since the antibody we had for capsid showed significant non-specific interactions (not shown), the capsid protein expressed from our construct was tagged with a hemagglutinin (HA) epitope at its N-terminus, allowing clean detection ("trans factor, " Fig. 3A) (36).
To confirm capsid's binding to the 6K/TF region of the reporter constructs and to distinguish this from capsid's known tendency for non-specific RNA interactions, we performed qIP in HEK293T cells transfected with the dual-luciferase constructs. RNA recovery from the IP was quantified and normalized for both WT HA-capsid (representing capsid binding to the RNA) and 9900 HA-capsid (which has disrupted capsid binding due to silent mutations). The retention of the capsid-RNA interaction was significantly reduced (approximately fourfold) in the 9900 construct compared to the WT, confirming that WT capsid efficiently binds to the RNA derived from the dual luciferase constructs, whereas the mutations in the 9900 construct impair this interaction (Fig. 3B).
To further validate these findings, we replaced the trans factor by expressing the SINV capsid protein, rather than CHIKV capsid (Fig. 3A). Consistent with the earlier results, SINV capsid did not efficiently bind the WT 9900 region as well as the CHIKV capsid (Fig. 3B), mirroring the reduced interaction seen with CHIKV 9900 HA-capsid and supporting the specificity of the capsid-RNA interaction.
Importantly, the levels of capsid protein pulled down in these experiments were similar for each construct (Fig. 3C), indicating that the difference in RNA binding was not due to differential protein expression but rather due to the disrupted binding capability of the capsid protein. This result supports our earlier findings, where we observed reduced RNA recovery following capsid IP in the 9900 virus compared to the WT virus in a full viral context (21). These data support CHIKV capsid binding to RNA containing the 6K/TF region, consistent with prior mapping (21), and the 9900 mutations effectively disrupt this interaction without altering capsid protein expression.
## Capsid suppresses -1 PRF via binding to the 6K/TF region
The binding of the capsid protein to our dual luciferase constructs confirms its interac tion with CHIKV dual luciferase-derived RNA specifically, establishing a foundation for further investigation into its modulatory effects on PRF. This binding interaction provides a direct means to assess how capsid influences PRF efficiency within the context of our reporter system.
Following transfection of the capsid-expressing reporter constructs in HEK293T cells, renilla luciferase activity (reflecting 0-frame translation) was consistent across all constructs with at least a 100-fold increase compared to background readings (Fig. 4A; Fig. S2A), confirming RNA integrity and proper translation. Firefly luciferase activity, indicating frameshift product formation, revealed differences in alternative reading frame translation among the constructs (Fig. 4B; Fig. S2B). Frameshifting efficiency, calculated as the firefly-to-renilla ratio and normalized by the respective CHIKV capsidcontaining in-frame control, was lowest for WT IRES-HA capsid (Fig. 4C; Fig. S2C-D) with an average of 2.67%. However, the 9900 IRES-HA capsid construct, which disrupts capsid binding through silent mutations in the RNA, showed significantly higher frameshifting efficiency (6.15%) on average compared to WT IRES-HA capsid. This suggests that the absence of capsid binding relieves suppression of PRF.
When SINV capsid, rather than CHIKV capsid, was present in the reporter system (Fig. 3A), frameshifting efficiency averaged 5.74% when normalized to the CHIKV firefly inframe construct, statistically indistinguishable from that seen for 9900 with CHIKV capsid (Fig. 4C). These findings demonstrate that CHIKV capsid binding to the WT 9900 region could play a key role in restricting PRF. Replacing CHIKV capsid with non-binding SINV capsid (Fig. 3B) or introducing silent mutations in the 9900 region that reduce capsid-RNA interaction consistently increases -1 PRF efficiency at the 6K/TF site. These findings indicate that capsid binding can suppress frameshifting, highlighting a modulatory role for CHIKV capsid in fine-tuning viral protein synthesis and potentially shaping replication dynamics and immune evasion strategies.
To ensure that the increased frameshifting observed in the 9900 construct was not due to changes in transcript abundance, we performed absolute quantification of renilla and firefly RNA levels by qRT-PCR using standard curves. RNA expression was equivalent between the CHIKV and 9900 dual luciferase constructs (Fig. S3). This confirms that changes in PRF efficiency reflect translational modulation rather than differences in transcript abundance.
## Disruption of capsid binding enhances replication in immune-competent cells
To assess the biological consequences of capsid-mediated PRF influence, we next examined viral replication kinetics of WT and 9900 CHIKV in both IFN-deficient and immune-competent cell types. To directly explore the consequences of capsid binding to the 6K/TF region in CHIKV and 9900 infectivity, we infected BHK cells with mKate-tagged WT and 9900 viruses at a high multiplicity of infection (5 PFU/cell). BHK cells are impaired in key antiviral pathways, namely type I interferon (37), which makes them an ideal model for studying viral replication and particle production without the influence of antagonistic host immune factors. We found that WT and 9900 viruses replicate with similar kinetics at a high multiplicity of infection (MOI), suggesting that the presence of the mutations does not significantly impact entry, replication efficiency, or assembly of virions compared to WT virus under these conditions (Fig. 5A).
After observing no notable differences in BHK cells, we chose to infect a cell line better equipped to mount an innate immune response in order to more accurately replicate a typical mammalian host environment. HEK293T cells were selected for this purpose due to their ability to initiate IFN-I signaling (38,39). Cells were infected at 5 PFU/cell with the mKate-tagged CHIKV and 9900 viruses to investigate how immune defense barriers are navigated when overwhelmed, and the resulting effects on viral replication. Notably, the replication of the 9900 mutant was observed 4 hours earlier than the WT virus, reaching its peak within 24 hours, while the WT had not yet done so (Fig. 5B). This suggests that the mutant virus may overcome immune-mediated barriers more rapidly.
Expectedly, infection with both viruses in BHK cells resulted in a comparable number of infectious virions in the medium at 24 hours post-infection (HPI), as determined by qRT-PCR detection of viral genomes and plaque assay titration on BHK-21 cells (Fig. 5C through E). Interestingly, in HEK293T cells under the same conditions, the number of virions was similar for both viruses (Fig. 5C). However, the 9900 mutant achieved ~1.5 log 10 higher titer (Fig. 5D), implying there were more infectious virions. As a result, the ratio of infectious particles to total particles increased from approximately 1:50 in WT infection to 1:5 in 9900 infection (Fig. 5E). This suggests that weakened capsid binding may play a critical role in infectivity in immunocompetent cells. These results suggest that enhanced TF synthesis may contribute to more efficient immune evasion or viral entry dynamics in HEK293T cells, potentially through elevated TF synthesis that improves immune evasion or alters viral entry dynamics.
## Ablation of PRF negates 9990 growth phenotype
To determine whether the observed phenotype depends on PRF-mediated TF produc tion, we next generated slip site knockout versions of both the WT and 9900 viruses. The SSKO mutation consisted of silent changes to the heptanucleotide slippery sequence required for -1 PRF, thereby aiming to prevent ribosomal frameshifting without altering the downstream TF coding region (Fig. 2A) (35). This design was meant to maintain all other aspects of the genomes explored previously while eliminating TF expression. The rationale for this approach was based on the proposed model in which reduced capsid binding in the 9900 mutant relieves WT levels of apparent suppression of PRF, thereby elevating TF levels and enhancing replication in the presence of an intact interferon response. If TF production is the key driver of the observed phenotype, abolishing PRF in the 9900 background should remove its replication advantage and yield kinetics comparable to WT SSKO virus.
Replication kinetics of the SSKO viruses were assessed in HEK293T cells at a high multiplicity of infection (5 PFU/cell). Both WT SSKO and 9900 SSKO viruses exhibited nearly identical replication curves, with similar onset times, slopes, and peak infection plateaus (Fig. 6A). Importantly, the accelerated spread seen for the 9900 mutant was absent in the SSKO background, confirming that the enhanced replication of 9900 requires an intact PRF site and is likely not attributable to other features of the 9900 mutations.
## Role for PRF in antagonizing interferon signaling
Given that the 9900-growth advantage is abolished when PRF is disrupted, we next tested whether its enhanced replication in immune-competent cells reflects resistance to type I IFN signaling. TF has been implicated as an antagonist of the IFN response in alphaviruses (27), and our model predicts that elevated TF production in the 9900 mutant could diminish the effectiveness of JAK-STAT-mediated antiviral pathways.
To directly assess this, we pharmacologically inhibited the JAK-STAT pathway with ruxolitinib, a potent and selective kinase inhibitor that blocks STAT1 phosphorylation and downstream interferon stimulated gene (ISG) induction (40,41). HEK293T cells were pretreated with 300 nM Ruxolitinib or vehicle (0.1% dimethyl sulfoxide [DMSO]) for 24 h prior to infection with WT CHIKV or the 9900 mutant (MOI 0.1 PFU/cell). After adsorption, treatments were maintained for the duration of infection. At 48 HPI, total RNA was collected for qRT-PCR analysis of interferon pathway activity, and infectious virus titers were determined by plaque assay.
Consistent with prior observations, WT CHIKV infection alone induced robust expression of ISG15, a canonical ISG (Fig. 6B) (42). Interestingly, while the 9900 mutant also induced ISG15 expression, its induction was markedly lower than that of CHIKV (by ~2.5-fold) and was further diminished by ruxolitinib, confirming that both viruses were responsive to JAK-STAT inhibition at the transcript level (Fig. 6B). These findings suggest that the 9900 mutant partially circumvents IFN-mediated transcriptional activation even in the absence of pharmacological inhibition.
The impact on viral replication mirrored these transcriptional patterns. Inhibition of the JAK-STAT pathway increased CHIKV infectious titers by twofold, consistent with the role of interferon signaling in restricting viral replication (Fig. 6C) (42). In contrast, ruxolitinib had no detectable effect on replication of the 9900 mutant, which remained high regardless of treatment (Fig. 6C). These results suggest that the 9900 mutant replicates efficiently despite intact IFN signaling. While TF expression is a potential contributor to this effect, how TF antagonizes the IFN response remains unclear. Elucidating this mechanism in future work would allow us to build directly on the findings presented here.
## DISCUSSION
In this study, we demonstrate that the CHIKV capsid protein modulates programmed ribosomal frameshifting at the 6K/TF region of the viral RNA by binding near the -1 PRF site. Disrupting this interaction through silent mutations in the 9900 region increases frameshifting efficiency, suggesting enhanced TF production, and confers a replication advantage in immune-competent cells. As we were unable to reliably detect TF using available antibodies, we used dual-luciferase reporters to estimate TF synthesis via frameshifting efficiency. These findings establish a novel post-transcriptional role for capsid in modulating the balance of structural and accessory protein expression during CHIKV infection.
## Capsid binding suppresses PRF at the 6K/TF site
A major conclusion of this work is that CHIKV capsid partially suppresses -1 PRF at the 6K/TF junction by binding to an upstream vRNA sequence (Fig. 7A, left panel). This capsid-mediated interaction is suggested here to reduce ribosomal slippage and TF synthesis, as evidenced by the increased frameshifting efficiency observed in the 9900 mutant (Fig. 7A, right panel). Our dual-luciferase reporter data show that WT capsid reduces PRF frequency compared to both the binding-deficient 9900 mutant and SINV capsid, supporting a model in which capsid-vRNA binding contributes to PRF modula tion in a virus-specific manner. Even modest changes in PRF (e.g., from ~3% to ~6%) can lead to significant differences in TF abundance, especially in the context of exponential translation amplification. As seen in other viruses, including HIV and coronaviruses (43,44), such differences can meaningfully affect host response modulation and viral fitness. This adds to growing evidence that alphavirus capsid proteins possess regulatory RNA-binding functions beyond structural roles, as has been observed in SINV and VEEV, where capsid binding to vRNA contributes to genome stability and translational control (16,17). Notably, a VEEV capsid-binding site has been identified in a comparable region of the genome (17), suggesting that this mode of translational modulation may be conserved across certain alphaviruses.
## Impact on viral replication and immune evasion
The increase in PRF and the inferred elevation in TF synthesis in the 9900 mutant correlates with enhanced replication in immune-competent cells. This is likely due to TF's known role in antagonizing type I IFN responses (27). Supporting this, the replication advantage of the 9900 mutant is lost in IFN-deficient BHK cells, and JAK-STAT inhibition boosts WT CHIKV replication but has no effect on the 9900 virus. Additionally, SSKO viruses, which aim to eliminate TF expression, replicate similarly in both WT and 9900 backgrounds, suggesting that the 9900 mutant's advantage depends on functional TF production. These findings imply that enhanced TF expression in the 9900 mutant interferes with the impact of interferon signaling, allowing replication to proceed even in the presence of intact immune responses. Importantly, these findings do not suggest that the WT virus is inherently less fit in natural settings; rather, they support a model in which CHIKV finely tunes PRF through capsid-RNA interactions to balance immune evasion with structural protein synthesis. Excessive TF production at the expense of 6K/E1 may disrupt glycoprotein stoichiometry or impair virion assembly, although direct evidence for these effects remains to be established. An additional consideration is the dual-host lifecycle of CHIKV. It remains plausible that capsid-mediated PRF modulation plays a more prominent role in the mosquito host, where different selective pressures may favor tight regulation of TF expression-an aspect not captured in mammalian cell culture models.
## Hypothesized model for PRF modulation
The mechanism by which CHIKV capsid suppresses PRF remains to be fully defined. One possibility is that capsid binding alters or stabilizes local RNA secondary structures near the PRF site, preventing efficient ribosome pausing or back slipping. Alternatively, capsid binding could sterically block ribosome movement at or near the PRF site. The precise nucleotide footprint of the CHIKV capsid on RNA is not yet defined; however, similar alphaviruses like SINV and VEEV suggest that capsid-RNA interactions typically span ~20-40 nucleotides (16, 17)-a range that could plausibly impact PRF efficiency if positioned near the stimulatory signals. We speculate that the functional consequence of this suppression is not to completely abolish TF expression, but rather to modulate it over the course of infection. Early in infection, when capsid levels are low, PRF efficiency may be higher, allowing transient TF production that could help suppress innate immune responses. As capsid accumulates, increasing capsid-RNA interactions could progressively reduce PRF, preventing excessive or prolonged TF expression that might disrupt viral protein stoichiometry or interfere with particle assembly. In this way, capsid-mediated PRF suppression could act as a feedback mechanism to fine-tune TF levels in a temporally regulated manner.
A mechanistically similar example is seen in SARS-CoV, where an upstream attenuator stem-loop positioned before the PRF structure physically impedes ribosome move ment and reduces back-slipping efficiency, thereby tuning frameshifting output (44,46). Whether CHIKV capsid directly alters RNA folding or exerts steric hindrance as a protein-based attenuator remains an open question-one that could be addressed using SHAPE-MaP or ribosome profiling approaches in future studies.
## Perspective and broader implications
Although disrupting capsid binding increases PRF and enhances replication in immunecompetent cells, the 9900 mutations are absent in circulating CHIKV isolates (Fig. 7B). Of note, we did not perform dN/dS or synonymous site conservation analysis and instead report that these mutations are not found in currently available sequences. This may indicate that CHIKV maintains a conserved balance between TF and 6K/E1 production to preserve glycoprotein stoichiometry (8). Alternatively, the absence of 9900 mutations could reflect host-specific or transmission-related constraints not captured in mamma lian cell culture. Further studies in Aedes mosquito cell lines could begin to answer some of these questions.
The involvement of a structural protein like capsid in influencing PRF highlights an additional layer of post-transcriptional control during the viral replication cycle. It also raises the possibility that similar capsid-RNA interactions may occur in other alphavi ruses, though further studies would be needed to determine their functional relevance.
## Conclusions
Our findings identify CHIKV capsid as a modulator of PRF at the 6K/TF region and demonstrate that disrupting this interaction likely enhances TF expression, replication, and some resistance to interferon signaling. This study broadens our understanding of alphavirus gene expression strategies and highlights capsid-vRNA interactions as potential regulatory checkpoints that balance immune evasion, structural protein production, and replication fitness.
## MATERIALS AND METHODS
## Cell culture
Baby hamster kidney fibroblasts (BHK-21) and modified human embryonic kidney (HEK293T) cells were grown at 37°C with 5% CO 2 in 1× minimal essential medium with 10% heat-inactivated fetal bovine serum, 1% L-glutamine, 1% non-essential amino acids, and 1% antibiotic-antimycotic solution (all purchased from Corning, Corning, NY, USA).
## Viruses
Plasmids encoding CHIKV strain 181/clone 25 (GenBank accession: MW473668.1), tagged with a mKate fluorophore and FMV2 element between capsid and E3, with either the wild type or 9900 mutated sequences and the respective SSKO viruses, were linearized with SacI (NEB, Ipswich, MA, USA). The inserted sequence includes the capsid coding region, followed by three amino acids from E3 to allow for autoproteolytic cleavage (ensuring mKate is not fused to capsid), the mKate reporter, and the FMV2 element, before the remaining portion of the structural polyprotein (47). The linearized plasmids were used for in vitro transcription (IVT) using SP6 RNA polymerase (NEB, Ipswich, MA, USA). IVTs were transfected into confluent BHK-21 cells using the LTX transfec tion reagent following the manufacturer's protocol (Invitrogen, Waltham, MA, USA) with Opti-MEM I reduced serum medium (Gibco, Waltham, MA, USA). Four hours after transfection, the medium was removed and replaced with fresh 1× MEM. Approximately 52 hours post-transfection, the supernatant was harvested and clarified by centrifugation at 7,000 × g for 5 minutes. Clarified viral supernatant was aliquoted and stored at -80°C.
Viruses were titered by standard plaque assay on confluent BHK-21 cells.
## Viral infection
BHK-21 and HEK293T cells grew to ~70% confluency in 12-well plates (Greiner Bio-One CellStar), then the original medium was discarded and replaced with 200 µL fresh 1× MEM. Cells were infected in triplicate with virus at an MOI of 5 PFU/cell and rocked for 1 hour at room temperature to allow for viral adsorption. The supernatant containing unbound virus and medium was aspirated, the cells were washed with 1× phosphatebuffered saline (PBS), and fresh 1× MEM was then added. The plates were then incubated at 37°C as described below.
## Viral spread by live cell imaging
Viral spread was measured via live cell imaging using the IncuCyte Live-Cell Analysis System (Essen Biosciences, USA) set to 37°C with 5% CO 2 . Briefly, BHK-21 and HEK293T cells were infected as described above, and images were captured from nine separate fields of view at 2-hour intervals over 24 hours. The results reflect the mean values from three independent biological replicates. The viral area measured by fluorescence generated by the mKate reporter was adjusted based on cell area.
## Specific infectivity
BHK-21 or HEK293T cells were infected with the indicated viruses at an MOI of 5 PFU/ cell and harvested 24 HPI. The supernatant was collected, clarified by centrifugation at 7,000 × g for 5 minutes, and either used immediately for cDNA synthesis or stored at 4°C for short-term use. To quantify total viral genomes, an equal volume of RNA was reverse transcribed into cDNA using MMuLV Reverse Transcriptase (NEB, Ipswich, MA, USA) with random hexamer primers (Integrated DNA Technologies, Coralville, IA, USA). A standard curve was generated from serial dilutions of the CHIKV 181/25 plasmid. Quantitative RT-PCR was conducted with the SensiFAST SYBR Hi-ROX Master Mix (Meridian Bioscience, Memphis, TN, USA) and primers specific to the CHIKV nsP2 coding region, following the manufacturer's protocol. Gene expression was measured using the Applied Biosystems StepOnePlus qRT-PCR system (Life Technologies, Carlsbad, CA, USA), with duplicate samples for both the standard curve and the experimental samples collected in biological triplicates. The Ct values from the standard curve were used to calculate the viral genomic copies in the samples. The infectious viral titer measured by standard plaque assay in BHK-21 cells was divided by the total genome copies to determine the PFU-to-particle ratio.
## Plasmid construction
Synthetic oligonucleotides were designed using SnapGene software (snapgene.com).
The dual-luciferase plasmid backbone, pJD2257 (32,48), was used for all constructs adjusted to prevent unintended splicing errors. The 6K/TF region of the CHIKV 181/25 genome (nt 9785-10022) with either the wild-type or mutant 9900 sequence was inserted between the renilla and firefly luciferase genes by PCR amplification with Q5 DNA polymerase (NEB, Ipswich, MA, USA), following the manufacturer's protocol. Amplified DNA was digested with DpnI (NEB) and purified using the DNA Clean and Concentrator-5 kit (Zymo Research, Irvine, CA, USA). Purified DNA was eluted in 10 µL of nuclease-free H 2 O and ligated with NEBuilder (NEB). The ligated products were transformed into in-house DH5α competent cells and grown on LB agar plates with ampicillin. Colonies were screened via PCR using Taq polymerase (NEB) with primers specific to the insert. Positive colonies were confirmed by whole plasmid sequencing through Plasmidsaurus using Oxford Nanopore Technology and custom annotation. The same approach was used to insert an EMCV-derived IRES element downstream of the dual luciferase coding region, in order to avoid any out-of-frame issues, followed by a hemagglutinin-tagged, CHIKV capsid sequence (nt 7541-8323), or SINV TE12 capsid sequence (NCBI accession: NC_001547.1; nt 7647-8630). Site-directed mutagenesis was performed to generate the SSKO viruses or mutated dual luciferase plasmids with PrimeSTAR HS DNA polymerase (Takara Bio USA Inc., San Jose, CA, USA), following the manufacturer's instructions, and transformed into homemade DH5α competent cells as described above.
## In vitro transcription/translation and dual luciferase assay
To evaluate translational efficiency and frameshifting dynamics, dual luciferase plasmids were incubated with the T7 TnT Quick Coupled Transcription/Translation System (Promega, Madison, WI, USA) according to the manufacturer's protocol, with a 90-minute incubation. A negative control reaction lacking plasmid DNA was included, and all samples and controls were performed in triplicate. The reaction products were subse quently analyzed using a dual luciferase assay. For the assay, 10 µL of translated products were combined with 50 µL of LARII (Promega) and incubated for 10 minutes before being transferred to a 96-well plate. Luminescence readings were taken with a BioTek Synergy LX Multimode Reader (Agilent) set for a 2-second pre-read delay followed by a 10-second measurement. Next, 50 µL of Stop and Glo Reagent (Promega) was added to each sample, and a second luminescence reading was taken.
Frameshifting ratios were calculated by first subtracting the raw firefly and renilla luciferase values from those of the mock controls, and the ratio was recorded using the formula: [(firefly RLU -mock firefly RLU)/(renilla RLU -mock renilla RLU)] 100. The resulting firefly-to-renilla ratio was then normalized to the corresponding ratio from the respective firefly in-frame control. All constructs were analyzed relative to their respective in-frame controls. Results represent the mean ± SEM from three biological replicates per sample.
## Transfection with dual luciferase plasmids
For cellular experiments, low-passage (<20 passages) HEK293T cells were seeded at ~40% confluency in six-well plates (Greiner Bio-One CellStar) and incubated overnight at 37°C. Midiprepped dual luciferase plasmids (Zymo Research) were prepared for transfection by combining 3,000 ng of DNA with 6 µL of FuGENE 4K Transfection Reagent (Fugene LLC) in 250 µL of Opti-MEM I reduced serum medium (Gibco). After a 15-minute incubation at room temperature, the transfection complexes were added dropwise to the cells in fresh 1× MEM. Cells were incubated at 37°C for 48 hours before harvesting.
At 48 hours post-transfection, cells were washed with 1× PBS and lysed using 1× passive lysis buffer (Promega). Lysates were rocked for 15 minutes at room temperature, transferred to 1.5 mL Eppendorf tubes, and clarified by centrifugation at 14,000 × g for 10 minutes. The supernatant was either stored at -20°C or immediately used for the dual luciferase assay exactly as described.
## Quantitative immunoprecipitation
HEK293T cells were transfected with the indicated HA-tagged constructs and incubated for 48 hours. Following incubation, cells were washed with 1× PBS, and a minimal volume of 1× PBS was added to the plates. Cells were irradiated with 5,700 × 100 μJ/cm 2 in a Stratalinker to cross-link RNA-protein complexes. The cross-linked complexes were solubilized in RIPA buffer (50 mM Tris [pH 7.6], 150 mM NaCl, 1.0% NP-40, 0.5% sodium deoxycholate, and 0.1% SDS) and lysed on ice for 7 minutes. Lysates were vortexed and clarified by centrifugation at 16,000 × g for 10 minutes at 4°C. The clarified supernatants were transferred to fresh tubes. The cross-linked lysates were pre-cleared by incubating with 30 µL of 50% slurry Protein A Agarose Beads (9863S, Cell Signaling) for 1 hour at 4°C under constant agitation. The tubes were spun down at 5,000 × g for 5 minutes, and the pre-cleared lysates were transferred to fresh tubes. A total of 10% of the sample volume was saved as input reference and stored at -80°C.
For immunoprecipitation, Pierce Anti-HA magnetic beads were washed in 0.05% TBS-T (Thermo Fisher) and incubated with lysates at 4°C overnight with constant agitation. Immunoprecipitated RNA-protein complexes were purified by magnetic separation of the beads. The beads were washed three times with RIPA buffer, followed by two washes with sterile 1× PBS. The beads were resuspended in 1× PBS, and 10% of the total volume was saved for western blot analysis. The remaining beads were used for RNA analysis as described below.
To verify the IP was specific to HA-tagged capsid, beads were boiled in 6× SDS dye and subjected to western blot analysis alongside the corresponding input and unbound lysate samples. For RNA analysis, the purified bead-bound RNA-protein complexes and input samples were treated with Proteinase K and RNAse Inhibitor Murine at 55°C for 30 minutes with gentle agitation to release RNA fragments. RNA was extracted using the Aurum Total RNA Mini Kit (Bio-Rad Laboratories) and resuspended in 20 µL elution buffer. cDNA was synthesized, and qRT-PCR was performed as described in the specific infectivity section with primers targeting the 9900 region. RNA-IP values were normal ized first to GAPDH and then to corresponding input samples to control for differences in starting material. Normalized values were made relative to WT HA-capsid. All samples were collected and analyzed in biological triplicate.
## Western blotting
Five micrograms of whole cell lysates from the immunoprecipitation protocol was separated by 10% SDS-PAGE (Bio-Rad Laboratories, Hercules, CA, USA) and run at 150 V for 1 hour. The gel was transferred to a polyvinylidene difluoride (PVDF) membrane and blocked in 5% TBSTM (Tris buffered saline 1×, 5% dry milk, 0.1% Tween-20) for 1 hour at room temperature. Following blocking, the membrane was incubated overnight with rabbit anti-HA tag polyclonal antibody (C29F4) and rabbit anti-beta-actin (#4967) antibody (both from Cell Signaling Technology) at 4°C in 2.5% TBSTM. The membrane was washed three times with TBST before incubation with goat anti-rabbit AlexaFluor 750 (#A-21039 ThermoFisher, Waltham, MA, USA) diluted with 2.5% TBSTM for 1 hour at room temperature in the dark. Following a minimum of three washes with TBST, the membrane was imaged using a Bio-Rad ChemiDoc MP Imaging System. All western blot experiments were repeated at least three times independently with similar results. Representative images are shown.
## JAK-STAT inhibition with ruxolitinib
HEK293T cells were seeded in 24-well plates and allowed to adhere for 24 hours. The medium was then removed and replaced with fresh 1× MEM. Pre-treatment of half of the wells was done by adding either 0.1% DMSO or 300 nM ruxolitinib (#83405, Cell Signaling). After 24 hours of pre-treatment, the medium was removed, and cells were infected at 0.1 PFU/cell with either mKate-tagged CHIKV or 9900 viruses. Following the infection period, virus-containing medium was discarded, cells were washed with 1× PBS, and fresh 1× medium containing 0.1% DMSO or 300 nM ruxolitinib was added to the cells that received prior treatment. Remaining wells received 1× MEM.
At 48 HPI, the cell culture medium was harvested and spun down, and the superna tant was titered on BHK-21 cells. The HEK293T cells were lysed, RNA was extracted, and cDNA synthesis followed by qRT-PCR was performed as above with minor modifications. The primers used for qRT-PCR were GAPDH and ISG15, and relative expression was calculated using the delta-delta Ct method, with GAPDH as the reference control. All data were obtained from biological triplicates, and qRT-PCR was performed in technical duplicates.
## Multiple sequence alignment
All available CHIKV sequences with complete nucleotide profiles and full virus sequence length (>11,000 bases) were downloaded from the NCBI Virus database (ncbi.nlm.nih.gov/labs/virus/vssi; taxid: 37124). These sequences were aligned using SnapGene with MAFFT, and the nucleotide region corresponding to positions 9869-9913 (the 9900 region) was extracted from the alignment. The extracted sequences were then used as input for WebLogo (https://weblogo.berkeley.edu/logo.cgi) to visualize nucleotide variation across strains. This analysis was used to determine whether the mutations introduced at the 9900 region are present in any naturally occurring CHIKV sequences.
## Statistical analyses
All statistical analyses were performed using GraphPad Prism 9 (GraphPad Software Inc., San Diego, CA, USA). Data reflect a minimum of three biological replicates per sample in each assay. Dual luciferase data are reflective of the guidelines highlighted in references 35, 49.
## References
1. Monteiro, Navegantes-Lima, De Lemos et al. (2019) "Aedes-Chikungunya virus interaction: key role of vector midguts microbiota and its saliva in the host infection" *Front Microbiol*
2. Coffey, Failloux, Weaver (2014) "Chikungunya virus-vector interactions" *Viruses*
3. Staples, Breiman, Powers (2009) "Chikungunya fever: an epidemiological review of a re-emerging infectious disease" *Clin Infect Dis*
4. Schwartz, Albert (2010) "Biology and pathogenesis of chikungunya virus" *Nat Rev Microbiol*
5. Fredericks, Fernandez-Sesma (2014) "The burden of dengue and chikungunya worldwide: implications for the southern United States and California" *Ann Glob Health*
6. Weaver (2014) "Arrival of chikungunya virus in the new world: prospects for spread and impact on public health" *PLoS Negl Trop Dis*
7. Leta, Beyene, Clercq et al. (2018) "Global risk mapping for major diseases transmitted by Aedes aegypti and Aedes albopictus" *Int J Infect Dis*
8. Strauss, Strauss (1994) "The alphaviruses: gene expression, replication, and evolution" *Microbiol Rev*
9. Rupp, Sokoloski, Gebhart et al. (2015) "Alphavirus RNA synthesis and non-structural protein functions" *J Gen Virol*
10. Chen, Mukhopadhyay, Merits et al. (2018) "ICTV virus taxonomy profile: togaviridae" *J Gen Virol*
11. Frolova, Frolov, Schlesinger (1997) "Packaging signals in alphaviruses" *J Virol*
12. Mendes, Kuhn (2018) "Alphavirus nucleocapsid packaging and assembly" *Viruses*
13. Choi, Tong, Minor et al. (1991) "Structure of Sindbis virus core protein reveals a chymotrypsinlike serine proteinase and the organization of the virion" *Nature*
14. Choi, Lu, Lee et al. (1997) "Structure of Semliki Forest virus core protein" *Proteins*
15. Brown, Anastasakis, Hafner et al. (2020) "Multiple capsid protein binding sites mediate selective packaging of the alphavirus genomic RNA" *Nat Commun*
16. Sokoloski, Nease, May et al. (2017) "Identification of interactions between sindbis virus capsid protein and cytoplasmic vRNA as novel virulence determinants" *PLoS Pathog*
17. Carey, Akhrymuk, Dahal et al. (2020) "Protein Kinase C subtype δ interacts with Venezuelan equine encephalitis virus capsid protein and regulates viral RNA binding through modulation of capsid phosphorylation" *PLoS Pathog*
18. Westcott, Isom, Karki et al. (2023) "Dancing with the devil: a review of the importance of host RNA-binding proteins to alphaviral RNAs during infection" *Viruses*
19. Lapointe, Gebhart, Meller et al. (2018) "Identification and characterization of sindbis virus RNA-host protein interactions" *J Virol*
20. Garcia-Moreno, Noerenberg, Ni et al. (2019) "System-wide profiling of RNA-binding proteins uncovers key regulators of virus infection" *Mol Cell*
21. Kiser, Sokoloski, Hardy (2021) "Interactions between capsid and viral RNA regulate Chikungunya virus translation in a host-specific manner" *Virology (Auckl)*
22. Caliskan, Peske, Rodnina (2015) "Changed in translation: mRNA recoding by -1 programmed ribosomal frameshifting" *Trends Biochem Sci*
23. Ramsey, Mukhopadhyay (2017) "Disentangling the frames, the state of research on the alphavirus 6K and TF proteins" *Viruses*
24. Firth, Chung, Fleeton et al. (2008) "Discovery of frameshift ing in alphavirus 6K resolves a 20-year enigma" *Virol J*
25. (2025) *Full-Length Text Journal of Virology*
26. Suhag, Borkotoky, Siddiqui et al. (2025) "Mechanistic Insights into the divergent membrane activities of a viroporin from chikungunya virus and its transframe variant" *ACS Infect Dis*
27. Melton, Ewart, Weir et al. (2002) "Alphavirus 6K proteins form ion channels" *J Biol Chem*
28. Rogers, Jones-Burrage, Mukhopadhyay (2020) "TF protein of Sindbis virus antagonizes host type I interferon responses in a palmitoylation-dependent manner" *Virology (Auckl)*
29. Taylor, Melton, Herrero et al. (2016) "Effects of an in-frame deletion of the 6k gene locus from the genome of ross river virus" *J Virol*
30. Snyder, Kulcsar, Schultz et al. (2013) "Functional characterization of the alphavirus TF protein" *J Virol*
31. Kendra, Advani, Chen et al. (2018) "Functional and structural characterization of the chikungunya virus translational recoding signals" *J Biol Chem*
32. Grentzmann, Ingram, Kelly et al. (1998) "A dualluciferase reporter system for studying recoding signals" *RNA*
33. Kelly, Dinman (2023) "Shiftless is a novel member of the ribosome stress surveillance machinery that has evolved to play a role in innate immunity and cancer surveillance" *Viruses*
34. Harger, Dinman (2003) "An in vivo dual-luciferase assay system for studying translational recoding in the yeast Saccharomyces cerevisiae" *RNA*
35. Loughran, Howard, Firth et al. (2017) "Avoidance of reporter assay distortions from fused dual reporters" *RNA*
36. Loughran, Andreev, Terenin et al. (2025) "Guidelines for minimal reporting requirements, design and interpretation of experiments involving the use of eukaryotic dual gene expression reporters (MINDR)" *Nat Struct Mol Biol*
37. Steel, Geiss (2015) "A novel system for visualizing alphavirus assembly" *J Virol Methods*
38. Stanwick, Hallum (1974) "Role of interferon in six cell lines persistently infected with rubella virus" *Infect Immun*
39. Kim, Park, Lee et al. (2023) "Interferon alpha and beta receptor 1 knockout in human embryonic kidney 293 cells enhances the production efficiency of proteins or adenoviral vectors related to type I interferons" *Front Bioeng Biotechnol*
40. Ferreira, Sumner, Rodriguez-Plata et al. (2020) "Lentiviral vector production titer is not limited in HEK293T by induced intracellular innate immunity" *Molecular Therapy -Methods & Clinical Development*
41. Gavegnano, Detorio, Montero et al. (2014) "Ruxolitinib and tofacitinib are potent and selective inhibitors of HIV-1 replication and virus reactivation In Vitro" *Antimicrob Agents Chemother*
42. Patel, Dash, Jacobson et al. (2019) "JAK/STAT inhibition with ruxolitinib enhances oncolytic virother apy in non-small cell lung cancer models" *Cancer Gene Ther*
43. Mostafavi, Abeyratne, Zaid et al. (2019) "Arthritogenic alphavirus-induced immunopathology and targeting host inflammation as a therapeutic strategy for alphaviral disease" *Viruses*
44. Dinman, Wickner (1992) "Ribosomal frameshifting efficiency and gag/gag-pol ratio are critical for yeast M1 double-stranded RNA virus propagation" *J Virol*
45. Kelly, Olson, Neupane et al. (2020) "Structural and functional conservation of the programmed -1 ribosomal frameshift signal of SARS coronavirus 2 (SARS-CoV-2)" *J Biol Chem*
46. Crooks, Hon, Chandonia et al. (2004) "WebLogo: a sequence logo generator: figure 1" *Genome Res*
47. Cho, Lin, Chou et al. (2013) "Regulation of programmed ribosomal frameshifting by co-translational refolding RNA hairpins" *PLoS One*
48. Sokoloski, Hayes, Dunn et al. (2012) "Sindbis virus infectivity improves during the course of infection in both mammalian and mosquito cells" *Virus Res*
49. Khan, Loughran, Steckelberg et al. (2022) "Evaluating ribosomal frameshifting in CCR5 mRNA decoding" *Nature*
50. Jacobs, Dinman (2004) "Systematic analysis of bicistronic reporter assay data" *Nucleic Acids Res*
51. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12803810&blobtype=pdf | # Molecular detection and sequencing of the NetB toxin Gene of clostridium perfringens and evaluation of its pathogenicity in broiler chicken
Doha Abd, Alrahman Ahmed, Ahmed Hassan, Doaa Ali, Abdallah Ali, Eman Abd, Elmenum Shosha, Ibrahim Eldaghayes, Mohamed Shimaa, Ahmed Ali, Khair
## Abstract
Clostridium perfringens is the primary cause of necrotic enteritis (NE), one of the most economically destructive illnesses in broiler production. The virulence of this pathogen has recently been linked to the NetB toxin gene, a pore-producing cytotoxin recognized as a key determinant in NE pathogenesis. Despite its global significance, molecular data on NetB-positive C. perfringens from Upper Egypt are extremely limited. Therefore, this investigation was carried out to isolate C. perfringens recovered from broiler chickens in Assiut Governorate, Egypt, with a focus on the detection, sequencing, and genetic analysis of the NetB gene. Out of 60 intestinal samples, 40 isolates (66.7 %) were identified as C. perfringens by cultural and biochemical methods. PCR screening revealed that 20 isolates (50 %) harbored both alpha (cpa) and NetB toxin genes. Two representative NetB-positive isolates (Assiut1 and Assiut2) were sequenced and compared with global reference strains. Phylogenetic analysis demonstrated that the Egyptian isolates were closely associated to those from Egypt (KJ724530) and Iran (GU581173), forming a distinct Middle Eastern clade. Multiple sequence alignment revealed several nucleotide substitutions and amino acid variations, particularly when compared with strains from India, Brazil, and Denmark, suggesting the occurrence of unique regional mutations within the NetB gene. Experimental infection of broiler chickens confirmed the pathogenic potential of these isolates, producing classical NE lesions including gas-filled intestines, necrotic patches, and hemorrhagic ceca in Eimeria-preinfected birds, along with a significant reduction in body weight gain and feed intake (p < 0.05). To our knowledge, this is the first comprehensive molecular investigation of NetB-positive C. perfringens isolates from Upper Egypt. The findings underscore the genetic diversity and virulence of circulating strains and highlight the urgent need for regional surveillance and preventive strategies against necrotic enteritis in poultry.
## Introduction
In the worldwide poultry business, necrotic enteritis (NE) continues to be one of the greatest economically destructive illnesses. leading to severe economic losses estimated at over US$ 6 billion annually, equivalent to approximately US$ 0.05-0.06 per bird (Moore, 2016;Ali and Islam, 2021). The disease is characterized by both acute and subclinical forms that compromise intestinal integrity, feed efficiency, and flock performance. The causative agent, Clostridium perfringens, is an anaerobic, Gram-positive, spore-forming bacillus which is ubiquitously found in the intestinal tract of diseased birds and healthy and in the environment (Ali et al., 2020). Under predisposing conditions such as coccidial infection, dietary imbalances, or poor hygiene, the organism proliferates rapidly and produces potent toxins that cause extensive necrosis of the intestinal mucosa. Clinically, infected birds exhibit depression, ruffled feathers, diarrhea, and sudden death. Gross lesions typically include intestinal distension, necrotic patches on the mucosa, and hemorrhagic ceca, particularly in cases associated with coccidiosis. Clostridium perfringens produces more than 17 extracellular toxins that contribute to its pathogenic potential (Rood et al., 2018). Based on its major toxin forming profile, the bacterium is categorized into seven toxinotypes (A-G) defined by six major exotoxins epsilon (ETX), iota (ITX), beta (CPB), alpha (CPA), enterotoxin (CPE), and NetB (McMullin, 2020). Among these, types A (CPA), C (CPB and CPA), and G (NetB) are most commonly related to necrotic enteritis in poultry (Rood et al., 2018;Anju et al., 2021). Historically, The cpa gene encodes the α-toxin which was considered the primary virulence factor of C. perfringens in NE. However, subsequent studies demonstrated that mutants lacking the cpa gene remained virulent in vivo, leading to the discovery of the netB gene, which encodes a novel pore-forming cytolysin responsible for inducing NE lesions (Keyburn et al., 2006(Keyburn et al., , 2008;;Paiva and McElroy, 2014). The NetB toxin causes osmotic lysis and tissue necrosis by creating holes in the phospholipid bilayer of host cell membranes, upsetting the ionic equilibrium (Na⁺, Cl⁻, and Ca²⁺) (Datta et al., 2014;Profeta et al., 2020). Despite the global recognition of netB as a key virulence gene, molecular and phylogenetic data on C. perfringens strains carrying this gene in Egypt, particularly in Upper Egypt, remain scarce. To date, only a single Egyptian sequence has been reported from poultry, and none from broiler flocks in Assiut Governorate. Therefore, the current study was conducted to isolate and molecularly identify C. perfringens from broiler chickens in Assiut, Egypt, focusing on netB-positive strains through PCR amplification, sequencing, and phylogenetic analysis, alongside pathogenicity evaluation. By bridging this local gap, this research sheds important light on genetic diversity of C. perfringens in Egyptian poultry and establishes a foundation for future epidemiological surveillance. These findings are expected to support the development of targeted preventive strategies, including vaccination and probiotic-based interventions, to control necrotic enteritis in broiler production systems.
## Materials and methods
## Ethical approval
The Institutional Animal Care and Use Committee (IACUC) of Assiut University reviewed and approved the experimental protocol under approval reference 06/2025/0373. This experiment was conducted in accordance with the ethical guidelines for the care and use of animals as outlined by the Faculty of Veterinary Medicine, Assiut University, Egypt.
## Sampling
From broiler chickens, 60 intestinal samples were aseptically obtained, showing clinical signs of diarrhea, reduced feed intake, depression, and poor growth performance. The birds were obtained from both commercial farms and backyard flocks located in Assiut Governorate, Egypt, during 2025. Postmortem examination revealed distended intestines containing foul-smelling brownish contents and necrotic patches typical of Clostridium perfringens infection. Samples were collected from the jejunum and ileum of freshly dead or euthanized birds (via cervical dislocation), placed in sterilized tubes, and transported on ice to the Avian and Rabbit medicine Laboratory, Faculty of Veterinary Medicine, Assiut University, for further bacteriological and molecular analyses.
## Bacteriological examination
Each intestinal specimen was first enriched in Cooked Meat Broth (Oxoid, UK) and incubated anaerobically at 37 • C for 48 hours using anaerobic jars and gas-generating kits to promote the growth of Clostridium perfringens. The enriched cultures were then streaked onto Reinforced Clostridial Agar (LAB M, UK), and well-developed colonies exhibiting the typical morphology of C. perfringens were subcultured on 7 % sheep blood agar (Oxoid, UK) and Egg Yolk Agar (Oxoid, UK) to assess hemolytic and lecithinase activities, respectively. Isolates showing double-zone hemolysis on blood agar and opaque lecithinase activity on egg yolk agar were presumptively identified as C. perfringens and subsequently purified on Tryptose Sulfite Cycloserine (TSC) agar (Oxoid, UK) for confirmation. Gram staining revealed large Grampositive bacilli with blunt ends, while litmus milk tests confirmed their biochemical characteristics. Verified isolates were preserved for subsequent molecular identification and pathogenicity assessment (Willis, 1977;MacFaddin, 2000).
## Molecular identification DNA Extraction and PCR Amplification
Following the manufacturer's instructions, the QIAamp DNA Mini Kit (Qiagen, Germany) was used to extract genomic DNA from the confirmed Clostridium perfringens isolates. The extracted DNA was then eluted in 50 µL of elution buffer and kept at -20 • C until needed. A NanoDrop (Thermo Fisher Scientific, USA) was used to measure the purity and concentration of DNA spectrophotometrically. Molecular detection of C. perfringens was carried out using two sets of primers targeting the NetB and alpha (cpa) toxin genes. The primer sequences and predicted amplicon sizes were as shown in Table 1. PCR reactions were performed in a total volume of 25 µL, consisting of 12.5 µL EmeraldAmp GT PCR Master Mix (Takara, Japan), 1 µL of each primer (20 pmol/µL), 5 µL DNA template, and 5.5 µL nuclease-free water. Amplification was carried out in a thermal cycler (Biometra, Germany) under the following conditions: Initial denaturation at 94 • C for 5 min, 35 cycles of denaturation at 94 • C for 30 s, annealing at 55 • C for 40 s, and extension at 72 • C for 40 s, Final extension at 72 • C for 10 min. PCR products were separated by electrophoresis on a 1.5 % agarose gel containing ethidium bromide (0.5 µg/mL) and visualized under a UV transilluminator. A 100 bp DNA ladder (Qiagen, Germany) was used to estimate the molecular size of the amplified fragments (Yoo et al., 1997;Datta et al., 2014).
## Sequencing of PCR Products
PCR products of the NetB-positive isolates were purified using the QIAquick PCR Product Extraction Kit (Qiagen, Germany) according to the manufacturer's instructions. Purified DNA was then subjected to sequencing using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA) and analyzed on an Applied Biosystems 3130 Genetic Analyzer (Hitachi, Japan). The sequencing reaction mixture (20 µL total volume) consisted of 2 µL BigDye Terminator mix, 1 µL primer, 1-10 µL purified PCR product (depending on DNA concentration), and nuclease-free water to the final volume. The sequencing reaction was performed according to the kit protocol. After the sequencing reaction, the products were purified using Centri-Sep spin columns (Princeton Separations, USA) following the manufacturer's protocol to remove unincorporated dyes and salts. The purified sequencing products were mixed with 10 µL of Hi-Di formamide, denatured at 95 • C for 3 min, rapidly chilled on ice, and loaded onto the sequencer for analysis. Both forward and reverse reads were generated for each isolate. Raw sequence chromatograms were visually inspected for quality and trimmed to remove low-quality bases (Phred score <20) and vector contamination. Consensus sequences for each isolate were then assembled using the CAP contig assembly algorithm integrated into the DNASTAR SeqMan Pro software. The obtained sequences were compared to published nucleotide sequences using the BLAST tool (Basic Local Alignment Search Tool; Altschul et al., 1990) available on the NCBI website to confirm gene identity (Altschul et al., 1990).
## Phylogenetic analysis
The obtained NetB gene sequences were edited and aligned using the CLUSTAL W algorithm integrated in the MegAlign module of Lasergene DNASTAR software (Madison, WI, USA) as described by Thompson et al. (1994). Comparative analysis was performed with reference Clostridium perfringens NetB gene sequences retrieved from the GenBank database. Using MEGA version 7.0 software and the neighbor-joining approach using the maximum composite likelihood model, a phylogenetic tree was created (Tamura et al., 2013). The reliability of the clustering was evaluated using 1000 bootstrap replications to assess the confidence of each branch. Sequence identity and genetic relationships were analyzed to determine the evolutionary proximity between the local isolates and other international C. perfringens strains (Thompson et al., 1994;Tamura et al., 2013).
## Pathogenicity evaluation
A total of 40 one-day-old broiler chicks were obtained from a reared and commercial hatchery under hygienic conditions with free access to water and feed. At 14 days of age, the birds were divided equally into two groups (n ¼ 20 each): a control group that remained uninfected and a challenged group that was orally infected with a mixed commercial culture of Eimeria species, including E. acervulina, E. maxima, and E. tenella, at a dose of (5 × 10 4 ) oocysts per bird followed by Clostridium perfringens challenge to induce necrotic enteritis (Davis et al., 1973). The C. perfringens inoculum was prepared from freshly cultured NetB-positive isolates grown in Cooked Meat Broth (Oxoid, UK) at 37 • C for 24 to 48hrs hours under anaerobic conditions, and the bacterial suspension was adjusted to approximately 1.8 £ 10 9 CFU/mL. Birds in the infected group received 1 mL of the inoculum daily for three consecutive days following Eimeria infection. As shown, Fig. 1. All birds were monitored for seven days post-infection for clinical signs and postmortem examination. The pathogenicity of the isolates was evaluated based on the degree of severity of clinical signs, gross lesions (Keyburn et al., 2008;Paiva and McElroy, 2014). At the end of the trial, all remaining birds were humanely euthanized by cervical dislocation, following the approved ethical protocol.
## Statistical analysis
Data for final body weight, daily gain, feed consumption, and feed conversion ratio (FCR) were analyzed using a one-way analysis of variance (ANOVA) followed by Duncan's multiple range test to determine significant differences between groups. All statistical analyses were conducted using SPSS software (version 25.0; IBM, Chicago, IL, USA). Differences were considered statistically significant at P < 0.05 (Steel and Torrie, 1980)
## Results
## Bacteriological and biochemical identification
Out of 60 intestinal samples collected from broiler chickens showing diarrhea, depression, and reduced feed intake, 40 isolates (66.7 %) were identified as Clostridium perfringens based on their cultural and biochemical properties. On Reinforced Clostridial Agar (LAB M, UK), colonies appeared large, round, and slightly opaque (Fig. 2), while growth on 7 % sheep blood agar (Oxoid, UK) produced a characteristic double zone of hemolysis (Fig. 3). Microscopically, Gram-stained smears showed large Gram-positive bacilli with blunt ends (Fig. 4). All isolates exhibited stormy fermentation in litmus milk medium and strong lecithinase activity on Egg Yolk Agar (Oxoid, UK), confirming their biochemical identity as C. perfringens (Figs. 5,6). These isolates were subsequently subjected to molecular identification for detection of the alpha (cpa) and NetB toxin genes (Willis, 1977;MacFaddin, 2000).
## Table 1
Primer sequences used for PCR amplification of the Clostridium perfringens alpha (cpa) and NetB toxin genes, along with their expected product sizes and references.
## Molecular characterization PCR Amplification
A total of 40 biochemically confirmed isolates were subjected to molecular detection using PCR. Among them, 20 isolates (50 %) showed positive amplification for both the alpha (cpa) and NetB toxin genes. Amplification of the cpa gene yielded a clear band at 402 bp, while the NetB gene produced a distinct fragment at 560 bp, matching the expected amplicon sizes. No amplification was detected in the negative control, confirming the specificity and accuracy of the assay. As shown in Fig. 7, electrophoresis on a 1.5 % agarose gel revealed sharp, single bands representing the two toxin genes. The concurrent detection of cpa and NetB in all isolates confirms their classification as C. perfringens type G, a strain strongly related to necrotic enteritis in poultry (Yoo et al., 1997;Datta et al., 2014).
## Sequencing and phylogenetic analysis
Two representative NetB-positive isolates, designated C. perfringens Assiut1 and Assiut2, were subjected to sequencing. The obtained NetB gene sequences were submitted in the GenBank database (accession numbers PX508548, PX508549) and were compared with twenty-three reference sequences retrieved from GenBank, including representative strains from Egypt (KJ724530), Iran (GU581173), Japan (MK902770), China (HQ585982), Italy (KY923245.1), Australia (FJ189481.1), USA (KY559052), Denmark (GU433326), India (MG600591), and Brazil (KT020614). BLAST analysis confirmed that both Assiut isolates were Clostridium perfringens NetB-positive, showing 100 % nucleotide identity with the Egyptian isolate (KJ724530) and 99.8-100 % identity with the Iranian strain (GU581173). High nucleotide identity (98-99.5 %) was also recorded with isolates from Japan (MK902770), China (HQ585982), and Italy (KY923245.1), while considerably lower values were observed with Denmark (44 %), India (52 %), and Brazil (47 %). A phylogenetic tree was constructed using the neighbor-joining method in MEGA version 7.0 (Tamura et al., 2013) with 1000 bootstrap replications. As shown in Fig. 8. The resulting topology (Fig. 8) revealed that the Egyptian isolates clustered tightly with the Egyptian (KJ724530) and Iranian (GU581173) reference strains, forming a distinct Middle Eastern clade. Isolates from Japan (MK902770) and China (HQ585982) appeared in a closely related subclade, while those from Denmark (GU433326), India (MG600591), and Brazil (KT020614) were positioned in distant branches, reflecting clear evolutionary divergence. These findings demonstrate that the NetB-positive C. perfringens isolates circulating in Assiut Governorate are genetically related to the virulent type G strains detected in neighboring countries, confirming the conserved molecular nature of this lineage and its potential involvement in necrotic enteritis outbreaks.
## Percent identity analysis
Pairwise sequence alignment revealed that the two local isolates (Assiut1 and Assiut2) were 100 % identical to each other. They exhibited 99-100 % nucleotide identity with Egyptian (KJ724530) and Iranian (GU581173) isolates, 98-99.5 % with Japan (MK902770), China (HQ585982), Italy (KY923245.1), Australia (FJ189481.1), and the USA The lowest identity values (44-52 %) were observed with isolates from Denmark (GU433326), India (MG600591), and Brazil (KT020614), suggesting regional genetic differentiation and the presence of geographically distinct evolutionary lineages. These results confirm strong regional genetic conservation among Middle Eastern NetB gene sequences, supporting the hypothesis of a shared ancestral origin and limited genetic drift within this region. As shown in Table 2 Multiple sequence alignment Nucleotide sequence alignment. Multiple sequence alignment of the NetB gene revealed a high degree of conservation between the two local isolates, with only two nucleotide substitutions, corresponding to a single amino acid change (R→G). The sequences were nearly identical to those of the Egyptian (KJ724530) and Iranian (GU581173) strains. However, extensive nucleotide variability was detected in distant isolates, particularly those from India, Brazil, and Denmark, where 180-190 base substitutions were identified, several of which occurred within conserved regions of the gene. In certain foreign sequences, premature stop codons were also observed, suggesting possible truncation or altered toxin functionality (Fig. 9A).
## Amino acid sequence alignment
Alignment of deduced amino acid sequences revealed almost complete homology among the Egyptian, Iranian, and Chinese isolates, whereas substantial amino acid substitutions (30-50 residues) were identified in Indian, Brazilian, and Danish isolates. These substitutions were primarily located within the central and C-terminal domains of the NetB protein, regions associated with pore formation and cytotoxic activity (Fig. 9B). Collectively, the alignment data confirm that the Egyptian isolates represent a genetically conserved lineage, closely related to other Middle Eastern strains, while more divergent variants circulate in geographically distant regions. These molecular findings are in full agreement with the phylogenetic analysis, confirming the presence of a genetically stable and virulent C. perfringens genotype circulating among Egyptian poultry flocks.
## Pathogenicity evaluation
Following experimental infection, the broiler chickens in the challenged group exhibited typical clinical signs of necrotic enteritis, while the control birds remained healthy throughout the trial. Starting from the third day post-infection, infected birds showed depression, ruffled feathers, decreased feed intake, diarrhea with brownish mucoid droppings, and reduced activity. Mortality reached approximately 25 %, whereas no deaths occurred in the control group. Gross pathological examination revealed that the small intestines of infected birds were distended with gases and contained foul-smelling brownish contents. The mucosa of the jejunum and ileum appeared congested, friable, and covered with and diphtheritic membranes. In addition, the ceca were markedly engorged with blood, in birds previously infected with Eimeria spp., confirming the synergistic effect of coccidial infection in predisposing to necrotic enteritis (Fig. 10). These findings confirm the high pathogenic potential of the NetB-positive Clostridium perfringens isolates, particularly when combined with Eimeria infection, which reproduced the classical clinical and pathological manifestations of necrotic enteritis in broiler chickens.
## Performance parameters
Experimental challenge significantly impaired the growth performance of broiler chickens (Table 3). The infected group (Group 2) exhibited a significantly lower final Body Weight (720.00 ± 114.08 b ) and Daily Gain (18.45 ± 3.07 b ) compared to the control group (Group 1). Furthermore, Feed Consumption was also significantly reduced in the challenged group (924.95 ± 0.00 b vs. 986.20 ± 0.00 in the control). Critically, the Feed Conversion Ratio (FCR) significantly worsened, increasing to 6.31 ± 1.23 a in the challenged group compared to 4.65 ± 0.53 in the control (P < 0.05), demonstrating severe intestinal damage and poor nutrient utilization.
## Discussion
Despite the global significance of NE, there is still a scarcity of molecular data on C. perfringens strains circulating in Egyptian poultry, especially in Upper Egypt. Therefore, the present study provides a comprehensive molecular and pathogenic characterization of C. perfringens isolates from broilers suffering from NE in Assiut, Upper Egypt. This research represents the first in Upper Egypt to sequence and phylogenetically analyze the netB gene, filling a critical regional knowledge gap. Out of 60 intestinal samples examined, 40 isolates (66.6 %) were confirmed as C. perfringens based on cultural and biochemical properties. The isolates produced characteristic lecithinase activity on egg yolk agar and double-zone hemolysis on blood agar, findings consistent with those described by Ammar et al. (2021) and Nguyen et al. (2017), who identified similar phenotypic traits in isolates from Egypt and Vietnam, respectively. These results confirm that the combination of hemolytic pattern, Gram staining, and lecithinase production remains a reliable diagnostic hallmark for C. perfringens. The isolation
## Table 2
Percent nucleotide identity of NetB gene sequences among different C. perfringens isolates. rate observed here (66.6 %) was slightly higher than that reported by Praveen Kumar et al. (2020) in India (52 %) and Mwangi et al. (2019) in Kenya (58 %), which might reflect regional variations in farm management practices, environmental contamination, and feeding systems.
Of the 40 confirmed isolates, 20 (50 %) were positive for both cpa and netB genes. The concurrent presence of both genes classifies these virulent isolates as C. perfringens type G. This finding strongly supports the role of netB as the primary virulence determinant in poultry NE, as previously established (Keyburn et al., 2008;Paiva and McElroy, 2014). The 50 % detection rate for netB is similar to reports from Jordan (47.6 %) and China (49 %) (Gharaibeh et al., 2016;Zhong et al., 2022) but is lower than the high prevalence seen in some European studies (70-80 %) (Rood et al., 2018). These differences likely stem from variations in diagnostic targets, rearing practices, or sampling methodology across regions. The concurrent detection of cpa and netB genes confirms that these isolates belong to C. perfringens type G. This agrees with the work of Keyburn et al. (2008) and Paiva and McElroy (2014), who demonstrated that netB-and not cpa alone-is responsible for lesion development and disease manifestation in poultry. Sequencing of the two representative isolates (Clostridium_Assiut1 and Clostridium_ Assiut2) revealed 99-100 % nucleotide identity with other regional strains (e.g., Iranian GU581173), underscoring the strong genetic conservation among Middle Eastern netB-positive isolates (Ammar et al., 2021). The phylogenetic analysis grouped the Assiut isolates within a distinct Middle Eastern clade, showing closer relation to strains from Japan and China than to distant isolates from Europe or Brazil. This clustering suggests a stable evolutionary lineage and limited gene flow in the Middle East, a pattern consistent with regional clustering observed in other studies (Zhong et al., 2022;Nguyen et al., 2017). Nucleotide alignment of the netB gene showed remarkable homology between the two Egyptian isolates, differing by only two nucleotides which resulted in a single amino acid substitution. This minimal variation supports a highly conserved functional role for netB in virulent strains (Profeta et al., 2020). Conversely, isolates from distant regions like India, Brazil, and Denmark displayed extensive variability up to 190 substitutions and 50 amino acid changes (Praveen Kumar et al., 2020). Such substantial non-synonymous changes, particularly in the pore-forming domains, suggest potential alterations in protein activity or host specificity. The genetic conservation among Egyptian isolates suggests that NE outbreaks in the region are driven by a relatively stable virulent genotype. The pathogenicity assay confirmed the virulent nature of the netB-positive isolates. Birds experimentally challenged developed typical clinical and pathological features of NE, including severe intestinal lesions and hemorrhagic ceca (Keyburn et al., 2008;Mwangi et al. (2019); Fasina and Lillehoj, 2019). Crucially, the co-infection model with Eimeria spp. successfully reproduced the disease, emphasizing the well-documented synergistic interaction required for full disease manifestation (Paiva and McElroy, 2014;Moore, 2016). The demonstrated severity, coupled with the high prevalence of netB strains, underscores the urgent need for robust surveillance and monitoring in Egyptian poultry production. This research establishes a genetic baseline for future investigations in Upper Egypt and the conserved netB sequences suggest potential targets for vaccine design or alternative control strategies (e.g., feed additives). These findings are highly relevant given the global pressure to reduce antibiotic usage, which risks increasing NE incidence (Gharaibeh et al., 2016).
## Conclusion
In brief, this research confirmed that C. perfringens type G strains harboring both cpa and netB genes are widespread in broiler flocks in Upper Egypt. Given that only one previous study in Egypt has reported partial NetB sequences, The genetic conservation of netB among these isolates and their strong pathogenicity reinforce its central role in necrotic enteritis. Future research should focus on large-scale genomic surveillance and supporting the development of effective preventive and control strategies against necrotic enteritis in Egypt to pose a major damage to poultry health and productivity.
## References
1. Ali, Rahman, Khan (2020) "Prevalence and toxinotyping of Clostridium perfringens isolated from poultry in Bangladesh" *Vet. World*
2. Ali, Islam (2021) "Necrotic enteritis in broiler chickens: a review of the disease, pathogenesis, and control measures" *Poult. Sci. J*
3. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J. Mol. Biol*
4. Ammar, Shalaby, El-Aziz et al. (2021) "Detection of NetB and beta2 toxin genes in Clostridium perfringens isolates obtained from chickens" *Egypt. J. Veterin. Sci*
5. Anju, Singh, Kumar (2021) "Molecular characterization of Clostridium perfringens type G isolated from broiler chickens" *Microb. Pathog*
6. Davis, Hammond, Long (1973)
7. Datta, Fernandez-Miyakawa, Poon et al. (2014) "Identification of the netB toxin gene in Clostridium perfringens isolates from chickens with necrotic enteritis" *Avian Pathol*
8. Datta, Niwa, Itoh (2014) "Prevalence of netB gene of Clostridium perfringens isolates obtained from broiler chickens in Japan" *Vet. Microbiol*
9. Datta, Rakha, Narang et al. (2014) "Prevalence of α, β and netB toxin producing strains of Clostridium perfringens in broiler chickens in Haryana" *Haryana Veterinarian*
10. Fasina, Lillehoj (2019) "Characterization of intestinal immune response to Clostridium perfringens infection in broiler chickens" *Poult. Sci*
11. Gharaibeh, Al-Rifai, Al-Majali (2016) "Molecular typing and antimicrobial susceptibility of Clostridium perfringens from broiler chickens" *Anaerobe*
12. Keyburn, Boyce, Vaz et al. (2008) "NetB, a new toxin that is associated with avian necrotic enteritis caused by Clostridium perfringens" *PLoS. Pathog*
13. Keyburn, Sheedy, Ford et al. (2006) "Alpha-toxin of Clostridium perfringens is not an essential virulence factor in necrotic enteritis in chickens" *Infect. Immun*
14. Macfaddin (2000) "Biochemical Tests for Identification of Medical Bacteria"
15. Mcmullin (2020) "A Pocket Guide to Poultry Health and Disease, 2nd Ed"
16. Moore (2016) "Necrotic enteritis predisposing factors in broiler chickens" *Avian Pathol*
17. Mwangi, Timmons, Fitz-Coy et al. (2019) "Characterization of Clostridium perfringens recovered from broiler chickens affected by necrotic enteritis" *Avian Dis*
18. Nguyen, Vu-Khac, Nguyen (2017) "Characterization of Clostridium perfringens isolated from chickens in Vietnam" *Vet. Res. Commun*
19. Paiva, Mcelroy (2014) "Necrotic enteritis: applications for the poultry industry" *J. Appl. Poult. Res*
20. Kumar, Vinod Kumar, Karthik (2020) "Molecular detection and characterization of Clostridium perfringens toxin genes causing necrotic enteritis in broiler chickens" *Vet. World*
21. Profeta, Di Francesco, Di Provvido et al. (2020) "Prevalence of netB-positive Clostridium perfringens in Italian poultry flocks by environmental sampling" *J. Veterin. Diagnost. Investig*
22. Rood, Adams, Lacey et al. (2018) "Expansion of the Clostridium perfringens toxin-based typing scheme" *Anaerobe*
23. Steel, Torrie (1980) "Principle and Procedures of Statistics a Biometrical Approach, 2nd ed"
24. Tamura, Stecher, Peterson et al. (2013) "MEGA6: molecular Evolutionary Genetics Analysis version 6.0" *Mol. Biol. Evol*
25. Thompson, Higgins, Gibson (1994) "CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positionspecific gap penalties, and weight matrix choice" *Nucleic. Acids. Res*
26. Willis (1977) "Anaerobic Bacteriology: Clinical and Laboratory Practice"
27. Yoo, Lee, Park et al. (1997) "Molecular typing and epidemiological survey of Clostridium perfringens by multiplex PCR" *J. Clin. Microbiol*
28. Zhong, Zheng, Wang et al. (2013) "Molecular characteristics and phylogenetic analysis of Clostridium perfringens from different regions in China" *Front. Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802245&blobtype=pdf | # Cathepsin B hyperactivation facilitates exosome release of CVB3 particles and exacerbation of acute pancreatitis by impairing lysosomal integrity and acidification
Tianming Liang, Zhipeng Zhang, Le Xu, Zhirong Sun, Wei Xu
## Abstract
Lysosomal cathepsin B (CTSB) exhibits diverse roles in physiological and pathological processes. Upregulation and trypsinogen-activating function of CTSB have been reported in experimental secretagogue-elicited AP. Whether CTSB regulates the lysosome pathway and viral release in viral acute pancreatitis (AP) remains obscure. In a murine model of Coxsackievirus 3 (CVB3)-induced AP, we tested effects of CTSB on lysosome integrity and exosome secretion. CVB3 infection does-and time-dependently upregulates expression and activity of CTSB, the most elevated CTS in pancreas, and induces lysosomal instability and extra-lysosomal translocation of CTSB. Overexpression of CTSB increases viral replication. Inhibition of CTSB with CA074Me reduces virion release via rescuing lysosome integrity and acidity. Mechanically, CTSB hyperactivation and inappropriate cytoplasmic translocation increase viral infections by decreasing LAMP-1 + lysosomes, exacerbating lysosomal membrane permeabilization (LMP), and enhancing exosomal release of virions. Pharmaceutical inhibition of CTSB improves AP pathology via reducing viral infection. Our study reveals a critical role of hyperactivated CTSB in disrupting lysosome integrity, facilitating exosomal release of CVB3 particles and exacerbation of pancreas pathology in AP. These findings offer new insights into the pathogenesis of viral AP and suggest that CTSB and exosome are potential therapeutic targets for viral AP. IMPORTANCE This study uncovers a critical role of lysosome cathepsin B (CTSB) in exacerbating viral acute pancreatitis (AP). CVB3 infection of acinar cells indu ces lysosomal membrane permeability and CTSB cytosolic translocation. Hyperactiva ted CTSB increases viral infections by decreasing lysosomes, exacerbating LMP, and enhancing exosomal release of virions. Pharmaceutical inhibition of CTSB protects mice against viral dissemination and AP pathology, suggesting that CTSB and exosome are potential therapeutic targets for viral AP.
yet the mechanistic basis involved in viral AP remains poorly characterized. No antivi ral treatment exists against CVB3, and no specific or effective treatment of AP is available.
Lysosomes play pivotal roles in macromolecule digestion, autophagy, and cellu lar homeostasis. Dysfunctional lysosomes, caused by altered acidification, defective enzymes, or compromised membrane integrity, are associated with various pathologies, including cancer, neurodegenerative diseases, inflammatory diseases, and infections (9). CVB3 infection impairs lysosome function by various means (10)(11)(12). CVB3 uses proteinase 3C to proteolyze TFEB (transcription factor EB), a master regulator of autophagy and lysosome biogenesis, into a loss-of-function cleavage fragment to disrupt host lysosomal function and enhance viral infection (10). CVB3 protein 2B-ER-inser tion initiated calcium outflow and CVB3-induced ROS increase, leading to lysosomal membrane damage (9,13). Lysosomes are the terminal destination for autophagy. Early endosomes mature into late endosomes or the multivesicular bodies (MVB). CVB3 induces double membrane (DMVs) in the host cells to serve as replication sites. Late in infection, CVB3 increases the formation of late endosomes/MVBs and acidified autophagosomes (amphisomes) in the cytoplasm of host cells, allowing viral particles to mature inside them (14). Lysosome fusion with endosomes/MVBs or autophago somes, forming highly acidic degradative endolysosomes or autophagolysosomes, is important in degradation and recycling of cellular waste (15). CVB3 evades autophagoso mal degradation (by blocking the downstream autophagosome-lysosome fusion steps via cleaving SNAP29) and misuses the endosomal/autophagosomal pathway for viral replication and release (10,16). Fusion of virion-included MVBs or amphisomes with the plasma membrane leads to non-lytic release of virus-associated exosome into the extracellular environment, increasing intracellular viral dissemination (17).
The exosomes, which originate from the MVB, are 30-150 nm in diameter and are released from the cell upon MVB-plasma membrane fusion, carrying cytoplasmic cargo to the extracellular environment (18). Compromised lysosomal activity leads to elevated exosome release due to disrupted endosome-lysosome fusion processes. Damaged lysosomes may alter molecular contents of exosomes, subsequently modulat ing their biological functions (15,19). Alterations in exosome release due to lysosomal dysfunction have been associated with the pathogenesis of Parkinson's disease (20), alcoholic liver disease (21), and non-alcoholic fatty liver disease (22). Enteroviruses, such as Coxsackievirus and enterovirus 71 (EV-A71), have been revealed to exploit the secretory autophagy-exosome pathway to exit cells (23). Although CVB3 hijacks cellular autophagy for pro-viral functions, the precise mechanisms by which viral proteins or cellular proteins disrupt autophagy or lysosome remain incompletely understood. How lysosomal components modulate the MVB/amphisome formation, membrane fusion, and exosome release remains to be clarified.
Cathepsin B (CTSB) is a powerful lysosomal hydrolase that plays a crucial role in modulating the autophagy process and lysosomal function, participating in various biological processes, including protein degradation, signal transduction, and cell death (24). CTSB is primarily localized within subcellular endosomal and lysosomal compart ments. Under homeostatic conditions, CTSB controls lysosomal dynamics and autophagy by cleaving the lysosomal calcium channel MCOLN1, which negatively regulates the efflux of calcium, activating PPP3. This dephosphorylates TFEB and initiates autophagy (24). CTSB was also detected in the cytosol and nuclear fraction of senescent microglial cells, suggesting its extralysosomal function in senescent cell (25). In the initial stage of experimental AP, CTSB plays a major role in pathological trypsinogen activation in the early course of cerulein-induced pancreatitis (26,27). Cytosol CTSB initiates apoptosis and necrosis of acinar cells in experimental cerulein-induced AP (28). CTSB is upregulated in cardiomyocytes and stimulates the aortic banding-induced activation of TNF-α and cytochrome c release, thereby enhancing pressure overload-induced cardiac hypertro phy and fibrosis (29). CTSB promotes the maturation and secretion of IL-1β and IL-18, thus exacerbating local pancreatic damage (28). Pharmacological inhibition and genetic deletion of CTSB reduce trypsin activation and alleviate the severity of experimental secretagogue-elicited AP (30). During CVB3-induced viral AP, whether hyperactivated and extra-lysosomal located CTSB regulates lysosome function, exosomal viral release, and AP pathology needs further investigation.
In this study, CVB3 infection-induced murine viral AP model was used to investigate the role of cathepsin B (CTSB) in the regulation of lysosome integrity, CVB3 parti cle exosome release, and viral AP pathology. We identified CTSB as the most signifi cantly upregulated cathepsin family member in the pancreas of CVB3-infected mice. CTSB aberrant activation and inappropriate cytoplasmic translocation led to lysosomal membrane permeabilization (LMP) and dysfunction as well as increased virus-associated exosome release. Inhibition of CTSB mitigated CVB3-induced AP pathology by suppress ing CVB3 replication and non-lytic exosomal release. Our findings highlight CTSB as a crucial modulator of lysosome function and exosome release during viral infection, providing a novel target for therapeutic intervention of CVB3-induced AP.
## MATERIALS AND METHODS
## Mice, virus, and cells
Male C57BL/6J mice were obtained from the Soochow University Animals Center and cultured in a pathogen-free environment. All animal experiments were conducted in compliance with the Institutional Animal Care and Use Committee of Soochow University. The study protocols were approved by the Animal Ethical Committee of Soochow University (SYXK2022-0128). CVB3 (Nancy strain) is a gift from Prof. Y. Yang (Key laboratory of viral heart diseases, Zhongshan Hospital, China) was propagated in a HeLa cells monolayer in DMEM (Life Technologies) supplemented with 10% fetal bovine serum (FBS, Sigma). CVB3 was titrated by 50% tissue culture infective dose (TCID 50 ) assay on HeLa cell monolayers according to the method of Reed and Muench (5). Acinar cell line 266-6 and pancreatic ductal adenocarcinoma (PDAC) cell line SW1990, gifts from Prof. H. Liu (IBMS, Soochow University, China), were cultured in Dulbecco's modified Eagle medium (DMEM, Sigma) supplemented with 10% FBS and penicillin and streptomycin (100 U/mL) at 37°C in a 5% CO2 incubator.
## CVB3-induced acute pancreatitis
Six-week-old male C57BL/6 mice were inoculated with 10 3 PFU CVB3 in 100 μL PBS by intraperitoneal injection (i.p.). The body weight of mice was recorded daily for 14 days following infection. The pancreas was removed on day 0, 3, or 7 post-infection (p.i.) for detection of viral load and histological analysis. Tissues were cut longitudinally, fixed in paraformaldehyde, embedded in paraffin, cut into 5 μm thick sections, and stained with H&E. Images were taken randomly using a TE2000-S microscope (Nikon). Inflammation and edema/necrosis scores were quantified using semiquantitative scoring criteria: 1 means 25% of tissue damage; 2 indicates 25%-50% tissue involvement; and 3 indicates that 50% tissue damage.
## Plaque assay
Pancreas was weighed and homogenized in DMEM containing 2% FBS. After centrifuga tion at 300 × g for 10 min, supernatants were titrated by plaque assay. Briefly, diluted homogenates or cell supernatants were applied onto 95% confluent HeLa cells and incubated for 1 h. After washing twice with PBS, cells were overlaid with a 0.7% agarose-DMEM with 0.2% FBS and cultured at 37°C for 96 h. After removing the agarose, the monolayer was fixed with 4% paraformaldehyde for 1 h and then stained with 0.1% crystal violet for 15 min to visualize and count the plaques.
## Cell transfection and treatment
Cells with 85% confluency were infected with CVB3 at a multiplicity of infection (MOI) of 1, 5, or 10 for 1 h. After washing with PBS, cells were maintained for 12 ~ 24 h. In experiments where CA074Me (Cathepsin B Inhibitor) (TargetMol T3420) was used, cells were treated with 10 µM CA074-Me dissolved in DMSO (Sigma) for 12 h after CVB3 infection (MOI of 1). For CTSB overexpression, cells were transfected with 1 µg pCTSB plasmid or pCDH vector plasmid for 24 h before infection with CVB3. For CTSB knockdown, cells were transfected with siRNA-CTSB using X-tremeGENE siRNA reagent (Roche) for 36 h before CVB3 infection.
## Subcellular fractionation
Cells were collected by centrifugation at 2000 × g and washed with ice-cold PBS. The cells were suspended in ice-cold fractionation buffer (20 mM HEPES [pH 7.4], 250 mM sucrose, 10 mM KCl, 1.5 mM MgCl2, 1 mM EDTA, 1 mM EGTA) containing protease and phosphatase inhibitors and passed through a 25 G needle 20 times to homogenize the cell suspension. The lysosome fraction was obtained by lysosome isolation kit (Bestbio, bb31452). Briefly, cells were homogenized with lysis buffer and then centrifuged at 4°C, 20,000 × g for 20 min to obtain the supernatant as cytoplasmic fraction. The resulting precipitates were resuspended with lysosome extraction reagent and centrifuged at 30,000 × g for 30 min to collect lysosomal fractions, which were suspended in RIPA lysis buffer (50 mM Tris HCl [pH 8], 150 mM NaCl, 1 mM EGTA, 1% NP-40, 0.1% SDS, 0.5% sodium deoxycholate, 5% glycerol) containing protease and phosphatase inhibitors.
## Western blot analysis
The cell pellet and pancreas tissue were lysed with RIPA buffer for 20 min on ice to obtain lysates and then centrifuged at 12,000 rpm for 15 min at 4°C. The cell pellet was resuspended in lysis buffer and incubated on ice for 10 min. The extracted proteins were separated using SDS-PAGE and subsequently blotted onto PVDF membrane (Amersham). The membranes were blocked with 5% non-fat dry milk in Trisbuffered saline containing 0.2% Tween 20 (TBST) for 1 h at room temperature and then incubated with anti-VP1 (Mediagnost; M47), anti-Tubulin (CST, D3U1W), anti-CTSB (CST, D1C7Y), anti-CTSD (CST, E5V4H), anti-LAMP1 (CST, C54H11), anti-GAPDH (CST, D16H11), anti-Annexin A1 (abcam, ab214486), anti-Calnexin (abcam, ab22595), anti-CD9 (ABclonal, A19027), anti-CD63 (ABclonal, A19023), or anti-ALIX (ABclonal, A25326) in 1% non-fat milk-TBST overnight at 4℃. Densitometric measurements were performed using Image J software (version 2.5) and were normalized to loading controls.
## Quantitative real-time PCR (qPCR)
Pancreatic tissues (20 mg) or cells were treated with 1 mL of RNAiso Reagent (AG RNAex Pro RNA, AGbio). Total RNA was extracted using TRIzol RNA isolation reagent (Thermo Fisher Scientific). Semiquantitative RT-PCR analyses with ransStart Green qPCR SuperMix UDG kit (Transgen, China) were conducted on 5 µg of total RNAs on a LightCycler 480 Real-Time PCR System (Roche, Basel, Switzerland) using the specific sets of primers (Table 1, Beijing Tsingke Biotech). The relative abundance of specific mRNA was normalized to that of GAPDH. Relative quantification was determined using the 2-ΔΔCt method.
## Cathepsin B activity
According to the manufacturer's protocol, a total of 50 µL of samples and 50 µL of reaction buffer were dispensed into black 96-well plates. To each well, 2 µL of CTSB substrate (Ac-RR-AFC; final concentration 200 µM) was added and incubated for 1 h at 37°C. Fluorescence was measured using a microplate reader with excitation and emission wavelengths set to 400 nm and 505 nm, respectively.
## Exosome purification
Exosome isolation medium was removed from cells and subjected to 3,000 × g centrifugation for 15 min to remove cell debris, then transferred to a new tube. Exosome was purified from cell culture supernatant using Isolation Reagent (Applygen Technolo gies Inc.), as described previously. Briefly, after filtration through 0.22 µm membrane, exosome purification reagent was added into sample at a 1:5 dilution and incubated at 4°C for 12 h. After 1,500 × g centrifugation for 15 min, exosomes were pelleted and subjected to 1,500 × g centrifugation for 5 min. Exosomes were resuspended with 100 µL PBS or lysed in RIPA buffer. The virus titer was determined using the plaque assay method.
## Immunofluorescent staining
266-6 cells, cultured on NuncChamber Slide (Nunc, Roskilde, Denmark), were infected with eGFP-CVB3 at a MOI of 1. Twelve hours post-infection, cells were processed for fluorescence microscopy. Briefly, cells were washed with PBS, fixed with 4% paraformal dehyde, and permeabilized with 0.5% Triton X-100. After blocking with 10% goat serum for 30 min, cells were incubated with anti-CTSB (Abcam, USA) for 1 h. For evaluating lysosomal CTSB leakage, 12 hpi, cells were stained with anti-CTSB and anti-LAMP1 (Santa Cruz Biotechnology) for 1 h. Following incubation with an FITC-or PE-labeled goat-antimouse antibody (Santa Cruz, USA) for 30 min in the dark, nuclei were counterstained with DAPI for nuclear visualization. Fluorescence images were captured with a confocal microscope (Olympus Corporate, FLUOVIEW FV3000).
## In vivo CTSB overexpression or CTSB inhibitor treatment
Mice were injected with 1.0 mL reagent containing 50 µg of mouse p-CTSB plasmid or vector plasmid using in vivo-JetPEITM-Gal transfection agent (Polyplus-transfection, USA) via intravenous injection according to the manufacturer's instruction. Mice received 3 doses of pCTSB at -1, 0, and 1 dpi. For CTSB inhibitor treatment, mice were intraperito neally (i.p.) administered CA074Me (10 mg/kg) at 0, 1, and 2 dpi with 10³ PFU of CVB3 on day 0. Sham control mice received an equivalent volume of vehicle via the same route. Pancreatic tissues were harvested at 3 or 7 dpi for viral titer quantification and histopathological analysis.
## Statistical analysis
Multiple group comparisons were performed by one-way ANOVA, followed by Bonfer roni post hoc tests (StatPlus). Results are presented as mean ± SD. A P value < 0.05 was considered statistically significant. The survival curves were plotted using Kaplan-Meier methods, and significance was determined using the log-rank test.
## RESULTS
## CTSB is hyperactivated in CVB3-induced AP
To investigate the role of CTSB in viral-induced AP, we employed an experimental model of CVB3-induced AP in which male C57BL/6 mice were i.p. injected with 10 3 PFU of CVB3. Within the 14-day period, mice experienced 50% mortality (Fig. 1A) and developed severe pancreatic necrosis and inflammatory injury, which began on day 3 and persisted until day 14 post-infection (p.i.) (Fig. 1B). To identify the most upregulated cathepsin in the pancreas during acute CVB3 infection, we screened the mRNA levels of 18 cathepsin family members in day 3 pancreas using real-time qPCR. At 3 dpi, pancreas tissue from infected mice showed a significant increase in mRNA expression of Ctsb and Ctsd, with their level rising approximately 20-fold and 10-fold, respectively (Fig. 1C), compared to their background expressions. Then, the time course of CTSB during viral AP was determined. The pancreatic Ctsb mRNA and protein expression increased starting at day 1, peaked at day 3 p.i., and declined by day 14 p.i., as shown by qPCR and Western blotting analysis (Fig. 1D andE). Notably, the activity of CTSB increased progressively within 14 days of infection (Fig. 1F). Analysis of CTSB levels in CVB3-infected murine acinar cell line 266-6 revealed a time-and dose-dependent increase in CTSB protein expression post-infection (p.i.) (Fig. 1G). Taken together, pancreatic CTSB was elevated and hyperactivated at early viral infection.
## CVB3 infection causes decreased lysosomes, lysosomal membrane permeabi lization (LMP), and CTSB relocalization
To see whether CVB3 induces lysosomal damage and translocation of cathepsins, 266-6 cells were infected with CVB3 (MOI = 1) for 1 h, maintained for 12 h, and then frac tionated into the cytosolic and lysosomal lysates. CTSB and CTSD protein levels in different fractions were assessed by immunoblotting. Notably, lysosomal CTSB level decreased from 8 hpi, while cytoplasmic CTSB expression started to increase from 8 h and maintained a high level until 18 h, indicating leakage of CTSB from lysosome into the cytoplasm, a marker of LMP (Fig. 2A). Accordingly, cytoplasmic CTSB activity increased progressively from 2 to 18 hpi (Fig. 2B).
To further clarify the impact of CVB3 infection on the abundance and subcellular location of CTSB, 266-6 cells were infected with eGFP-CVB3 for 12 h and stained for CTSB. 12 hpi, when over 70% of acinar cells are infected and viral particles are produced, CTSB (red) abundance was markedly increased compared to its low expression in uninfected cells, with most upregulated CTSB localized in CVB3 + cells (Fig. 2C). LMP is often associated with a decrease in lysosomal-associated membrane protein 1 (LAMP1) and the release of lysosomal cathepsins into the cytoplasm (31). The subcellular location of CTSB was determined by confocal analysis of 266-6 cells 12 hpi labeled with CTSB and LAMP-1. Confocal imaging of the lysosomal marker, LAMP1, and CTSB in virus-infected cells showed that the LAMP-1 + lysosome numbers were markedly decreased, indicating compromised lysosomal membrane integrity. Moreover, although the CTSB puncta (red) increased after infection, the LAMP1-CTSB colocalization (yellow puncta) was signifi cantly reduced in infected cells, compared with high colocalization in control cells (Fig. 2E), indicating that most of elevated cytosolic CTSB is extra-lysosomal at 12 hpi.
The subcellular location of CTSB was also determined by confocal analysis of cells probed against CTSB (green) and stained with DAPI. As shown in Fig. 2D, CTSB level is low and distributed near the nucleus in control cells, and no nuclear localization was visible. In contrast, at 12 hpi, abundant CTSB proteins were more diffusely distributed in the cytoplasm, with some CTSB puncta (green) observed in the nucleus, as reflected by the ~10-fold shift in the nuclear/cytosolic ratio of CTSB MFI. It indicates that CVB3 infection triggers CTSB nuclear translocation.
Immunofluorescent analysis using Lysotracker indicated a significant decrease in lysosome acidity in infected acinar cells, indicating lysosome acidification impairment and instability (Fig. 2F). qPCR analysis revealed that CVB3 decreased RNA expressions of lysosomal biogenesis genes (Lamp-1, Mcoln1, Atp6v1h, and M6pr) in 266-6 cells 12 hpi (Fig. 2G), indicating reduced lysosomal biogenesis and repair. These results suggest that CVB3 induces lysosome instability, LMP, and CTSB cytosolic translocation by reducing lysosome integrity through decreased LAMP1 expression.
## CTSB increases viral replication and virion release in pancreatic acinar cells
To confirm the pro-viral role of CTSB in the pancreas, acinar cell 266-6 and pancreatic ductal adenocarcinoma (PDAC) cell line SW1990 cells were analyzed for viral level upon treatment with specific CTSB inhibitor CA074Me (5-40 μM) for 12 h. CA074Me dose-dependently reduced VP1 expression in both cells (Fig. 3A). Immunofluorescence imaging revealed that 12 h after eGFP-CVB3 (MOI = 1) infection, CA074Me (20 µM) treatment led to a marked reduction of eGFP + cells compared to DMSO treatment (Fig. 3B). Plaque assay on the cell supernatant and qPCR assay revealed a substantial reduction in extracellular viral titer and viral mRNA expression after CA074Me treat ment (Fig. 3C). Using small interfering RNAs targeting CTSB (si-CTSB) to reduce CTSB expression (Fig. 3D) significantly decreased CVB3 VP1 protein (Fig. 3D), extracellular viral titer, and viral mRNA levels (Fig. 3E). While overexpression of CTSB (Fig. 3F) led to an increase in extracellular viral titer and viral RNA expression (Fig. 3G), these data demonstrate that CTSB increases CVB3 replication and virion release in pancreatic cells.
## Inhibition of CTSB blocks exosome release of CVB3 via rescuing lysosomal integrity and function
To ascertain the effect of CTSB hyperactivation on LMP, 266-6 cells were infected with CVB3 for 1 h, then treated with CTSB inhibitor, CA074Me for 12 h, after which the cytoplasmic and lysosomal fractions were subjected to immunoblotting. CTSB inhibi tor treatment significantly reduced cytoplasmic but increased lysosomal CTSB/CTSD expression (Fig. 4A), and markedly reduced the cytosolic CTSB activity (Fig. 4B). The Lysotracker Red staining marks the luminal acidity degree necessary for lysosomal protease function. Compared to uninfected cells, virus-infected cells had significantly reduced LTR intensity, indicating impaired lysosome function. Inhibition of CTSB activity largely rescued lysosomal acidity (Fig. 4C). To further evaluate lysosomal biogenesis gene expressions, qPCR analysis revealed that CVB3 infection markedly decreased mRNA expression of Lamp-1, Mcoln1, Atp6v1h, and M6pr (Fig. 4D), which were restored by CTSB inhibition, indicating recovery of stability and self-repair function of lysosome. Cell viability was increased by CTSB inhibitor treatment (Fig. 4E). Collectively, these data indicate that inhibition of CTSB rescues lysosome integrity and function.
When lysosomal function is impaired, the degradation of endosome/MVB or autophagosome through lysosomes is compromised. This leads to the accumulation of MVB and autophagosome/amphisome, and the release of exosomes (contained in MVB) could be promoted (31). To clarify whether CTSB hyperactivation increases exosome release in acinar cells, exosomes in 266-6 culture supernatant were purified at different infection times. Total protein expression and levels of exosomal markers (CD9, CD63, and Alix), the non-exosomal ER marker (calnexin), and the MV marker (Annexin A1) in exosomes were characterized by immunoblotting. CVB3 infection increased exosomal protein expression in a time-dependent manner (Fig. 5A) and markedly increased CD9, CD63, and Alix exosome marker expressions compared to those in uninfected cells (Fig. 5B). Treating cells with the V-ATPase (acidification) inhibitor bafilomycin A1 (BAF) led to higher exosome production (Fig. 2A andB). Many studies reveal that positive-sense RNA, including Coxsackievirus, exploit the secretory autophagy pathway to exit cells (32). To explore whether CVB3-induced lysosomal dysfunction affects viral release, we infected 266-6 with eGFP-CVB3, and we observed that eGFP + particles in the exosomes were continuously increased from 4 h, peaked at 24 h, and were maintained at 36 hpi (Fig. 5C), showing kinetics similar to that of exosome release.
To examine whether CTSB inhibitor influences exosome release, exosomes were purified from 266 to 6 cell culture supernatant 24 hpi. The increased exosomal protein quantity after infection was largely reduced by CTSB inhibitor (Fig. 5D). Immunoblot ting of cell lysates against exosome markers revealed that CVB3 increased exosome production in acinar cell supernatant compared to non-infected cells, which was partially suppressed by CTSB inhibitor (Fig. 5E). To directly show that exosomes contain eleva ted levels of virions, 266-6 cells were infected with eGFP-CVB3 (MOI = 1) for 1 h and then treated with CA074Me for 24 h, and the viral progeny in exosomes was detected by fluorescent microscopy. The high mean fluorescence intensity (MFI) of exosomal eGFP-CVB3 from infected 266-6 cells was markedly blocked by CA074Me (Fig. 5F). Further, plaque assay on 1 mL culture supernatant-derived exosomes revealed that CVB3 increased exosomal virions release, and CTSB activation mainly increases exosomal virion release, not free virion release (Fig. 5H). Additionally, 0.1 µg purified exosome (from infected 266-6 cells) was examined for infectivity on 266-6 or SW1990 cells, revealing that a high number of infective virions were contained in the exosome, which was reduced by CTSB inhibition (Fig. 5G). It indicates that the antiviral mechanism of CA074Me is primarily mediated through interfering exosome release rather than the virion itself. Downregulating CTSB by transfecting cells with CTSB-siRNA 36 h prior to infection reduced VP1 protein in the exosomes (Fig. 5I). The antiviral effects of CTSB inhibitor or CTSB-siRNA mainly rely on targeting exosome-associated virus, not the free virus (Fig. 5J). Taken together, CTSB increases exosome release of viral particles via disrupting lysosomal integrity and function in acinar cells.
## In vivo CTSB overexpression aggravates CVB3 infection and viral AP pathol ogy
We next evaluate the effect of CTSB overexpression in mice using an in vivo-JetPEITM strategy. Mice were intravenously (i.v.) injected with 50 µg p-CTSB or p-vector on day -1, 0, and 1 dpi using the in vivo-Jet PEI reagent (Fig. 6A). Susceptibility to CVB3-AP and viral titer were evaluated within 7 days of infection. CTSB-overexpressing mice exhibited worse disease condition, increased mortality (75 vs 50%, pCTSB vs vector), and increased weight loss by day 7 p.i. (37 vs 19%, pCTSB vs vector, Fig. 6B). Histopathology analysis revealed that CTSB overexpression led to a significantly aggravated acinar cell necrosis and increased immune infiltration in the pancreas at day 3 (Fig. 6C) compared to shamor p-CTSB-treated mice. Consistent with the pathology, levels of pancreatic inflammatory cytokines (Tnf-α, Il-6, Il-1β, Il-17a) were significantly higher in CTSB-overexpressing mice than in control mice (Fig. 6D). To test whether the above results were due to differences in viral load, pancreatic CVB3 burden was measured. At 3 dpi, peaking time for viral replication, CTSB-overexpressing mice showed increased VP1 protein expression (Fig. 6E), CVB3 mRNA levels, and viral titers (Fig. 6F) in the pancreas compared to shamor p-CTSB-treated control mice. It indicates that CTSB overexpression promotes viral replication and viral AP pathology in vivo.
## Therapeutic administration of pharmaceutical CTSB inhibitor reduces viral infection and AP severity
To evaluate the treating effect of CTSB inhibitor on CVB3-induced AP, mice were i.v. injected with 10 mg/Kg of CA074Me on day 0 and day 1 p.i. (Fig. 7A). By day 7 p.i., CA074Me-treated mice exhibited higher survival rates and weight recovery compared to untreated mice, suggesting improved disease progression (Fig. 7B). CTSB inhibitor significantly reduced acini necrosis and immune infiltration in the pancreas of infected mice on days 3 and 7 p.i. (Fig. 7C). A significant decrease in pancreatic mRNA levels of proinflammatory cytokines (Tnf-α, Il-6, Il-1β, Il-17a) was confirmed (Fig. 7D). CTSB inhibition led to marked reduction in VP1 protein, viral titer, and viral mRNA expression in the pancreas of treated mice compared to control mice (Fig. 7E andF). Taken together, targeting CTSB exhibits good treating effect on viral-induced AP by decreasing viral infection and pancreatic inflammatory injury.
## DISCUSSION
The current study tests the effects of CTSB hyperactivation on lysosome integrity and exosome secretion in pancreatic acinar cells using an experimental murine model of CVB3-induced viral AP. We found that CVB3 impairs lysosomal function by decreasing LAMP-1 expression, lysosome biogenesis, and acidification, while markedly upregulat ing the expression and activity of lysosomal CTSB, as well as its cytosolic transloca tion. CTSB facilitates exosome release of viral particles through impairing lysosomal integrity and function. Notably, inhibition of CTSB improves lysosomal function and decreases exosome release, thus mitigating pancreatic acini necrosis and inflammation. We highlight the correlation between CTSB hyperactivation, decreased LAMP-1, and exosome secretion, which provide CTSB and exosome as potential targets for the treatment of viral AP.
Known viral pathogens associated with pancreatitis include hepatitis viruses, Coxsackie viruses, and coronaviruses (33,34). Direct cytopathic effects, trypsinogen activation, and immune infiltration contribute to the pathogenesis of viral AP, but the mechanism is far from illustrated. Apart from trypsin-mediated autodigestion in the early stage of AP, multiple parallel mechanisms, including endoplasmic reticulum stress, autophagy flux impairment, mitochondrial dysfunction, and inflammatory cascade activation in acinar cells, collectively drive the severe systemic inflammatory response and extensive pancreatic damage characteristic of AP (35).
Lysosomal cathepsins (CTSs) comprise a family of serine, aspartic (CTSD), and mainly cysteine (CTSB) proteases, which are important for lysosomal degradative functions (36). Early studies have proposed a critical role of CTSB in intrapancreatic trypsinogen activation and the onset of acute pancreatitis (27,30). CTSB-dependent conversion of premature trypsinogen into trypsin triggers a subsequent cascade of digestive enzyme activation (elastase, phospholipase A2), leading to pancreatic autodigestion and tissue damage (35). Trypsinogen activation also occurs in pancreaticinfiltrated macrophages, which depends on endocytosis of zymogen-containing vesicles and CTSB activation, leading to macrophage NF-κB activation and systemic inflammation (37). In addition, cytosolic CTSB activates the intrinsic pathway of apoptosis through cleavage of Bid and activation of Bax in acinar cells, leading to necrotic cell death (26). Excessive CTSB activation leads to various forms of cell death, including apoptosis, necrosis, pyroptosis, and ferroptosis (11,28,38,39).
CTSB is a cysteine proteolytic enzyme widely expressed in various cells and mainly located in the lysosomes. In cerulein-or Arg-1-induced experimental AP, activated CTSB protein levels and its activity have been found to be upregulated (27,28,38). In a mice model of CVB3-induced viral myocarditis and pressure overload-induced cardiac hypertrophy, CTSB was upregulated in cardiomyocytes and promoted cardiac inflammation and dysfunction through activating the inflammasome or TNF-α/ASK1/JNK apoptotic signaling pathway (11,29). The current study evaluates the time course of CTSB expression and activity using a CVB3-induced murine AP model. We confirm a significant increase in pancreatic CTSB/CTSD expression both in vitro (Fig. 2A through C) and in vivo (Fig. 1C, E andF). Interestingly, although CVB3 infection induces over expression and activation of pancreatic CTSB, it also reduces the expression of lysoso mal biogenesis genes (Lamp-1, Mcoln1, Atp6v1h, and M6pr), as well as Lamp-1 protein expressions (Fig. 2C through E and4D) and lysosomal acidity (Fig. 2F and4C). This suggests a significant lysosome instability and LMP, leading to the translocation of CTSB from lysosome to the cytoplasm and the nucleus (Fig. 2A, C andE). The upregulated CTSB further disrupts lysosomal integrity, whereas CTSB inhibitor restores lysosomal acidity (Fig. 4C). CTSB hyperactivation increases viral titer by enhancing exosome release of viral particles (Fig. 5). Pharmaceutical inhibition of CTSB improves lysosomal function and decreases viral release, eventually leading to mitigated AP pathology (Fig. 4, 5 and 7), demonstrating a detrimental role of CTSB activation in viral AP. Halangk et al. first found that Ctsb-KO mice exhibited partly reduced pancreas damage during cerulein-induced pancreatitis, as assessed by serum amylase activity, pancreas edema, and acinar cell necrosis. However, CTSB deletion had no impact on proinflammatory changes during cerulein-pancreatitis (27). A recent study found that Ctsb-KO mice, which were deprived of intrapancreatic trypsin activity, did not exhibit a mitigated cerulein-pancreatitis (40). Therefore, CTSB activation and function may be versatile in various models of experi mental AP. We highlight a pivotal role for hyperactivated and cytosolic translocated CTSB in disrupting lysosome integrity by decreasing the expression of LAMP-1, a key lysosomal membrane protein whose loss of expression is associated with LMP (31). We demonstrate that CVB3 infection significantly reduced lysosome numbers at 12 hpi by decreasing mRNA and LAMP-1 protein expression (Fig. 4D and2E). Notably, LAMP1-CTSB colocaliza tion was significantly reduced in infected cells compared to the high co-localization in sham cells (Fig. 2E), indicating that the upregulated CTSB proteins are not localized in lysosomes but are distributed in the cytosol. A previous study has demonstrated that cytosolic-translocated CTSB directly degrades LAMP-1 (31). We find that cytosolic CTSB activity was greatly increased (Fig. 2B and4B), which may accelerate the reduction of LAMP-1, exacerbating lysosome membrane instability.
Lysosomes are membranous organelles that play pivotal roles in macromolecule digestion, signal transduction, autophagy, and cellular homeostasis. Lysosome instability is associated with various pathologies, including cancer, neurodegenerative diseases, inflammatory diseases, and infections (41). Serving as the terminal stations of the endocytic pathway, lysosomes have indispensable roles in the degradation of endog enous and exogenous macromolecules, damaged or superfluous organelles, and pathogens. The autophagy-lysosome pathway is essential for maintaining cellular proteostasis and is associated with viral restriction and viral disease progression. Upon infection, the host initiates the autophagy-lysosome pathway to eliminate damaged organelles, proteins, and exogenous pathogens (42). Impaired autophagy/autophago lysosome has been implicated in experimental and human pancreatitis (43). Positivestrand RNA viruses have employed various mechanisms to antagonize and evade host's autophagy-lysosome degradation system (32). First, CVB3 exploits endosomal/autopha gosomal membrane and remodels them into special structures (closed single-membrane tubules to DMVs and multilamellar structures) for viral RNA synthesis (16). Moreover, upon endocytosis-mediated entry into the cell, instead of accessing endosome-lyso some or autophagy-lysosome degradation pathway, CVB3 and other viruses have evolved diverse mechanisms to avoid interactions with lysosome, thereby evading lysosomal degradation (11,44). Vesicular stomatitis virus (VSV) nucleocapsid regulates the dynamics of multivesicular endosomes by transferring itself to LBPA-containing intraluminal vesicles (ILVs) and releasing virus into the cytoplasm via fusion with the limiting membrane (45). CVB3 uses the viral protease 2A to cleave p62 and inhibit autophagic degradation of virions (virophagy) (46). HIV-1 Nef interacts with Beclin 1 to sequester TFEB in the cytosol, thus inhibiting maturation of autophagosomes and lysosome biogenesis (47). SARS-CoV-2 virulence factor ORF3a blocks lysosome func tion by modulating TBC1D5-dependent Rab7 GTPase cycle and impedes the fusion of autophagosomes with lysosomes, causing accumulation of membranous vesicles for replication (48). CVB3 proteinase 3C disrupts host lysosomal function via proteolysing transcription factor EB (TFEB), a master regulator of autophagy and lysosome biogenesis, into a lower-molecular-mass, loss-of-function cleavage fragment, thereby enhancing viral infection (10). During CVB3 infection, fasting/restricted feeding (FR) treatment increased viral RNA levels in multiple organs and accelerated pathology via CTSB-dependent intact autophagy-lysosome pathway, indicating that viral replication and release require functional autophagy and lysosomal pathways (12).
In the current study, we provide new evidence of how CVB3-hyperactivated lysosomal CTSB enhances exosome-mediated viral dissemination through disrupting lysosome integrity and function. Exosomes, a subset of extracellular vesicles generating through the fusion of specific endosomes (MVB) with the plasma membrane, are critical for cellcell communication, viral dissemination, and cancer metastasis (15,18). A major source of exosomes is intraluminal vesicles (ILVs) within MVBs. The autophagosome-lysosome system modulates the biogenesis and degradation of exosomes (49). MVBs containing ILVs can either be targeted for degradation in lysosomes or released as exosomes into the extracellular space (49). Lysosomes are essential for the degradation of MVBs, ILVs, and autophagosomes (15). Hypoxia impairs lysosomal degradation by downregulat ing ATP6V1A expression, leading to the reduced fusion of MVBs with lysosomes and enabling the secretion of ILVs as EVs (50). PTEN facilitates lysosome biogenesis and acidification by dephosphorylating TFEB. PTEN deficiency increases exosome secre tion by reducing lysosome-mediated degradation of MVBs (51). Endolysosomal fusion BORC-ARL8-HOPS pathway is a critical determinant of the amount of exosome secretion (52). ER stress activating PERK and IRE1α (UPR) could repress the acidification and catabolic activity of lysosomes, leading to MVB-lysosome fusion block and redirecting MVBs from lysosomal degradation to plasma membrane fusion, resulting in exosome release (53).
Viruses harness the autophagy-lysosome pathway to secrete EVs (exosomes containing viral particles), thereby enhancing infectivity (47). Pathogens, such as enteroviruses, compromise lysosomal membrane integrity by inducing membrane permeabilization, causing cytosolic leakage of proteases and cations, which further induce cell death pathways (54). CVB3 activates TFEB by inactivating mTORC1 signaling, promoting the non-lytic release of CVB3 via a secretory autophagic pathway during the early stages of infection (55). Poliovirus uses phosphatidylserine (PS)-enriched autopha gosome vesicles for non-lytic release. Vaccinia virus (VACV) hijacks ESCRT-mediated MVB formation to facilitate virus egress and spread (56). SARS-CoV-2 ORF3a activates the SNARE complex (STX4-SNAP23-VAMP7), inducing fusion of lysosomes with the plasma membrane for viral release in exosomes (48,57). Viruses exploit these intricate lyso somal-exosomal connections to manipulate incomplete autophagy, enhancing their escape from the exosomal pathway (33). Our findings reveal an intriguing relation ship between lysosome inhibition (LMP, biogenesis, and acidification impairment) and exosome secretion in acinar cells during CVB3 infection, which is exacerbated by CTSB. Acinar CTSB hyperactivation facilitates exosome viral secretion and acinar cell necrosis by impairing lysosome integrity and acidification. These data indicate that intact lysoso mal integrity and degradation function are critical for blocking exosome release of virions, which is highly efficient for the systemic and receptor-independent infection and dissemination of CVB3 in animals (17). Sustained lysosome dysfunction contributes to chronic disease pathologies.
Our findings reveal a critical role for hyperactivated CTSB-mediated disruption of lysosomal integrity and function in promoting non-lytic viral release and acinar cell necrosis in AP. Pharmaceutical inhibition of CTSB protects mice against viral dissemina tion and AP pathology by blocking exosome release. Our results suggest targeting CTSB or exosome as alternative therapies for viral AP. Better understanding of CTSB-mediated lysosome disruption and exosome formation mechanisms is essential for developing therapeutic interventions against viral-induced AP.
## References
1. Boxhoorn, Voermans, Bouwense et al. (2020) *Lancet*
2. Habtezion (2015) "Inflammation in acute and chronic pancreatitis" *Curr Opin Gastroenterol*
3. Rafaqat, Patoulias, Behnoush et al. (2024) "Interleukins: pathophysiological role in acute pancreatitis" *Arch Med Sci*
4. Szatmary, Grammatikopoulos, Cai et al. (2022) "Acute pancreatitis: diagnosis and treatment" *Drugs (Abingdon Engl)*
5. Yan, Yang, Qian et al. (2019) "Pathogenic role of an IL-23/γδT17/neutrophil axis in coxsackievirus B3-induced pancreatitis" *J Immunol*
6. Huber, Ramsingh (2004) "Coxsackievirus-induced pancreatitis" *Viral Immunol*
7. Geisler, Hazini, Heimann et al. (2021) "Coxsackievi rus B3-its potential as an oncolytic virus" *Viruses*
8. Liang, Zhang, Bai et al. (2024) "STAT3 increases CVB3 replication and acute pancreatitis and myocarditis pathology via impeding nuclear translocation of STAT1 and interferon-stimulated gene expression" *Int J Mol Sci*
9. Zhang, Vooijs, Keulers (2025) "Key mechanisms in lysosome stability, degradation and repair" *Mol Cell Biol*
10. Mohamud, Tang, Xue et al. (2021) "Coxsackievirus B3 targets TFEB to disrupt lysosomal function" *Autophagy*
11. Wang, Jia, Shen et al. (2018) "Cathepsin B aggravates coxsackievirus B3-induced myocarditis through activating the inflammasome and promoting pyroptosis" *PLOS Pathog*
12. Alirezaei, Flynn, Garcia et al. (2021) "A foodresponsive switch modulates TFEB and autophagy, and determines susceptibility to coxsackievirus infection and pancreatitis" *Autophagy*
13. Stefan (2019) "A kidnapping story: how coxsackievirus B3 and its host cell interact" *Cell Physiol Biochem*
14. Corona, Mohamud, Jackson et al. (2018) "Oh, SNAP! How enteroviruses redirect autophagic traffic away from degradation" *Autophagy*
15. Raghu, Lebleuvs (2020) "The biology, function, and biomedical applications of exosomes" *Science*
16. Limpens, Van Der Schaar, Kumar et al. (2011) "The transformation of enterovirus replication structures: a three-dimensional study of single-and doublemembrane compartments" *mBio*
17. Fu, Xiong (2023) "Exosomes mediate coxsackievirus B3 transmission and expand the viral tropism" *PLoS Pathog*
18. Van Niel, Angelo, Raposo (2018) "Shedding light on the cell biology of extracellular vesicles" *Nat Rev Mol Cell Biol*
19. Villarroya-Beltri, Baixauli, Mittelbrunn et al. (2016) "ISGylation controls exosome secretion by promoting lysosomal degradation of MVB proteins" *Nat Commun*
20. Alvarez-Erviti, Seow, Schapira et al. (2011) "Lysosomal dysfunction increases exosome-mediated alpha-synuclein release and transmission" *Neurobiol Dis*
21. Babuta, Furi, Bala et al. (2019) "Dysregulated autophagy and lysosome function are linked to exosome production by micro-RNA 155 in alcoholic liver disease" *Hepatology*
22. Zhang, Tan, Wang et al. (2021) "Lipid-induced DRAM recruits STOM to lysosomes and induces LMP to promote exosome release from hepatocytes in NAFLD" *Sci Adv*
23. Mutsafi, Bonnet (2018) "Enterovirus transmission by secretory autophagy" *Viruses*
24. Xie, Zhao, Yan et al. (2023) "Cathepsin B in programmed cell death machinery: mechanisms of execution and regulatory pathways" *Cell Death Dis*
25. Meng, Liu, Xie et al. (2020) "Nucleus distribution of cathepsin B in senescent microglia promotes brain aging through degradation of sirtuins" *Neurobiol Aging*
26. Talukdar, Sareen, Zhu et al. (2016) "Release of cathepsin B in cytosol causes cell death in acute pancreatitis" *Gastroenterology*
27. Halangk, Lerch, Brandt-Nedelev et al. (2000) "Role of cathepsin B in intracellular trypsinogen activation and the onset of acute pancreatitis" *J Clin Invest*
28. Wang, Wang, Zhang et al. (2021) "Cathepsin B aggravates acute pancreatitis by activating the NLRP3 inflammasome and promoting the caspase-1-induced pyroptosis" *Int Immunopharmacol*
29. Wu, Xu, Yuan et al. (2015) "Cathepsin B deficiency attenuates cardiac remodeling in response to pressure overload via TNF-α/ASK1/JNK pathway" *Am J Physiol Heart Circ Physiol*
30. Van Acker, Saluja, Bhagat et al. (2002) "Cathepsin B inhibition prevents trypsinogen activation and reduces pancreatitis severity" *Am J Physiol Gastrointest Liver Physiol*
31. Tanaka, Warner, Michael et al. (2022) "LAMP3 inhibits autophagy and contrib utes to cell death by lysosomal membrane permeabilization" *Autophagy*
32. Mohamud, Shi, Qu et al. (2018) "Enteroviral infection inhibits autophagic flux via disruption of the SNARE complex to enhance viral replication" *Cell Rep*
33. Simons-Linares, Imam, Chahal (2021) "Viral-attributed acute pancreatitis: a systematic review" *Dig Dis Sci*
34. Bonney, Gao, Chew et al. (2020) "SARS-COV-2 associated acute pancreatitis: cause, consequence or epiphenomenon?" *Pancreatol ogy*
35. Saluja, Dudeja, Dawra et al. (2019) "Early intra-acinar events in pathogenesis of pancreatitis" *Gastroenterology*
36. Vizovišek, Vidak, Javoršek et al. (2020) "Cysteine cathepsins as therapeutic targets in inflammatory diseases" *Expert Opin Ther Targets*
37. Sendler, Weiss, Golchert et al. (2018) "Cathepsin B-mediated activation of trypsinogen in endocytosing macrophages increases severity of pancreatitis in mice" *Gastroenterology*
38. Sendler, Maertin, John et al. (2016) "Cathepsin B activity initiates apoptosis via digestive protease activation in pancreatic acinar cells and experimental pancreatitis" *J Biol Chem*
39. Kuang, Liu, Li et al. (2020) "Cathepsin B is a mediator of organellespecific initiation of ferroptosis" *Biochem Biophys Res Commun*
40. Geisz, Tran, Orekhova et al. (2023) "Trypsin activity in secretagogue-induced murine pancreatitis is solely elicited by cathepsin B and does not mediate key pathologic responses" *Gastroenterology*
41. Settembre, Perera (2024) "Lysosomes as coordinators of cellular catabolism, metabolic signalling and organ physiology" *Nat Rev Mol Cell Biol*
42. Lőrincz, Juhász (2020) "Autophagosome-lysosome fusion" *J Mol Biol*
43. Wang, Ni, Chao et al. (2019) "Impaired TFEB-mediated lysosomal biogenesis promotes the development of pancreatitis in mice and is associated with human pancreatitis" *Autophagy*
44. (2019)
45. Gruenberg, Van Der Goot (2006) "Mechanisms of pathogen entry through the endosomal compartments" *Nat Rev Mol Cell Biol*
46. Van Der Goot, Gruenberg (2006) "Intra-endosomal membrane traffic" *Trends Cell Biol*
47. Shi, Wong, Piesik et al. (2013) "Cleavage of sequestosome 1/p62 by an enteroviral protease results in disrupted selective autophagy and impaired NFKB signaling" *Autophagy*
48. Choi, Bowman, Jung (2018) "Autophagy during viral infection -a double-edged sword" *Nat Rev Microbiol*
49. Shariq, Malik, Sheikh et al. (2023) "Regulation of autophagy by SARS-CoV-2: the multifunctional contribu tions of ORF3a" *J Med Virol*
51. Wang, Ren, Bai et al. (2023) "Exosome-autophagy crosstalk in enveloped virus infection" *Int J Mol Sci*
52. Wang, Wu, Zhai et al. (2023) "Hypoxia promotes EV secretion by impairing lysosomal homeostasis in HNSCC through negative regulation of ATP6V1A by HIF-1α" *J Extracellular Vesicle*
53. Jiang, Shi, Cui et al. (2023) "PTEN deficiency facilitates exosome secretion and metastasis in cholangiocarcinoma by impairing TFEBmediated lysosome biogenesis" *Gastroenterology*
54. Shelke, Williamson, Jarnik et al. (2023) "Inhibition of endolysosome fusion increases exosome secretion" *J Cell Biol*
55. Zhou, Zhou, Chen et al. (2025) "PERK and IRE1α promote exosome secretion via blocking lysosomal degradation of multiple vesicular body" *J Biol Chem*
56. Radulovic, Yang, Stenmark (2025) "Lysosomal membrane homeostasis and its importance in physiology and disease" *Nat Rev Mol Cell Biol*
57. Giansanti, Strating, Defourny et al. (2020) "Dynamic remodelling of the human host cell proteome and phosphoproteome upon enterovirus infection" *Nat Commun*
58. Huttunen, Samolej, Evans et al. (2021) "Vaccinia virus hijacks ESCRT-mediated multivesicular body formation for virus egress" *Life Sci Alliance*
59. Zhou, Hu, Castro-Gonzalez (2023) "Bidirectional interplay between SARS-CoV-2 and autophagy" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12719184&blobtype=pdf | # Hepatitis B viral DNA integration occurs within three days of infection and is enhanced by ATR inhibition
Dong Li, Jochen Wettengel, Harout Ajoyan, Vikki Ho, Gabriela Wu, Sarah Bae, Henrik Zhang, Delgerbat Boldbaatar, Jacob George, Mark Douglas, Thomas Tu
## Abstract
Chronic hepatitis B virus (HBV) infection is a major risk factor for hepatocellular carcinoma, with viral DNA integration into the host genome playing a pivotal role in oncogenesis. While HBV integration has been historically considered an event occurring late in a chronic infection, sensitive assays have detected integrations early infection. This study investigates the specific timing and molecular mechanisms of HBV DNA integration using a replication deficient HBV reporter system (HBV-Zeo) in HepG2-NTCP cells. Infection of this virus followed by positive selection led to cellular colony formation, showing that the input virus is the substrate that undergoes integration. By inducing DNA double-strand breaks via X-ray irradiation at specific timepoints after HBV infection, we observed a 2-3-fold increase in integration frequency when cells are irradiated between 16 and 76 h post-infection. Pharmacological inhibition of DNA repair pathways in this specific time window revealed that suppression of homologous recombination (HR) via ATR inhibitors significantly enhances integration rates (2.4-2.8-fold), while microhomology-mediated end joining (MMEJ) inhibition reduced integration to 17 % of untreated controls. These findings suggest that MMEJ plays a key role in HBV DNA integration occurring within hours of HBV infection. Together, our results advance understanding of HBV-associated hepatocarcinogenesis and may inform therapeutic strategies to disrupt viral integration and mitigate HBV-associated liver cancer risk.
## 1. Background
Chronic hepatitis B virus (HBV) infection affects ~360 million worldwide [1], increases the risk of hepatocellular carcinoma (HCC, i.e. primary liver cancer) by ~100-fold [2] and is the single most common driver of liver cancer (responsible for nearly 60 % of HCC cases). HCC is the third leading cause of cancer-related death worldwide, with approximately 830,000 deaths annually [3]. Due to the absence of specific symptoms during cancer development, HCC is often diagnosed at an advanced stage, resulting in poor therapeutic outcomes and a dismal 5-year survival rate of less than 20 % [4].
Integration of HBV DNA into the host cell is highly associated with liver cancers, as ~90 % of HBV-associated liver cancers contain HBV DNA integrations, compared to 1 in ~10 4 cells in the non-tumour tissue [5][6][7][8][9][10][11]. Integration events have been reported to occur when breaks in the double-stranded DNA cellular genome are repaired through Microhomology-Mediated End Joining (MMEJ) and Non-Homologous End Joining (NHEJ) pathways, using HBV double-stranded linear DNA (dslDNA) as a substrate [12][13][14]. While NHEJ dominates as the primary double stranded break (DSB) repair pathway across all cell cycles, its error-prone nature facilitates HBV integration by introducing small indels at break sites. In contrast, MMEJ is considered a backup pathway and plays a critical role in rapid DSB repair during replication stress [15,16]. MMEJ involves using short tracts of homology to stabilise the ligation of two exposed DNA strands. Consistent with the use of MMEJ and NHEJ pathways, the cellular sites of HBV integrations are distributed widely across the entire human genome with short sequence homology between viral and host DNA in some integrations and indels evident in others [17][18][19].
Early studies of human HCC tissues using low sensitivity assays (e.g.
Southern blot) suggested HBV integration occurs decades after initial infection [8,9,20]. This led to the prevailing model that integration is a late event, requiring long-term viral replication and cumulative genomic damage. However, emerging data challenge this paradigm. Studies using more sensitive PCR-based assays have detected HBV DNA integrations shortly after acute HBV infection [21,22] and in children with chronic hepatitis B [6]. In the woodchuck hepatitis virus (WHV) model, viral DNA integration has been reported 15 min after infection [21,23], although this is difficult to reconcile with human cell studies showing formation of cccDNA -the earliest marker of HBV entry into the nucleusis not detectable until 12-16 h post-infection [24,25].
Although HBV DNA integration may occur within hours of infection, liver cancer may not occur until decades later. HBV DNA integration has been reported to drive liver cancer through two mechanisms: (1) cismediated effects (including insertional mutagenesis disrupting host gene function or regulation [26][27][28]; and (2) trans-mediated effects (e.g., persistent production of truncated and mutated viral proteins). However, the specific role of HBV DNA integrations in HBV-associated HCC (HBV-HCC) remains incompletely understood, as previous models do not recapitulate viral integration through authentic infection [29][30][31].
Here, we established an in vitro system to easily quantify and investigate factors associated with HBV DNA integration. Using a replication-competent reporter virus, we could assess the integrations resulting from first round HBV infection. We used this system to show that X-ray irradiation increases the rate of HBV DNA integration and precisely determined that integration occurs between 12-and 76-h postinfection. Moreover, we measured the effect of inhibitors of various DNA repair factors and found that blocking Ataxia Telangiectasia and Rad3related protein (ATR) function (necessary for HR) induced a significant increase in integration rates.
## 2. Methods
## 2.1. Cell culture
A replication-deficient reporter HBV (HBV-Zeo) was generated by replacing the entire hepatitis B surface open reading frame (PreS1/ PreS2/S) with a transthyretin promoter driven zeocin-resistance cassette. This HBV-Zeo construct maintains: 1) Core and HBx (but not polymerase and HBsAg) ORFs; 2) an intact ε packaging signal and direct repeats (DR1/DR2) for pgRNA packaging and 3) native HBV promoter/ enhancer elements to preserve transcriptional regulation [32]. The zeocin marker enables selection of stable integrations in HepG2-NTCP cells. HBV-Zeo viral particles were produced and secreted by transfecting a helper cell line [33] trans-complementing polymerase and surface proteins through by the HBV-Zeo construct. Viral particles were purified by a streamlined heparin affinity chromatography and sucrose ultracentrifugation method as previously described [34].
HepG2-NTCP cells were infected with the reporter HBV-Zeo virus following a previously established protocol [35]. As a negative control, cells were treated with 100 μM Myrcludex B (Myr-B, an HBV entry inhibitor) from 15 min before infection to 18 h post-inoculation [36]. After inoculation, cells were washed three times with 1x DPBS (14,040,141, Gibco), and fresh DMEM media supplemented with 10 % v/v Fetal Bovine Serum (10099, Sigma-Aldrich), 20 mM L-glutamine (G7513, Sigma-Aldrich) and 2.5 % (v/v) DMSO (D2650, Sigma-Aldrich), and cells were incubated for 3 days. HepG2-NTCP cells were trypsinised using TrypLE Enzyme (12604021, ThermoFisher), re-seeded into 6-well plates and incubated for approximately 25 days to select HBV-infected cells in DMEM supplemented with 0.5 mg/mL Zeocin Selection Reagent (phleomycin D1; R25001, Invitrogen) to select HBV-infected cells. As cccDNA is cleared through mitosis [37], expanded colonies under these conditions therefore contain only cells with viral integrations (confirmed in our previous characterisation of these clones [38]).
To dissect the role of host DNA repair pathways in HBV integration, infected HepG2-NTCP cells were treated with inhibitors targeting distinct repair enzymes (Table 1). All compounds were prepared and applied according to manufacturer protocols. Concentration ranges were selected based on previously reported effective doses and cytotoxicity profiles [39][40][41][42].
Inhibitors were applied at two critical timepoints: 1) Pre-integration treatment: cells were pretreated with inhibitors for 24 h prior to HBV-Zeo infection, and at 72 h post-infection (hpi); 2) Co-integration treatment: inhibitors were added at 4 hpi and maintained until 72 hpi to target the window when integration occurs.
## 2.2. Quantification of HBV DNA copy number per cell
Total cellular DNA was extracted from cell pellets using the QIAGEN DNeasy Blood and Tissue Kit (69506, Germany) according to the manufacturer's instructions and then eluted in 20 μL of elution buffer. DNA concentration was quantified using NanoDrop™ 2000 spectrophotometer and stored at -30 • C for further experiments.
Digital droplet PCR (ddPCR) was used to quantify the copy number of HBV DNA integrations. 500 ng of total DNA was first digested with EcoRI-HF (R3101S, New England Biolabs) in a 20 μL reaction containing 1x Cutsmart Buffer and 10 units EcoRI-HF. 25 ng of digested DNA was added to a 20 μL of ddPCR mastermix consisting of 1X ddPCR Multiplex Supermix for Probes (12005910, Bio-rad), and target-specific primers and probes (sequences and final concentration listed in Table 2). The samples underwent droplet generation using the Bio-rad QX200™ Droplet Generator Cartridges (1864008, Bio-rad), Oil (1863004, Biorad) and QX200™ Droplet Generator. 40 μL of oil-emulsion droplets were transferred into a ddPCR plate and amplified by PCR (initial denaturation at 95 • C for 10 min, followed by 40 cycles of a denaturation at 95 • C for 10 s, an annealing at 54 • C for 15 s and an extension at 68 • C for 20 s, deactivation at 95 • C for 10 min). Droplets were analysed by a Bio-rad QX200™ Droplet Reader and resultant data was analysed in Quantasoft. The concentration of HBV DNA was normalized against RNase P, known to be present as two copies per cell [43].
## 2.3. X-ray irradiation
X-ray irradiation was performed with a cabinet X-ray irradiation system X-RAD320 Biological Irradiator (Precision X-Ray Inc., North Branford, CT) at the Westmead Institute for Medical Research. X-ray was generated with an operating voltage of 320 kVp and a tube current of 12.5 mA, using a 2.0 mm Al filter, at a dose rate of approximately 1.15 Gy/min. Cells were exposed to X-ray irradiation at doses of 0.08, 0.4, 2 or 4 Gy, as described previously [44].
## 2.4. Quantifying HBV infection rate
Total HBV DNA or HBeAg secretion were used to quantify HBV infection. To quantify total HBV DNA DNA from HBV-Zeo infected HepG2-NTCP cells, digital PCR (dPCR) was performed using the QIAcuity One platform (Qiagen). Total DNA was extracted and digested as above, and 50 ng of digested DNA was added to 40 μL of dPCR mastermix reaction containing 1X QIAcuity Probe PCR Master Mix (250015, QIAGEN) and primers specific for HBx and RNaseP (sequences and final concentration listed in Table 2). Reactions were loaded into QIAcuity Nanoplates (26k, 24-well format, 250001) and sealed. PCR amplification was carried out on the QIAcuity One instrument (5plex) using the following thermal cycling conditions: Initial denaturation at 95 • C for 2 min, followed by 40 cycles of denaturation at 95 • C for 10 s, annealing at 54 • C for 15 s, extension at 68 • C for 20 s, and deactivation at 95 • C for 10 min. After amplification, the partitioned reactions were imaged and analysed using the QIAcuity Software Suite (v3.1). HBV DNA copy numbers per cell were calculated using Poisson distribution-based analysis provided by the QIAcuity software (v3.1).
Secreted Hepatitis E Antigen (HBeAg) from cells was measured using the HBeAg ELISA Assay kit (WB-2496, Wantai) according to the manufacturer's instructions. Standards were generated using serum from an HBV-positive patient (Genotype B, immune-tolerant, HBeAgpositive, NA treated, HBeAg measured against WHO standard = 1888 ng/μL). Absorbance measurements (Optical Density, OD) of the samples were obtained using the SpectraMax iD5 Multi-Mode Microplate Reader (Molecular Devices) at two wavelengths: 450 nm to quantify HBeAg and 630 nm to measure background absorbance. Background absorbance was subtracted to normalise [OD (450 nm)-OD (630 nm)] and the blank-corrected values used to determine the lower limit of detection (LLOD). Absolute HBeAg concentration was calculated based on the standard curve. For samples exceeding the linear range, a 1:5 initial dilution in assay buffer was performed, followed by iterative dilutions (1:2 to 1:10) until OD values fell within the standard curve's linear range (20-1000 ng/μL). Final concentrations were adjusted for dilution factors.
## 3. Results
## 3.1. Establishment of a system to quantify de novo HBV DNA integration rates
HepG2-NTCP cells were infected with HBV-Zeo and selected with Zeocin to enrich cell colonies harbouring de novo HBV DNA integrations as previously described [38] (Fig. 1A). As a negative control, cells were pre-treated for 15 min with Myrcludex B (Myr-B) (100 mM), an HBV entry inhibitor [36]. Myrcludex B was also included in the infection media for 18-20 h during inoculation. No colonies were observed in Myr-B-treated controls, indicating that HBV entry is essential for integration and colony formation (Fig. 1B).
Assuming a starting population of 5 × 10 5 HepG2-NTCP cells and ~30 % infection, we estimated ~1.5 × 10 5 infected cells were present in each well [35,45]. Each well generated on average 20 integration-containing clones, corresponding to an observed integration frequency of ~1 in 7500 infected cells, consistent with previously reported integration rates of ~1 per 10,000 infected cells [11,12,46].
Colonies from multiple infections were isolated (106 single cellderived clones, in total). ddPCR analysis showed that each clone contained 1 or 2 integrations per cell with no HBV DNA detected in parental cell controls (Fig. 1C), normalized to the diploid reference gene RNase P, is consistent with a pattern of genomic integration. This and our previous data characterizing clones developed using this method suggest there is limited possibility that these clones contain significant levels of cccDNA after selection [38]. We have previously shown that the level of HBV DNA detected by direct PCR is the same level as measured by either ddinvPCR or invPCR (integration-specific PCR assays), showing they are not derived from cccDNA. Targeted HBV DNA sequencing of these clones was able to find the exact virus-cell junction in all clones, directly showing that these sequences are integration derived. Thus, the quantified colonies are very likely the result of stable genomic integration events.
## 3.2. HBV DNA integration occurs between 16 and 76 hpi post-infection
We used X-ray irradiation to rapidly induce double stranded DNA breaks, which have been reported to occur within 20 s of irradiation [47] and peak ~30 min post-irradiation [48]. We found that a dose of 0.4 Gy was able to induce DNA breaks in HepG2-NTCP cells while not affecting downstream cell mitosis (Supplemental materials).
As cellular double stranded DNA breaks are the substrates for HBV DNA integration [12], we posited that increasing the number of DNA breaks by X-ray irradiation would increase the number of integration events when these two events coincide. Thus, to define the exact timepoint when HBV DNA integration occurs post-infection, HepG2-NTCP cells were infected with HBV-Zeo for a short duration (4 h) to increase temporal resolution (Fig. 2A for experimental outline) and then irradiated with X-rays at various time points after infection: 0, 15 min, 1, 2, 4, 8, 12, 16, 20, 28, 52, or 76 h post-infection (hpi). Colony formation was used as a measure of HBV DNA integration frequency.
The rate of HBV infection appeared unaffected by X-ray irradation: intracellular total HBV DNA levels (quantified by dPCR) were not significantly different between cells irradiated at different time points (Fig. 2B). Thus, time points could be directly compared with one another without needing to consider differences in infection rate.
Cells irradiated at times between 0 and 12 hpi did not show significantly different rates of colony formation compared to non-irradiated controls. (Fig. 2C). A small increase in colony formation rate was observed in cells irradiated at 16 hpi, increasing between 28 and 52 hpi (showing a 2-3-fold increase compared to non-irradiated controls). A sharp decline was observed in cells irradiated at 76 hpi (~30-50 % of the non-irradiated cells). Thus, we find that HBV DNA integration occurs in a specified window of 12-76 hpi.
## 3.3. HBV DNA integration increases with inhibition of ATR
Pharmacological inhibition of DNA repair pathways may modulate HBV DNA integration by either prolonging the persistence of DNA DSBs or redirecting repair toward error-prone mechanisms during the integration window. To evaluate which pathways may be involved in the HBV DNA integration process, we employed small-molecule inhibitors targeting distinct repair enzymes (Fig. 3A): ART558 (10 nM) inhibiting POLQ, a key molecule in MMEJ promoting error-prone repair of DSBs; AZD6738 (0.5 μM) and VE822 (50 nM) inhibiting ATR, a key molecule in HR that regulates DNA damage signalling; and NU7441 (0.25 μM) inhibiting DNA-PKcs, a core component of classical NHEJ. These concentrations were selected based on cytotoxicity assays, ensuring minimal impact on cell viability while effectively inhibiting their respective pathways (Supplemental materials). It is important to note that when a particular DNA repair pathway is inhibited, the other two may increase to repair the increased number of DNA breaks [16,49,50].
To dissect the role of DNA repair pathways in integration, infection of HepG2-NTCP cells with HBV-Zeo was repeated and inhibitors was applied at two critical timepoints (Fig. 3B).
1) Pre-integration (24 h before infection to time 0 -defined as the time of inoculation): this accounted for any changes due to reductions in cell viability or ability to undergo mitosis due to application of the drugs (assessing baseline effects on cell fitness and viral entry). 2) Co-integration (4-72 hpi): this measured the effect of inhibitors directly at the peak times of integration (targeted the window in potentially influence the absolute frequency of integration, the clear temporal profile of increased colony formation (occurring between 16 and 52 hpi), specifically identifies the window during which the input viral DNA is competent to serve as a substrate for DSB repair. This timeline is further corroborated by the finding that inhibition of the MMEJ pathway is most effective at suppressing integration when applied during this same window (4-72 hpi, Fig. 3F). Together, this data shows that integration is coincident with the formation of cccDNA (12-72 hpi) [25]. This supports the conclusion that integration originates from input viral genomes rather than from second-round replicative intermediates, consistent with prior studies using HBc-deficient HBV that is unable to undergo productive replication [24]. We observed no evidence of integration occurring within 15 min of infection, as previously reported in the woodchuck model [21,23,53], potentially due to species-specific differences.
The apparent lack of integration beyond 72 hpi in our system should not be interpreted as an absolute limit for integration events in vivo. It is important to note that our reporter virus is replication-defective, restricting our results to assessing a single round of infection; once the input viral genomes are processed, no new viral DNA substrates for integration are generated. In contrast, chronic HBV infection in patients is likely characterized by continuous cycles of hepatocyte infection, which would provide ongoing chances of HBV DNA integration. Thus, while our data demonstrate that integration occurs early after initial entry, the ongoing hepatocellular infection in vivo is likely to cumulatively drive integration events over time and contribute to the high integration burden observed in patient livers.
Exactly how much integration occurs over the course of a chronic infection is difficult to calculate, even with this new data. For example, there is little known to what extent super-infection exclusion occurs with the chronically infected liver [54]. For human HBV, this could range from complete to non-existent. Given the strong dependence of new infection events on the integration rate, this single fact makes it very difficult to even theoretically calculate the ongoing rate of integration. Integration rates may also depend on the level of inflammation (driving clonal expansion of cells with integrations), the extent to which cells with integrations are recognised and cleared by the immune system, whether integration rates linearly increase with viral load, etc. Our results provide a single figure in this highly-complex system which may be used in future to calculate the level of integration rate over the course of a chronic or acute HBV infection.
We showed that pharmacological inhibition of ATR with AZD6738 and VE822 significantly increases integration rates. Pharmacological inhibition experiments revealed two key mechanistic findings: (a) ATR inhibition significantly increased integration rates when applied at preintegration timepoints, consistent with ATR promoting homologous recombination and thereby limiting error-prone repair, and (b) POLQ inhibition strongly suppressed integration, suggesting that MMEJ is a likely pathway involved in HBV DNA integration. Indeed, this is consistent with previous sequencing data showing that HBV DNA integrations have a higher than expected chance to show seqeunce microhomology at the virus-cell DNA junction [17][18][19]. In addition, the increased integration rate under ATR inhibition could reflect an elevated abundance of nuclear viral DNA free ends, as ATR inhibition has been linked to increased deproteinized HBV rcDNA forms [52]. Whether these forms actively participate in integration remains unclear, future analysis of viral integration ends under such conditions may resolve this question.
Collectively, our findings suggest that MMEJ is a likely pathway for HBV DNA integration and that it occurs very early in an infection. These insights not only advance our understanding of HBV persistence but also reveal actionable therapeutic targets. First, these results underscore the importance of early intervention (e.g. vaccination and antiviral treatment) to prevent ongoing infection events and therefore reducing integration rates. Since integration begins almost immediately after infection and likely accumulates over time, early suppression of viral replication with NA could significantly reduce the long-term integration burden and cancer risk. Indeed, this supports recent calls for expansion of NA treatment for both health benefits and limiting anxiety over disease progression [55,56], even in apparently healthy younger people (not currently recommended in many clinical management guidelines).
Second, our identification of MMEJ a likely pathway for HBV DNA integration reveals novel therapeutic strategies. The development of anti-integration agents, such as POLQ inhibitors or other MMEJ antagonists, holds promise. These would likely be used in combination with existing antivirals to target both new integration events and ongoing viral replication. Furthermore, our replication-deficient reporter system provides an ideal platform for drug discovery pipelines to screen for selective anti-integration compounds.
Third, the landscape of HBV therapy is evolving. Novel epigenetic therapeutics designed to silence both cccDNA and integrated HBV DNA have entered clinical trials. If effective, these agents could target the transcriptional activity of established integrations, addressing a key driver of oncogenesis and potentially contributing to a functional cure.
Several limitations of our experimental system should be acknowledged. Our model depends on the zeocin resistance gene to be included in the integration. Long-read sequencing studies of human liver tissues from HBV patients have shown that the majority of integrations include HBsAg ORF (which is replaced by the zeocin resistance gene in our system), there is a subset of integrations (~20 %) that do not [57]. Whether these represent integrations occurring through a different mechanism is unknown, but cannot be studied using our platform.
Secondly, our reporter HBV lacks HBsAg expression. In natural infections, integrated HBV DNA often drives constitutive HBsAg production, which may induce endoplasmic reticulum stress and promote hepatocyte apoptosis [58,59]. However, we have shown in previous work that the expression of HBsAg is relatively low during the time period in which integration is occurring and is only detetable by highly sensitive RNAscope assays 72hr post-infection. Thus, we suggest that the HBsAg expression would not have a strong contribution on integration rates in this system.
The suggestion that MMEJ is the dominant pathway is based on pharmacological evidence and sequencing results from previous studies. While the significant suppression of integration upon POLQ inhibition and its potential enhancement upon ATR blockade provide robust functional evidence, this study lacks direct mechanistic evidence. Knocking out MMEJ pathways in future studies (e.g. using CRISPR approaches) may allow confirmation of our hypotheses.
Looking forward, our findings provide a framework for several research directions: (1) investigating extrinsic triggers of integration (e. g., inflammation or genotoxic stress) to inform adjunctive therapies that lower integration risk; and (2) developing therapeutic strategies aimed at mitigating HBV-driven genomic instability and ultimately preventing HBV-associated hepatocarcinogenesis. Future work could translating these insights into strategies to mitigate HBV-driven genomic instability and liver cancer.
## CRediT authorship contribution statement
## References
1. Resources "Funding acquisition. Thomas Tu: Writingreview & editing, Writingoriginal draft, Supervision, Project administration, Methodology, Investigation, Funding acquisition, Data curation, Conceptualization"
2. (2024) "Global hepatitis report 2024: action for access in lowand middle-income countries"
3. Kew (2010) "Epidemiology of chronic hepatitis B virus infection, hepatocellular carcinoma, and hepatitis B virus-induced hepatocellular carcinoma" *Pathol. Biol*
4. Sung (2021) "Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries" *CA Cancer J. Clin*
5. Singal (2022) "HCC surveillance improves early detection, curative treatment receipt, and survival in patients with cirrhosis: a meta-analysis" *J. Hepatol*
6. Chow (2023) "Effect of antiviral treatment on hepatitis B virus integration and hepatocyte clonal expansion" *Clin. Infect. Dis*
7. Mason (2016) "HBV DNA integration and clonal hepatocyte expansion in chronic hepatitis B patients considered immune tolerant" *Gastroenterology*
8. Tu (2015) "Clonal expansion of hepatocytes with a selective advantage occurs during all stages of chronic hepatitis B virus infection" *J. Viral Hepat*
9. Sung (2012) "Genome-wide survey of recurrent HBV integration in hepatocellular carcinoma" *Nat. Genet*
10. Mason (2010) "Clonal expansion of normal-appearing human hepatocytes during chronic hepatitis B virus infection" *J. Virol*
11. Mason, Jilbert, Summers (2005) "Clonal expansion of hepatocytes during chronic woodchuck hepatitis virus infection" *Proceedings of the National Academy of Sciences -PNAS*
12. Tu (2018) "Hepatitis B virus DNA integration occurs early in the viral life cycle in an in vitro infection model via sodium taurocholate cotransporting polypeptidedependent uptake of enveloped virus particles" *J. Virol*
13. Bill, Summers (2004) "Genomic DNA double-strand breaks are targets for hepadnaviral DNA integration" *Proceed. National Academy. Sci. -PNAS*
14. Mcvey, Lee (2008) "MMEJ repair of double-strand breaks (director's cut): deleted sequences and alternative endings" *Trends Genet*
15. Zhao (2016) "Genomic and oncogenic preference of HBV integration in hepatocellular carcinoma" *Nat. Commun*
16. Truong (2013) "Microhomology-mediated end joining and homologous recombination share the initial end resection step to repair DNA double-strand breaks in mammalian cells" *Proceed. National Academy Sci. -PNAS*
17. Dutta (2017) "Microhomology-mediated end joining is activated in irradiated human cells due to phosphorylation-dependent formation of the XRCC1 repair complex" *Nucleic Acids Res*
18. Seeger, Mason (2015) "Molecular biology of hepatitis B virus infection" *Virology*
19. Budzinska (2018) "Cellular genomic sites of hepatitis B virus DNA integration" *Genes*
20. Budzinska (2018) "Sequence analysis of integrated hepatitis B virus DNA during HBeAg-seroconversion" *Emerg. Microb. Infect*
21. Edman (1980) "Integration of hepatitis B virus sequences and their expression in a human hepatoma cell" *Nature*
22. Chauhan, Michalak (2020) "Kinetics of DNA damage repair response accompanying initial hepadnavirus-host genomic integration in woodchuck hepatitis virus infection of hepatocyte" *Cancer Genet*
23. Kimbi, Kramvis, Kew (2005) "Integration of hepatitis B virus DNA into chromosomal DNA during acute hepatitis B" *World J. Gastroenterol. : WJG*
24. Chauhan (2017) "Initial sites of hepadnavirus integration into host genome in human hepatocytes and in the woodchuck model of hepatitis B-associated hepatocellular carcinoma" *Oncogenesis*
25. Tu (2021) "D e novo synthesis of hepatitis B virus nucleocapsids is dispensable for the maintenance and transcriptional regulation of cccDNA" *JHEP Rep*
26. Tu (2020) "A novel method to precisely quantify hepatitis B virus covalently closed circular (ccc)DNA formation and maintenance" *Antivir. Res*
27. Koch (1984) "Amplification and rearrangement in hepatoma cell DNA associated with integrated hepatitis B virus DNA" *EMBO J*
28. Matsubara, Tokino (1990) "Integration of hepatitis B virus DNA and its implications for hepatocarcinogenesis" *Mol. Biol. Med*
29. Steinemann (2006) "Assessment of differentiation and progression of hepatic tumors using array-based comparative genomic hybridization" *Clin. Gastroenterol. Hepatol*
30. Neuveut, Wei, Buendia (2010) "Mechanisms of HBV-related hepatocarcinogenesis" *J. Hepatol*
31. Tu (2017) "HBV DNA integration: molecular mechanisms and clinical implications" *Viruses*
32. Zoulim (2024) "Hepatitis B virus DNA integration: implications for diagnostics, therapy, and outcome" *J. Hepatol*
33. Wing (2019) "A dual role for SAMHD1 in regulating HBV cccDNA and RTdependent particle genesis" *Life Sci. Alliance*
34. Rust "Liver-specific transgenic expression of human NTCP in rhesus macaques confers HBV susceptibility on primary hepatocytes" *Proceedings of the National Academy of Sciences -PNAS*
35. Wettengel (2021) "Rapid and robust continuous purification of high-titer hepatitis B virus for in vitro and in vivo applications" *Viruses*
36. Ni (2014) "Hepatitis B and D viruses exploit sodium taurocholate Co-transporting polypeptide for species-specific entry into hepatocytes" *Gastroenterol*
37. Cheng (2021) "Clinical effects of NTCP-inhibitor myrcludex B" *J. Viral Hepat*
38. Tu (2022) "Mitosis of hepatitis B virus-infected cells in vitro results in uninfected daughter cells" *JHEP reports*
39. Li (2025) "Novel digital droplet inverse PCR assay shows that natural clearance of hepatitis B infection is associated with fewer viral integrations" *Emerg. Microb. Infect*
40. Zatreanu (2021) "Polθ inhibitors elicit BRCA-gene synthetic lethality and target PARP inhibitor resistance" *Nat. Commun*
41. Yin (2021) "VE-822, a novel DNA Holliday junction stabilizer, inhibits homologous recombination repair and triggers DNA damage response in osteogenic sarcomas" *Biochem. Pharmacol*
42. Bradbury (2020) "Targeting ATR as Cancer therapy: a new era for synthetic lethality and synergistic combinations?" *Pharmacology & therapeutics (Oxford)*
43. Tavecchio (2012) "Further characterisation of the cellular activity of the DNA-PK inhibitor, NU7441, reveals potential cross-talk with homologous recombination" *Cancer Chemother. Pharmacol*
44. Baer (1990) "Structure and transcription of a human gene for H1 RNA, the RNA component of human RNase P" *Nucleic Acids Res*
45. Wang (2015) "Radiosensitization to X-ray radiation by telomerase inhibitor MST-312 in human hepatoma HepG2 cells" *Life Sci*
46. Iwamoto (2014) "Evaluation and identification of hepatitis B virus entry inhibitors using HepG2 cells overexpressing a membrane transporter NTCP" *Biochem. Biophys. Res. Commun*
47. Mason (2009) "Detection of clonally expanded hepatocytes in chimpanzees with chronic hepatitis B virus infection" *J. Virol*
48. Rogakou (1998) "DNA double-stranded breaks induce histone H2AX phosphorylation on serine 139" *J. Biol. Chem*
49. Sak, Stuschke (2010) "Use of γH2AX and other biomarkers of double-strand breaks during radiotherapy" *Semin. Radiat. Oncol*
50. Turk, Wisinski (2018) "PARP inhibition in BRCA-mutant breast cancer" *Cancer*
51. Bouwman (2010) "53BP1 loss rescues BRCA1 deficiency and is associated with triple-negative and BRCA-Mutated breast cancers" *Nat. Struct. Mol. Biol*
52. Ning (2023) "Inhibition of DNA-dependent protein kinase catalytic subunit boosts rAAV transduction of polarized human airway epithelium" *Mol. Ther., Methods Clin. Dev*
53. Luo (2020) "Involvement of host ATR-CHK1 pathway in hepatitis B virus covalently closed circular DNA Formation" *mBio*
54. Chauhan (2019) "Retrotransposon elements among initial sites of hepatitis B virus integration into human genome in the HepG2-NTCP cell infection model, Cancer Genet"
55. Walters (2004) "Superinfection exclusion in duck hepatitis B virus infection is mediated by the large surface antigen" *J. Virol*
56. Cohen (2025) "Patient and public health perspectives to inform expansion of hepatitis B treatment guidelines. The lancet" *Gastroenterol. Hepatol*
57. Kennedy (2025) "Scientific and medical evidence informing expansion of hepatitis B treatment guidelines. The lancet" *Gastroenterol. Hepatol*
58. Van Buuren (2022) "Targeted long-read sequencing reveals clonally expanded HBV-associated chromosomal translocations in patients with chronic hepatitis B" *JHEP reports*
59. Wang (2022) "Endoplasmic reticulum stress promotes HBV production by enhancing use of the autophagosome/multivesicular body axis" *Hepatol*
60. Li (2011) "Subversion of cellular autophagy machinery by hepatitis B virus for viral envelopment" *J. Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12328836&blobtype=pdf | # Letter to the editor Comments on "An unusual presentation of a gastric gastrointestinal stromal tumor as a chronic contained perforation with concomitant Helicobacter pylori infection: A case report and literature review"
Omid Pouresmaeil, Masoud Keikha, Hadi Farsiani
## Abstract
We read with great interest the case report by Lee et al. describing an atypical presentation of a gastric gastrointestinal stromal tumor (GIST) manifesting as a chronic contained gastric perforation in the presence of concomitant Helicobacter pylori infection [1]. This report is undoubtedly thought-provoking, merging two significant gastrointestinal pathologies into a single diagnostic scenario and underlining the complexity of the differential diagnosis of chronic gastric symptoms. This case, while rare and educational, highlights the diagnostic challenges that persist in distinguishing chronic perforation from submucosal tumors, particularly when overlapping with a high-prevalence infection such as H. pylori. However, the report presents several methodological and interpretative issues that warrant further discussion to clarify the clinical implications and ensure optimal learning value for a broader readership.
## 1. Insufficient histopathological and immunohistochemical detailing
Although the diagnosis of GIST was ultimately confirmed, the report lacked a comprehensive histopathologic discussion, particularly regarding tumor risk stratification. Details such as the mitotic index, tumor size, Ki-67 labeling index, and immunohistochemical markers (e. g., DOG1 and CD34) were either minimally addressed or omitted. These factors are essential for guiding prognosis and decisions regarding adjuvant therapy with imatinib. Additionally, the authors could have further elaborated on whether H. pylori presence influenced the histological interpretation of the surrounding gastric mucosa, given the chronic inflammatory context [2].
## 2. Limited discussion of diagnostic delay and clinical reasoning
The unusual presentation with a chronic contained perforation raises the question of a potential diagnostic delay. However, the article lacks critical reflections on initial clinical suspicions, imaging interpretations, or endoscopic findings prior to surgery. Did earlier imaging show submucosal thickening, pneumoperitoneum, or inflammatory reactions misattributed to ulcer disease? A timeline of clinical evolution and decision-making would have added significant educational value [3].
## 3. Overlooked pathophysiological link between GIST and H. pylori
The authors briefly mentioned H. pylori infection but did not explore the possible interactions, such as mechanistic or coincidental, between H. pylori-induced inflammation and GIST pathogenesis or symptom modulation. Although current evidence does not support a causal relationship, the inflammatory milieu induced by chronic H. pylori infection may obscure or mimic GIST-related presentations, potentially contributing to misdiagnosis or delayed detection of GIST. This potential confounder could have been examined more critically in the discussion [4].
## 4. Literature review not sufficiently exhaustive
Although the case is rare, the literature review appears limited in scope. At least two similar cases of perforated or ulcerated gastric GISTs presenting with atypical features have been reported in the past decade, but were not referenced. A more systematic approach to the literature would strengthen the claim of uniqueness and aid clinicians in recognizing comparable scenarios [5].
## 5. Conclusion
Lee et al. presented a clinically intriguing case that bridges gastric oncology and infectious pathology in a unique anatomical context. While the novelty of the presentation is evident, the discussion would benefit from a more robust integration of histopathological precision, clinical decision analysis, and literature contextualization. Such enhancements would significantly increase the educational and translational value of this compelling case report.
## References
1. Lee, Bowman (2025) "An unusual presentation of a gastric gastrointestinal stromal tumor as a chronic contained perforation with concomitant Helicobacter pylori infection: a case report and literature review" *Int. J. Surg. Case Rep*
2. Sözütek, Yanık, Akkoca et al. (2014) "Diagnostic and prognostic roles of DOG1 and Ki-67, in GIST patients with localized or advanced/metastatic disease" *Int. J. Clin. Exp. Med*
3. Timbang, Sinigayan, Molla (2023) "A missed bowel perforation-the importance of diagnostic reasoning"
4. Kagihara, Matsuda, Young et al. (2020) "Novel association between Helicobacter pylori infection and gastrointestinal stromal tumors (GIST) in a multi-ethnic population" *Gastrointest. Stromal Tumor*
5. Chaudhary, Kumar (2025) "Rare diseases: a comprehensive literature review and future directions" *J. Rare Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548385&blobtype=pdf | # Baculovirus 25K hijacks host UAP56 to facilitate nuclear export of viral mRNA in insect cells
Sixuan Xiao, Huizhen Guo, Jiayi Liu, Lihua Wei, Qingqing Yang, Enyu Xie, Bingbing Wang, Qingyou Xia, Liang Jiang
## Abstract
Bombyx mori nucleopolyhedrovirus (BmNPV) is a baculovirus that hijacks host genes to cause viral infections. UAP56 is highly conserved in different species. Previous studies have shown that UAP56 is involved in several viral infections. However, its role in insect-baculovirus interactions remains unknown. In this study, we aimed to identify which BmNPV proteins interact with host UAP56 and to characterize the associated mechanism underlying viral infection using a BmNPV-silkworm model. Our data indicated that CCT018159, an inhibitor of UAP56, could suppress the proliferation of BmNPV and that the addition of CCT018159 within 12 h post-infection had a significant protective effect on BmE cells. To identify the interacting viral proteins, recombinant UAP56-GST was constructed through prokaryotic expression for pull-down screen ing. Further, immunofluorescence, co-immunoprecipitation, and pull-down analyses demonstrated that the late viral protein, 25K, directly binds to UAP56. CCT018159 did not affect the co-localization of 25K and UAP56 but disrupted the interaction between them, resulting in significant upregulation of viral mRNA content in the nucleus and opposite trend in the cytoplasm. Overexpression of 25K and UAP56 caused a significant reduc tion in viral mRNA content in the nucleus and a significant increase in the cytoplasm. The addition of CCT018159 counteracted this effect. Overall, our data show that the baculovirus 25K protein hijacks host UAP56 to facilitate the nuclear export of viral mRNA to cause infections. IMPORTANCE Nuclear export of viral mRNA is essential for viral proliferation. UAP56 is highly conserved among species and is involved in multiple viral infections. In this study, we found that the Bombyx mori nucleopolyhedrovirus 25K protein hijacks host UAP56 to facilitate viral mRNA nuclear export, and disruption of their interactions can inhibit viral proliferation. Our results provide novel insights into the mechanism of insect-baculovirus interaction and emphasize the important role that 25K plays in baculovirus infection. This research not only deepens our understanding of the transcription and translation mechanisms of baculoviruses but also provides potential targets for antiviral research.
occlusion-derived virion (ODV) (7). After the virus invades the host cell, it hijacks host genes for transcription and translation. Baculovirus gene expression is strictly time-regu lated and can be divided into four categories according to the transcription initiation time: immediate-early (<4 h post-infection [hpi]), delayed-early (5-7 hpi), late (8-18 hpi), and very late genes (>18 hpi) (8). This tightly regulated temporal program ensures the ordered progression of viral replication from DNA synthesis to progeny virion assem bly and egress. Early gene expression activates viral DNA replication and late gene expression (9), followed by assembly to produce new virions. A total of 28 early and 78 late BmNPV genes have been identified (10)(11)(12). When a large number of late genes are transcribed, the nucleation and transport efficiency of viral mRNA affect the number of progeny viruses and the entire infection process.
Baculoviruses regulate the host in two ways. On the one hand, the virus inhibits the host antiviral signaling pathway to achieve immune escape. For example, baculoviruses induce Bmserpin2 to inhibit the polyphenol oxidase pathway-mediated melaninization response (13); BmNPV induces BmPGRP2-2 to inhibit PTEN expression, resulting in an increase in p-Akt levels and the inhibition of apoptosis and autophagy (14); BmNPV downregulates BmSpry, activates the ERK pathway, and promotes viral proliferation (15). On the other hand, viruses hijack host genes to participate in multiple processes, such as viral transcription, translation, and replication, and promote virus proliferation. For example, the capsid protein, VP39, promotes viral replication by binding to host phosphatase 38K and DNA-binding protein P6.9 (16,17); E25, a key component of the ODV envelope, regulates viral DNA release and intranuclear replication by interacting with host membrane proteins and nucleoporins (18); the virus-encoded BmNPV-mir-1 and BmNPV-mir-3 enhance BmNPV infection by regulating expression of the output protein 5 cofactor, Ran, and viral P6.9, respectively (19,20). A large number of studies on baculovirus-insect interactions have focused on the function of baculovirus genes through sequence mutations; however, few have investigated the host proteins involved in baculovirus interactions.
UAP56 (NCBI reference sequence NP_004631.1) is known as an ATP-dependent RNA helicase. As a multifunctional RNA metabolism regulator, UAP56 participates in various RNA metabolic processes in eukaryotic cells. Its functions span through transcription, splicing, and quality control. First, in terms of regulating the assembly and fidelity of pre-mRNA spliceosomes, UAP56 assists in the binding of U2 small nuclear ribonucleopro teins to SF3b complexes by binding to the branch site region of pre-mRNA to ensure the precise assembly of spliceosomes (21). Second, in terms of RNA quality control for transcriptional coupling, UAP56 dynamically interacts with RNA polymerase II to monitor the integrity of nascent RNA during transcriptional elongation (22). In addition, UAP56 is a core component of transcription/export complex (TREX). TREX is a large multipro tein complex involved in cellular bulk mRNA nuclear export (23). UPA56 is hijacked by multiple viruses during viral infection to achieve efficient replication (24). One of its core functions is to promote viral RNA splicing and transcription complex assembly; for example, influenza A virus directly recruits UAP56 through nuclear proteins (NPs) to form TREX-NP complexes, enhancing viral RNA polymerase activity and regulating mRNA nuclear export (25). In addition, UAP56 participates in the infection process by mediating the nucleocytoplasmic transport of viral mRNA. For example, human cytomegalovirus binds to UAP56 using the UL69 protein, bypassing the host splicing mechanism to drive the nuclear export of intron-free virus mRNA (26). Hepatitis B virus regulates viral RNA splicing and transport through HBx protein interactions and inhibits the inter feron signaling pathway to maintain persistent infection (27). Evolutionary conservation studies have shown that the RNA transport function of UAP56 is highly conserved in eukaryotes, such as Drosophila and yeast, and the deletion of its homologous genes (Hel25E and yeast Sub2) leads to RNA retention in the nucleus, providing a molecular basis for the mechanism of cross-species host hijacking by viruses (28,29).
We speculated that UAP56 may be involved in baculovirus-insect interactions, and its function during these interactions was investigated using the BmNPV-silkworm model.
We examined the effect that UAP56 has on the proliferation of BmNPV, identified the viral proteins that interact with UAP56, and analyzed the molecular mechanism of action of UAP56 in viral infection.
## RESULTS
## CCT018159 targets UAP56 to inhibit BmNPV proliferation
In the literature, the small molecule compound, CCT018159 (CCT), has been repor ted to exhibit inhibitory effects on both UAP56 and HSP90 (30), and 17-DMAG is an alternative inhibitor of HSP90. BmE cells were treated with CCT and 17-DMAG at different concentrations (0, 2.5, 5, 10, 20, 40, and 80 µM), and cell viability was assessed using 3-(4,5-dimethyl-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium (MTS) assays 72 h later, which showed that 17-DMAG did not exhibit significant cytotoxicity at 80 µM (Fig. 1Aa), whereas CCT exhibited dose-dependent cytotoxicity (Fig. 1Ba). After BmNPV infection, different doses of drugs were added, and total DNA was extracted for quantitative PCR (qPCR) detection 24 h later. Results showed that the virus content decreased with an increase in CCT dose (Fig. 1Bb). Fluorescence observation and Western blotting were performed 72 h after infection, showing that virus fluorescence (Fig. 1Bc) and GFP protein content (Fig. 1Bd) decreased with an increase in CCT dose (Fig. 1Bc). However, no notable trends were observed in the 17-DMAG treatment group (Fig. 1Ab-d). These results suggest that CCT targets UAP56 rather than HSP90 to inhibit BmNPV.
After BmE cells were infected with BmNPV, 10 µM CCT was added, and total RNA was extracted at 3, 6, 12, and 24 hpi for qPCR detection. Results showed that the viral mRNA content in the CCT-treated group was significantly lower than that in the control at 24 hpi (Fig. 1C). Total protein and DNA were extracted at 24 hpi, and the viral protein content in the CCT-treated group was found to be significantly lower than that in the control group, as determined via Western blotting (Fig. 1D). Moreover, qPCR showed that CCT caused a significant decrease in the number of viral DNA copies (Fig. 1E). These results suggest that CCT significantly inhibited the proliferation of BmNPV.
## CCT acts during the late gene expression stage of BmNPV
BmE cells were treated with 10 µM CCT at different time points after BV-GFP (a GFP-tag ged virus) infection, and dimethyl sulfoxide (DMSO) was added as a control. Total DNA of the cells was extracted 48 h after infection, and the reverse transcription PCR (RT-PCR) results showed that the virus content in the cells was significantly lower than that in the DMSO control group at 0, 4, 8, and 12 h after infection with the addition of CCT (Fig. 2A).
The qPCR results showed that the addition of CCT at 0 and 12 h after infection signifi cantly reduced the viral content, and the inhibitory effect exerted by both treatments on BmNPV was nearly identical (Fig. 2B). Fluorescence observation was performed 72 h after infection and showed that virus fluorescence was significantly weaker than that in the DMSO group in cells treated for 0, 4, 8, and 12 h (Fig. 2C). These results suggest that CCT addition within 12 h after infection significantly inhibited BmNPV, whereas CCT addition at 24 hpi had a poor viral inhibitory effect. These findings suggest that the inhibitory effect of CCT on BmNPV occurred mainly during the late viral gene expression stage.
## BmUAP56 interacts with BmNPV 25k
We cloned BmUAP56 and found an 89% amino acid similarity (Fig. S1A) between BmUAP56 and human HsUAP56, which both contain DEXDc and HELIc domains (Fig. S1B). The RT-PCR results showed that BmUAP56 was expressed in silkworms at all developmental stages (Fig. S1C), and qPCR showed that the gene was expressed in all silkworm tissues (Fig. S1D). A prokaryotic expression vector of BmUAP56 was construc ted, and its expression was induced at 25°C with 0.1% isopropyl β-D-1-thiogalactopyrano side (IPTG). The target protein band (Fig. S1E) was obvious in the supernatant, and the purified GST-BmUAP56 recombinant protein (Fig. S1FG) was obtained after GST affinity chromatography and molecular sieve and ion exchange purification. At 48 h after the ingestion of BmNPV and phosphatebuffered saline, total midgut protein was extracted as the input, the lysate was used as the control for pull-down of GST-BmUAP56 and the control GST. Electrophoresis showed that BmNPV had a specific binding band with GST-BmUAP56, with a size of approximately 25 kDa (Fig. 3Aa). The band was mined for mass spectrometry analysis (Table S1), and five viral proteins were identified: polyhedrin, orf44, E25, E56, and 25K (Fig. 3Ab). Among them, only E25 and 25K were late essential genes of BmNPV, which were thus cloned to construct incremental expression vectors. BmE cells were co-transfected with vectors encoding HA-tagged E25 and His-tagged 25K with BmUAP56-Flag. Immunofluorescence showed that E25 and BmUAP56 proteins localized to different locations in the cell (Fig. 3Ba), whereas 25K co-localized with BmUAP56 in the nucleus (Fig. 3Ca). The co-immunoprecipitation (Co-IP) results showed that E25 did not bind to BmUAP56 or control GFP (Fig. 3Bb), but 25K specifically bound to BmUAP56 (Fig. 3Cb). These results suggest that BmUAP56 binds to the late essential protein, 25K, of the baculovirus. The recombinant protein GST-UAP56-Flag and SUMO-25K-His were purified (Fig. S1H). Pull-down results indicated that SUMO-25K bound to GST-UAP56 but not GST (Fig. 3D), and SUMO-25K but not SUMO bound to GST-UAP56 (Fig. S2), suggesting that BmUAP56 and 25K are directly interacted.
## 25K binding to BmUAP56 promotes nuclear export of viral mRNA
BmE cells were co-transfected with UAP56-Flag and 25K-His expression vectors and treated with 10 µM CCT for 48 h. CCT did not affect the co-localization of BmUAP56 and 25K proteins (Fig. 4A). Co-IP experiments showed that CCT inhibited the binding of BmUAP56 to 25K, and the inhibitory effect increased with an increase in drug concentra tion (Fig. 4B). After BmE cells were infected with BmNPV, 10 µM CCT was added, and the cells were extracted at 24 hpi for nucleoplasmic isolation. Antibodies for H4 and GAPDH were detected in the nucleus and cytoplasmic protein samples, respectively, indicat ing that nucleoplasmic isolation was successful (Fig. 4Ca). The qPCR detection results showed that the mRNA content of the virus in the nucleus of the CCT treatment group significantly increased (Fig. 4Cb). The opposite result was observed in the cytoplasm (Fig. 4Cc). The transient expression vector of viral gene VP39 was constructed and transfec ted into BmE cells, and nuclear and cytoplasmic RNA was extracted after nucleoplasm isolation (Fig. 4Da). The mRNA content of VP39 in the nucleus of BmE cells co-transfected with VP39, UAP56, and 25K expression vectors was significantly lower than that in the control group (only basic 1180 and VP39 expression vectors were transfected) (Fig. 4Db), while a contrary result was detected in the cytoplasmic RNA (Fig. 4Dc). These trends could be reversed by the addition of CCT (Fig. 4Db andc). These results suggest that 25K binds to BmUAP56 and promotes the nuclear export of viral mRNA and that CCT inhibits this process by blocking the binding of BmUAP56 and 25K.
## DISCUSSION
Viruses commonly hijack key genes in hosts to achieve efficient infection. UAP56 is highly conserved in different species and has important biological functions. Therefore, it has become a target of viral regulatory hosts (31). Our study revealed that UAP56 is involved in baculovirus infection. CCT targets UAP56, and 17-DMAG is an alternative inhibitor of HSP90 (30). CCT inhibited the proliferation of BmNPV, and the inhibitory effect increased with an increase in drug concentration; however, 17-DMAG exhibited no similar phenomenon (Fig. 1). The addition of CCT within 12 hpi significantly inhibited viral proliferation (Fig. 2), indicating that CCT exerts its function in the late gene expression stage of BmNPV. Therefore, if a BmNPV protein binds to BmUAP56, it should be considered as a late essential gene of the virus. Through protein expression, pull-down, immunofluorescence, and Co-IP analysis, the late essential gene of BmNPV, 25K, was demonstrated to directly bind to BmUAP56 (Fig. 3), whereafter they promote the nuclear export of viral mRNA (Fig. 4). CCT can block the binding of BmUAP56 and 25K, resulting in the abnormal accumulation of viral mRNA in the nucleus, causing a decrease in viral content in the cell, and ultimately inhibiting viral proliferation.
As a central hub of host RNA metabolism, UAP56 plays a key role in mRNA tran scription, splicing, and nuclear export (32). It is one of the key targets of virus-host interactions and is involved in the infection of a variety of viruses. After influenza virus infects the host, the viral NP protein directly binds to UAP56 to maintain stability of the vRNA-NP complex, promote the formation of ribonucleoproteins, and regulate viral RNA synthesis through helicase activity (33). The Rev protein of HIV-1 binds to UAP56 and assists in the export of unspliced/partially spliced mRNA via the CRM1 noncanonical pathway, which may be involved in helicase-dependent gRNA dimerization (34). The SM protein of Epstein-Barr virus binds to UAP56 to promote the stability and export of mRNA during the late stage of the virus (35). The ICP27 protein of herpes simplex virus recruits UAP56 and Aly/REF to bypass host splicing checkpoints and directly exports incompletely spliced viral mRNA (36,37). Our results suggest that the baculovirus 25K protein binds to UAP56 to promote the nuclear export of viral mRNA, thereby ensur ing viral proliferation. The baculovirus genome does not contain introns, which would necessitate the hijacking of UAP56 for efficient nuclear export of viral mRNA. Hijacking UAP56 may be a conservative strategy used by viruses to regulate their hosts, although its role varies among different viral infections.
The 25K protein plays an important role during multiple stages of baculoviral infection. Previous studies have found that 25K regulates both replication and structural stability, as observed through viral gene knockout, Co-IP, cryo-electron microscopy, and other methods. Baculovirus 25K forms a functional replication complex with DNA polymerase, helicase (P143), and the LEF-1/2 cofactor (38), which plays a scaffolding role in the initial stage of DNA replication; deletion of 25K leads to replication arrest, as well as loose nucleocapsid structure and VP39 arrangement (39). Notably, the function of 25K is spatiotemporally specific: early deletions result in the defective generation of ODVs, whereas late deletions do not affect BV release (40). Transcriptome analysis revealed that 25K balances the viral transmission strategy by regulating the expression of ODV genes, such as p74, and BV genes, such as gp64 (41,42). In the present study, the binding of viral proteins to UAP56 was blocked by inhibitors without missing 25K, and 25K was found to demonstrate a new function of hijacking UAP56 to promote the nuclear export of viral mRNA. Deletion of 25K results in higher levels of BV (43), but inhibition of binding with UAP56 appears to have the opposite effect. Key sites for the binding of UAP56 to 25K could be identified in future experiments and targeted through gene editing to enhance the antiviral capacity of silkworms.
In conclusion, we confirmed that the baculovirus 25K protein hijacks host UAP56 to facilitate the nuclear export of viral mRNA and consequently cause infections. This research not only deepens our understanding of the transcription and translation mechanisms used by baculoviruses but also provides potential targets for antiviral research.
## MATERIALS AND METHODS
## Silkworms, cells, and viruses
The B. mori Dazao strain was maintained at the Silkworm Gene Resource Bank of Southwest University (Chongqing, China). BmE cells were cultured at 27°C. BmNPV and GFP-expressing BmNPV were collected from the hemolymph of infected silkworm larvae and BmE cells, respectively.
## Viral challenge and toxicity assay
BmE cells were seeded in 24-well plates and infected with BmNPV at a multiplicity of infection of 1 by adding 0.1 µL viral suspension per well, followed by incubation for 1 h. After infection, the viral medium was replaced with a drug-containing medium, and the cells were cultured for an additional 48 h. Genomic DNA was extracted using an OMEGA kit, and viral proliferation quantified via qPCR targeting the late viral gene, GP41, and the reference gene, GAPDH. Viral fluorescence was observed under green and white light at 72 hpi. The total protein was extracted for Western blotting using antibodies against GFP and GAPDH at 72 hpi. Student's t-test was used to analyze the statistical data.
## Drug cytotoxicity assay
Cell viability was measured using an MTS-based CellTiter 96 AQueous One Solution Cell Proliferation Assay (Promega). Cells were seeded in 96-well plates with 80 µL cell suspension or medium per well (four replicates per drug concentration), alongside four background control wells containing medium only. Each well received 20 µL drug-sup plemented complete medium, with DMSO as a control. After 72 h at 27°C, 20 µL Cell Titer 96 AQueous One Solution Reagent was added, followed by a 3 h incubation in the dark. Absorbance at 490 nm was measured using a microplate reader, and backgroundsubtracted values were used to calculate the relative cell viability as a percentage of the control group.
## Plasmid construction, protein expression, and purification
A PGEX-4T-BmUAP56 vector for the fusion expression of GST-BmUAP56 was constructed and transformed into competent cells. Single colonies were expanded in culture for 4-6 h at 37°C and 220 rpm, followed by induction with 0.1% IPTG under varying conditions (4 h at 37°C, 10 h at 25°C, or 20 h at 16°C). Uninduced controls were cultured at 37°C for 4 h. Cells were harvested using centrifugation (8,000 × g, 5 min), resuspended in equilibration buffer A, and lysed via sonication. The lysate was loaded onto a GST column pre-equilibrated with buffer A, and the bound proteins were eluted using 1 and 10 mM reduced glutathione. The column was washed and stored in 20% ethanol at 4°C.
## Pull-down and mass spectrometry
Fifth-instar Dazao silkworms were orally infected with BmNPV at a density of 1 × 10 9 polyhedra per larva. Midguts were dissected at 48 hpi, and proteins were then extrac ted. GST beads (30 µL) were mixed with 1 mL equilibration buffer A, centrifuged, and incubated with 0.9 mg purified protein (double the GST bead-binding capacity). After overnight rotation at 4°C, beads were washed five times with buffer A. Bound proteins were eluted, resolved via sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and specific bands were excised for mass spectrometry analysis.
## References
1. Matsumoto, Sekimizu (2019) "Silkworm as an experimental animal for research on fungal infections" *Microbiol Immunol*
2. Jiang (2021) "Insights into the antiviral pathways of the silkworm Bombyx mori" *Front Immunol*
3. Shi, Zhang, Zhu et al. (2021) "Response to Bombyx mori nucleopolyhedrovirus infection in silkworm: gut metabolites and microbiota" *Developmental & Comparative Immunology*
4. Fei, Xia, Mehmood et al. (2024) "Autophagy promotes replication of Bombyx mori nucleopolyhedrovirus in insect cells" *Int J Biol Macromol*
5. Jiang, Xia (2014) "The progress and future of enhancing antiviral capacity by transgenic technology in the silkworm Bombyx mori" *Insect Biochem Mol Biol*
6. Xia, Fei, Huang et al. (2024) "Single-nucleus sequencing of silkworm larval midgut reveals the immune escape strategy of BmNPV in the midgut during the late stage of infection" *Insect Biochem Mol Biol*
7. Jia, Gao, Huang et al. (2023) "Architecture of the baculovirus nucleocapsid revealed by cryo-EM" *Nat Commun*
8. Jiang, Zhao, Wang et al. (2013) "Comparison of factors that may affect the inhibitory efficacy of transgenic RNAi targeting of baculoviral genes in silkworm, Bombyx mori" *Antiviral Res*
9. Jiang, Goldsmith, Xia (2021) "Advances in the arms race between silkworm and baculovirus" *Front Immunol*
10. Gomi, Majima, Maeda (1999) "Sequence analysis of the genome of Bombyx mori nucleopolyhedrovirus" *J Gen Virol*
11. Zhao, He, Yang et al. (2019) "Nucleocapsid assembly of baculoviruses" *Viruses*
12. Van Beek, Davis (2007) "Baculovirus insecticide production in insect larvae" *Methods Mol Biol*
13. Toufeeq, Wang, Zhang et al. (2019) "Bmserpin2 is involved in BmNPV infection by suppressing melanization in Bombyx mori Insects"
14. Jiang, Liu, Guo et al. (2019) "Distinct functions of Bombyx mori peptidoglycan recognition protein 2 in Immune responses to bacteria and viruses" *Front Immunol*
15. Guo, Sun, Wang et al. (2019) "Spry is downregulated by multiple viruses to elevate ERK signaling and ensure viral reproduction in silkworm" *Developmental & Comparative Immunology*
16. Zhang, Li, Zhao et al. (2020) "Identification of A functional region in Bombyx mori nucleopolyhedrovirus VP39 that is essential for nuclear actin polymerization" *Virology (Auckl)*
17. Benning, Garcia, Nguyen et al. (2024) "Helical reconstruction of VP39 reveals principles for baculovirus nucleocapsid assembly" *Nat Commun*
18. Chen, Hu, Xiang et al. (2012) "Autographa californica multiple nucleopolyhedrovirus odv-e25 (Ac94) is required for budded Full-Length Text Journal of Virology October"
19. "virus infectivity and occlusion-derived virus formation" *Arch Virol*
20. Singh, Singh, Nagaraju (2014) "Bmnpv-miR-3 facilitates BmNPV infection by modulating the expression of viral P6.9 and other late genes in Bombyx mori" *Insect Biochem Mol Biol*
21. Singh, Singh, Bhavani et al. (2010) "Discovering microRNAs from Bombyx mori nucleopolyhedrosis virus" *Virology (Auckl)*
22. Kondo, Oubridge, Van Roon et al. (2015) "Crystal structure of human U1 snRNP, a small nuclear ribonucleoprotein particle, reveals the mechanism of 5' splice site recognition"
23. Jimeno-González, Reyes (2016) "Chromatin structure and pre-mRNA processing work together" *Transcription*
24. Belz, Wherry (2010) "Current opinion in immunology. Editorial overview" *Curr Opin Immunol*
25. Zhou, Luo, Straesser et al. (2000) "The protein Aly links pre-messenger-RNA splicing to nuclear export in metazoans" *Nature*
26. Morris, Wang, Ivey et al. (2020) "Cellular mRNA export factor UAP56 recognizes nucleic acid binding site of influenza virus NP protein" *Biochem Biophys Res Commun*
27. Lischka, Toth, Thomas et al. (2006) "The UL69 transactivator protein of human cytomegalovirus interacts with DEXD/H-Box RNA helicase UAP56 to promote cytoplasmic accumulation of unspliced RNA" *Mol Cell Biol*
28. Hu, Yu, Zhu et al. (2020) "Cellular UAP56 interacts with the HBx protein of the hepatitis B virus and is involved in viral RNA nuclear export in hepatocytes" *Exp Cell Res*
29. Li, Du, Huai et al. (2022) "The RNA helicase UAP56 and the E3 ubiquitin ligase COP1 coordinately regulate alternative splicing to repress photomorphogenesis in Arabidopsis" *Plant Cell*
30. Rehwinkel, Herold, Gari et al. (2004) "Genome-wide analysis of mRNAs regulated by the THO complex in Drosophila melanogaster" *Nat Struct Mol Biol*
31. Schumann, Jackson, Whitehead et al. (2016) "Targeting the ATP-dependent formation of herpesvirus ribonucleoprotein particle assembly as an antiviral approach" *Nat Microbiol*
32. Guo, Zhu, Ma et al. (2023) "Virus infection and mRNA nuclear export" *IJMS*
33. Yellamaty, Sharma (2024) "Critical cellular functions and mechanisms of action of the RNA helicase UAP56" *J Mol Biol*
34. Hu, Gor, Morikawa et al. (2017) "Cellular splicing factor UAP56 stimulates trimeric NP formation for assembly of functional influenza viral ribonucleoprotein complexes" *Sci Rep*
35. Tomezsko, Corbin, Gupta et al. (2020) "Determi nation of RNA structural diversity and its role in HIV-1 RNA splicing" *Nature*
36. Fu, Verma, Burton et al. (2019) "Cellular RNA helicase DHX9 interacts with the essential epstein-barr virus (EBV) protein SM and restricts EBV lytic replication" *J Virol*
37. Tunnicliffe, Tian, Storer et al. (2018) "Overlapping motifs on the herpes viral proteins ICP27 and ORF57 mediate interactions with the mRNA export adaptors ALYREF and UIF" *Sci Rep*
38. Tang, Patel, Krause (2016) "Herpes simplex virus ICP27 regulates alternative pre-mRNA polyadenylation and splicing in a sequencedependent manner" *Proc Natl Acad Sci*
39. Giri, Feiss, Bonning et al. (2012) "Production of baculovirus defective interfering particles during serial passage is delayed by removing transposon target sites in fp25k" *J Gen Virol*
40. Giri, Li, Sandgren et al. (2010) "Removal of transposon target sites from the Autographa californica multiple nucleopolyhedrovirus fp25k gene delays, but does not prevent, accumulation of the few polyhedra phenotype" *J Gen Virol*
41. Nakanishi, Goto, Kobayashi et al. (2010) "Comparative studies of lepidopteran baculovirusspecific protein FP25K: development of a novel Bombyx mori nucleopolyhedrovirus-based vector with a modified FP25K gene" *J Virol*
42. Li, Wang, Shen et al. (2015) "The FP25K acts as a negative factor for the infectivity of AcMNPV budded virus" *PLoS One*
43. Cheng, Hillman, Zhang et al. (2013) "Reduction of polyhedrin mRNA and protein expression levels in Sf9 and Hi5 cell lines, but not in Sf21 cells, infected with Autographa californica multiple nucleopolyhedrovirus fp25k mutants" *J Gen Virol*
44. Harrison, Summers (1995) "Mutations in the Autographa californica multinucleocapsid nuclear polyhedrosis virus 25 kDa protein gene result in reduced virion occlusion, altered intranuclear envelopment and enhanced virus production" *J Gen Virol*
45. Butt, Edavettal, Hall et al. (2005) "SUMO fusion technology for difficulttoexpress proteins" *Protein Expr Purif*
46. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12613983&blobtype=pdf | # Complete genomes of two Chlamydia psittaci isolated from ducks and pigeons in China
Xiaoxue Wang, Yihan Wang, Yanyan Wang, Junkai Zhang, Chunguo Liu, Ping Liu, Zhaocai Li, Jizhang Zhou
## Abstract
Chlamydia psittaci is an important zoonotic pathogen that can transmit from avian to human. Herein, we report complete genome sequences of two C. psittaci strains isolated from ducks and pigeons in China. The duck strain CPS-QD/LS was identified as ompA genotype A, while the pigeon strain CPS-BY/JY was identified as genotype B.
KEYWORDS Chlamydia psittaci, genome, psittacosisC hlamydia psittaci is a unique zoonotic intracellular bacterial pathogen that can infect avian species, mammals, and humans (1, 2). Based on ompA gene sequences, the C. psittaci isolates can be classified into nine classical (A to F, E/B, M56, and WC) and several atypical genotypes (3). Previous studies have shown that C. psittaci infections in avian species are common in China. However, few genome data are available for analysis of its genetic diversity and host specificity. During routine detection of Chlamydia species in our laboratory, genomic DNA was isolated from pure cultures of C. psittaci using the TIANamp Genomic DNA Kit (TIANGEN, Beijing) (4). We isolated two C. psittaci strains from real-time PCR-positive duck lung tissue and pigeon cloacal swab samples by inoculating the samples into specific-pathogen-free embryonated chicken eggs (5). The isolates obtained were confirmed by real-time PCR (6) and were named as CPS-QD/LS (duck strain) and CPS-BY/JY (pigeon strain) according to their host origin region. The isolates were then inoculated into the L929 cell line for propagation and purification for genome sequencing.Purified genome DNA was prepared for sequencing using a combination of Illumina NovaSeq 6000 and Nanopore PromethION platforms. The Illumina library was prepared using the TruSeq DNA Library Prep Kit (Illumina, San Diego, USA) and amplified within the flow cell on an Illumina cBOT instrument for cluster generation (NovaSeq 6000 PE Cluster Kit, Illumina). The clustered flow cell was loaded onto a NovaSeq 6000 sequencer (Illumina) for paired-end sequencing. The purified libraries were loaded onto primed R9.4 Spot-On flow cells and sequenced on a PacBio sequencer (Pacific Biosciences, Menlo Park, USA). Raw reads were filtered with fastp (v.0.23.2) (7), yielding 929,758,353 and 1,829,539,148 bp of clean data for strains CPS-QD/LS and CPS-BY/JY, respectively. In parallel, libraries were prepared using the NEBNext Ultra II DNA Library Prep Kit (NEB, USA) and sequenced on the PromethION platform (Oxford Nanopore Technologies, Oxford, UK) (8). Raw reads were quality-filtered (Q ≥ 7) and length-filtered (≥1,600 bp), generating 1,087,063,417 and 2,137,810,282 bp of pass reads for CPS-QD/LS and CPS-BY/JY, respectively, with N50 values of 9.04 and 5.55 kb. Hybrid genome assembly of Illumina and Nanopore reads was performed using SOAPdenovo (v.2) (9). The repeat sequence analysis was performed using the software RepeatMasker (4.1.2-p1) (10). Circularization of contigs and re-orientation to the dnaA start were assessed and corrected using BLAST. Gene prediction was carried out with Prokka (v.1.14.6) (11). Promoter regions were predicted using PromPredict (v.V1) (12), and transmembrane
proteins were identified with TMHMM (v.2.0c) (13). All software was run with default parameters.
The final genomes of C. psittaci were complete, and both genomes comprised a circular chromosome and a plasmid, with GenBank accession numbers CP103952.1, CP103953.1, CP184490.1, and CP184489.1. The genomes were annotated by National Center for Biotechnology Information Prokaryotic Genome Annotation Pipeline (v.6.2/ v.6.9). The genome characteristics are summarized in Table 1. Sequence analysis of the ompA gene of the two obtained isolates showed that duck strain CPS-QD/LS was identical to that of a duck isolate, SZ15, and two human isolates in China (14), which could be classified to the ompA genotype A group (Fig. 1), while the ompA gene sequence of CPS-BY/JY matched the C. psittaci strains of the typical genotype B group (15), as shown in Fig. 1. These data may improve our understanding of the genetic diversity and host specificity of this Chlamydia species.
## References
1. Kaleta, Taday (2003) "Avian host range of Chlamydophila spp. based on isolation, antigen detection and serology" *Avian Pathol*
2. Li, Liu, Hou et al. (2020) "Detection of Chlamydia psittaci and Chlamydia ibidis in the endangered crested ibis (Nipponia nippon)" *Epidemiol Infect*
3. Ravichandran, Anbazhagan, Karthik et al. (2021) "A comprehensive review on avian chlamydiosis: a neglected zoonotic disease" *Trop Anim Health Prod*
4. Genomic, Kit
5. Pantchev, Sting, Bauerfeind et al. (2010) "Detection of all Chlamydophila and Chlamydia spp. of veterinary interest using species-specific real-time PCR assays" *Comp Immunol Microbiol Infect Dis*
6. Braukmann, Sachse, Jacobsen et al. (2012) "Distinct intensity of host-pathogen interactions in Chlamydia psittaciand Chlamydia abortus-infected chicken embryos" *Infect Immun*
7. Chen, Zhou, Chen et al. (2018) "Fastp: an ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
8. Kolmogorov, Billingsley, Mastoras (2023) "Scalable Nanopore sequencing of human genomes provides a comprehensive view of haplotype-resolved variation and methylation" *Nat Methods*
9. Li, Zhu, Ruan et al. (2010) "De novo assembly of human genomes with massively parallel short read sequencing" *Genome Res*
10. Tarailo-Graovac (2009) "Using RepeatMasker to identify repetitive elements in genomic sequences" *Curr Protoc Bioinform Chapter*
11. Seemann (2014) "Prokka: rapid prokaryotic genome annotation" *Bioinformatics*
12. Huerta, Serra, Gabaldón (2019) "eggNOG 5.0: a hierarchical, functionally -and phylogenetically -aware orthology resource for bacteria and archaea" *Nucleic Acids Res*
14. Krogh, Larsson, Von Heijne et al. (2001) "Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes" *J Mol Biol*
15. Lin, Chen, Liao et al. (2019) "A parrot-type Chlamydia psittaci strain is in association with egg production drop in laying ducks" *Transbound Emerg Dis*
16. Mitchell, Wolff, Thacker et al. (2009) "Genotyping of Chlamydophila psittaci by real-time PCR and high-resolution melt analysis" *J Clin Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12180221&blobtype=pdf | Virology Journal (2025) 22:91 h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 / s 1 2 9 8 5 -0 2 5 -0 2 6 9 7 -8
In this article [1], Figs. 1 and 2 captions had been interchanged. Caption of Fig. 1 inadvertently give in as caption for Fig. 2 and vice versa. The figure(s) should have appeared as shown below.
The original article has been corrected.
## Virology Journal
The original article can be found online at h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 / s 1 2 9 8 5 -0 2 5 -0 2 6 9 7 -8.
## References
1. Zhang, Wang, Zhou (2025) "Characterization of the epidemiology, susceptibility genes and clinical features of viral infections among children with inborn immune errors: a retrospective study" *Virol J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645915&blobtype=pdf | # A conserved RNA structure at the capsid-coding sequence of Zika virus genome is required for viral replication in a hostdependent manner
Guadalupe Navarro, Horacio Pallarés, María González, López Ledesma, Luana De Borba, Romina Mazzolenis, Andrea Gamarnik
## Abstract
Flaviviruses are emerging and re-emerging pathogens causing widespread epidemics worldwide. Their RNA genomes play multiple roles during infection, folding into dynamic structures that regulate viral processes. To understand the mechanisms of flavivirus infection and to design genetic tools for viral countermeasures, it is important to dissect functional RNA structures present in viral genomes. Here, we investigate RNA structures within the open reading frame of the Zika virus (ZIKV) genome that regulate viral replication. We identified a functional stem-loop structure, SL1, located within the conserved C1 element in the capsid protein coding sequence of mosquito-borne flavivirus genomes. The integrity of the SL1 structure was crucial for viral RNA amplifi cation in mosquito cells and enhanced ZIKV replication in vertebrate cells. Evolution experiments in mosquito cells with lethal SL1-disrupting mutants revealed reversions and pseudo-reversions that restored SL1 structure, confirming its role as a cis-acting RNA element. We also found that a sequence within SL1 contributes to a novel genome cyclization element unique to ZIKV. This sequence folds locally into SL1 or hybridizes with a 3' UTR sequence to extend the conserved cyclization sequence (CS1), which is known to be essential for RNA synthesis. Although the C1 element is conserved among mosquito-borne flaviviruses, the RNA structures and long-range interactions in this element required for ZIKV replication differ from those reported for dengue virus. Our studies highlight the presence of a conserved RNA element operating through distinct mechanisms in related flaviviruses. These findings offer insights into the dynamic nature of the ZIKV genome and provide information for rational flavivirus attenuation. IMPORTANCE Flaviviruses are important human pathogens mainly vectored by arthropods. They contain RNA genomes that fold into complex structures with biological functions in viral infection. Zika virus is a flavivirus that has caused significant outbreaks and epidemics around the world. In this study, we used Zika virus to identify functional RNA structures present in the viral coding sequence. We manipulated an infectious clone from an Argentinean Zika virus isolate to dissociate protein-coding sequences from cis-acting RNA structures and discovered an RNA element in the capsid coding region that is essential for Zika virus replication in mosquito cells. Point mutations, disrupt ing the identified structure, impaired infection in mosquito cells and rendered viral attenuation in mammalian cells. Selection of revertant viruses in cell culture restored the RNA structure and the viral replication capacity. Our studies provide a basic understand ing of the flavivirus genome organization, which is necessary for designing rational antiviral strategies.
V iral RNA genomes play a myriad of functions during infection. In addition to coding for viral proteins, these molecules contain RNA regulatory elements in the coding and non-coding regions that govern viral replication and participate in processes to escape or trigger host antiviral responses (1)(2)(3)(4). A combination of RNA structure evolutionary conservation studies, high-throughput RNA structure mapping, and functional analysis highlights the complexity of the spatial organization of viral RNAs (5)(6)(7)(8). RNA folding includes local and long-range RNA-RNA interactions that determine the overall organization of viral genomes (9). Viral RNAs exist as dynamic ensembles including multiple conformations (6,10). The composition of these ensembles depends on genome engagement with the translation or RNA replication machinery, the cell environment, and the interaction with viral and cellular proteins. For instance, an equilibrium between different conformations of flavivirus genomes is crucial for viral RNA synthesis (11,12). In this regard, RNA structures with alternative conformations were found to be essential for viral replication, indicating the flexibility and the multiple folding properties of RNA genomes in the infected cell (11,13).
Recent studies have provided important information regarding RNA genome folding using global mapping of RNA-RNA interactions in different pathogenic viruses (6)(7)(8)14). The challenge is to define their roles and mechanisms involved in viral infection. Here, we used the Zika virus (ZIKV) to assess the relevance of RNA structures present in the viral coding sequence. ZIKV is a mosquito-borne flavivirus that emerged in 2015 as a pathogen of global concern that spread across South and Central America (15,16). The flavivirus genus includes a number of important human pathogens such as dengue virus (DENV), West Nile virus (WNV), Japanese encephalitis virus (JEV), tick-borne encephalitis virus (TBEV), and yellow fever virus (YFV) (17). Flaviviruses contain a positive-strand RNA genome of about 11 kb, encoding at least ten viral proteins in a single open reading frame. A number of conserved RNA structures present in the coding and non-coding regions of flavivirus genomes have been identified to modulate RNA replication and counteract antiviral responses (13,(18)(19)(20)(21)(22)(23)(24). In this regard, flavivirus replication requires long-range RNA-RNA interactions that lead to genome cyclization, a conformation that is essential for polymerase initiation during viral RNA synthesis (12). Long-range interac tions are mediated by complementary sequences, including the conserved sequence 1 (CS1) (25), the upstream AUG region (UAR) (26,27), the downstream AUG region (DAR) (28,29), and other sequences that are virusspecific. Also, RNA structures at the 3'UTR of flavivirus genomes participate in both RNA replication and immune evasion mecha nisms. Complex folding of RNA structures, known as xrRNAs, inhibits 5' to 3' degradation of the viral genome (30,31). This leads to the production of non-coding RNAs, sfRNAs, which accumulate during infection and counteract host antiviral responses. Recent data on the mechanistic aspects of how ZIKV sfRNA evades immune responses have been reported (32,33).
RNA structures that modulate flavivirus replication have also been described in protein-coding sequences. Downstream of the cyclization elements in the coding sequence of the capsid protein, the conserved capsid-coding region hairpin (cHP) has been proposed to modulate translation initiation (34) and to be essential for RNA replication in DENV and WNV (22); however, the mechanism by which the cHP partic ipates in viral RNA replication is still unclear. More recently, using DENV as a model, an additional structure downstream of the cHP, named C1 or DCS-PK, has been described to enhance viral replication by modulating genome cyclization (20,35).
Flaviviruses naturally alternate between mosquitoes and humans, replicating efficiently in these two very different hosts. In this regard, there are flavivirus RNA structures that are essential for infection in both hosts, whereas there are RNA elements that play hostspecific functions (3,(36)(37)(38). Adaptation of flaviviruses to a specific host often results in the emergence of viral variants that replicate more efficiently in that host, at the expense of reduced fitness in the alternative host. Some of these adaptive variations involve alterations in viral RNA structures, highlighting a tradeoff where the RNA adopts a suboptimal folding state that balances the requirements of both hosts without being fully optimized for either. For example, we have previously demonstrated in DENV that restricting viral replication in mosquitoes or mosquito cells leads to the accumulation of mutations in the 3' UTR of the viral genome. These mutations enhance replication in mosquitoes but diminish infection in human cells (13,39).
Here, we identified a stem-loop structure, SL1, located within the coding sequence of the capsid protein, as essential for ZIKV infection in mosquito cells while having a minor impact on replication in mammalian cells. SL1 is part of the C1 structure (also referred to as DCS-PK), previously described as relevant for DENV infection. However, our findings reveal that the functional elements within C1 differ between ZIKV and DENV. Mutations that disrupt and reconstitute the SL1 structure in ZIKV, without altering the encoded protein, demonstrate the critical role of SL1 in viral RNA replication. In addition, lethal mutations targeting SL1 produced revertant and pseudo-revertant viruses that restored the structure, indicating strong selective pressure to maintain its integrity. In DENV2, C1 interacts with a conserved 3'UTR structure known as DB1, which is absent in the ZIKV genome. In this study, we found that a sequence within SL1 is complementary to a sequence at the viral 3'UTR present upstream of 3'CS1. Mutational analysis and reconstitution of this potential long-range interaction confirmed its requirement for ZIKV RNA synthesis in a host-dependent manner. Together, these findings provide new insights into the functional RNA elements of the ZIKV genome and underscore the distinct roles of conserved RNA structures in closely related flaviviruses.
## RESULTS
Flavivirus infectious clones and reporter viruses have been instrumental tools to dissect functions and mechanisms of viral RNA elements. To identify RNA structures in the ZIKV coding sequence that modulate viral RNA replication, we used an infectious clone constructed from an Argentinean ZIKV clinical isolate (36). In this system, a luciferase gene was included, and the capsid protein coding region was duplicated to dissociate potential cis-acting RNA elements that regulate RNA replication from the protein coding sequence (ZIKV-Luc, Fig. 1A). The capsid sequence immediately following the 5' UTR was used to interrogate cis-acting RNA elements that could modulate the SLA promoter activity present at the 5' end of the genome, whereas the second copy of capsid followed by prM-E provided the structural proteins for viral encapsidation. Using this construct, we generated a deletion mutant (ZIKV-Luc ΔC, Fig. 1A), retaining only the first 62 nucleotides of the first capsid-coding sequence, containing the known cyclization sequence 1 (CS1) and the conserved cHP.
Genomic RNA from the ZIKV-Luc WT and ΔC was quantified, and equal amounts of RNA were transfected into mosquito C6/36 and mammalian BHK or Huh-7 cells together with a control that was replication impaired, containing a mutation in the polymerase NS5 catalytic site (NS5 Mut). Luciferase activity was evaluated as a function of time (Fig. 1B). At 4 h post-transfection (hpt), the luciferase activity measured indicates translation of the input RNA. The difference in luciferase activity between WT and NS5 Mut after 24 hpt represents amplification of the viral genome by the viral NS5 polymerase (compare WT and NS5Mut, Fig. 1B). The ZIKV-Luc WT replicates efficiently in mosquito and vertebrate cells (BHK and human Huh-7). The ZIKV-Luc ΔC shows efficient translation of input RNA but fails replication in mosquito cells, displaying luciferase levels comparable with those of the NS5 Mut replication-impaired control (Fig. 1B). In BHK cells, ZIKV-Luc ΔC replicates but produces significantly lower luciferase levels at 24 h compared with the WT. In human Huh-7 cells, ZIKV-Luc ΔC supports efficient translation of the input RNA but shows a slight reduction in RNA amplification at 12 and 24 hpt (Fig. 1B). The data show that deletion of the capsid coding region reduces viral replication in hamster and human cells while abolishing replication in mosquito cells.
To identify functional RNA sequences or structures missing in the ZIKV-Luc ΔC mutant that may be relevant for viral RNA replication, we systematically deleted each of the previously predicted RNA elements in this region (7,14,20,35) (C1, K, L, and M) in the context of ZIKV-Luc (Fig. 1C). Viral RNAs were generated from ZIKV-Luc ΔC1, ΔK, ΔL, and ΔM mutants, and equal amounts of quantified RNA were transfected into cells alongside the WT and the NS5 Mut control. At 4 hpt, luciferase levels for all mutants were compara ble to WT, indicating that the transfected RNAs were efficiently translated (Fig. 1D).
Reporter viruses with deletions in the K, L, or M structures produced viral RNAs that were successfully amplified in both mosquito and mammalian cells, as evidenced by luciferase levels after 24 hpt. In contrast, deletion of the C1 element impaired viral RNA replication in mosquito cells and significantly delayed RNA amplification in mammalian cells. In both cell types, the phenotypes of ZIKV-Luc ΔC1 were comparable with those observed with the mutant carrying the large deletion, ZIKV-Luc ΔC. The C1 region has been previously shown to be conserved across various mosquitoborne flaviviruses (20). SHAPE analysis performed in our laboratory and others indicates that C1 of DENV folds into a short SL1, a longer SL2, and a pseudoknot (PK) interaction with downstream nucleotides (20,35) (Fig. 2A). More recently, chemical probing using ZIKV predicted a C1 region (also named DCS-PK) (9), which accommodates the same 2D structure as the one proposed using SHAPE for DENV and other flaviviruses (Fig. 2B). Using IPKnot as a tool for predicting RNA secondary structures that can model pseudo knot interactions, we obtained a C1 structure similar to the one previously proposed (Fig. S1). In addition, we analyzed the secondary structure projection of the 3D prediction of C1 obtained using AlphaFold3, which also predicted the formation of SL1, SL2, and a PK (Fig. S2).
Although it might be expected that the C1 RNA structure serves a similar function in different flaviviruses, its requirement for ZIKV replication was unclear. This is because ZIKV lacks the DB1 structure in the 3'UTR, which contains complementary sequences that enhance genome cyclization through hybridization with the C1 sequence in DENV (13,35). To address this question, we decided to further investigate the elements of C1 that are required for ZIKV replication.
We first analyzed the role of sub-elements of C1, assessing the function of SL2 and the PK interaction on ZIKV replication. To this end, we generated ZIKV mutants designed to disrupt and reconstitute the SL2 stem (SL2M1 and SL2M2) and created a mutant lacking the entire SL2 structure (ΔSL2). Additionally, we constructed a mutant disrupting the PK complementarity (PKM1) and another reconstituting it via mutations in the complemen tary strand (PKM2, Fig. 2C). Predictions of the RNA structures of SL2 and PK mutants suggest that the overall C1 was maintained and PKM2 restored PK structure (Fig. S2).
RNAs corresponding to SL2M1, SL2M2, ΔSL2, PKM1, and PKM2, along with WT and NS5 Mut control, were quantified, and equal amounts were transfected into mosquito and mammalian cells. The data show that RNAs from ZIKV-Luc mutants with SL2 disruptions or deletion were able to replicate in both host cell types (Fig. 2D). The ΔSL2 mutant exhibited a slight but significant replication delay compared with the WT reporter ZIKV but ultimately reached near WT levels by 48 hpt. Similarly, transfection of the PKM1 and PKM2 mutant RNAs showed efficient translation and replication in both hosts. These data suggest that the SL2 and the PK do not play a relevant role as cis-acting elements for RNA replication, which differs from that previously reported by our lab and others in DENV (20,35). It has been previously shown that the SL2 and PK elements of the C1 structure enhance DENV RNA amplification. To further investigate a possible role of SL2, we examined whether the SL2 structure present in the second copy of the capsid protein could provide a function when the upstream copy was deleted. To accomplish this, we generated a new ZIKV mutant with a deletion of both SL2 copies (Fig. 2E). Viral RNA levels from WT, NS5Mut, and ΔSL2 mutants, either with one or both copies deleted, were transfected, and luciferase activity was measured over time. The results show that the ZIKV-Luc mutant with double SL2 deletion replicates in both cell types, suggesting that this structure does not play a critical role in RNA amplification. It is important to mention that mutants lacking one or both SL2 copies exhibit a slight reduction in luciferase levels at 36 hpt compared with the WT in mosquito cells (Fig. 2E).
To investigate the potential role of SL1, we generated a series of ZIKV mutants carrying substitutions on either side of the stem that disrupt the structure (SL1-A, SL1-B, SL1-C, SL1-D, and SL1-E, Fig. 3A). Secondary structure predictions of mutant RNAs were analyzed using IPKnot and AlphaFold3, which indicated that SL2 and the PK were maintained (Fig. S1 andS2). Transfection of the corresponding viral RNAs showed that disrupting SL1 stem abolishes ZIKV replication in mosquito cells while causing a delay in viral replication in mammalian cells (Fig. 3B andC). We also used the mutant SL1-E to introduce additional substitutions to restore base pairing (SL1-ERec). Transfection of ZIKV-Luc SL1-ERec resulted in a significant increase in luciferase with respect to the SL1-E mutant (Fig. 3B). Although the reconstituted mutant did not fully restore WT levels of RNA amplification, the results support the importance of the SL1 structure for ZIKV replication, particularly in mosquito cells.
To further study the role of different elements of C1, instead of using the reporter ZIKV-Luc, we used the ZIKV WT infectious clone that contains a single capsid coding region. In this case, we introduced synonymous substitutions to disrupt SL1 structure, maintaining the capsid protein sequence (SL1-D and SL1-B) and included a mutant disrupting the stem of the SL2 structure (SL2M1) (Fig. 4A). Equal amounts of viral RNAs were transfected into the two host cell types, and viral propagation was evaluated by immunofluorescence assay (IFA) using antibodies against ZIKV NS3 protein and by evaluating viral particle secretion at each time using RT-qPCR. In mosquito-transfected cells, no viral signal was detected at 1 day post-transfection (dpt). IFA positive was observed at 2 dpt, and the complete monolayer was antigen positive at 3 dpt for the WT virus (Fig. 4B), whereas no IFA signal was detected for the two mutants disrupting the SL1 structure. In contrast, disruption of stem 2 in SL2M1 mutant resulted in viruses that propagate like the WT ZIKV (Fig. 4B). To evaluate viral particle secretion in a quantitative manner, viral RNA for each mutant was quantified by RT-qPCR (Fig. 4B, right panel). The results indicate that the amount of secreted viral RNA of mutant SL2M1 was similar to that of the WT. In contrast, no secreted viral RNA was detected for the SL1-D and SL1-B mutants (Fig. 4B). In mammalian cells, replication of the ZIKV WT showed IFA positive at 1 dpt, with the complete monolayer infected at 2 dpt and cytopathic effect observed at 3 dpt (Fig. 4C). The two mutants in the SL1 structure showed delayed propagation with a few ZIKV antigen-positive cells at 1 dpt, but after 2 days, the monolayer was widely infected (Fig. 4C). In addition, the mutant SL2M1 propagated as efficiently as the WT. Quantification of the secreted viral RNA showed a delay of RNA accumulation for the mutants ZIKV SL1-D and SL1-B, but not for the SL2M1, compared with the WT ZIKV (Fig. 4C, right panel).
It is important to mention that although the nucleotide substitutions in SL1-D and SL1-B were synonymous, we cannot rule out an impact of those sequence changes on a viral function. However, since the mutants only carry one or two synonymous changes, and the phenotype recapitulates that observed in the reporter system (without nucleo tide changes in the capsid coding sequence), we propose that the drastic impairment of viral replication in mosquito cells can be attributed to the disruption of the RNA struc ture.
The data support that the SL1 structure is critical for ZIKV replication in mosquito cells. To further investigate the importance of this RNA element, we tested infections of mosquito cells with viral stocks of mutants produced in mammalian cells. Viral stocks (~10⁷ pfu/mL) of WT, SL1-A, and SL1-B mutants were produced in BHK cells (Fig. 5A). The mutants exhibited small plaque phenotypes in these cells (Fig. 5B ), in agreement with their delayed RNA replication (Fig. 3 and4). To rule out reversions or selection of additional mutations in mammalian cells, viruses were used to extract viral RNA for sequencing. This analysis confirmed that the original mutations were retained, and no additional changes were observed. Next, mosquito cells were infected with WT, SL1-A, or SL1-B, using a multiplicity of infection (MOI) of 1, and followed by successive passages of supernatants every 7 days (P1, P2, and P3) (Fig. 5A).
Viral replication was monitored by IFA (Fig. 5C). Mutant SL1-A showed signs of propagation at P2 (14 days), whereas mutant SL1-B propagated at P3 (21 days). To determine whether these mutants are associated with inefficient replication in mosquito cells or if adaptive mutations enhancing viral fitness were selected during passages, the harvested viruses were sequenced. Viral RNA from supernatants collected at P3 was extracted, amplified by RT-PCR, and the capsid-coding region was cloned and sequenced. Sequencing 20 clones revealed that SL1-A reverted to the WT sequence: the input mutation at position 162A reverted to 162C in all clones (Fig. 5D). Interest ingly, sequencing of SL1-B clones revealed a pseudo-reversion in 15 out of 20 clones (75%), where the mutation 176A was replaced by 176C, creating a new CG base pair and restoring the stem. The remaining 25% of clones reverted to the WT sequence (176U, forming a UG base pair) (Fig. 5D). To confirm that the replicative phenotype observed was explained by the detected substitution and not by additional spontaneous mutations outside the capsid-coding region in the pseudo-revertant virus, the mutation 176C was introduced into the ZIKV-Luc construct and its ability to replicate in mosquito cells was evaluated. RNA from this mutant, along with WT and NS5 Mut controls, was transfected. The reporter ZIKV-RNA carrying the pseudo-reversion resulted in efficient viral RNA amplification in mosquito cells (Fig. 5E). The spontaneous emergence of a virus restoring the SL1 stem with an alternative sequence highlights a selective pressure to preserve this RNA structure, supporting its role in ZIKV replication in mosquito cells. To investigate the mechanism by which SL1 contributes to ZIKV replication, we compared its role with that previously reported for DENV (20,35). In DENV2, most of the C1 sequence hybridizes with the DB1 sequence present at the 3'UTR (Fig. 6A), and this interaction was shown to promote genome cyclization and RNA synthesis. However, ZIKV 3'UTR lacks a DB1 structure (Fig. 6A). Analysis of the predicted circular and linear forms of the ZIKV genome suggests a potential interaction between SL1 sequence and nucleotides located upstream of CS1 at the 3' end of the genome (Fig. 6A). This longrange interaction requires unfolding of SL1, extending the known 5'-3' CS1 interaction by three base pairs.
To assess the relevance of this potential long-range interaction, we designed a ZIKV mutant (SL1-F) with substitutions in SL1 disrupting the predicted complementarity with the 3' end, without affecting SL1 structure. Complementarity was restored in a sepa rate mutant (SL1-F 3'rec) by introducing compensatory mutations at the 3' end of the genome. An additional mutant was included (3'CSMut), carrying a substitution in the 3'CS1 sequence, without affecting C1 (Fig. 6B). The mutated viral RNAs (SL1-F, SL1-F3'rec, and 3'CSMut) were transfected into mosquito and mammalian cells, alongside the WT and the NS5-Mut as controls.
The mutation disrupting CS1 (3'CSMut) impaired replication in both hosts, supporting the essential role of this element in genome cyclization (Fig. 6C). The SL1-F mutation, which disrupts the formation of the extended CS1 interaction, impaired replication in mosquito cells but caused only a delay in replication in BHK cells. Notably, the mutant RNA, in which complementarity was restored by substituting 3' end nucleotides (SL1-F3'rec), achieved approximately 100-fold higher RNA amplification level compared with SL1-F in mosquito cells. In mammalian cells, this mutant exhibited a slight improvement in replication (Fig. 6C). These findings support a role of the extended 5'-3' CS for viral replication in a host-dependent manner.
## DISCUSSION
Here, we search for functional RNA structures present in the coding sequence of the ZIKV genome. Our data show that the first 100 nucleotides of the capsid-coding sequence contain cis-acting RNA elements for viral replication. This information is relevant for constructing reporter systems and for designing attenuated viruses. Specifically, we identified an RNA stem-loop structure, SL1, present in the coding sequence that is important for viral RNA replication in a hostspecific manner. The integrity of SL1 was found to be crucial for ZIKV infectivity, and disrupting the stem structure was sufficient to abolish viral replication in mosquito cells. The SL1 structure is formed in the linear conformation of the genome, but its sequence also hybridizes with a sequence upstream of the conserved 3'CS1 at the 3'UTR, generating a long-range interaction. It is important to mention that disruption of the stem of SL1 without altering complementarity with the 3' end of the genome impairs ZIKV replication, supporting a role of the SL1 structure in the linear form of the genome. Our data suggest that both RNA structures, SL1 and 5'-3' extended CS, are necessary for viral RNA replication, supporting dynamic conformations of the ZIKV genome during infection. The SL1 structure is part of a larger conserved RNA element originally described in DENV as DCS-PK or C1 (20,35). This element contains several structures: SL1, SL2, and a pseudoknot interaction (PK). For clarity, the previously reported structures using DENV4 and DENV2 are included in the supplemental material. A structure homologous to SL2 of DENV and ZIKV was recently described to be relevant for YFV replication (40). In the context of DENV4, the conserved DCS-PK structure was systematically studied, and it was reported that mutations altering the PK and the stems of SL1 and SL2 reduce viral replication in mammalian cells, whereas this was not assessed in mosquito cells. We have previously reported that the sequence of C1 was necessary for optimal DENV2 replication, and that nucleotides of this structure hybridize with the sequence of the conserved DB1 structure at the viral 3'UTR (13,35). Here, we present data that supports substantial differences in the role of C1 in ZIKV and DENV. First, the SL1 structure was found to be essential for ZIKV infection in mosquito cells. The crucial role of SL1 was supported by using reporter viruses and infectious clones with mutations that disrupted the stem of SL1 or restored the structure by artificial mutations or by naturally emerging reversions (Fig. 3 to 5). Second, using ZIKV with mutations disrupting SL2 or the PK interaction, we showed that these elements were dispensable for ZIKV RNA replication in both hosts (Fig. 2). Even the deletion of the complete SL2 showed a minor effect on viral RNA replication, in contrast to that shown for DENV (20,35). Third, a sequence within SL1 was found to be involved in a long-range interaction with a sequence at the 3' end of the genome located upstream of 3'CS, enhancing viral RNA replication, which also differs from that previously reported for DENV (Fig. 6).
It is noteworthy that the closely related DENV and ZIKV share a conserved RNA structure within the capsid-coding sequence that modulates genome cyclization, but through the interaction with different regions of the 3' UTR. We speculate that the topology of the RNA around the 5' end of the genome is critical because it is near the RNA promoter for polymerase binding (3), which is conserved in all flaviviruses. The data suggest that the location of complementary sequences at the 3' end is less critical. From the mosquito-borne flavivirus groups (MBFV), the members of the Spondweni (including ZIKV) and the YFV groups contain a single DB structure at the 3'UTR, whereas a conserved duplication of the DB structure is found in most of the other MBFV (37). It is important to mention that the two DB structures in the DENV genome are not redun dant; instead, they play different functions during viral replication (13). It is possible that ZIKV lost one copy of the DB structure, forcing evolution of alternative sites for cyclization at the 3' end of the genome. In agreement with the idea of diversification of regulatory mechanisms for genome cyclization among MBFV, a recent report highlights differential requirements for YFV. In this case, a regulatory RNA element at the coding sequence of the capsid, homologous to the SL2 of DENV and ZIKV, assists a balanced equilibrium between linear and circular forms of the YFV genome (40), a process that has been previously shown to be critical for DENV infectivity (11).
Our studies, together with findings from other groups, have demonstrated that disrupting cyclization elements results in a more severe impact on flavivirus replication in mosquito cells than in human cells. For example, in the case of DENV, disruption of complementarity between UAR and DAR elements demonstrated that single mismatches impaired replication in mosquito cells without significantly affecting replication in mammalian cells (38). Similarly, disruption of the C1-DB1 interaction has been reported to have a differential impact on viral replication, with a greater effect observed in mosquito than in mammalian cells (35). In the case of ZIKV, disruption of the comple mentary DAR sequence was also shown to have minimal impact on viral replication in mammalian cells but caused a dramatic impairment in viral propagation in mosquito cells (29). Consistent with these findings, we observed here that alterations in the 5'-3' extended CS complementary region had a critical effect on ZIKV replication in mosquito cells, whereas the impact on replication in mammalian cells was relatively minor. These observations suggest that the equilibrium between the linear and circular conformations of the viral genome displays different requirements in the two hosts, with a stronger constraint in insect cells. This may be influenced by factors such as the distinct tempera tures at which the virus replicates in the two hosts or differences in host proteins that interact with the viral RNA.
The conserved location of functional RNA elements in flavivirus capsid-coding sequences is intriguing. A selective pressure to position critical RNA elements near the 5' end of the genome may have shaped the capsid-coding sequence to be more adaptable, allowing it to accommodate these essential regulatory structures without compromising protein function. In this regard, the capsid protein is the least conserved of the flavivirus proteins, likely reflecting an evolutionary tolerance for variability in this region. Alignments of capsid-coding sequences from different flaviviruses show less than 40% sequence identity (41). For DENV, a large flexibility in the amino acid sequence was found in the N-terminal of the capsid, provided that the nucleotide sequence of the cis-acting elements was conserved (42). A remarkable flexibility in the capsid protein of TBEV was also found, which tolerated large deletions ranging from 19 to 30 residues (43). The related avian flavivirus, tembusu virus, is another example, where large deletions of capsid spanning almost the entire C-terminal helix were tolerated, and infectious particles were produced (44). This coding sequence flexibility could be associated with the need to accommodate cis-RNA sequence/structure requirements.
In summary, we describe a functional RNA element in the capsid-coding sequence of the ZIKV genome with distinct requirements in the two hosts. A detailed understanding of relevant RNA signals in the viral genome of important human pathogens will help design genetic tools for rational antiviral strategies.
## MATERIALS AND METHODS
## Cell culture
Baby hamster kidney cells (BHK-21) were cultured in Minimum Essential Medium α (MEMα) (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS) (Gibco, Thermo Fisher Scientific) and 100 U/mL penicillin-streptomycin in 100 mm cell culture plates and incubated at 37°C with 5% CO2. Huh-7 cells (human hepatocyte cell line) were cultured in Dulbecco's modified Eagle's high-glucose medium (4,500 mg/L) supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. Mosquito C6/36 HT cells were cultured in Leibovitz's L-15 Medium (Gibco, Thermo Fisher Scientific) supplemented with 10% FBS, 100 U/mL penicillin-streptomycin, 0.3% tryptose phosphate broth, 0.02% glutamine, 1% minimal essential medium (MEM) nonessential amino acids solution, and 0.25 µg/mL amphotericin B (amphotericin B) in T75 cell culture flasks, and incubated at 33°C.
## Construction of recombinant Zika viruses
For constructing the recombinant full-length ZIKV luciferase reporter containing mutations in the capsid-coding region, we modified a ZIKV reporter infectious clone that we have previously described (36). For every mutant, we replaced the MluI-XhoI fragment of the clone with a PCR product obtained by overlapping PCR, generated using primers containing the desired mutations. For the non-reporter mutants, we employed the same overlapping PCR methodology and replaced the MluI-AvrII fragment of the ZIKV full-length infectious clone previously reported (36). The ligation products were transformed into XL1-Blue bacteria, and several clones for each mutant were obtained. The resulting plasmids were sequenced, and the positive clones were used for in vitro RNA transcription. The sequence of the oligonucleotides used is included in Table S1.
## RNA transcription and transfection
WT or recombinant ZIKV infectious clone plasmids were linearized by digestion with KpnI and used as templates for in vitro transcription by Ambion T7 RNA Polymerase (Ambion, Thermo Fisher Scientific) in the presence of m7GpppA cap structure analog (New England Biolabs), and incubated for 2 h at 37°C. RNA integrity was analyzed by agarose gel electrophoresis, and the concentration was measured using Qubit RNA HS Assay Kits (Thermo Fisher Scientific).
RNA transfections were performed using Lipofectamine 2000 (Thermo Fisher Scientific) and Opti-MEM medium (Gibco) according to the manufacturer's instructions. For 24-well cell culture plates, 100 ng of viral RNA per well was used. Renilla luciferase assays were performed using the Renilla luciferase assay system kit (Promega).
## Immunofluorescence assay
BHK-21 or C6/36 HT cells were grown in 24-well cell culture plates containing 12 mm glass coverslips. At the corresponding times post-transfection or infection, the coverslips were collected, and the cells were fixed with methanol for 15 min at -20°C. The coverslips were blocked using gelatin 0.2% (Sigma) in PBS and incubated with rabbit anti-NS3 polyclonal antibody diluted 1/500 in blocking solution. Goat anti-rabbit antibody Alexa Fluor 488 conjugate (Thermo Fisher Scientific) was employed to detect the primary antibody, and DAPI (Thermo Fisher Scientific) was added to visualize the nuclei. The counterstained nuclei of the immunofluorescence shown in Fig. 4 and 5 are included in Fig. S3 and S4, respectively. Images were obtained using an Axio Observer 3 (Zeiss) inverted fluorescence microscope, with 10× and 20× objectives.
## Viral infections and plaque assays
C6/36 HT cells were seeded in 24-well cell culture plates and grown overnight. Transfec ted BHK-21 cells supernatants were used to infect C6/36 HT monolayers for 1 h at 33°C. Then, the inoculum was removed, and 500 µL of Leibovitz's L-15 Medium supplemented with 5% FBS was added to each well. For serial passages, 200 µL of infection supernatants were employed to infect new C6/36 monolayers.
For viral plaque assays, BHK-21 cells were seeded in 24-well cell culture plates and grown overnight. Transfected BHK-21 cells supernatants were serially diluted, and 200 µL of the inoculum was used to infect BHK-21 monolayers for 1 h at 37°C. Afterward, 1 mL of overlay medium (MEMα, 0.8% methyl cellulose, supplemented with 5% FBS) was added to each well. Cells were fixed 7 days post-infection with 10% formaldehyde and stained with crystal violet.
## RNA secondary structure prediction
RNA secondary structures of WT and mutants were predicted using two independent models: IPKnot (45) and AlphaFold 3 (46). IPKnot is a model that accounts for secondary structures, including pseudoknots. In the case of Alphafold 3 predictions, secondary structures were extracted from the three-dimensional models using RNApdbee 3.0 (47). Visualization of three-dimensional structures was performed using ChimeraX v1.10.1 (48), and FORNA (49) was used for visualizing secondary structures.
## RNA extraction and quantification
For qRT-PCR of viral RNA, TRIzol reagent (Invitrogen) was used for RNA extraction of supernatants. cDNA was synthesized with random decaoligonucleotide primer mix by M-MLV Reverse Transcriptase (Promega) as previously described (36). qPCRs were performed in duplicates in 96-well plates using 2 µL of the RT reaction mixture as template, 5 µL of FastStart SYBR green Master 2 x mix (Roche), a 600 nM concentration of each primer, and RNase-free water up to 10 µL. The primers AVG2098 (5′-GCCGCCACCA AGATGAACTGATTG-3′) and AVG2099 (5′-GCAGTCTCCCGGATGCTCCATC-3′) were targeted to amplify nucleotides 9,854 to 9,927 within the NS5 coding sequence.
## References
1. Gebhard, Filomatori, Gamarnik (2011) "Functional RNA elements in the dengue virus genome" *Viruses*
2. Hyde, Gardner, Kimura et al. (2014) "A viral RNA structural element alters host recognition of nonself RNA" *Science*
3. Filomatori, Carballeda, Villordo et al. (2017) "Dengue virus genomic variation associated with mosquito adaptation defines the pattern of viral non-coding RNAs and fitness in human cells" *PLoS Pathog*
4. Roby, Pijlman, Wilusz et al. (2014) "Noncoding subgenomic flavivirus RNA: multiple functions in West Nile virus pathogenesis and modulation of host responses" *Viruses*
5. Morandi, Van Hemert, Incarnato (1722) "SHAPE-guided RNA structure homology search and motif discovery" *Nat Commun*
6. Huber, Lim, Ng et al. (2019) "Structure mapping of dengue and Zika viruses reveals functional long-range interactions" *Nat Commun*
7. Dethoff, Boerneke, Gokhale et al. (2018) "Pervasive tertiary structure in the dengue virus RNA genome" *Proc Natl Acad Sci*
8. Ziv, Gabryelska, Lun et al. (2018) "COMRADES determines in vivo RNA structures and interactions" *Nat Methods*
9. Nicholson, White (2015) "Exploring the architecture of viral RNA genomes" *Curr Opin Virol*
10. Schroeder (2018) "Challenges and approaches to predicting RNA with multiple functional structures" *RNA*
11. Villordo, Alvarez, Gamarnik (2010) "A balance between circular and linear forms of the dengue virus genome is crucial for viral replication" *RNA*
12. Filomatori, Lodeiro, Alvarez et al. (2006) "A 5′ RNA element promotes dengue virus RNA synthesis on a circular genome" *Genes Dev*
13. De Borba, Villordo, Marsico et al. (2019) "RNA structure duplication in the dengue virus 3′ UTR: redundancy or host specificity? mBio"
14. Li, Wei, Mei et al. (2018) "Integrative analysis of Zika virus genome RNA structure reveals critical determinants of viral infectivity" *Cell Host Microbe*
15. Paho (2020) "Integrated management strategy for arboviral disease prevention and control in the Americas"
16. Shragai, Tesla, Murdock et al. (2017) "Zika and chikungu nya: mosquito-borne viruses in a changing world" *Ann N Y Acad Sci*
17. Pierson, Diamond (2020) "The continued threat of emerging flaviviruses" *Nat Microbiol*
18. Akiyama, Graham, O′donoghue et al. (2021) "Three-dimensional structure of a flavivirus dumbbell RNA reveals molecular details of an RNA regulator of replication" *Nucleic Acids Res*
19. Friebe, Harris (2010) "Interplay of RNA elements in the dengue virus 5′ and 3′ ends required for viral RNA replication" *J Virol*
20. Liu, Li, Jiang et al. (2013) "Novel cis-acting element within the capsid-coding region enhances flavivirus viral-RNA replication by regulating genome cyclization" *J Virol*
21. Villordo, Gamarnik (2009) "Genome cyclization as strategy for flavivirus RNA replication" *Virus Res*
22. Clyde, Barrera, Harris (2008) "The capsid-coding region hairpin element (cHP) is a critical determinant of dengue virus and West Nile virus RNA synthesis" *Virology (Auckl)*
23. Göertz, Abbo, Fros et al. (2018) "Functional RNA during Zika virus infection" *Virus Res*
24. Hahn, Hahn, Rice et al. (1987) "Conserved elements in the 3' untranslated region of flavivirus RNAs and potential cyclization sequences" *J Mol Biol*
25. Khromykh, Meka, Guyatt et al. (2001) "Essential role of cyclization sequences in flavivirus RNA replication" *J Virol*
26. Alvarez, Lodeiro, Ludueña et al. (2005) "Long-range RNA-RNA interactions circularize the dengue virus genome" *J Virol*
27. Zhang, Dong, Stein et al. (2008) "Co-selection of West Nile virus nucleotides that confer resistance to an antisense oligomer while maintaining long-distance RNA/RNA base pairings" *Virology (Auckl)*
28. Friebe, Shi, Harris (2011) "The 5′ and 3′ downstream AUG region elements are required for mosquito-borne flavivirus RNA replication" *J Virol*
29. Li, Deng, Yuan et al. (2020) "Different degrees of 5'-to-3' DAR interactions modulate Zika virus genome cyclization and hostspecific replication" *J Virol*
30. Pijlman, Funk, Kondratieva et al. (2008) "A highly structured, nuclease-resistant, noncoding RNA produced by flaviviruses is required for pathogenicity" *Cell Host Microbe*
31. Akiyama, Laurence, Massey et al. (2016) "Zika virus produces noncoding RNAs using a multi-pseudoknot structure that confounds a cellular exonuclease" *Science*
32. Slonchak, Wang, Aguado et al. (2022) "Zika virus noncoding RNA cooperates with the viral protein NS5 to inhibit STAT1 phosphorylation and facilitate viral pathogenesis" *Sci Adv*
33. Pallarés, López Ledesma, Oviedo-Rouco et al. (2024) "Zika virus non-coding RNAs antagonize antiviral responses by PKR-mediated translational Full-Length Text Journal of Virology November"
34. *Nucleic Acids Res*
35. Clyde, Harris (2006) "RNA secondary structure in the coding region of dengue virus type 2 directs translation start codon selection and is required for viral replication" *J Virol*
36. Jvi (2006)
37. De Borba, Villordo, Iglesias et al. (2015) "Overlapping local and long-range RNA-RNA interactions modulate dengue virus genome cyclization and replication" *J Virol*
38. Pallarés, Navarro, Villordo et al. (2020) "Zika virus subgenomic flavivirus RNA generation requires cooperativity between duplicated RNA structures that are essential for productive infection in human cells" *J Virol*
39. Villordo, Carballeda, Filomatori et al. (2016) "RNA structure duplications and flavivirus host adaptation" *Trends Microbiol*
40. Villordo, Gamarnik (2013) "Differential RNA sequence requirement for dengue virus replication in mosquito and mammalian cells" *J Virol*
41. Villordo, Filomatori, Sánchez-Vargas et al. (2015) "Dengue virus RNA structure specialization facilitates host adaptation" *PLoS Pathog*
42. Li, Lu, Ding et al. (2023) "Specialized cis-acting RNA elements balance genome cyclization to ensure efficient replication of yellow fever virus" *J Virol*
43. Byk, Gamarnik (2016) "Properties and functions of the dengue virus capsid protein" *Annu Rev Virol*
44. Samsa, Mondotte, Caramelo et al. (2012) "Uncoupling cis-acting RNA elements from coding sequences revealed a requirement of the N-terminal region of dengue virus capsid protein in virus particle formation" *J Virol*
45. Kofler, Heinz, Mandl (2002) "Capsid protein C of tick-borne encephalitis virus tolerates large internal deletions and is a favorable target for attenuation of virulence" *J Virol*
46. He, Wang, Guo et al. (2021) "Replication/assembly defective avian flavivirus with internal deletions in the capsid can be used as an approach for living attenuated vaccine" *Front Immunol*
47. Sato, Kato, Hamada et al. (2011) "IPknot: fast and accurate prediction of RNA secondary structures with pseudoknots using integer programming" *Bioinformatics*
48. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
49. Antczak, Zok, Popenda et al. (2014) "RNApdbee-a webserver to derive secondary structures from pdb files of knotted and unknotted RNAs" *Nucleic Acids Res*
50. Kerpedjiev, Hammer, Hofacker (2015) "Forna (force-directed RNA): simple and effective online RNA secondary structure diagrams" *Bioinformatics*
51. Pettersen, Goddard, Huang et al. (2021) "UCSF ChimeraX: structure visualization for researchers, educators, and developers" *Protein Sci*
52. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548431&blobtype=pdf | # Developing an eVLP mRNA vaccine for respiratory syncytial virus with enhanced pre-fusion targeting humoral responses
Lei Sun, Mengting Huang, Simin Feng, Wei Zhang, Yun Quan, Ruyi Chen, Yupeng Yang, Haidong Xu, Wansheng Li, Qianyu Pan, Xinwen Chen, Danyang Zhang, Bin Yuan, Jincun Zhao, Zhongfang Wang, Jinzhong Lin, Wei Peng, Martin Ludlow, Qiong Zhang
## Abstract
Respiratory syncytial virus (RSV) is a leading cause of lower respiratory tract disease (LRTD) in infants, the elderly, and immunocompromised individuals. In this study, we leveraged the innovative EABR (ESCRT-and ALIX-binding region) vaccine technology to develop a pre-fusion (Pre-F)-based RSV mRNA vaccine (Pre-F-EABR) that encodes self-assembling enveloped virus-like particles (eVLPs). In vitro expression of Pre-F-EABR mRNA manifested pronounced Pre-F protein expression, particularly on cell membranes and in cell culture supernatant, compared to classical Pre-F and Pre-F-Ferritin nanoparti cle (Pre-F-Fe) mRNAs, suggesting efficient eVLP self-assembly. Immunization of mice with lipid nanoparticle-encapsulated, nucleoside-modified mRNAs (mRNA-LNPs) demonstra ted that the Pre-F-EABR vaccine elicited enhanced neutralizing activities and more robust cellular immunity when compared to Pre-F or Pre-F-Fe mRNA vaccines. Notably, the Pre-F-EABR mRNA vaccine induced a significantly higher level of protective pre-fusionspecific antibodies in contrast to post-fusion-targeting antibodies, dramatically reducing pulmonary viral load following RSV challenge in vaccinated mice. Furthermore, the Pre-F-EABR mRNA vaccine demonstrated significantly improved efficacy against the contemporary clinical ON1 and BA9 isolates. Overall, these findings demonstrate that the Pre-F-EABR mRNA vaccine induces more robust immune responses, highlighting its potential in preventing RSV infection in vulnerable populations. IMPORTANCE Respiratory syncytial virus (RSV) is a major cause of lower respiratory tract disease in infants, the elderly, and immunocompromised individuals. It is wellestablished that enhancing the neutralizing antibody levels and Th1-biased cellular immune responses can potentially improve the efficacy and safety of RSV vaccines. In this study, we developed an RSV pre-fusion protein-based mRNA vaccine that encodes self-assembling enveloped virus-like particles. Mice immunized with this vaccine showed significantly enhanced pre-fusion protein-targeted humoral responses and improved protection against RSV infection compared to the conventional RSV mRNA vaccine. Additionally, this vaccine demonstrated a considerably stronger neutralizing ability against contemporary clinical RSV isolates and induced more robust Th1-biased cellular immune responses, suggesting its potential as a promising RSV vaccine candidate. KEYWORDS mRNA vaccine, respiratory syncytial virus, EABR, eVLPs, immune responses R espiratory syncytial virus (RSV), a single-stranded negative-sense RNA virus, is a worldwide leading cause of severe lower respiratory tract disease (LRTD) among infants, the elderly, and immunocompromised individuals. As one of the primary contributors to childhood (≤5 years of age) morbidity and mortality (1-5), RSV infects nearly every child by the age of two, with recurrent infections occurring throughout
life (1)(2)(3)(4)(5). Current global estimates indicate approximately 33 million annual cases of RSV-associated LRTD in young children, including 3 million hospitalizations (1,3,4).
While vaccination remains the most effective strategy for preventing RSV infection and severe LRTD (6), vaccine development has long been hampered by concerns over potential vaccine-associated enhanced respiratory disease (VAERD) first observed in RSV-naive infants vaccinated with a formalin-inactivated RSV (FI-RSV) in the 1960s (7). Although the precise mechanisms underlying VAERD remain incompletely character ized, current evidence implicates that a T helper (Th) 2-biased CD4+ T-cell response, poor induction of neutralizing antibodies, a lack of CD8+ cytotoxic T-cell responses, and pulmonary eosinophilia following RSV challenge were the main attributes (8)(9)(10). Consequently, optimal RSV vaccine candidates should promote the induction of Th1-polarized cellular immunity coupled with superior neutralizing activities to ensure comprehensive protection (6,11).
The RSV fusion glycoprotein (F) serves as the primary target of neutralizing antibodies in human sera (12,13) and represents the most promising antigen for RSV vaccine development. This metastable protein mediates the fusion of viral particles to the host cell membranes through structural rearrangement from its prefusion (Pre-F) conforma tion to the more stable postfusion (Post-F) state (14)(15)(16). Although clinical trials of RSV Post-F-based vaccines demonstrated only modest protective efficacy (17), Pre-F antigen formulations have consistently shown enhanced immunogenicity and superior protection in clinical evaluations (15,(18)(19)(20), primarily due to their capacity to elicit highly potent neutralizing antibodies and memory B cell responses targeting Pre-F-spe cific antigenic sites (Ø and V) (6,12,15,21). The clinical translation of Pre-F vaccines has been hindered by the protein's conformational instability (6). Structure-based stabiliza tion strategies, particularly through rational design of disulfide bonds and cavity-filling mutations (14,15,22), have overcome this limitation and revolutionized RSV vaccine development (6,19,20,23). DS-Cav1, the most prominent stabilized Pre-F prototype, elicits high-affinity antibodies with strong neutralizing ability in mice, non-human primates, and seropositive adults (14,15,21,24,25). These advances have paved the way for three approved RSV vaccines, including GSK's Arexvy, Moderna's mResvia, and Pfizer's Abrysvo (18,19,26), as well as other vaccine candidates currently under preclinical and clinical development (23,27,28).
The widespread deployment of COVID-19 mRNA vaccines developed by Moderna and Pfizer/BioNTech has demonstrated the safety, efficacy, and reliability of lipid nanoparticle (LNP)-encapsulated, nucleoside-modified mRNA vaccines (29)(30)(31). Unlike conventional vaccines, mRNA-based vaccines leverage host cellular machinery to produce antigens, thereby activating specific and efficient humoral immune responses and T-cell immun ity (32). This attribute is particularly advantageous for RSV vaccine development, as mRNA platforms preferentially induce robust humoral responses and Th1-skewed cellular immunity (20,33). Accordingly, a single dose of the mRNA-based RSV vaccine mResvia conferred an 83.7% efficacy against RSV-associated LRTD in adults aged 60 and older over 3.7 months (20). However, the efficacy of the vaccine against LRTD declined to ~50% at 18 months, underscoring the need to reassess the current RSV mRNA vaccine designs for durable protection in vulnerable populations.
A recent study introduced an innovative mRNA vaccine strategy that integrates the mRNA vaccine platform with nanoparticle technology by incorporating an ESCRT (endosomal sorting complex required for transport)-and ALIX (ALG-2-interacting protein X)-binding region (EABR) into the cytoplasmic tail of the severe acute respiratory syndrome coronavirus 2 spike protein (34). This Spike-EABR mRNA vaccine leverages host cell ESCRT machinery to assemble enveloped virus-like particles (eVLPs) that display densely arrayed spike proteins on their surface, mimicking the natural budding process of enveloped viruses. Compared to conventional COVID-19 mRNA vaccines, the Spike-EABR construct induced enhanced neutralizing activities and more potent T-cell immune responses in preclinical models. However, the broader applicability of this platformparticularly for other high-priority pathogens like RSV-remains unexplored.
Herein, we employed the eVLP vaccine technology to develop a novel RSV Pre-F-EABR mRNA vaccine. Comparative immunization studies in mice demonstrated that lipid nanoparticle-encapsulated, nucleoside-modified Pre-F-EABR mRNA elicited enhanced neutralizing activities and more robust T-cell responses when compared to both the mRNA vaccine encoding DS-Cav1 (Pre-F) and the Pre-F-Ferritin nanoparticle mRNA vaccine. Notably, the Pre-F-EABR mRNA vaccine elicited significantly higher titers of protective Pre-F-specific antibodies in contrast to Post-F-directed antibodies, resulting in a dramatic reduction in pulmonary viral loads following RSV challenge in vaccina ted mice. Moreover, the Pre-F-EABR vaccine exhibited significantly improved breadth of protection against contemporary clinical RSV-A (ON1 genotype) and RSV-B (BA9 genotype) strains. Overall, the Pre-F-EABR mRNA vaccine induces robust immune responses and holds great potential in preventing RSV infection.
## RESULTS
## mRNA vaccine design and the in vitro expression characterization
A recent study has demonstrated the ability of the "EABR technology" to elicit potent humoral and T-cell responses (34), prompting speculation about its applicability to vaccines against other life-threatening pathogens, such as RSV, via the production of self-assembled eVLPs. Therefore, we designed a Pre-F-EABR mRNA vaccine using a strategy similar to the Spike-EABR vaccine: the EABR segment, derived from the human CEP55 protein, was fused to the C terminus of the prefusion stabilized DS-Cav1 construct (Fig. 1A). To prevent coated pit localization and endocytosis, an endocytosis preven tion motif (EPM) derived from the murine Fc gamma receptor FcgRII-B1 cytoplasmic tail was inserted upstream of the EABR sequence (34). Vaccinations with DS-Cav1 or its corresponding mRNA have consistently elicited robust neutralizing antibody and T-cell responses in both animal models and human clinical trials (15,21,24,28). For comparison, we also included the pre-F-ctm vaccine design (a full-length RSV F protein containing the four point mutations present in DS-Cav1) (33). Additionally, an mRNA construct encoding a DS-Cav1 displaying nanoparticle was formulated for comparative analysis by introducing the ferritin sequence from Helicobacter pylori at the C terminus of DS-Cav1 (Pre-F-Fe) (Fig. 1A). Swanson et al. previously validated the self-assembly of the ferritin nanoparticle by in vitro overexpression of the Pre-F-Fe construct (35).
Following synthesis, mRNA integrity was evaluated by TBE RNA gel electrophoresis (Fig. 1B) and capillary gel electrophoresis (CGE) (Fig. 1C andD). CGE analysis revealed single, homogeneous peaks corresponding to the expected sizes for all RSV mRNA constructs, with measured integrity values >90% for all constructs (Fig. 1C). The transla tional efficiency of the synthesized mRNAs was assessed by transfection of mRNAs into BHK-21 cells. Intracellular antigen expression was analyzed at 24 h post-transfection via indirect immunofluorescence staining using the Pre-F-specific antibody D25, followed by a fluorophore-conjugated secondary antibody (Fig. 1E). Robust and comparable protein expression was demonstrated across all constructs. Beyond intracellular expression, cell surface antigen presentation was evaluated by staining with the same antibody (D25). Unlike the intracellular expression pattern, Pre-F-EABR mRNA demonstrated significantly enhanced membrane localization (>8-fold enrichment) compared with other constructs (Fig. 1F andG). Notably, Pre-F-Fe was completely absent from the cell surface. The significant enrichment of Pre-F-EABR on the cell membrane aligns with our vaccine design strategy and enables the assembly and budding of eVLPs that can activate immune cells. The release of eVLPs into culture supernatants of HEK293T and HEK293F cells was evaluated by western blot analysis following the ultracentrifugation of transfec ted cell supernatants on a 20% sucrose cushion to concentrate eVLPs (34). In accordance with the cell surface expression results, Pre-F and Pre-F-Fe transfections did not generate detectable eVLPs in supernatants, whereas eVLPs were readily detected in supernatants from Pre-F-EABR mRNA-transfected cells (Fig. 1H andI). Collectively, the mRNA expres sion results demonstrate that the mRNA-encoded Pre-F-EABR construct enables the pronounced expression of Pre-F protein, the enrichment of antigens on cell surfaces, and the release of eVLPs.
Prior to evaluating the vaccine efficacy in murine models, we formulated Pre-F, Pre-F-ctm, Pre-F-EABR, and Pre-F-Fe mRNAs using the classical SM102-based four-com ponent LNPs (Fig. 2A). The formulated mRNA-LNPs exhibited uniform hydrodynamic diameters ranging from 60 to 70 nm with a polydispersity index (PDI) ≤0.07 (Fig. 2B andC). Cryo-electron microscopy (Cryo-EM) analysis of the mRNA-LNP confirmed the formation of uniformly arrayed spherical particles with regular morphology (Fig. 2E). Additionally, all mRNA-LNPs manifested >90% encapsulation efficiency as quantified by RiboGreen assay and verified through agarose gel electrophoresis of Triton X-100treated versus untreated mRNA-LNPs (Fig. 2B andD). Overall, these comprehensive characterizations indicate proper preparation of mRNA-LNPs with optimal physicochemi cal properties and biological activity.
## Pre-F-EABR mRNA vaccine induced enhanced Pre-F-specific IgG antibody responses in murine models
To characterize the dose-response relationship of the mRNA vaccine, female BALB/c mice were intramuscularly (I.M.) immunized with escalating doses (0.1, 1, and 5 µg) of the pre-F-EABR mRNA vaccine in a prime-boost regimen with a 3-week interval (Fig. 3A). Serum samples collected 3 weeks post-immunization were analyzed for Pre-F protein-specific IgG titers by enzyme-linked immunosorbent assay (ELISA) (Fig. S1A andC). Both prime and boost vaccinations induced dose-dependent increases in Pre-F-spe cific antibody titers. Neutralization assays against multiple RSV subtypes (A2, ON1, and BA9) revealed similarly dose-dependent enhancement of neutralizing antibody activity in both prime and boost sera (Fig. S1B andD). Notably, the 5 µg dose consistently generated the highest Pre-F-specific IgG titers and most potent neutralizing responses across all tested subtypes, supporting its selection for subsequent immunization studies.
To evaluate mRNA vaccine-elicited humoral responses, serum samples collected post-prime and post-boost immunization (5 µg dose) were analyzed by ELISA for Pre-Fand Post-F-specific antibody responses (Fig. 3B through E; Fig. S2A andB). Following the prime immunization, all mRNA-LNP vaccines elicited moderate levels of Pre-F-specific binding antibodies; however, the titers induced by the Pre-F-EABR were significantly higher (2.2-to 5.7-fold increase) than those produced by other formulations (Fig. 3B; Fig. S2A). Following the boost immunization, all mRNA-LNP vaccines elicited robust Pre-F-specific binding antibody responses but low Post-F-specific binding antibody titers, confirming that in vivo-expressed antigens predominantly maintained the prefusion conformation (Fig. 3C andD; Fig. S2B). Remarkably, the Pre-F-EABR construct consistently elicited more favorable antibody responses, achieving 2.0-to 6.1-fold higher Pre-F-spe cific binding antibody titers than other mRNA-LNPs (Fig. 3C; Fig. S2B) while maintaining the lowest Post-F-binding antibody titers (1.9-fold lower than Pre-F and 1.3-fold lower than Pre-F-Fe mRNA-LNPs; Fig. 3D). This superiority extended to the 1 µg dose, where Pre-F-EABR vaccination induced 3.5-to 4.9-fold higher Pre-F-specific antibody titers after prime immunization and 8.7-to 17.8-fold higher titers after boost immunization compared to other constructs (Fig. S2C andD). Additionally, Pre-F-EABR at 1 µg induced 4.8-and 5.8-fold higher Pre-F-specific antibody titers than Pfizer's subunit vaccine (18,36,37) after prime and boost immunization, respectively. These results suggest that the EABR segment may enhance the stability of the Pre-F conformation, thereby enhancing the antibody responses. Furthermore, we assessed Th1/Th2 immune polarization by determining Pre-F-spe cific IgG1 and IgG2a subclasses in post-boost sera (Fig. 3E). In mice, type-1 cytokine interferon γ (IFN-γ) promotes IgG2a secretion while inhibiting IgG1, whereas type-2 cytokine interleukin-4 (IL-4) promotes IgG1 secretion while inhibiting IgG2a (38). As depicted, all three mRNAs elicited high titers of both IgG2a and IgG1 Pre-F-specific binding antibodies compared to the control, resulting in IgG2a/IgG1 ratios approaching 1.0, indicative of a balanced Th1/Th2 response. Nevertheless, the Pre-F-EABR group displayed a modest but statistically significant increase in IgG2a/IgG1 ratio (P < 0.01 by two-tailed unpaired t-test) compared to the Pre-F mRNA-LNP, suggesting that Pre-F-EABR may have greater potential in stimulating Th1-skewed immune response.
## Pre-F-EABR mRNA vaccine induced elevated cross-neutralizing antibody responses against diverse viral strains
Beyond F protein-targeting IgG responses, neutralization profiles against multiple RSV subtypes-including the prototype RSV A2 strain and two contemporary clinical isolates representing current circulating ON1 (RSV-A-0594 strain) and BA9 (RSV-B-9671 strain) genotypes were comprehensively evaluated in mice immunized with 1 and 5 µg doses of mRNA-LNPs (Fig. 4). Primary immunization with either dose elicited modest yet detectable neutralizing antibody titers against all tested RSV strains (Fig. 4A andC). Notably, the Pre-F-EABR vaccine consistently elicited stronger neutralizing activities, exhibiting statistically significant higher titers against all tested RSV strains-RSV A2 (2.2to 2.6-fold), RSV ON1 (1.8-to 2.7-fold), and RSV BA9 (1.4-to 1.7-fold)-compared to other vaccine formulations at both dose levels. Booster immunization with 5 µg mRNA-LNPs significantly boosted the neutralizing antibody responses (Fig. 4B). In line with pri mary immunization results, the Pre-F-EABR mRNA-LNP consistently outperformed other vaccine candidates in stimulating neutralizing antibody responses across all tested strains. Against the RSV A2 strain, geometric mean neutralization titers were 3.8-, 3.0-, and 3.7-fold higher than those elicited by Pre-F, Pre-F-ctm, and Pre-F-Fe vaccines, respectively (Fig. 4B, top panel). This enhanced neutralization capacity extended to contemporary clinical isolates, with Pre-F-EABR showing 3.7-, 3.5-, and 4.2-fold increase against RSV subtype ON1 (39) and 2.9-, 2.7-, and 2.9-fold elevations against RSV subtype BA9 (39) compared to the control vaccines (Pre-F, Pre-F-ctm, and Pre-F-Fe constructs, respectively) (Fig. 4B, middle and bottom panels). While immunization with a 1 µg dose of mRNA vaccines generally induced substantially lower neutralization titers compared to the 5 µg dose, the Pre-F-EABR mRNA vaccine demonstrated more pronounced advantages when compared to Pre-F and Pre-F-ctm mRNA vaccines (Fig. 4D). At this reduced antigen dose, Pre-F-EABR demonstrated 5.8-and 4.9-fold higher neutralization titers against RSV A2 (top panel), 3.7-and 3.6-fold greater activity against RSV ON1 (middle panel), and 3.4-and 2.5-fold increased neutralization capacity against RSV BA9 subtypes (Fig. 4D, bottom panel) compared to the Pre-F and Pre-F-ctm vaccines, respectively. Notably, neutralization assays further demonstrated that the Pre-F-EABR mRNA vaccine elicited neutralization antibody responses that were 1.5-to 2.2-fold stronger than those induced by the Pfizer subunit vaccine across all tested RSV subtypes (Fig. 4C andD), consistent with the observed Pre-F-specific antibody titers (Fig. S2C andD). No significant differences in serological immune responses were observed among the Pre-F, Pre-F-ctm, and Pre-F-Fe constructs.
Collectively, these findings demonstrate that mRNA-LNP-generated Pre-F-EABR eVLPs elicited superior Pre-F-targeted immune responses in mice compared to conven tional RSV pre-F antigen-encoding mRNA vaccines and pre-F-displaying nanoparticle formulations. Notably, the Pre-F-EABR vaccine induced more potent and broadly cross-reactive neutralizing antibodies, achieving both enhanced neutralization potency and significantly improved breadth against diverse RSV strains.
## Pre-F-EABR vaccine induces potent T-cell responses
T-cell immune responses were evaluated in mice immunized with a 5 µg dose of mRNA-LNP vaccines in a prime-boost regimen (Fig. 3A) (40). Splenocytes were isolated 3 weeks after the boost immunization, stimulated with a peptide pool spanning the RSV A2 F protein, and analyzed by intracellular cytokine staining (ICS) (Fig. 5). The flow cytometry gating strategy for ICS is detailed in Fig. S3. Compared to the GFP mRNA-LNP control, all mRNA vaccines elicited robust CD4+ (Fig. 5A) and CD8+ (Fig. 5B) T-cell responses, as evidenced by the production of IFN-γ, TNF-α, and IL-2 cytokines (Fig. 5; Fig. S3B). These findings align with prior studies demonstrating that mRNA-LNP-encoded Pre-F antigens conferred strong CD4+ and CD8+ T-cell responses in mice (23,33,41). Of note, the Pre-F-EABR mRNA vaccine exhibited greater capacity to elicit T-cell immunity compared to control vaccines (Pre-F, Pre-F-ctm, and Pre-F-Fe mRNAs), as evidenced by significantly enhanced cytokine production profiles across both CD4+ and CD8+ T-cell subsets. Among CD4+ T cells, the Pre-F-EABR vaccination resulted in 1.3-to 1.6-fold increase in IFN-γ, ~1.3-fold higher TNF-α, and 1.1-to 1.4-fold enhanced IL-2 expression relative to control vaccines (Fig. 5A). Similarly, CD8+ T cells from Pre-F-EABR immunized mice showed 1.5-to 1.6-fold greater IFN-γ, 1.4-to 1.7-fold higher TNF-α, and 1.3-to 1.5-fold increased IL-2 expression compared to control groups (Fig. 5B). In contrast to Beyond the ICS analysis, we further assessed T-cell responses by ELISpot, quantifying IFN-γ, TNF-α, IL-2, and IL-4 secretion using the same splenocyte preparations (Fig. 6). Mirroring the ICS findings, the Pre-F-EABR mRNA vaccine elicited significantly stronger T-cell responses compared to control formulations. Specifically, the Pre-F-EABR vaccine induced 1.4-to 2.5-fold more IFN-γ-secreting cells, 1.7-to 3.3-fold more TNF-α-producing cells, and 1.7-to 2.9-fold more IL-2-secreting cells than controls, while IL-4 remained undetectable (Fig. 6B). These observations further support the robust Th1-skewed immune profile induced by the Pre-F-EABR vaccine. While ICS analysis revealed no statistically significant differences in T-cell responses between the Pre-F mRNA vaccine (encoding secreted DS-Cav1) and Pre-F-ctm mRNA vaccine (encoding membrane-anch ored DS-Cav1) (Fig. 5), ELISpot assays showed significantly higher percentages of IL-2and TNF-α-secreting T cells (P < 0.05 by two-tailed unpaired t-test; Fig. 6B). This finding aligns with previous ICS data demonstrating enhanced IL-2 and TNF-α production in CD4+ T cells in response to membrane-anchored DS-Cav1 versus the secreted form (33). This discrepancy may reflect the enhanced sensitivity of ELISpot in detecting low-frequency cytokine-secreting cells. Together, these findings demonstrate that the Pre-F-EABR mRNA vaccine effectively enhanced both the magnitude and polyfunctional ity of T-cell responses in vaccinated mice.
## Pre-F-EABR vaccine provides enhanced protection against RSV challenge
To assess the protective efficacy of mRNA vaccines against RSV infection, BALB/c mice were vaccinated with 5 µg of mRNA-LNPs and challenged with RSV A2 3 weeks after the boost immunization (Fig. 7A). Four days post-challenge, lung and nasal tissues were harvested, and viral loads were quantified to assess vaccine-mediated protection efficacy. All tested RSV mRNA-LNP vaccines conferred significant protection against RSV A2 challenge, with markedly reduced viral titers in both lung and nose tissues (Fig. 7B). This protection profile correlated with the robust PreF-binding antibody responses, neutralizing activity, and T-cell immunity elicited by the vaccines (Fig. 3 to 6), and was consistent with previous reports (33,41). Notably, the protective effect conferred by the Pre-F-EABR mRNA vaccine significantly outperformed other vaccine formulations, with viral loads reaching the limit of detection in both lung and nasal tissues (Fig. 7B).
In contrast, mice receiving Pre-F, Pre-F-ctm, or Pre-F-Fe vaccines exhibited significantly higher viral loads in both lung (5.7-, 4.1-, and 5.7-fold increase, respectively; P < 0.05) and nasal tissue (7.2-, 4.3-, and 4.8-fold increase, respectively; P < 0.05) compared to the Pre-F-EABR group (Fig. 7B). This exceptional protection conferred by the Pre-F-EABR vaccine was consistent with its enhanced humoral and cellular immune responses.
## DISCUSSION
Herein, we designed an mRNA-based RSV vaccine, Pre-F-EABR, leveraging the recently reported EABR-based vaccine approach, and evaluated its immunogenicity and protective efficacy in mice. In vitro studies confirmed efficient production of eVLPs in cell culture supernatants following Pre-F-EABR mRNA transfection. Immunization with the Pre-F-EABR mRNA-LNP elicited significantly enhanced immune responses against both RSV A and B subtypes and conferred more efficient protection in mouse challenge studies. In comparison with the mRNA-LNP encoding Pre-F, the Pre-F-EABR construct elicited significantly more prominent RSV neutralizing antibody titers and robust T-cell responses. Notably, the viral loads in the lung and nose were also significantly lowered in mice that received the Pre-F-EABR mRNA-LNP (Fig. 7). Collectively, these results demonstrate the potential of Pre-F-EABR as a promising RSV vaccine candidate worthy of further development. RNA-LNP-mediated vaccines, including those targeting RSV, are well-documented for their capacity to elicit robust cellular immune responses in both animal models and humans (20,23,32,33,40,42,43). In our study, the Pre-F-based mRNA vaccines also induced potent CD4+ and CD8+ T-cell immune responses in mice (Fig. 5 and6), especially for the Pre-F-EABR vaccine, which elicited significantly stronger T-cell immune responses than the other candidates (Fig. 5). Mechanistically, the EABR platform not only facilitates antigen presentation on cell surfaces but also drives the assembly and release of eVLPs, mimicking the budding process of enveloped viruses (34). The secreted eVLPs may subsequently disseminate to lymph nodes distal from the injection site, thereby increasing the overall antigen exposure to the immune system and promoting more robust immune cell activation. This likely contributes to the observed enhancement in neutralizing antibody titers and T-cell responses elicited by the Pre-F-EABR vaccine. While correlates of protection against RSV infection remain incompletely defined, numerous studies have highlighted the critical roles of both neutralizing antibodies and cellular immune responses (27,(44)(45)(46). Consistent with this paradigm, Pre-F-EABR-vaccinated mice exhibited significantly reduced viral loads in the lung and nose after RSV challenge (Fig. 7), correlating with the elevated neutralizing activities and T-cell response profiles. However, further validation is needed to confirm in vivo eVLP assembly, and interac tions between eVLPs and host immune cells also require further investigations. Such investigations will provide a more thorough insight into the mechanisms underlying the enhanced efficacy of the Pre-F-EABR mRNA vaccine.
Apart from older adults, there is an urgent need for vaccines that protect young infants against RSV-linked LTRD (1,26). However, no pediatric RSV vaccines are currently accessible. The legacy of VAERD, which has been linked to Th2-biased immune response, remains a major obstacle in RSV pediatric vaccine development (26). In the BALB/c mouse model, IgG subclass profiles serve as key immunological markers, with IgG1 and IgG2a representing Th2-and Th1-associated responses, respectively, making the IgG2a/IgG1 ratio a hallmark of Th1/Th2 skewing (47). Our evaluation of IgG2a/IgG1 values demonstrated that all constructed mRNA-LNPs in this study induced balanced Th1/Th2 responses, though the Pre-F-EABR vaccine exhibited a comparatively higher IgG2a/IgG1 ratio compared to the Pre-F mRNA vaccine, suggesting a greater Th1 skewing (Fig. 3E). Unlike the IgG subclass results, ICS analysis demonstrated Th1-dom inant immune responses across all RSV mRNA vaccine formulations, as evidenced by elevated expression of Th1-associated cytokines, including IFN-γ, TNF-α, and IL-2, and silenced expression of Th2-associated cytokine IL-4 (Fig. 5; Fig. S3B). This Th1-skewed cytokine profile was further validated by ELISpot assays, which not only confirmed the ICS findings but also revealed quantitatively enhanced, yet proportionally consis tent, Th1-type response patterns (Fig. 6). Remarkably, the Pre-F-EABR vaccine induced significantly higher expression of IFN-γ, TNF-α, and IL-2 in both CD4+ and CD8+ T-cell populations compared to other vaccine formulations, indicating a more potent Th1biased cellular immunity. This enhanced Th1 bias was further substantiated by ELISpot analysis, which showed significantly greater frequencies of IFN-γ+, TNF-α+, and IL-2+ T cells in Pre-F-EABR-vaccinated groups compared to controls (Fig. 6). The capacity of the Pre-F-EABR vaccine to elicit robust neutralizing antibody responses coupled with Th1-polarized cellular immunity suggests its potential as a candidate for pediatric RSV vaccine development. However, further preclinical studies are warranted to comprehen sively assess the immunogenicity, protective efficacy, and safety profile of the Pre-F-EABR mRNA vaccine in proper animal models. This need is underscored by recent clinical trials of Moderna's mRNA-1345 (RSV vaccine) and mRNA-1365 (RSV/human metapneumovirus combination vaccine), which reported increased incidence of severe respiratory disease among vaccinated young children ages 5 to 7 months (43). This finding underscores the complexities in pediatric RSV vaccine development and highlights the critical need for rational vaccine design coupled with rigorous preclinical assessment (43). Another key challenge in RSV mRNA vaccine development lies in achieving durable protection, as Moderna's licensed mResvia vaccine demonstrated a significant decline in efficacy against RSV-associated LRTD in adults ages 60+ by 18 months post-vaccination (48). To address this limitation, future studies will focus on evaluating the longevity of immune protection conferred by the Pre-F-EABR mRNA vaccine and elucidating the mechanisms underlying its sustained efficacy.
Histopathological evaluation of RSV-challenged vaccinated mice in this study revealed a notable dissociation between virological and pathological outcomes. Histopathological examination of H&E-stained lung sections (Fig. S5) revealed similar pulmonary pathology profiles across all vaccinated groups, despite markedly different viral loads observed between the RSV mRNA vaccinated mice and the GFP control (Fig. 7B). These observations align with established literature demonstrating that BALB/c mice, despite their widespread use in RSV vaccine research, exhibit only semi-permissive characteristics for human RSV replication (49,50). Previous studies have shown that even with high viral inoculum doses, this animal model typically exhibited minimal clinical disease manifestations and generally mild pulmonary pathology (50). To address this constraint, we have implemented a parallel experimental approach incorporating cotton rats (Sigmodon hispidus), which demonstrate enhanced permissiveness to RSV infection and more robust pathological manifestations (27,33,41,51). These complementary studies are currently in progress, and we plan to report these additional findingsincluding any potential VAERD effects induced by the Pre-F-EABR vaccine-in future publications.
In this study, we have observed a dose-dependent humoral immune response elicited by the pre-F-EABR mRNA vaccine in BALB/c mice, with the 5 µg dosage consistently eliciting the highest serum antibody titers (Fig. S1A). Consequently, the 5 µg regimen was selected for subsequent immunization studies. Our selection is consistent with established mRNA vaccine research, where effective immunization doses in BALB/c mice typically fall within the 0.1-10 µg range (33,34,40,41). However, when adjusted for body weight, these murine doses are substantially higher than the 30-100 µg doses used in human clinical trials for prophylactic mRNA vaccines (20, 30-32, 42, 43). To bridge this translational gap, future preclinical studies should evaluate the pre-F-EABR vaccine at clinically relevant doses in more physiologically appropriate models, such as non-human primates, which better recapitulate human RSV infection and immune responses.
In summary, we have developed a novel RSV mRNA vaccine that efficiently gener ates eVLPs. Our preclinical evaluation demonstrated that the Pre-F-EABR mRNA vaccine elicited significantly enhanced humoral responses and induced more robust T-cell responses in murine models. These compelling immunogenicity findings warrant further evaluation of this vaccine in additional preclinical models and subsequent clinical trials. This work provides critical insights into both the magnitude and quality of immune responses induced by this new and promising vaccine approach, offering valuable guidance for current and future RSV vaccine development programs.
## MATERIALS AND METHODS
## Viruses, cells, and antibodies
BHK-21 and HEK293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM; BasalMedia, Shanghai, China) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Sigma-Aldrich) and 10 U/mL penicillin-streptomycin (Gibco #15140122) at 37°C and 5% CO 2 . Expi293F cells were kindly provided by Prof. Xuepeng Wei from Guangzhou National Laboratory and cultured in Expi293 expression medium (Sino Biological, Shanghai, China) at 37°C and 8% CO 2 with shaking at 135 rpm. All cell lines were confirmed to be mycoplasma-free using MycoBlue Mycoplasma Detector (Vazyme #D101-02). Antibodies used in this study include two anti-RSV F recombinant monoclonal antibodies, motavizumab (PDB: 3IXT) and D25 Fab (PDB: 6S3D), and an anti-GAPDH mouse monoclonal antibody (Proteintech #HRP-60004). Both motavizumab and D25 Fab were custom-ordered from Biointron (Nanjing, China). Escherichia coli DH5α cells (AlpalifeBio, Guangzhou, China) were used for plasmid cloning and were cultured in LB broth (Sango #A507002) supplemented with or without ampicillin (100 µg/mL) or kanamycin (50 µg/mL).
All RSV strains were propagated in HEp-2 cells in DMEM containing 2% FBS (D2). RSV A2 (ATCC VR-1540) was used for mouse challenge experiments. RSV A2, rRSV-A-0594-EGFP (ON1 genotype, rRSV-ON1-GFP) and rRSV-B-9671-EGFP (BA9 genotype, rRSV-BA9-GFP) were used in serum neutralization assays. Both rRSV-ON1-GFP and rRSV-BA9-GFP were generated using previously characterized optimized reverse genetics systems (39). Briefly, the pSMART BAC vector (1.6 µg) containing the full-length cDNA copy of the RSV-ON1-GFP or RSV-BA9-GFP antigenome was co-transfected with helper plasmids (pcDNA3.1(+) vector) encoding the corresponding N (1.6 µg), P (1.2 µg), M2-1 (0.8 µg), and L (0.4 µg) open reading frames into HEp-2 cells with Neon Transfection System (Thermo Fisher Scientific). Prior to the electroporation, HEp-2 cells were infected with MVA-T7 (52) (MOI = 5) for 1 h at 37°C. Electroporated cells were seeded in 6-well plates (1 × 10 6 cells/well) and monitored for fluorescent cell foci or syncytium formation. Rescued rRSVs were harvested in 5-6 days post-transfection by two freeze-thaw cycles. Viral titers were determined by plaque assays. Briefly, virus stocks were serially diluted in D2 and added in duplicate to 90% confluent HEp-2 cell monolayers in 12-well plates. After 2-4 h incubation at 37°C, 3 mL of 0.75% carboxymethyl cellulose solution (prepared in D2) was added per well. Following 5 days of incubation at 37°C, cells were fixed with 4% paraformaldehyde (PFA) and stained with the Crystal Violet Staining Solution (Beyotime #C0121). Viral plaques were counted, and titers (PFU/mL) were determined.
## mRNA synthesis
RSV Pre-F-expressing mRNAs were designed based on the Ds-Cav1-stabilized F protein derived from the A2 strain (14). The T4-fibritin trimerization domain (foldon) and amino acid substitutions were incorporated as shown in Fig. 1A. Specifically, mRNA encod ing Pre-F-EABR was designed by fusing the Ds-Cav1 sequence to the transmembrane domain (residues 514-551 of RSV F from A2 strain), the EPM motif (residues 243-290 of mouse FcgRII-B1), and the EABR domain (residues 160-217 of human CEP55 protein). The EPM and EABR motifs were separated by a 4-residue GS linker. The Pre-F-Fe encoding sequence was generated by fusing the Ds-Cav1 sequence (without foldon) to H. pylori ferritin DNA, separated by a 3-residue GS linker inserted between them. All sequences were codon-optimized and integrated into our mRNA production plasmid containing the T7 RNA polymerase promoter, 5′ UTR, open reading frame, and 3′ UTR. Linearized DNA templates were PCR-amplified using Phanta Flash Master Mix (Vazyme #P510-03), gel-purified, and extracted with the GeneJET Gel Extraction Kit (Thermo Fisher Scientific #K0692). mRNAs were synthesized by T7 RNA polymerase-mediated in vitro transcrip tion reaction using an Enzyme Mix (Hzymes Biotech #HBP000331-2), the ribonucleo side triphosphates mix (100 mM each: ATP, GTP, CTP, and N1-methyl-pseudouridine triphosphate; Henovcom #HN3004), and the Cap1 analog LZCap (Henovcom #HN3004). Post-IVT, DNA templates were digested by DNase I (NEB #M0303L), and mRNAs were purified by LiCl precipitation. After dissolution in DNase/RNase-free water (Beyotime #ST876), concentrations were quantified spectrophotometrically. mRNA integrity was verified by agarose gel electrophoresis and capillary electrophoresis (Agilent 5200 Fragment Analyzer, DNF-471 kit). Data were analyzed using ProSize software (Agilent). All mRNAs were stored at -80°C.
## LNP encapsulation of mRNAs
The mRNAs were encapsulated in LNPs via microfluidic mixing (RNACure) by combining the mRNA and lipid solutions at a 3:1 vol ratio and a flow rate of 12 mL/min. Specifically, mRNAs were diluted in 50 mM citrate buffer (pH 4.0) to 150 µg/mL, while the lipid mixture-consisting of SM-102 ionizable lipid (AVT, Shanghai, #O02010), distear oylphosphatidylcholine (DSPC; AVT #S01005), cholesterol (AVT #57-88-5), and poly(eth ylene glycol)2000-dimyristoylglycerol (PEG2000-DMG; AVT #O02005) in a 50:38.5:10:1.5 molar ratio-was dissolved in ethanol to 8 mM concentration. The resulting LNPs were dialyzed against PBS to remove ethanol and displace the acidic solution. LNPs were characterized by measuring the hydrodynamic size and polydispersity index by dynamic light scattering (Malvern Nano-ZS zetasizer), while RNA concentration and encapsula tion efficiency were determined using the Quant-iT RiboGreen assay (Thermo Fisher Scientific) following the manufacturer's protocol.
## Cryo-EM characterization of mRNA-LNP
For cryo-EM microscopy, Quantifoil R1.2/1.3 300-mesh holy carbon gold grids with 2 nm ultrathin carbon support films were glow-discharged prior to sample application. A 3 µL aliquot of mRNA-LNP solution was transferred onto each grid and incubated for 1 min at room temperature. Excess solution was blotted for 3 s in a Vitrobot Mark IV (Thermo Fisher Scientific) chamber maintained at 4°C and 100% humidity, followed by rapid plunging into liquid ethane. Cryo-preserved grids were loaded into a 300 kV Titan Krios G4 microscope (Thermo Fisher Scientific) equipped with a Selectris-X energy filter (10 eV slit width) and Falcon 4i direct electron detector. Data collection was performed using EPU2 software in counting mode at 1.94 Å/pixel resolution, with 3-4 μm defocus and a total electron dose of 20 e/Å 2 .
## mRNA transfection and protein expression
BHK-21 cells were seeded in 12-well plates and incubated at 37°C overnight to reach 70%-90% confluency at transfection. mRNAs were complexed with Lipo8000 Transfection Reagent (Beyotime #C0533) in Opti-MEM I Reduced Serum Medium (Gibco #31985070) according to the manufacturer's protocol. For western blot analysis, samples were mixed with LDS buffer (Thermo Fisher #NP0008) and resolved on 4%-20% FuturePAGE precast gels (ACE Biotechnology, Changzhou, China). Proteins were transferred to PVDF membrane (Millipore #IPVH00010) and probed overnight at 4°C with either anti-RSV F (motavizumab) or anti-GAPDH antibodies. After washing, membranes were incubated for 1 h at room temperature with HRP-conjugated goat anti-human IgG (Beyotime #A0201) or anti-mouse IgG (Sangon #D110087), followed by detection using eECL Western Blot Kit (CoWin Biotech, Jiangsu, China, #CW0049S) on an iBright CL750 Imaging System (Thermo Fisher Scientific). For immunofluorescence detection of RSV Pre-F protein expression, transfected cells were fixed with 4% PFA at 24 h post-transfec tion and washed with PBS. Fixed cells were permeabilized with 0.2% Triton X-100, then blocked with 5% BSA for 30 min at room temperature. Following blocking, cells were incubated with D25 Fab (1 µg/mL in PBS) for 1 h at 37°C, then stained with Alexa Fluor 488-conjugated goat anti-human IgG (1:2,000 dilution in PBS, Thermo Fisher #A48276) for 1 h at 37°C. For cell surface antigen detection, the protocol was modified by omitting the Triton X-100 permeabilization step while maintaining all other conditions. Fluores cence imaging was performed using an Operetta CLS high-content analysis system (PerkinElmer) following final PBS washes.
## Production of EABR eVLPs
The production and purification of eVLPs followed a previously described protocol (34) with minor modifications. HEK293T cells were seeded in 10 cm dishes (Corning #430167) and incubated at 37°C overnight to reach 70%-90% confluency at transfection. Each dish was transfected with 10 µg mRNA with Lipo8000 Transfection Reagent in Opti-MEM I Reduced Serum Medium. Following 72 h incubation at 37°C, cell supernatants were collected, centrifuged, filtered through 0.45 µm membranes, and concentrated using Amicon Ultra-15 centrifugal filters (100 kDa MWCO; Millipore). The concentrated samples were then ultracentrifuged at 135,000 × g for 2 h at 4°C over a 20% wt/vol sucrose cushion using a Sorvall WX100+ ultracentrifuge (Thermo Fisher Scientific). Supernatants were removed, and pellets were re-suspended in 200 µL sterile PBS at 4°C overnight. Residual cell debris was removed by centrifugation at 10,000 × g for 10 min, and the final supernatants were collected for analysis. eVLP production was confirmed by western blot. The production of eVLPs was also tested in Expi293F cells. The day before transfec tion, Expi293F cells were seeded at a density of 2 × 10 6 cells/mL in Expi293 expression medium and adjusted to 2.5 × 10 6 cells/mL at transfection. Cells were transfected with 2 µg mRNA per milliliter culture using Lipo8000 Transfection Reagent. Seventy-two hours post-transfection, supernatants were collected, and eVLPs were purified using the same protocol as for HEK293T cells.
## Mouse immunization
All mouse immunization procedures were conducted in accordance with the approved protocol (GZLAB-AUCP-2022-10-A02) from the Animal Care and Use Committee of Guangzhou National Laboratory. Female BALB/c mice (5-7 weeks old; GemPharmatech, Guangzhou, China) were divided into groups (n = 5, 8, or 16) and received two intramus cular immunizations (0.1, 1, or 5 µg mRNA-LNP in 100 µL) in the hind leg at 3-week intervals. The Pfizer subunit vaccine (RSVpreF; ABRYSVO) (18,36,37) was provided as a gift by Professor Jinzhong Lin's group at Fudan University, who purchased it from the pharmacy, and a dose of 1 µg was administered to each mouse. Blood samples were collected via retro-orbital bleeding under 2%-3% isoflurane anesthesia 21 days post-each immunization for antibody titer determination. For ICS assays and ELISpot assays, mice were euthanized by cervical dislocation under 2%-3% isoflurane anesthesia 21 days post-boost, followed by aseptic spleen collection.
## Enzyme-linked immunosorbent assay
ELISAs were performed to quantify serum antibody titers against RSV Pre-F or post-F proteins. Ninety-six-well ELISA plates (Corning #9018) were coated overnight at 4°C with 1 µg/mL recombinant RSV Pre-F (Sino Biological #11049-VNAS) or Post-F (Sino Biologi cal #11049-V08H5) in carbonate/bicarbonate buffer. After washing with PBS containing 0.05% Tween-20 (PBS-T), plates were blocked for 1 h at 37°C with 5% non-fat milk in PBS. Mouse sera were initially diluted 100-fold, followed by threefold serial dilutions. Diluted samples were incubated on coated plates for 1 h at 37°C, washed three times with PBS-T, and then incubated with HRP-conjugated goat anti-mouse IgG (1:2,000; Sangon #D110058) for 1 h at 37°C. Following additional washes, reactions were developed with TMB substrate solution (Beyotime #P0209) for 15 min at 37°C and stopped with stop solution (Beyotime #P0215). Absorbance was measured at 450 nm using an Ensight Multimode Plate Reader (PerkinElmer). Endpoint titers were defined as the highest reciprocal dilution yielding an optical density ≥2× background. For IgG subclass analysis, identical procedures were performed using HRP-conjugated goat anti-mouse IgG1 (Abcam #ab97240) or IgG2a (Abcam #ab97245).
## Neutralization assays
Serum neutralization assays were conducted as previously described (21,27,53,54) with modifications. Briefly, sera were heat-inactivated at 56°C for 30 min and serially diluted (initial 1:50 dilution, followed by threefold dilutions in D2 medium) in 96-well plates. Equal volumes of diluted sera were mixed with RSV at optimized titers (A2/ON1: 800 PFU/well; BA9: 600 PFU/well) and incubated at 37°C for 1 h. The serum-virus mixtures were then transferred to HEp-2 cell monolayers (3 × 10 4 cells/well in 96-well plates, seeded 24 h prior) and incubated at 37°C for 30 h before PBS washing and fixation (4% PFA). For RSV A2 neutralization, fixed cells were processed for immunostaining. Briefly, fixed cells were permeabilized with 0.2% Triton X-100, then blocked with 5% BSA for 30 min at room temperature. After blocking, cells were incubated with D25 Fab (1 µg/mL in PBS) for 1 h at 37°C, followed by staining with Alexa Fluor 488-conjuga ted goat anti-human IgG (1:2,000 dilution in PBS) for 1 h at 37°C. Between each step, cells were washed three times with PBS. GFP fluorescence was quantified directly for rRSV-ON1-GFP and rRSV-BA9-GFP strains. All plates were analyzed with an Operetta CLS high-content analysis system, and neutralization titers were calculated using four-param eter non-linear regression analysis (GraphPad Prism 6). A representative picture of the RSV A2 neutralizing assay is shown in Fig. S4.
## Intracellular cytokine staining
Mouse spleens were homogenized through a 40 µm cell strainer using syringe plung ers in 5 mL of complete I10 medium (Iscove's Modified Dulbecco's Medium, Gibco #12440053) supplemented with 10% FBS, 50 µM β-mercaptoethanol, and 1 U/mL penicillin-streptomycin. After centrifugation, erythrocytes were lysed using 5 mL ACK lysis buffer (Beyotime #C3702) for 3 min at room temperature, followed by neutralization with 10 volumes of PBS. The resulting splenocytes were filtered through 40 µm strainers, pelleted, and resuspended in 2 mL I10 medium. The splenocytes were then enumerated, and cell concentrations were adjusted to ~1 × 10 7 /mL for subsequent assays.
For intracellular cytokine staining, 100 μL of splenocytes were stimulated in 96-well U-bottom plates with an equal volume of RSV A2 F0 peptide pool (15-mers overlapping by 11 amino acids, 1 µg/mL in I10, Sino Biological) or control treatments (DMSO vehicle or 1 µg/mL concanavalin A; InvivoGen #inh-cona) for 1 h at 37°C. Protein transport inhibitors (GolgiStop/GolgiPlug, BD Biosciences #554724/550583; 1:200 dilution in I10) were then added (50 µL/well), followed by 7 h incubation at 37°C. Following two PBS washes, cells were sequentially stained with Fixable Viability Stain 780 (BD #565388) for 15 min at room temperature, then with a surface marker antibody cocktail (all antibodies diluted 1:100 in MACS buffer [PBS with 2% BSA and 1 mM EDTA]) containing: anti-CD3e FITC (BD #553061), anti-CD4 PE-Cy7 (BD #552775), and anti-CD8a BV510 (BD #563068) for 30 min at 4°C in the dark. Following washing with MACS buffer, cells were fixed and permeabilized with Cytofix/Cytoperm solution (BD #554722, 100 µL/ well, 20 min, 4°C).
For intracellular cytokine detection, cells were washed with Perm/Wash buffer (BD #554723) and stained with a panel of anti-cytokine antibodies (all 1:100 diluted in Perm/ Wash buffer), including anti-IFN-γ BV711 (BD #564336), anti-TNF BB700 (BD #566510), anti-IL-2 PE (BD #554428), and anti-IL-4 BV786 (BD #564006), for 30 min at 4°C in the dark. After a final Perm/Wash buffer wash, cells were resuspended in 200 µL MACS buffer and acquired on a Novocyte Advanteon flow cytometer (Agilent Technologies). Data analysis was performed using NovoExpress software, with the gating strategy and representative results shown in Fig. S3.
## ELISpot assay
ELISpot assays were performed to quantify F protein-specific T-cell responses by measuring IFN-γ, TNF-α, IL-2, and IL-4 secretion in splenocytes obtained from the above ICS experiments. Cells were stimulated for 30 h with either RSV A2 F0 peptide pool (15-mers overlapping by 11 amino acids at 1 µg/mL in I10), DMSO vehicle control (0.1%), or concanavalin A (1 µg/mL positive control). Cytokine production was detected using mouse ELISpot Plus kits (MabTech: IFN-γ #3321-4HST-2, IL-2 #3441-4HPW-2, TNF-α #3511-4HPW-2, and IL-4 #3311-4HPW-2) according to the manufacturer's instructions. Spot-forming units (SFU) were quantified using an automated ELISpot reader, normalized to SFU per 10⁶ viable cells. Background levels were calculated as the 95% percentile of the SFU observed in non-stimulated splenocytes.
## RSV challenge
The RSV challenge study was approved by the Animal Care and Use Committee of Guangzhou National Laboratory (GZLAB-AUCP-2023-10-A7) and conducted under ABSL2 conditions. Female BALB/c mice (n = 8 per group, aged 5-7 weeks, GemPharmatech) were immunized twice I.M. with 5 µg mRNA at 3-week intervals. Three weeks postboost immunization, mice were anesthetized under isoflurane (2%-3%) and intranasally challenged with 1 × 10 6 PFU RSV A2 in 50 µL, while control groups received PBS immunization and mock challenged. Four days post-infection, mice were euthanized by cervical dislocation under isoflurane anesthesia (2%-3%). Lungs and nose were then harvested, homogenized, and clarified by centrifugation (3,000 × g, 10 min). Viral loads were assessed via plaque assays on HEp-2 cell monolayers in 12-well plates and expressed as PFU/g.
## Statistical analysis
Statistical analyses were conducted using GraphPad Prism 6.0. Antibody responses (ELISA and neutralization assays, Fig. 3 to 4; Fig. S1 andS2) and respiratory tract viral titers (Fig. 7) were expressed as geometric mean with 95% confidence intervals (95% CI). Cellular immune responses, including cytokine production (ICS, Fig. 5) and antigen-specific T-cell frequencies (ELISpot, Fig. 6), were expressed as mean ± standard deviation (SD). Differences between immunized groups were statistically evaluated using the two-tailed unpaired t-test that passed the Anderson-Darling normality test or by the Mann-Whitney test that included non-parametric populations. Statistical significance was determined by a P value below 0.05, and the significant differences are indicated with asterisks in each graph (*P < 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001).
## References
1. Li, Wang, Blau et al. (2022) "Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in children younger than 5 years in 2019: a systematic analysis" *The Lancet*
2. Kenmoe, Nair (2024) "The disease burden of respiratory syncytial virus in older adults" *Curr Opin Infect Dis*
3. Curns, Rha, Lively et al. (2020) "Respiratory syncytial virus-associated hospitalizations among children <5 years old: 2016 to" *Pediatrics*
4. Zhang, Yuan, Zhang et al. (2015) "Burden of respiratory syncytial virus infections in China: systematic review and meta-analysis" *J Glob Health*
5. Piralla, Chen, Zaraket (2023) "An update on respiratory syncytial virus" *BMC Infect Dis*
6. Ruckwardt, Morabito, Graham (2019) "Immunological lessons from respiratory syncytial virus vaccine development" *Immunity*
7. (2025) *Full-Length Text Journal of Virology*
8. Kim, Canchola, Brandt et al. (1969) "Respiratory syncytial virus disease in infants despite prior administration of antigenic inactivated vaccine" *Am J Epidemiol*
9. Acosta, Caballero, Polack (2015) "Brief history and characteriza tion of enhanced respiratory syncytial virus disease" *Clin Vaccine Immunol*
10. Knudson, Hartwig, Meyerholz et al. (2015) "RSV vaccineenhanced disease is orchestrated by the combined actions of distinct CD4 T cell subsets" *PLoS Pathog*
11. Delgado, Coviello, Monsalvo et al. (2009) "Lack of antibody affinity maturation due to poor Toll-like receptor stimulation leads to enhanced respiratory syncytial virus disease" *Nat Med*
12. Van Der Fits, Bolder, Heemskerk-Van Der Meer et al. (2020) "Adenovector 26 encoded prefusion conformation stabilized RSV-F protein induces long-lasting Th1-biased immunity in neonatal mice" *NPJ Vaccines*
13. Ngwuta, Chen, Modjarrad et al. (2015) "Prefusion F-specific antibodies determine the magnitude of RSV neutralizing activity in human sera" *Sci Transl Med*
14. Magro, Mas, Chappell et al. (2012) "Neutralizing antibodies against the preactive form of respiratory syncytial virus fusion protein offer unique possibili ties for clinical intervention" *Proc Natl Acad Sci*
15. Mclellan, Chen, Joyce et al. (2013) "Structure-based design of a fusion glycoprotein vaccine for respiratory syncytial virus" *Science*
16. Crank, Ruckwardt, Chen et al. (2019) "A proof of concept for structure-based vaccine design targeting RSV in humans" *Science*
17. Battles, Mclellan (2019) "Respiratory syncytial virus entry and how to block it" *Nat Rev Microbiol*
18. Falloon, Yu, Esser et al. (2017) "An adjuvanted, postfusion F protein-based vaccine did not prevent respiratory syncytial virus illness in older adults" *J Infect Dis*
19. Walsh, Marc, Zareba et al. (2023) "Efficacy and safety of a bivalent RSV prefusion F vaccine in older adults" *N Engl J Med*
20. Papi, Ison, Langley et al. "AReSVi-006 Study Group. 2023. Respiratory syncytial virus prefusion F protein vaccine in older adults" *N Engl J Med*
21. Wilson, Goswami, Baqui et al. (2023) "Efficacy and safety of an mRNA-based RSV PreF vaccine in older adults" *N Engl J Med*
22. Chang, Phung, Crank et al. (2022) "A prefusion-stabilized RSV F subunit vaccine elicits B cell responses with greater breadth and potency than a postfusion F vaccine" *Sci Transl Med*
23. Krarup, Truan, Furmanova-Hollenstein et al. (2015) "A highly stable prefusion RSV F vaccine derived from structural analysis of the fusion mechanism" *Nat Commun*
24. Vijayan, Vogels, Groppo et al. (2024) "A self-amplifying RNA RSV prefusion-F vaccine elicits potent immunity in pre-exposed and naïve non-human primates" *Nat Commun*
25. Joyce, Chen, Georgiev et al. (2019) "Crystal structure and immunogenicity of the DS-Cav1-stabilized fusion glycoprotein from respiratory syncytial virus subtype B" *Pathog Immun*
26. Ruckwardt, Morabito, Phung et al. (2021) "Safety, tolerability, and immunogenicity of the respiratory syncytial virus prefusion F subunit vaccine DS-Cav1: a phase 1, randomised, open-label, dose-escalation clinical trial" *Lancet Respir Med*
27. Ruckwardt (2023) "The road to approved vaccines for respiratory syncytial virus" *NPJ Vaccines*
28. Saeland, Van Der Fits, Bolder et al. (2023) "Combina tion Ad26.RSV.preF/preF protein vaccine induces superior protective immunity compared with individual vaccine components in preclinical models" *NPJ Vaccines*
29. Marcandalli, Fiala, Ols et al. (2019) "Induction of potent neutralizing antibody responses by a designed protein nanoparticle vaccine for respiratory syncytial virus" *Cell*
30. Walsh, Frenck, Jr et al. (2020) "Safety and immunogenicity of two RNA-based COVID-19 vaccine candidates" *N Engl J Med*
31. Polack, Thomas, Kitchin et al. (2020) "Safety and efficacy of the BNT162b2 mRNA COVID-19 vaccine" *N Engl J Med*
32. Baden, Sahly, Essink et al. (2021) "Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine" *N Engl J Med*
33. Chaudhary, Weissman, Whitehead "2021. mRNA vaccines for infectious diseases: principles, delivery and clinical translation" *Nat Rev Drug Discov*
34. Espeseth, Cejas, Citron et al. (2020) "Modified mRNA/ lipid nanoparticle-based vaccines expressing respiratory syncytial virus F protein variants are immunogenic and protective in rodent models of RSV infection" *NPJ Vaccines*
35. Hoffmann, Yang, Huey-Tubman et al. (2023) "ESCRT recruitment to SARS-CoV-2 spike induces virus-like particles that improve mRNA vaccines" *Cell*
36. Swanson, Rainho-Tomko, Williams et al. (2020) "A respiratory syncytial virus (RSV) F protein nanoparticle vaccine focuses antibody responses to a conserved neutralization domain" *Sci Immunol*
37. Kampmann, Madhi, Munjal et al. (2023) "Bivalent prefusion F vaccine in pregnancy to prevent RSV illness in infants" *N Engl J Med*
38. Che, Gribenko, Song et al. (2025) *Full-Length Text Journal of Virology*
39. Martinez, Roopchand, Han et al. (2023) "Rational design of a highly immuno genic prefusion-stabilized F glycoprotein antigen for a respiratory syncytial virus vaccine" *Sci Transl Med*
40. Coffman, Lebman, Rothman (1993) "Mechanism and regulation of immunoglobulin isotype switching"
41. Jo, Schadenhofer, Habierski et al. (2021) "Reverse genetics systems for contemporary isolates of respiratory syncytial virus enable rapid evaluation of antibody escape mutants" *Proc Natl Acad Sci*
42. Laczkó, Hogan, Toulmin et al. "Oguin TH 3rd, et al. 2020. A single immunization with nucleoside-modified mRNA vaccines elicits strong cellular and humoral immune responses against SARS-CoV-2 in mice" *Immunity*
43. Lin, Yin, Zhao et al. (2025) "A truncated pre-F protein mRNA vaccine elicits an enhanced immune response and protection against respiratory syncytial virus" *Nat Commun*
44. Nussbaum, Cao, Railkar et al. (2023) "Evaluation of a stabilized RSV pre-fusion F mRNA vaccine: preclinical studies and Phase 1 clinical testing in healthy adults" *Vaccine (Auckl)*
45. Snape, Moore, Du et al. (2024) "Safety and immunogenicity of an mRNA-based RSV vaccine and an RSV/ hMPV combination vaccine in children 5 to 23 months of age"
46. Leroux-Roels, Davis, Steenackers et al. (2023) "Safety and immunogenicity of a respiratory syncytial virus prefusion F (RSVPreF3) candidate vaccine in older adults: phase 1/2 randomized clinical trial" *J Infect Dis*
47. De, Pickles, Yao et al. (2023) "Human T cells efficiently control RSV infection" *JCI Insight*
48. Graham, Bunton, Wright et al. (1991) "Role of T lympho cyte subsets in the pathogenesis of primary infection and rechallenge with respiratory syncytial virus in mice" *J Clin Invest*
49. Abbas, Murphy, Sher (1996) "Functional diversity of helper T lymphocytes" *Nature*
50. Britton, Roper, Kotton et al. (2024) "Use of respiratory syncytial virus vaccines in adults aged ≥60 years: updated recommendations of the advisory committee on immunization practices -United States" *MMWR Morb Mortal Wkly Rep*
51. Taylor (2017) "Animal models of respiratory syncytial virus infection" *Vaccine (Auckl)*
52. Sacco, Durbin, Durbin (2015) "Animal models of respiratory syncytial virus infection and disease" *Curr Opin Virol*
53. Boukhvalova, Prince, Blanco (2009) "The cotton rat model of respiratory viral infections" *Biologicals*
54. Kovacs, Parks, Vasilakis et al. (2003) "Enhanced genetic rescue of negative-strand RNA viruses: use of an MVA-T7 RNA polymer ase vector and DNA replication inhibitors" *J Virol Methods*
55. Sun, Zhang, Zhao et al. (2018) "An optimized high-throughput fluorescence plate reader-based RSV neutralization assay" *J Virol Methods*
56. Van Haren, Pedersen, Kumar et al. (2022) "CAF08 adjuvant enables single dose protection against respiratory syncytial virus infection in murine newborns" *Nat Commun*
57. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12570505&blobtype=pdf | # Intra-host variation and transmission dynamics of SARS-CoV-2 Omicron outbreaks in Shandong, China
Xuemin Wei, Qi Gao, Yuhao Wang, Xinyi Gao, Zengqiang Kou, Xiujun Li, Yifei Xu
## Abstract
Investigating the intra-host diversity of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is essential for understanding its transmission and the emergence of new variants. However, there is limited insight into SARS-CoV-2 intra-host diversity and the extent to which shared intra-host single nucleotide variants (iSNVs) occur among samples without epidemiological links. To characterize intra-host diversity, we analyzed sequencing data from 803 samples across four Omicron transmission clusters. The potential co-mutation patterns formed by shared iSNVs contributed to regions in the genome with elevated iSNV density. Most samples did not share iSNVs. Even among the sample pairs that did share at least one iSNV, 24.4% originated from different transmission clusters. For shared iSNV sites that can become fixed as single nucleotide polymorphisms (SNPs), iSNVs cluster within the phylogenetic tree, with branches supporting the same variants as SNPs. This observation suggests that iSNVs likely serve as reservoirs for SNPs. Additionally, the BA.1.1 samples carried iSNVs identical to the characteristic mutations of BA.2 and BA.2.3. These findings provide important insights into the evolution and transmission inference of SARS-CoV-2. IMPORTANCE Understanding the mechanisms behind viral evolution and transmission is crucial, as novel SARS-CoV-2 variants continue to emerge and spread worldwide. Viral evolution is driven not only by variants that circulate globally but also by mutations arising within individual hosts, resulting in the emergence of iSNVs. The role of iSNVs in shaping SARS-CoV-2 evolution and transmission remains poorly characterized. Our results showed a significant enrichment of shared iSNVs in high-density genomic regions, potentially contributing to the formation of co-mutation patterns. However, the presence of shared iSNVs in samples lacking epidemiological links indicates that they alone are insufficient for accurately reconstructing transmission routes. Instead, iSNVs may act as a reservoir for the emergence of single nucleotide polymorphisms. Our study offers new insights into the evolution of SARS-CoV-2 and the interpretation of transmission from sequencing data. KEYWORDS intra-host diversity, iSNV, SARS-CoV-2, co-mutation pattern, evolution S ince its emergence in November 2019, severe acute respiratory syndrome coronavi rus 2 (SARS-CoV-2) has continuously evolved through mutations and recombination events, leading to the emergence of multiple variants of concern such as Alpha, Delta, and Omicron (1-3). These variants are characterized by distinct mutations that enhance transmissibility, immune evasion, or pathogenicity, resulting in successive waves of COVID-19 infections worldwide (1, 4). The continuous emergence and global spread of such variants highlight the critical need to understand the mechanisms driving viral evolution and transmission.The evolution of the virus occurs not only through globally circulating variants but also through mutations within individual hosts, leading to the emergence of intra-host
single nucleotide variants (iSNVs) (4). Compared to single nucleotide polymorphisms (SNPs), iSNVs encompass a broader spectrum of genetic mutations, offering deeper insights into viral diversity and evolutionary dynamics. For example, the D614G of SARS-CoV-2, which improved viral replication and transmission, was initially detected as an iSNV in samples before becoming recognized as an SNP (5)(6)(7)(8)(9). While intra-host diversity of SARS-CoV-2 has been observed, research on this aspect remains limited (10).
Analyzing transmission patterns during outbreaks is crucial for understanding pathogen spread and designing targeted public health measures. Combining phyloge netic analysis with epidemiological investigations allows for the identification of the origins and transmission clusters of pathogens (11)(12)(13). However, consensus sequences, which reflect only the dominant viral lineage within a host, offer limited resolution for transmission analysis. In contrast, the analysis of intra-host diversity may provide a more precise approach to studying viral transmission (14,15). Despite its potential significance, the characteristics of SARS-CoV-2 intra-host diversity and the extent to which shared iSNVs occur among samples without epidemiological links remain insufficiently explored (16,17).
Here, we analyzed sequencing data from 803 samples across four transmission clusters that occurred during the Omicron epidemic in Shandong Province in March 2022. We aimed to characterize the features of intra-host diversity, investigate the likelihood of shared iSNVs arising among samples without epidemiological links, and explore the transition of iSNVs into SNPs. Our findings enhance our understanding of SARS-CoV-2 variation in individual hosts and highlight key considerations for inferring transmission using shared iSNVs.
## MATERIALS AND METHODS
## Data collection
The sequencing data of 803 SARS-CoV-2 Omicron samples, collected from public health laboratories in Shandong as part of routine surveillance, were obtained from our previous report (18). These samples were obtained from four transmission clusters, SD-1-SD-4, during the Omicron epidemic in Shandong between 1 March and 27 March 2022. These four local transmission clusters were identified by a previous study based on genome phylogeny and public health investigation results. These four clusters belonged to three genetic sublineages of the Omicron variant: BA.1.1 (SD-1), BA.2 (SD-2), and BA.2.3 (SD-3 and SD-4).
## Genome sequencing
All samples were collected by personnel certified through unified biosafety training and operational protocols. Viral RNA extraction, cDNA synthesis, and sequencing were performed as previously described (18). Viral RNA was extracted using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany). The extracted RNA was reverse transcribed and amplified using the ULSEN 2019-nCoV Whole-Genome Capture Kit (MicroFuture Technology, Beijing, China). The resulting cDNA was purified using AMPure XP beads (Beckman Coulter, USA) at a 1:1 ratio and quantified with the Qubit 1X dsDNA HS Assay Kit (Thermo Fisher Scientific, USA), following the manufacturer's instructions. Libraries were prepared from 0.2 ng/µL cDNA using the Nextera XT Library Prep Kit with Nextera XT Indexes (Illumina, San Diego, CA, USA). Sequencing was performed on Illumina NextSeq 2000 or MiSeq platforms.
## Calling of iSNVs
Sequencing reads were mapped to the reference genome (Wuhan-Hu-1, NCBI NC_045512.2) using BWA-MEM version 0.7.17 (19). Only properly paired reads were retained for the downstream analysis. PCR-duplicated reads were removed using Picard MarkDuplicates (version 2.10.10) (http://broadinstitute.github.io/picard). Base composition of each position was summarized from the mapping profile using pysamstats (version 1.1.2, https://github.com/alimanfoo/pysamstats). To enable accurate iSNV detection while minimizing false positives from sequencing or alignment errors, the following filtering criteria were applied, consistent with established practices: (i) bases with a Phred-scaled base quality score <20 were excluded to improve the specificity of low-frequency variant calling (20,21). (ii) Sites required a minimum depth ≥100× to ensure sufficient power for detecting variants (22). (iii) The variants required ≥10 supporting reads to distinguish them from sequencing noise (22). (iv) iSNVs with strand bias greater than 10-fold were removed to avoid strand-specific artifacts (20). (v) Only variants with allele frequencies (AFs) between 3% and <70% were retained, ensuring that low-frequency variants (AF ≥ 3%) exceed typical sequencing error rates (17,(22)(23)(24). (vi) Genomic regions containing ≥5 iSNVs within a 50 bp window were excluded to minimize false positives from local misalignment (25). To avoid recurring sequencing errors, iSNVs with consistently low AF identified in multiple samples were excluded (Fig. S1) (23). SNPs were defined as variants meeting the following criteria: (i) depth of coverage ≥ 10; (ii) AF ≥ 70%. Using these rigorous, literature-supported criteria, iSNVs and SNPs were identified from our sample set. To rule out cross-sample contamination, we specifically compared the genomic positions of identified iSNVs with SNP sites from our samples. Contamination typically results in spurious iSNVs overlapping SNP positions due to cross-sample mixing. In our data set, only 8.37% (123 iSNVs) overlapped with SNP positions-a much lower proportion than previously reported (26)-indicating that contamination did not appreciably affect our iSNV calls. The iSNV density in a given region was calculated by dividing the total number of identified iSNVs by the number of samples and then further dividing by the length of the region.
The consensus sequences were derived from prior research (18). These sequences were generated with a minimum of 10-fold mapping coverage and supported by at least 70% of reads at each position. Consensus sequences with more than 90% genome coverage were used for downstream analysis. The genetic clade for each SARS-CoV-2 sequence was determined using Nextclade version 1.11.0 (https://clades.nextstrain.org).
## Evaluating iSNV genomic distance distributions
iSNV positions were simulated in R using the runif function, which utilizes a uniform distribution based on the number of iSNVs. Both actual and simulated iSNVs were ranked, and genomic distances between neighboring mutant pairs were calculated. It was found that the distances between simulated iSNVs followed a Poisson distribution. To evaluate the distribution of distances between adjacent iSNVs, the Kolmogorov-Smir nov test was used to compare the observed distribution with the expected distribution under a Poisson model.
## Identification of co-infection cases
Co-infection cases must satisfy three criteria: (i) a sample must contain featured mutations from more than one sublineage; (ii) the AF of mutations within the same sublineage should be similar; (iii) the total AF across all detected sublineages should be close to 100% (27)
## Analysis of potential co-mutation patterns
A network of the shared iSNVs was constructed using Cytoscape version 3.9.1 (28). An edge-weighted, spring-embedded model was used to determine the layout. In the network, each node represented a sample, and two nodes were connected through an edge if samples shared ≥1 iSNVs. Samples that shared more iSNVs were closer in network distance. There were 229 nodes and 994 edges.
Potential co-mutation patterns were further evaluated using the Pearson correlation coefficient (PCC) and the significance level (P-value) obtained from linear regression analysis. The coefficients of linear correlation between two shared iSNVs were evaluated by PCC, with a value between -1.0 and 1.0. There was a strong correlation between the two iSNVs when the absolute value of PCC was between 0.7 and 1. The relationship between one or more independent and dependent variables was modeled by linear regression.
## Phylogenetic association of iSNVs and SNPs
If an iSNV was detected at a specific genomic site in a sample, and a correspond ing SNP (with the same base pair substitution, not just the same genomic location) was found at the same site in other samples, these sites were designated potential fixed sites. To prevent misclassification of lineage-defining mutations (relative to the Wuhan-Hu-1) as SNPs in subsequent analyses, all genomic positions corresponding to the featured mutations defining the Omicron sublineages observed in our study (BA.1.1, BA.2, and BA.2.3) were systematically excluded. These featured mutations were predominantly curated from Outbreak.info (https://outbreak.info/situation-reports). The consensus sequences obtained from GISAID (https://gisaid.org/) and those from the Omicron outbreak in Shandong, China, were aligned using MAFFT version 7.310 (29). A maximum likelihood phylogenetic tree was constructed based on SARS-CoV-2 coding region sequences using IQ-TREE version 2.0.3 with the best-fit substitution model determined by the software (30). Branch support was assessed with ultrafast bootstrap ping. The final tree was rooted using the reference sequence, after which the reference sequence, all GISAID isolates, and sequences not included in this study were pruned. When an iSNV corresponded to an SNP (defined by the base pair involved, not just the site), ancestral state reconstruction was performed on the consensus trees using ClonalFrameML to identify all branches on which that substitution had occurred (31). The patristic distance from each tip in the tree to the midpoint of the closest one of these branches was calculated. A one-tailed Mann-Whitney U test was then used to assess the association between the presence of iSNV in a sample and this distance. Multiple testing was controlled for each using the Benjamini-Hochberg adjustment. These analyses were done both on an individual site level and across all sites of interest.
## RESULTS
## Distribution of iSNVs among samples
We retrieved sequencing data for 803 SARS-CoV-2 Omicron samples from the Shandong epidemic between 3 March and 27 March 2022. The number of iSNVs in each sample was not affected by sequencing depth or Ct values (R-square [R 2 ] = 0.14 for sequencing depth; R 2 = 0.10 for Ct values of ORF1ab gene; R 2 = 0.08 for Ct values of N gene; Fig. S1a through c), suggesting that the identified iSNVs were unbiased to sequencing data. We identified 1,470 iSNVs at 1,133 sites. The median number of iSNVs per sample was one (Fig. 1a). About 54.05% of samples (434 of 803) harbored at least one iSNV in comparison with the reference genome. We found a steady increase in the density of iSNVs over time during the epidemic, increasing from 0.03 to 0.23 iSNVs/kb within 24 days (Fig. 1b).
## Distribution of iSNVs across the genome
We investigated the distribution of iSNVs across the SARS-CoV-2 genome and found an overall iSNV density of 0.06 iSNVs/kb (Fig. 1d). The iSNV density of the 5′ UTR, 3′ UTR, and intergenic region was 0.08, 0.07, and 0.05 iSNVs/kb, respectively (Fig. 1d). We identified 1,436 (97.69%) iSNVs in the coding regions, which cover 97.85% of the genome. Most iSNVs (1,062, 73.96%) were present in the ORF 1ab, followed by the S gene (165 iSNVs, 11.49%) and N gene (66 iSNVs, 4.60%). After normalization by gene length, the ORF7a gene showed the highest iSNV density at 0.09 iSNVs/kb (Fig. 1e). To determine if selection pressure influenced the distribution of iSNVs across the genome, we calculated the genomic distances between allele pairs. The fitted density curve showed a significant difference from randomly generated mutations (Kolmogorov-Smirnov test, P < 0.001; Fig. 1c), indicating a non-stochastic distribution of iSNVs. We also analyzed iSNV distribution across codon positions for all genes. The third codon position of ORF1ab had the highest iSNVs, followed by the first, with the second having the lowest (Fisher's exact test, P < 0.05; Fig. 1f). The N gene showed more iSNVs at the third codon position than at the second (Fisher's exact test, P < 0.05) and exhibited a lower nonsy nonymous/synonymous ratio compared to other regions of the viral genome (1.13 vs 1.91, Fisher's exact test, P = 0.03; Fig. 1g). This suggested that the N gene may be subject to stronger negative selection than other genomic regions.
Interestingly, we identified two distinct regions (referred to as high-density regions) in the ORF1ab gene at positions 5,700-5,800 and 16,700-16,800. The iSNV density in these two regions (1.12 and 0.42 iSNVs/kb) was 19 and 7 times higher than the overall iSNV density in the whole genome. In these high-density regions, the number of iSNVs at the first codon position was higher than at the second and third codon positions (Fisher's exact test, P < 0.05; Fig. 1d). Moreover, the non-synonymous/synonymous iSNV ratio in these regions (23.8) was significantly higher than that of other regions of the viral genome (1.7, Fisher's exact test, P < 0.001). While we did not observe a significant difference in AF between non-synonymous and synonymous iSNVs (Mann-Whitney U-tests, P = 0.367), the median AF of non-synonymous iSNVs (0.10) was higher than that of synonymous iSNVs (0.06, Fig. 1h). These results suggested that the iSNVs in high-density regions might be under stronger selection pressure.
## Potential co-mutation patterns in high-density regions
We analyzed the number of iSNV sites and the proportion of shared iSNVs within high-density regions. The number of iSNV sites in high-density regions (0.06 iSNV sites/kb) was similar to that in other regions of the viral genome (0.05 iSNV sites/kb). Notably, the proportion of shared iSNVs (96.8%) among all iSNVs in high-density regions was significantly higher compared to other regions of the viral genome (22.9%; Fisher's exact test, P < 0.001). The majority of these shared iSNVs within the high-density regions formed potential co-mutation patterns, as revealed by the shared iSNV network (Fig. 2a; Fig. S3), with shared nonsynonymous iSNVs-G5743C, T5744C, G5765A, G5766C, and C16708G-forming a significant cluster.
We then characterized co-occurring mutations at four genomic positions (5,743, 5,744, 5,765, and 5,766) across 752 high-coverage SARS-CoV-2 samples (≥100× at all target sites). After excluding reads with no mutations at these sites (the GTGG haplo type), the most common pattern observed was CCAC, corresponding to simultaneous mutations G5743C, T5744C, G5765A, and G5766C. This indicated that these shared iSNVs appeared together in the same virus strain (Fig. 2b). The reads exhibiting the CCAC pattern at four sites originated from 18 samples, all of which also carried the C16708G. To further evaluate the potential co-mutation pattern, we performed linear regression analysis and calculated Pearson correlation coefficients for iSNVs at these five positions. The results showed a strong correlation (r > 0.7) with high statistical significance (P < 2.5e-87; Fig. 2c andd). These findings indicated that these iSNVs contributed to a potential co-mutation pattern.
## Shared iSNVs across samples
We investigated patterns of shared intra-host diversity between individuals. Most (1,041, 70.82%) iSNVs appeared in only a single sample. Among all iSNVs, 429 (29.18%) were present in at least two samples (Fig. 3a). The 197 iSNVs that occurred in more than five samples were not located at common sequencing error sites or homoplasious sites (32,33). The shared iSNVs observed at seven positions had high AF across multiple samples, with only G13109A being a nonsynonymous iSNV (Fig. 3b). The G13109A was located in the ORF1ab gene and caused the amino acid aspartic acid (D) to be replaced by asparagine (N; D4282N). Additionally, among these seven mutations, the C7303T was present across different transmission clusters, indicating that shared iSNVs may arise independently in samples without epidemiological links.
To further explore the occurrence of shared iSNVs in sample pairs lacking epidemio logical connections, we analyzed all unique sample pairs with detected iSNVs (n = 93,961). The majority of iSNVs were unique to samples, resulting in most sample pairs not sharing any iSNVs. Only 1.06% of the pairs (994 pairs) shared at least one iSNV. Of these pairs, 243 involved samples that belonged to different transmission clusters (Fig. 3c andd). Within these 243 pairs, 15 pairs showed SNP differences of just two to three and were collected within a ≤2-day interval. Additionally, 74% of these pairs were sequenced in different batches, lowering the likelihood that shared iSNVs were due to cross-contami nation. Co-infection cases can also confound transmission analysis based on shared iSNVs. An examination of AF at polymorphic sites in each sample confirmed that no co-infections were present in this study. This is likely because each transmission cluster was predomi nantly driven by a single sublineage strain.
## The evolution patterns from iSNVs to SNPs
To avoid misclassifying Omicron-featured mutations (BA.1.1, BA.2, and BA.2.3) as SNPs, we excluded all sites containing featured mutations prior to subsequent analysis. Among 1,253 genomic sites harboring mutations with AF ≥ 0.03, we identified 37 sites where an iSNV in one sample corresponded to an identical base pair substitution reaching SNP status in at least one other distinct sample. We define these sites as potential fixed sites. The fixed sites displayed a higher AF compared to non-fixed sites (Fig. 4a). The proportion of nonsynonymous iSNVs in fixed sites was lower than that of synonymous iSNVs (Fig. 4b). Meanwhile, the nonsynonymous iSNVs in fixed sites were only distributed in ORF1ab, S, and ORF3a genes.
We further investigated whether iSNVs can be used to resolve phylogenies and transmission clusters. If an iSNV evolves into an SNP through transmission, we would expect a phylogenetic association where samples containing the iSNV cluster with branches showing an SNP at the same sites. Of 113 sites shared across ≥2 samples, only 20 were fixed sites and located within protein-coding regions. For these 20 fixed iSNV sites, we observed that the patristic distance between a given iSNV position in one sample and the nearest sample with the same SNP was significantly lower than that of samples without the iSNV (P < 0.001; Fig. 4c). This suggested that iSNVs cluster within the phylogenetic tree, with branches supporting the same variants as SNPs.
When analyzing each site individually, 11 showed a significant association after applying the Benjamini-Hochberg correction (P < 0.05). For example, in site 13,051, the blue branches represent the SNP change from a C to T, with nearby iSNVs appearing as minor transitions at the ancestral nodes of the changing branches (Fig. 4d). Among 11 significant iSNV sites, we observed a gradual AF increase in C23978T and C20404T within a single transmission cluster (Fig. 4e). These findings suggested that iSNVs may act as a mutation reservoir for SNPs. We found that samples from the BA.1.1 sublineage
## DISCUSSION
The emergence of the divergent Omicron lineage pandemic is of worldwide concern. Characterizing the intra-host diversity of SARS-CoV-2 is crucial for identifying potential variants and inferring transmission dynamics (5,34,35). Our analysis characterized the genetic diversity of iSNVs and provided insights into the likelihood of shared iSNVs independently emerging in epidemiologically unlinked individuals, as well as the transition of iSNVs into SNPs.
Since the emergence of SARS-CoV-2, there have been numerous mutant strains, with a few exhibiting significantly enhanced transmissibility and infectivity (36)(37)(38). Studying the mutation patterns in the SARS-CoV-2 genome helps identify specific regions or genes that have undergone positive selection. Our results showed that the distribution of iSNVs was relatively uniform across most regions of the whole genome, while uneven in some regions, consistent with previous studies (25). Among all the genes, the N gene appears to be the most conserved gene during this outbreak, making it a potential target for vaccine development (39,40). Specifically, we found that two regions, located in nsp3 and nsp13, accumulated more iSNVs and were subject to more selection pressure than other regions in the genome. They have a higher iSNV density compared to other regions primarily due to the presence of shared iSNVs, which form a potential co-muta tion pattern, including G5743C, T5744C, G5765A, G5766C, and C16708G. The first four mutations were located in the papain-like protease (PLpro) encoded by nsp3, which cleaves the viral polyprotein for assembly and exhibits deubiquitination activity to evade the host's innate immune response (41,42). The C16708G was located on nsp13, which inhibits interferon production in vivo (43,44). The accumulation of iSNVs on PLpro and nsp13 may affect the pathogenicity and immune escape ability of SARS-CoV-2.
Evaluating within-host viral diversity is crucial for understanding the transmission process. A narrow transmission bottleneck allows only a limited number of variants to successfully transmit between hosts. Our results revealed limited within-host diversity of SARS-CoV-2, suggesting insufficient selective pressure for the virus to adapt to the host environment. This limited diversity, together with our finding that most samples lack shared iSNVs, supports the existence of a narrow transmission bottleneck (10,23). Such a bottleneck not only influences viral evolution but also restricts the utility of iSNVs for high-resolution transmission inference.
Shared iSNVs observed across genomes separated by time and distinct evolutionary lineages indicate their convergent emergence, complicating their use in transmission inference (17). In contrast to previous studies (35,45), which faced challenges in establishing epidemiological connections, our samples, derived from four transmission clusters, allow us to better understand the extent of the emergence of convergent iSNVs. Our findings revealed that 24.4% of sample pairs sharing at least one iSNV originated from different transmission clusters. Moreover, even in cases where consensus sequen ces were highly similar and sampling dates were within 2 days apart, iSNV sharing can still occur between samples with no epidemiological links. Given that convergent evolution can significantly contribute to the emergence of shared iSNVs, caution should be exercised when using these shared iSNVs to infer transmission (46).
While convergent evolution may limit direct transmission inference from shared iSNVs, our findings demonstrate that, in certain cases, these iSNVs can still be transmit ted successfully and significantly influence the evolution of phylogenetic consensus sequences (22,23). Additionally, the frequencies of shared iSNV can be observed to gradually increase over time, becoming fixed in subsequent infections. These observa tions underscore the critical role of shared iSNVs in elucidating transmission dynamics. The iSNVs serve as a reservoir of mutations for potential epidemic variants (5,47). Our results indicate that iSNVs detected in samples from earlier sublineages correspond to characteristic mutations observed in later-emerging sublineages, highlighting their important role in the formation of SNPs. This process may drive the evolution of new viral sublineages, especially those with enhanced transmissibility or virulence. Under specific high selective pressures, iSNVs can undergo adaptive evolution, transitioning from iSNVs to SNPs, which facilitates the virus's ability to adapt and spread (48).
Our study has several limitations. First, although we detected iSNVs in early subline age samples that are characteristic mutations of later sublineages, we did not observe the progression of these iSNVs becoming fixed SNPs. Therefore, future studies should increase the sample size and sampling duration to further investigate how variants of concern mutations and lineages become fixed as SNPs. Second, the biological functions of the identified potential co-mutations were not validated. Further studies are needed to determine whether these co-occurring iSNVs are adaptive mutations and how they affect viral function.
## Conclusions
In summary, our findings revealed that the significant enrichment of shared iSNVs in high-density regions drives the formation of potential co-mutation patterns. Although shared iSNVs are important, they alone are insufficient to reliably reconstruct trans mission pathways due to the interference of convergent evolution. The iSNVs serve as mutation reservoirs, with iSNVs from early sublineages potentially developing into characteristic mutations in later sublineages. These insights enhanced our understand ing of the mutational processes, evolutionary dynamics, and transmission mechanisms of SARS-CoV-2.
## References
1. Carabelli, Peacock, Thorne et al. (2023) "SARS-CoV-2 variant biology: immune escape, transmission and fitness" *Nat Rev Microbiol*
2. Wang, Møhlenberg, Wang et al. (2023) "Immune evasion of neutralizing antibodies by SARS-CoV-2 Omicron" *Cytokine Growth Factor Rev*
3. Wolf, Wolf, Bello et al. (2019) "Molecular evolution of SARS-CoV-2 from December" *J Med Virol*
4. Yamasoba, Kimura, Nasser et al. (2022) "Virological characteristics of the SARS-CoV-2 Omicron BA.2 spike" *Cell*
5. Sun, Wang, Tan et al. (2021) "SARS-CoV-2 quasispecies provides an advantage mutation pool for the epidemic variants" *Microbiol Spectr*
6. Hou, Chiba, Halfmann et al. (2020) "SARS-CoV-2 D614G variant exhibits efficient replication ex vivo and transmission in vivo" *Science*
7. Yurkovetskiy, Wang, Pascal et al. (2020) "Structural and functional analysis of the D614G SARS-CoV-2 spike protein variant" *Cell*
8. Su, Zeller, Cronin et al. (2025) "Rapid emergence and evolution of SARS-CoV-2 intrahost variants among COVID-19 patients with prolonged infections" *Emerg Infect Dis*
9. Pavia, Quirino, Marascio et al. (2024) "Persistence of SARS-CoV-2 infection and viral intra-and inter-host evolution in COVID-19 hospitalized patients" *J Med Virol*
11. Hannon, Roychoudhury, Xie et al. (2022) "Narrow transmission bottlenecks and limited within-host viral diversity during a SARS-CoV-2 outbreak on a fishing boat" *Virus Evol*
12. Da Silva, Shepherd, Williams et al. (2021) "Genomic epidemiology reveals multiple introductions of SARS-CoV-2 from mainland Europe into Scotland"
13. Lu, Du Plessis, Liu et al. (2020) "Genomic epidemiology of SARS-CoV-2 in Guangdong Province" *Cell*
14. Lane, Sherry, Porter et al. (2021) "Genomics-informed responses in the elimination of COVID-19 in Victoria, Australia: an observational, genomic epidemiological study" *Lancet Public Health*
15. Song, Li, Kang et al. (2021) "Genomic epidemiology of SARS-CoV-2 in Pakistan" *Genomics Proteomics Bioinformatics*
16. San, Ngcapu, Kanzi et al. (2021) "Transmission dynamics of SARS-CoV-2 within-host diversity in two major hospital outbreaks in South Africa" *Virus Evol*
17. Popa, Genger, Nicholson et al. (2020) "Genomic epidemiology of superspreading events in Austria reveals mutational dynamics and transmission properties of SARS-CoV-2" *Sci Transl Med*
18. Braun, Moreno, Wagner et al. (2021) "Acute SARS-CoV-2 infections harbor limited within-host diversity and transmit via tight transmission bottlenecks" *PLoS Pathog*
19. Xu, Liu, Li et al. (2023) "Transmission of SARS-CoV-2 Omicron variant under a dynamic clearance strategy in Shandong" *Microbiol Spectr*
20. Li, Durbin (2010) "Fast and accurate long-read alignment with Burrows-Wheeler transform" *Bioinformatics*
21. Wang, Wang, Zhang et al. "2021. Intra-host variation and evolutionary dynamics of SARS-CoV-2 populations in COVID-19 patients" *Genome Med*
22. Du, Ding, Li et al. (2020) "Genomic surveillance of COVID-19 cases in Beijing" *Nat Commun*
23. Li, Deng, Li et al. (2022) "Viral infection and transmission in a large, well-traced outbreak caused by the SARS-CoV-2 Delta variant" *Nat Commun*
24. Lythgoe, Hall, Ferretti et al. (2021) "SARS-CoV-2 within-host diversity and transmission" *Science*
25. Schirmer, 'amore, Ijaz et al. (2016) "Illumina error profiles: resolving fine-scale variation in metagenomic sequencing data" *BMC Bioinformatics*
26. Li, Du, Yang et al. (2022) "Two-step fitness selection for intra-host variations in SARS-CoV-2" *Cell Rep*
27. Khatib, Benslimane, Elbashir et al. (2020) "Within-host diversity of SARS-CoV-2 in COVID-19 patients with variable disease severities" *Front Cell Infect Microbiol*
28. Zhou, Cheng, Xu et al. (2022) "Genomic evidence for divergent co-infections of cocirculating SARS-CoV-2 lineages" *Comput Struct Biotechnol J*
29. Otasek, Morris, Bouças et al. (2019) "Cytoscape automation: empowering workflow-based network analysis" *Genome Biol*
30. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
31. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol Biol Evol*
32. Didelot, Wilson (2015) "ClonalFrameML: efficient inference of recombination in whole bacterial genomes" *PLoS Comput Biol*
33. Van Dorp, Acman, Richard et al. (2020) "Emergence of genomic diversity and recurrent mutations in SARS-CoV-2" *Infect Genet Evol*
34. (2020) "Issues with SARS-CoV-2 sequencing data"
35. Lin, Drysdale, Snape et al. (2021) "Distinct patterns of within-host virus populations between two subgroups of human respiratory syncytial virus" *Nat Commun*
36. Pathak, Mishra, Uppili et al. (2022) "Spatio-temporal dynamics of intra-host variability in SARS-CoV-2 genomes"
37. Cheng, Song, Zhou et al. (2021) "Impact of the N501Y substitution of SARS-CoV-2 Spike on neutralizing monoclonal antibodies targeting diverse epitopes" *Virol J*
38. Wang, Zhang, Li et al. (2022) "Reduced sensitivity of the SARS-CoV-2 Lambda variant to monoclonal antibodies and neutralizing antibodies induced by infection and vaccination" *Emerg Microbes Infect*
39. Li, Nie, Wu et al. (2021) "SARS-CoV-2 501Y.V2 variants lack higher infectivity but do have immune escape" *Cell*
40. Abbasi, Tabaraei, Hosseini et al. (2022) "Real-time PCR Ct value in SARS-CoV-2 detection: RdRp or N gene?" *Infection*
41. Oronsky, Larson, Caroen et al. (2022) "Nucleocapsid as a next-generation COVID-19 vaccine candidate" *Int J Infect Dis*
42. Klemm, Ebert, Calleja et al. (2020) "Mechanism and inhibition of the papain-like protease, PLpro, of SARS-CoV-2" *EMBO J*
43. Ferreira, Villanueva, Adem et al. (2024) "Identification of novel allosteric sites of SARS-CoV-2 papain-like protease (PLpro) for the development of COVID-19 antivirals" *J Biol Chem*
44. Feng, Zhang, Min et al. (2023) "SARS-CoV-2 NSP13 interacts with host IRF3, blocking antiviral immune responses" *J Med Virol*
45. Yuen, Lam, Wong et al. (2020) "SARS-CoV-2 nsp13, nsp14, nsp15 and orf6 function as potent interferon antagonists" *Emerg Microbes Infect*
46. Tonkin-Hill, Martincorena, Amato et al. (2021) "Patterns of within-host genetic diversity in SARS-CoV-2. Elife 10:e66857"
47. Valesano, Rumfelt, Dimcheff et al. (2021) "Temporal dynamics of SARS-CoV-2 mutation accumulation within and across infected hosts" *PLoS Pathog*
48. Lemieux, Siddle, Shaw et al. (2021) "Phylogenetic analysis of SARS-CoV-2 in Boston highlights the impact of superspreading events" *Science*
49. Nijhuis, Van Maarseveen, Boucher (2009) "Antiviral resistance and impact on viral replication capacity: evolution of viruses under antiviral pressure occurs in three phases" *Handb Exp Pharmacol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12414411&blobtype=pdf | # Thermal tolerance and inactivation of Ebola virus
Xiaoxiao Gao, Cheng Peng, Rongjuan Pei, Virologica Sinica
## Dear Editor,
Viruses of the genus Orthoebolavirus cause sporadic outbreaks of severe haemorrhagic fever, with case fatality rates ranging from 25% to 90% (Mahanty and Bray, 2004). Six species of the virus (Orthoebolavirus zairense, sudanense, bundibugyoense, taiense, restonense, and bombaliense) have so far been identified (Biedenkopf et al., 2023). Among these, Orthoebolavirus zairense, commonly known as Ebola virus (EBOV), stands out as the most virulent. Given its high contagiousness and lethality, EBOV must be manipulated under biosafety level 4 (BSL-4) conditions, as stipulated by the the People's Republic of China's list of human pathogenic microorganisms (National Health Commission of the People's Republic of China, 2023). Prior to being removed from a BSL-4 laboratory, it is imperative that infectious EBOV undergoes complete inactivation. Here we systematically evaluate viral thermostability under BSL-4 containment conditions, demonstrating EBOV's marked thermotolerance.
The EBOV stock (Mayinga, 1976 strain) was dispensed into 0.2 mL aliquots within 2 mL hermetic screw-cap tubes for thermal challenge experiments. Samples allocated for 4 • C, ambient temperature (22 • C-25 • C), and 37 • C treatments were maintained under refrigerated conditions, ambient bench conditions, and an incubator, respectively. Comparative validation studies were performed using both dry-bath and water-bath heating systems at 60 • C and 100 • C. Temperature monitoring was conducted throughout the experiment using a calibrated thermometer in a non-spiked medium sample run concurrently. Post-treatment protocols involved immediate thermal arrest through flash-cooling on ice (15 minutes), followed by refrigerated storage (4 • C) until completion of all experimental conditions. The infectivity of the treated samples was determined using median tissue culture infectious dose (TCID 50 ) assay. Briefly, monolayers of Vero E6 cells grown in 96-well plates were inoculated with 100 μL of serial 10-fold diluted samples in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 2% fetal bovine serum (FBS). Viral titers were calculated by the Reed & Muench method, based on the cytopathic effect (CPE) observed at 10 days post-infection (Cao et al., 2023).
A one-week storage of EBOV at 4 • C resulted in a significant reduction in viral infectivity, dropping from an initial titer of 8.38 log 10 TCID 50 /mL to 6.62 log 10 TCID 50 /mL. However, the decrease was not marked for storage periods of 1 or 2 days at 4 • C (Fig. 1A). This suggests that EBOV stocks can be safely stored at 4 • C for at least two days without a significant loss in infectivity, thus avoiding the need for repeated freeze-thaw cycles. In contrast, the infectivity of the samples stored at room temperature and 37 • C decreased significantly and progressively over time. After one day, a reduction of approximately 1 log 10 TCID 50 /mL in viral titer was observed, with further reductions of 1.6 log 10 TCID 50 /mL at room temperature and 2 log 10 TCID 50 /mL at 37 • C after two days. Notably, viral infectivity became undetectable after a seven-day incubation period (Fig. 1A). Dry-bath treatment at 60 • C reduced viral titers from 8.38 to 3.82 log 10 TCID 50 /mL within 15 minutes, yet residual infectivity (2.63 log 10 TCID 50 /mL) persisted after 2 hours (Fig. 1B). Comparable water-bath treatment showed similar kinetics, with titers declining from 7.5 to 2.7 log 10 TCID 50 /mL in 15 minutes and remaining at 1.7 log 10 TCID 50 /mL after 2 hours (Fig. 1B). Control samples maintained at room temperature showed no titer reduction. Critical differences emerged at 100 • C: dry-bath treatment required 10 minutes for complete inactivation, with detectable virus (1.62 log 10 TCID 50 /mL) remaining after 5 minutes (Fig. 1C). Water-bath treatment achieved complete inactivation within 5 minutes (Fig. 1C). Three sequential blind passages in Vero E6 cells confirmed complete inactivation, with no cytopathic effect or detectable EBOV RNA in thirdpassage supernatants (Fig. 1D). Methodological analysis revealed that incomplete tube immersion in dry-bath systems compromised inactivation efficacy. When 30 μL aliquots were deliberately placed in unheated tube regions (extending beyond the heating block), 90% of samples (9/10) retained high infectivity (7 log 10 TCID 50 /mL) after 10min 100 • C treatment (Fig. 1C), demonstrating the critical importance of complete sample immersion for reliable thermal inactivation.
Our study demonstrates the notable thermostability of EBOV. Existing literature presents variable outcomes across experimental conditions. Mitchell et al. (1984) demonstrated effective viral inactivation in serum samples through 60 • C exposure for 60 minutes while maintaining biochemical stability of glucose, blood urea nitrogen, and electrolytes. Conversely, Smither et al. (2015) documented thermal resistance in murine whole blood specimens, with EBOV-Kikwit remaining infectious following 60 • C exposure for 15 minutes. Haddock et al. (2016) achieved EBOV inactivation by boiling the virus stock at either 100 • C for 10 minutes or 120 • C for 5 minutes, though 65 • C treatment for 30 minutes retained viral infectivity in cellular assays. Olejnik et al. (2023) reported effective inactivation (60 • C/30 min) for samples containing ≤ 1.67 × 10 6 infectious particles, while demonstrating infectious residual for samples with higher viral concentrations.
Based on these observations and our experimental validations, we categorically advise against dry-bath methodologies and advocate for complete sample submersion in water-bath systems. Our optimized protocol specifies immersion in boiling water (100 • C) for ≥ 15 minutes. Critical parameters influencing thermal efficacy include: 1) viral strain characteristics, 2) aggregation state, 3) matrix composition, 4) initial viral titer, and 5) thermal transfer dynamics mediated by equipment specifications (Smither et al., 2015;Olejnik et al., 2023;Chung et al., 2007;Negovetich et al., 2010;Tuladhar et al., 2012;Pastorino et al., 2020;Feng et al., 2021). This multifactorial dependence necessitates institution-specific validation of inactivation protocols through rigorous quantitative assays prior to implementation in standard operating procedures. Strict adherence to validated thermal parameters remains imperative for biosafety compliance across all subsequent applications.
## References
1. Biedenkopf, Bukreyev, Chandran et al. (2023) "Renaming of genera Ebolavirus and Marburgvirus to Orthoebolavirus and Orthomarburgvirus, respectively, and introduction of binomial species names within family Filoviridae" *Arch. Virol*
2. Chung, Wang, Tang (2007) "Influence of heat transfer with tube methods on measured thermal inactivation parameters for Escherichia coli" *J. Food Protect*
3. Feng, Li, Lin et al. (2021) "Stability of SARS-CoV-2 on the surfaces of three meats in the setting that simulates the cold chain transportation" *Virol. Sin*
4. Haddock, Feldmann, Feldmann (2016) "Effective chemical inactivation of Ebola virus" *Emerg. Infect. Dis*
5. Mahanty, Bray (2004) "Pathogenesis of filoviral haemorrhagic fevers" *Lancet Infect. Dis*
6. Mitchell, Mccormick (1984) "Physicochemical inactivation of Lassa, Ebola, and Marburg viruses and effect on clinical laboratory analyses" *J. Clin. Microbiol*
7. (2023) "National Health Commission of the People's Republic of China"
8. Negovetich, Webster (2010) "Thermostability of subpopulations of H2N3 influenza virus isolates from mallard ducks" *J. Virol*
9. Olejnik, Leon, Michelson et al. (2023) "Establishment of an inactivation method for Ebola virus and SARS-CoV-2 suitable for downstream sequencing of low cell numbers" *Pathogens*
10. Pastorino, Touret, Gilles et al. (2020) "Heat inactivation of different types of SARS-CoV-2 samples: what protocols for biosafety, molecular detection and serological diagnostics?" *Viruses*
11. Smither, Weller, Phelps et al. (2015) "Buffer AVL alone does not inactivate Ebola virus in a representative clinical sample type" *J. Clin. Microbiol*
12. Tuladhar, Bouwknegt, Zwietering et al. (2012) "Thermal stability of structurally different viruses with proven or potential relevance to food safety" *J. Appl. Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12530122&blobtype=pdf | # From activator to suppressor: PACT is joining the company of PKR negative regulators
Francisco Acosta, Christian Pfaller
## Abstract
080743.125. Freely available online through the RNA Open Access option.
In this issue of RNA, Young et al. identify a new regulatory role of the protein activator of the interferon-induced protein kinase (PACT, PRKRA gene) on the activation of protein kinase R (PKR) (Young et al. 2025). PKR is one of four cellular stress-activated kinases inducing global protein translation shutdown through phosphorylation of their natural substrate, the eukaryotic initiation factor 2-α (eIF2α). PKR activation occurs through binding to double-stranded RNA (dsRNA), a hallmark of many viral infections. As such, PKR is an integral part of the cell-intrinsic innate immune response system against viral pathogens (Pfaller et al. 2011). However, tight regulation of PKR activation is essential for maintaining cell viability and health. A large body of evidence published in the past decade has revealed that endogenous dsRNA structures can activate PKR as well as the dsRNA sensors melanoma differentiation-associated gene 5 (MDA-5) and retinoic acid-inducible gene I (RIG-I), thereby causing aberrant inflammatory responses. Hyperactive MDA-5 is associated with type-I interferonopathies such as Aicardi-Goutières syndrome (AGS) (Crow and Manel 2015). Likewise, a growing number of studies have linked self dsRNA-mediated PKR activation to the onset of neurological disorders such as Parkinson's disease, Alzheimer's disease, and Huntington's disease (Mohan et al. 2025). In cancer cells, PKR activation can lead to autoinflammation and cell death; however, this is efficiently suppressed in many tumors. Cells have evolved multiple redundant and synergistic mechanisms to prevent these autoinflammatory responses. One central regulatory enzyme, the IFN-inducible 150 kDa isoform of adenosine deaminase acting on RNA 1 (ADAR1 p150), suppresses both MDA-5 and PKR activation by competitively binding to dsRNA molecules, as well as by editing of dsRNAs, which permanently alters dsRNA secondary structures and removes immunostimulatory capacities (Fig. 1A; Li and Walkley 2025).
In their study, Young et al. find that many triple-negative breast cancer (TNBC) cell lines express high levels of another dsRNA binding protein, PACT (Young et al. 2025). PACT depletion in a subset of TNBC cell lines led to PKR activation and induced cell death, indicating that PACT acted as a PKR suppressor. However, a different subset of TNBC cell lines required simultaneous depletion of both PACT and ADAR1 for efficient PKR activation, suggesting that both mechanisms to suppress PKR are redundant in these cells. PACT possesses two functional dsRNAbinding domains (dsRBD1/2) and dsRBD3 that does not bind dsRNA but allows PACT to form homodimers. The authors show through mutational analyses that dsRNA binding and dimer formation are required for the suppressive activity of PACT on PKR under physiological conditions.
In two other simultaneous studies, PACT was identified as a suppressor of PKR activation against self-RNAs ( Ahmad et al. 2025;Manjunath et al. 2025). Manjunath et al. used a CRISPR-based screening method to identify genes necessary to activate PKR in the context of a viral infection. PACT was identified as a suppressor of PKR activation, and, remarkably, depletion of both PACT and ADAR1 led to synthetic lethality in uninfected cells dependent on PKR activation (Manjunath et al. 2025). Ahmad et al. provide detailed mechanistic insights into the mode of PKR suppression by PACT. While PACT does not prevent PKR binding to endogenous dsRNA substrates such as inverted Alu repeats, it prevents efficient scanning of PKR along longer dsRNA elements, a prerequisite for PKR activation (Ahmad et al. 2025). In addition, PACT dsRBD1/2 directly and independently of dsRNA interacted with the PKR kinase domain and served as a decoy substrate for phosphorylation, preventing PKR autophosphorylation.
Together, the findings of these three studies provide strong evidence for the suppressive activity of PACT on PKR activation. This is in stark contrast to PACT being originally described as activator of PKR (Patel and Sen 1998) but is in line with the observed rescue of embryonic lethality of Prkra knockout in mice by additional depletion of PKR (Dickerman et al. 2015). Key to these observed opposing functions seems to be the availability of dsRNA, as well as the concentration of PACT. In the absence of dsRNA, low amounts of PACT can directly interact with PKR, causing its autophosphorylation and activation; however, when dsRNA is present, or in the absence of dsRNA but high PACT concentrations, PACT suppresses PKR activation. Through the combined findings of all three studies, we are now able to establish a clear molecular mechanism of the suppressive function of PACT (Fig. 1B): PACT dimers occupy dsRNA molecules; while this does not per se prevent PKR binding to the same dsRNA molecules, it prevents sliding of PKR along the dsRNA scaffold, which is required for PKR dimerization, autophosphorylation, and hence activation. PACT is now joining this increasing list of PKR-controlling factors (Cottrell et al. 2024a). It is now clear that regulating PKR is a main priority for the cell. Understanding how these mechanisms function across different tissue types will provide significant insights into strategies aimed at manipulating these regulatory effectors for therapeutic development against cancers, such as TNBC, viral infection, and neurological disorders. However, the goal to manipulate PKR activity may not be easily achievable, as we may have to target multiple redundant controlling mechanisms in parallel.
## References
1. Ahmad, Zou, Hwang et al. (2025) "PACT prevents aberrant activation of PKR by endogenous dsRNA without sequestration" *Nat Commun*
2. Bendavid, Pfaller, Pan et al. (2022) "Host 5 ′ -3 ′ exoribonuclease XRN1 acts as a proviral factor for measles virus replication by downregulating the dsRNA-activated kinase PKR" *J Virol*
3. Bendavid, Yang, Zhou et al. (2024) "Host WD repeat-containing protein 5 inhibits protein kinase R-mediated integrated stress response during measles virus infection" *J Virol*
4. Cottrell, Andrews, Bass "2024a. The competitive landscape of the dsRNA world" *Mol Cell*
5. Cottrell, Ryu, Pierce et al. "2024b. Induction of viral mimicry upon loss of DHX9 and ADAR1 in breast cancer cells" *Cancer Res Commun*
6. Crow, Manel (2015) "Aicardi-Goutieres syndrome and the type I interferonopathies" *Nat Rev Immunol*
7. Junior, Sampaio, Rehwinkel (2019) "A balancing act: MDA5 in antiviral immunity and autoinflammation" *Trends Microbiol*
8. Dickerman, White, Kessler et al. (2015) "The protein activator of protein kinase R, PACT/RAX, negatively regulates protein kinase R during mouse anterior pituitary development" *FEBS J*
9. Li, Walkley (2025) "Leveraging genetics to understand ADAR1mediated RNA editing in health and disease" *Nat Rev Genet*
10. Manjunath, Santiago, Ortega et al. (2025) "Cooperative role of PACT and ADAR1 in preventing aberrant PKR activation by selfderived double-stranded RNA" *Nat Commun*
11. Mohan, Mannan, Singh (2025) "Unravelling the role of protein kinase R (PKR) in neurodegenerative disease: a review" *Mol Biol Rep*
12. Patel, Sen (1998) "PACT, a protein activator of the interferon-induced protein kinase, PKR" *EMBO J*
13. Pfaller, Li, George et al. (2011) "Protein kinase PKR and RNA adenosine deaminase ADAR1: new roles for old players as modulators of the interferon response" *Curr Opin Immunol*
14. Young, Juhler, Pierce et al. (2025) "PACT suppresses PKR activation through dsRNA binding and dimerization, and is a therapeutic target for triple-negative breast cancer" *RNA* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12775852&blobtype=pdf | # A monoclonal antibody W1 blocks mesothelin-mediated tumor progression
Qingguang Wang, Guangcan Cao, Mingxin Li, Rui Gong
## Abstract
Mesothelin (MSLN) represents an attractive and potentialbiomarker for targeted cancer therapy. Although MSLN has been implicated in mediating tumorigenesis, its precise physiological function remains incompletely elucidated. Moreover, only a limited number of antibodies have been reported to directly inhibit its function in tumor progression. In this study, we confirmed that the N-terminal domain 1 (D1) of MSLN serves as the key region responsible for mediating binding to tumor cells and promoting migration and invasion through the activation of the matrix metalloproteinase-7 (MMP7) pathway. We further identified two monoclonal antibodies, W1 and A12H, which specifically bind to D1 and domain 3 (D3) of MSLN, respectively. W1 effectively inhibits the interaction between D1 and tumor cells, thereby suppressing cell migration and invasion via interference with MSLNmediated signaling pathways. In contrast, A12H, which targets D3, does not exhibit a similar functional blockade. Additionally, W1 significantly reduces the tumor burden in mice bearing MSLN-positive tumor cells. Together, these findings provide mechanistic insights into the role of D1 in driving tumor progression and reinforce the potential of MSLN as a therapeutic biomarker. The antibody W1 emerges as a promising candidate, offering novel perspectives for developing immunotherapeutic strategies against MSLN-related cancers.
## INTRODUCTION
Mesothelin (MSLN) is a glycosylphosphatidylinositol (GPI)-anchored glycoprotein with an approximate molecular weight of 40 kDa, processed from a precursor protein through proteolytic cleavage by Furin. 1 Additionally, MSLN can be cleaved from the cell surface by various proteases, resulting in a soluble form that enters the systemic circulation. 2 While MSLN expression is relatively restricted in normal tissues-primarily limited to mesothelial cells of the pleura, pericardium, and peritoneum-it is significantly upregulated in a variety of malignancies. These include mesothelioma, ovarian cancer, pancreatic ductal adenocarcinoma, triple-negative breast cancer, non-small cell lung cancer, and other solid tumors, [3][4][5] as well as acute myeloid leukemia. 6 This renders MSLN a valuable tumor marker.
Epithelial-mesenchymal transition (EMT) is widely recognized as an initial step in cancer metastasis. 7 This process involves significant alterations in the expression of multiple molecules, such as the downre-gulation of E-cadherin and the upregulation of mesenchymal factors like Snail and Twist, as well as various matrix metalloproteinases (MMPs). 5,8 It is also associated with the activation of several signaling pathways, including the Wnt and Notch signaling pathways. 8 MSLN contributes to EMT by upregulating mesenchymal markers and enhancing the migratory and invasive capabilities of MSLN-positive tumor cells, facilitating metastasis to distant organs. 5,9 According to its structure, MSLN can be divided into three domains: domain I (D1), domain II (D2), and domain III (D3). 10,11 These functional domains of MSLN remain an active area of investigation. D1 has been implicated in receptor binding and may represent a key functional domain. 10 However, its precise role is not fully elucidated, and the functions of D2 and D3 are even less understood. To date, the primary identified receptor for MSLN is cancer antigen 125 (CA125, also known as MUC16). 12,13 MSLN binds with high affinity (5-10 nM) to multiple sites within the sea urchin sperm protein, enterokinase, and the agrin (SEA) 10 domain of CA125. 14 Nevertheless, extensive glycosylation of CA125 has complicated the full structural characterization of this interaction, leaving the complete spectrum of binding sites between MSLN and CA125 unresolved. Through receptor engagement, MSLN activates multiple downstream signaling pathways, such as wingless-related integration site (Wnt), protein kinase B (AKT), and mitogen-activated protein kinase (MAPK), 3,15 which in turn stimulate the expression of various MMPs and promote tumor progression. [16][17][18] Additionally, MSLN has been demonstrated to regulate cholestatic liver fibrosis via the TGF-β1 pathway, 19 highlighting its broader role in fibrotic disease and cancer metastasis. MSLN-targeted immunotherapies represent a rapidly advancing frontier in preclinical oncology research, currently encompassing a range of strategies. These include direct effector mechanisms such as antibody-dependent cellular cytotoxicity (ADCC) mediated by monoclonal antibodies (e.g., m912 20 with a half maximal effective concentration [EC 50 ] of 20 nM for MSLN), T cell-mediated killing by bispecific antibodies (e.g., M9657 [MSLN×CD137)] 21 with a dissociation constant (KD) of 0.34 nM for MSLN, and BsAb5 [MSLN×CD3] 22 with a KD of 2 nM for MSLN), and antibody-drug conjugates (e.g., 3C9 23 with a KD of 2 nM, and NAV-001 24 with a KD of 70 pM). More advanced cellular immunotherapies include chimeric antigen receptor T-cells (CAR-Ts), such as M1xx 25 and h15B6. 26 While showing promising affinity and antitumor activity in preclinical models, the majority of these MSLN-targeted candidates are still in early development, and their successful translation into clinical trials demands further validation. This underscores the critical need to elucidate the mechanistic contributions of MSLN and to develop function-blocking monoclonal antibodies capable of inhibiting signaling pathways, which is a prerequisite for creating novel therapeutic strategies against MSLN-associated cancers.
In this study, we demonstrated that D1 of MSLN promotes tumor cell migration and invasion by upregulating MMP7 via activation of the AKT signaling pathway. We further identified a high-affinity monoclonal antibody, W1, with a KD of 3.88 pM, which specifically binds to D1 and effectively blocks MSLN-mediated signaling pathways. Treatment with W1 suppressed migration and invasion in MSLN-positive ovarian cancer cells and reduced tumor burden in mouse xenograft models. These findings not only deepen the understanding of the role of MSLN in oncogenesis and underscore D1 as a precise immunotherapeutic target but also provide new insights for the treatment of MSLN-related cancers.
## RESULTS
## Identification of the binding region in MSLN to its receptor
It has been established that region I of MSLN is responsible for binding to its receptor on OVCAR3 cells. To further elucidate the MSLNmediated signaling pathway, we divided the mature MSLN protein into distinct domains. 10 Specifically, the segments spanning amino acid (aa) residues 296-390 (D1), 391-486 (D2), and 487-581 (D3) were individually fused to the Fc fragment of human immunoglobulin G1 (IgG1) for expression and purification (Figure 1A; Figure S1). Strong binding of full-length MSLN and D1 to OVCAR3 cells was observed (Figure 1B), while the binding to SKOV3 cells was moderate (Figure 1C). In contrast, the binding of D2 and D3 to these two cell lines was negligible. No binding of MSLN or any of the three domains (D1, D2, and D3) was detected in PANC-1 cells (CA125-negative 17 ) (Figure 1D). These results indicate that D1 is likely the primary region mediating MSLN binding to its receptor.
## D1 promotes cell migration and invasion but not proliferation
To further investigate the functional role of MSLN, we generated MSLN-knockout OVCAR3 cells (OVCAR3 MSLN KO ) to abolish endogenous MSLN expression (Figure S2). Compared to parental OVCAR3 cells, OVCAR3 MSLN KO cells exhibited significantly reduced migration (Figure 1E) and invasion (Figure 1F) capabilities. The addition of soluble full-length MSLN or its individual domains did not enhance the proliferation of OVCAR3 MSLN KO or SKOV3 cells (Figure S3). However, wound-healing assays revealed that the migratory capacity of OVCAR3 MSLN KO cells was enhanced in a concentration-dependent manner upon treatment with MSLN-Fc (Figure S4). In subsequent comparative experiments, significant recovery of the scratch was observed in OVCAR3 MSLN KO (Figure 2A) and SKOV3 (Figure 2B) cells treated with either MSLN-Fc or D1-Fc, whereas D2 and D3 showed no notable effect. Similarly, both full-length MSLN and D1 markedly stimulated the invasion of OVCAR3 MSLN KO (Figure 2C) and SKOV3 (Figure 2D) cells. Collectively, these findings demonstrate that MSLN promotes migration and invasion in ovarian cancer cells primarily through its D1.
## D1 promotes cell migration and invasion through the MMP7mediated signal pathway
According to previous reports, [16][17][18]27 MMP7 and MMP9 are implicated in facilitating migration and invasion across multiple cancer types, including pancreatic, ovarian, and breast cancer, as well as mesothelioma. To elucidate the mechanism by which MSLN promotes cell migration and invasion, we examined the activation of MSLN-mediated signaling by measuring Mmp7 and Mmp9 mRNA expression levels. Using RT-qPCR, we observed that in OVCAR3 MSLN KO cells, the mRNA level of Mmp7 increased in a time-dependent manner upon treatment with 100 nM MSLN-Fc (Figure 2E) and also rose with increasing concentrations of MSLN-Fc after 24 h of incubation (Figure 2F). In contrast, Mmp9 expression remained largely unchanged. These findings suggest that MSLN may enhance migration and invasion in OVCAR3 cells primarily through activation of the MMP7 pathway rather than the MMP9 pathway. This result was further corroborated at the protein level by western blotting, which showed that MSLN upregulates MMP7 expression in a dose-dependent manner (Figure 2G). Subsequent experiments treating OVCAR3 MSLN KO cells with MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, and Fc revealed that both full-length MSLN and D1 markedly enhanced AKT phosphorylation and downstream MMP7 expression, while concurrently reducing E-cadherin levels (Figure 2H). Similar results were observed in SKOV3 cells (Figure 2I). Collectively, these results indicate that MSLN, particularly through its D1, activates AKT phosphorylation and MMP7 expression to promote cell migration and invasion.
## Figure 1. Functional validation of MSLN domains
(A) Schematic representation of MSLN domain truncations. Domain 1 (D1, aa 296-390), domain 2 (D2, aa 391-486), and domain 3 (D3, aa 487-581) were individually fused to the human IgG1 Fc tag for recombinant expression. Binding activity of MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, and Fc control to human ovarian cancer cell lines OVCAR3 (B) and SKOV3 (C), as well as the CA125-negative pancreatic cancer cell line PANC-1 (D), was assessed by flow cytometry. Comparative analysis of migration (E) and invasion (F) capabilities between OVCAR3 and OVCAR3 MSLN KO cells. Data were normalized to the OVCAR3 group and are presented as mean ± SD from three biologically independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001; data were analyzed by ANOVA followed by Tukey's multiple comparison test.
## Identification of monoclonal antibodies W1 and A12H targeting D1 and D3
To identify monoclonal antibodies capable of targeting functional domains of MSLN, such as D1, mice were immunized three times with MSLN-Fc (Figure S5A). Following B cell sorting via fluorescence-activated cell sorting (FACS) (Figure S5B), a phage display library was constructed. Through successive rounds of panning against MSLN-Fc and OVCAR3 cells, two Fabs were isolated: W1 Fab, which targets D1, and A12H Fab, directed against D3. The W1 Fab was subsequently converted into a mouse IgG1κ-type antibody (W1 mIgG1), while A12H was engineered into a chimeric mouse/human antibody by grafting onto a human IgG1 and kappa constant region (A12H cIgG1). Both antibodies were produced using the HEK293F expression system (Figure S5C).
Both W1 mIgG1 and A12H cIgG1 exhibited strong binding to MSLN-Fc, with EC 50 values of 48 pM and 166 pM, respectively. W1 mIgG1 specifically recognized D1 (EC 50 = 160 pM), while A12H cIgG1 bound specifically to D3 (EC 50 = 3.12 nM) (Figure 3A). Surface plasmon resonance (SPR) analysis further confirmed the high binding affinity of both antibodies toward MSLN, with KD values of 3.88 pM for W1 mIgG1 and 64.1 pM for A12H cIgG1 (Figure 3B). To further evaluate binding specificity, MSLN-overexpressing A431 cells (A431-MSLN + ) were generated (Figure S6). Cell-based binding assays demonstrated that both antibodies strongly bound to MSLN-positive cell lines, including OVCAR3 (high MSLN expression 28 ), SKOV3 (moderate expression 28 ), and A431-MSLN + , but not to negative controls such as OVCAR3 MSLN KO and A431 (MSLN-negative cells 29 ) (Figure 3C). Quantitative analysis revealed EC 50 values of 0.3 nM for W1 mIgG1 and 0.65 nM for A12H cIgG1 when binding to OVCAR3 cells (Figure 3D). Together, these results indicate that both W1 mIgG1 and A12H cIgG1 display high affinity and specific binding to cell surface-expressed MSLN.
## W1 mIgG1 blocks the binding of MSLN to its receptor on cells
Treatment with W1 mIgG1 effectively inhibited the binding of both full-length MSLN and its D1 domain to the receptor on OVCAR3 MSLN KO cells. In contrast, A12H cIgG1 showed no such inhibitory activity (Figure 3E). To further assess whether these antibodies could disrupt the interaction between membrane-anchored MSLN and its cellular receptor, we generated a luciferase-expressing A431 cell line stably overexpressing MSLN (A431-MSLN + -luciferase). 30 When A431-MSLN + -luciferase cells were co-cultured with immobilized OVCAR3 cells, the adhesion between the two cell types was progressively inhibited with increasing concentrations of W1 mIgG1 (Figure 3F), yielding a half maximal inhibitory concentration (IC 50 ) of approximately 1.8 nM. Conversely, A12H cIgG1 had no detectable effect on cell-cell adhesion. These results demonstrate that W1 mIgG1 specifically blocks the interaction between MSLN and its receptor and suppresses MSLN-mediated cell adhesion by targeting D1.
## W1 mIgG1 inhibits cell migration and invasion via the MMP7mediated signal pathway
To further explore the role of W1 mIgG1, wound-healing and invasion assays were performed. In both OVCAR3 MSLN KO and SKOV3 cells, treatment with 300 nM W1 mIgG1 effectively suppressed the migration and invasion induced by 100 nM exogenous MSLN-Fc or D1-Fc after 48 h of incubation. In contrast, 300 nM A12H cIgG1 showed no inhibitory effect. The addition of these two antibodies exerted no substantial influence on PANC-1 cells (Figures 4A and4B). Furthermore, the presence of W1 mIgG1 led to a decrease in MMP7 expression and a recovery of E-cadherin levels that had been altered by exogenous MSLN-Fc or D1-Fc stimulation in both cell lines (Figure 4C). W1 mIgG1 also inhibited the migration and invasion of native OVCAR3 cells in a dose-dependent manner (at concentrations of 100 nM, 300 nM, and 600 nM), whereas A12H cIgG1 again exhibited no such activity (Figures 5A and5B). Consistent with these findings, W1 mIgG1 treatment resulted in a dose-dependent downregulation of Mmp7 at both mRNA and protein levels, which was not observed with A12H cIgG1 (Figures 5C and5D). To further assess whether W1 acts through inhibition of the MSLN-induced MMP7 pathway, we used a selective MMP inhibitor (MMPi, Marimastat, 1,000 nM). Both W1 mIgG1 and MMPi significantly inhibited MSLN-enhanced migration and invasion in OVCAR3 MSLN KO cells (Figures 5E and5F). These results suggest that W1 mIgG1 suppresses MSLNand D1-mediated cell migration and invasion primarily by blocking the MMP7 signaling pathway.
## W1 mIgG1 reduces the tumor burden of MSLN-positive tumors in vivo
To simulate the process of tumor dissemination, we established an OVCAR3-luciferase cell line and administered intraperitoneal (i.p.) injections of these cells into BALB/c-nu mice, along with W1 mIgG1 at doses of 10 mg/kg and 25 mg/kg, or PBS as a control. Tumor burden within the abdominal cavity was monitored over time (Figure 6A). In the PBS-treated group, OVCAR3derived tumors progressively increased in burden and exhibited widespread dissemination. In contrast, both the 10 mg/kg and 25 mg/kg W1 mIgG1 treatment groups showed significant tumor regression, as quantified by bioluminescence signal analysis (Figure 6B). By Week 11, three of six mice in the 10 mg/kg W1 mIgG1 group displayed marked tumor remission, while four of six mice in the 25 mg/kg group had undetectable tumor bioluminescence signals compared to the PBS group. Dissection of peritoneal organs further confirmed these results: strong bioluminescence signals were observed in PBS-treated mice, whereas only minimal signals were detected in the 10 mg/kg W1 mIgG1 group, and no signals were visible in the 25 mg/kg W1 mIgG1 group (Figure 6C).
In the therapeutic model, OVCAR3-luciferase cells were first injected i.p. into BALB/c-nu mice to allow tumor establishment. W1 mIgG1 (10 mg/kg or 25 mg/kg) was then administered at weeks 1, 2, and 3, with bioluminescence imaging performed periodically for observation and statistical analysis (Figure 6D). Results indicated that tumors in the PBS group (n = 6) grew progressively and disseminated over time. In the W1 10 mg/kg group (n = 6), most tumors regressed, and in the W1 25 mg/kg group (n = 6), tumors were nearly completely eradicated (Figure 6E). Histological examination via hematoxylin and eosin (H&E) staining revealed normal intestinal tissue (negative control) architecture with uniform cell morphology and no signs of atypia. In contrast, tumor masses showed high cellular density, disrupted tissue organization, and evident cellular atypia, clearly distinguishing them from normal tissues (Figure 6F). Subsequent immunohistochemical analysis indicated higher expression of MMP7 and lower expression of E-cadherin in tumor tissues compared to normal intestinal tissues, where the opposite pattern was observed (Figure 6G). These findings suggest that OVCAR3-derived tumors promote malignancy primarily through MMP7-mediated mechanisms and highlight the efficacy of W1 mIgG1 in reducing tumor burden in vivo, underscoring its translational potential.
## Analysis of the principal sites for binding of MSLN to its receptor and W1
Since D1 of MSLN is a functional region mediating receptor binding and is specifically recognized by the antibody W1, we sought to identify the precise binding sites involved. We first performed structural simulations using AlphaFold3 31 to predict the putative binding regions of W1 mIgG1 and A12H cIgG1 on MSLN (Figure 7A). Consistent with previous findings, W1 was predicted to bind to D1, whereas A12H primarily interacted with D3. To further delineate the binding epitope within D1, we generated a series of truncated variants: D1-1 (aa 296-329), MSLN ΔD1-middle-1 (Δaa 312-581), MSLN ΔD1-1 (Δaa 329-581), and MSLN ΔD1-1-2 (Δaa 359-581), each fused to the human IgG1 Fc fragment (Figure 7B). Binding assays revealed that D1-1-Fc retained strong binding affinity to W1 mIgG1, comparable to fulllength MSLN-Fc and D1-Fc, while the other truncations showed markedly reduced binding (Figure 7C). In contrast, A12H cIgG1 recognized all MSLN variant-Fc proteins except D1-1-Fc. Furthermore, D1-1-Fc also bound effectively to the receptor on OVCAR3 cells, similar to MSLN-Fc (Figure 7D), indicating that this minimal domain contains key residues essential for both receptor engagement and antibody recognition. We next performed alanine scanning mutagenesis within the D1-1 region to identify critical binding residues. The D312A mutation completely abolished MSLN binding to W1 mIgG1, and the K319A mutation also significantly reduced binding (Figure 7E; Figure S7A). In addition, E313A and F317A markedly impaired MSLN binding to its receptor (Figure 7F; Figure S7B). These findings align with previously reported key binding sites for MSLN-CA125 interaction, 11,14 which include E313, F317, I316, R338, N340, A341, and P343, among others. Collectively, these results suggest that W1 mIgG1 could inhibit the interaction between MSLN and its receptor through steric hindrance, targeting critical functional residues within the D1-1 subdomain (Figure 7G).
## DISCUSSION
Prior research has established that D1 of MSLN mediates its binding to the receptor. 10 However, the specific contribution of the D1 domain to MSLN-mediated tumor progression remains to be fully elucidated. In this study, we did not observe a significant promotive effect of MSLN on tumor cell proliferation, a finding that remains controversial according to results from different reports. 28,32,33 It is well-documented that MSLN plays a role in the EMT process 5 and activates signaling pathways such as PI3K-AKT 34 and MAPK-ERK. 16 Consistent with these mechanisms, our results demonstrate that D1 recapitulates key functions of full-length MSLN: it induces AKT phosphorylation, upregulates MMP7, and downregulates E-cadherin expression. Collectively, these changes enhance the migratory and invasive capacities of tumor cells. Our findings provide deeper insight into MSLN-driven oncogenesis and highlight D1 as a promising and specific target for immunotherapeutic strategies against MSLN-positive cancers.
A number of marketed monoclonal antibodies function by blocking signaling pathways implicated in tumorigenesis. For instance, trastuzumab inhibits the HER2-mediated PI3K/AKT pathway, suppressing tumor cell proliferation. 35 These antibodies impede tumor immune evasion, proliferation, and metastasis by interfering with the corresponding cellular pathways. Consequently, developing monoclonal antibodies capable of inhibiting MSLN-mediated signaling represents a promising therapeutic direction. It has been reported that the monoclonal antibody amatuximab targets the MSLN-mediated p-MET signaling pathway. 36,37 Additionally, it has demonstrated excellent affinity (KD of 1.5 nM) and clinical potential, having completed phase 2 trials. Our study has demonstrated that the W1 antibody exhibits high affinity (KD of 3.88 pM) and suppresses tumor migration and invasion by blocking MSLN-receptor interaction, downregulating MMP7 expression, and disrupting downstream signaling pathways. These findings suggest that W1 may hold potential for achieving superior clinical outcomes in controlling tumor metastasis and dissemination. Furthermore, W1 mIgG1 could be engineered into antibody-based derivatives, such as ADCs, that simultaneously inhibit oncogenic signaling pathways and directly kill tumor cells, analogous to the mechanism of disitamab vedotin. 38 Such strategies may enhance antitumor efficacy in MSLN-positive cancers. Our results not only contribute to a deeper understanding of the role of MSLN in tumor progression but may also support the development of novel therapeutic strategies for MSLN-associated malignancies. Molecular Therapy: Oncology
## MATERIALS AND METHODS
## MSLN and truncated antigen construction, expression, purification, and SDS-PAGE
All MSLN-related antigens (MSLN, D1, D2, D3, D1-1, MSLN ΔD1-middle-1 , MSLN ΔD1-1 , and MSLN ΔD1-1-2 ) were cloned into pSecTag2A expression vectors, with a human IgG Fc fragment at the C terminus. In addition, the DNA sequences of W1 mIgG1 and A12H cIgG1 were cloned into pVitro expression vectors. HEK293F cells were transfected with plasmids using polyethyleneimine (PEI-25 kDa, Polysciences). After 5 days of culture, the supernatant was collected, and the soluble proteins were purified using protein A/G resin (GE Healthcare/Genscript). The target proteins were analyzed by SDS-PAGE under both reducing and nonreducing loading conditions to evaluate protein bands.
## Cell culture
OVCAR3 cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and 1:1000 mycoplasma prevention reagent (Beyotime, cat#FM5-01), while SKOV3 and A431 cells were cultured in DMEM with 10% FBS and 1:1000 mycoplasma prevention reagent (Beyotime, cat#FM5-01). FreeStyle 293-F cells (cat#R79007) were obtained from Thermo Fisher Scientific and maintained in Freestyle 293 expression medium (Thermo Fisher Scientific) at 37 • C in an 8% CO 2 atmosphere.
## MSLN KO cell line generation
The design of gRNAs for CRISPR-Cas9 and vector cloning is described in a previous article. 39 Cas9-and gRNA-containing vectors (lenti-CRISPR v2) were delivered into OVCAR3 cells via lentiviral infection. MSLN KO cell populations were selected using puromycin (1 μg/mL). Expanded single-cell clones lacking MSLN were identified and sorted by FACS (BD Melody). The lentiCRISPR v2 vector lacking gRNA was used to generate a mock cell line control as described above.
## Establishment of stably transduced cell lines
MSLN was cloned into pCDNA3.1+ and transfected into A431 cells, followed by selection with G418 (600 μg/mL). Single cells were then sorted by FACS (BD FACSMelody cell sorter) to establish the A431-MSLN + cell line. Luciferase was introduced into A431-MSLN + or OVCAR3 cells using the lentiCRISPR v2 vector, accompanied by puromycin selection for 1 week, to generate stably expressing A431-MSLN + -luciferase and OVCAR3-luciferase cell lines.
## Flow cytometry
Samples were analyzed on a BD LSRFortessa and sorted by FACS (BD FACSMelody cell sorter). The general procedure involves di-gesting the sample into a single-cell suspension, incubating it with the primary antibody at 4 • C for 1 h, and then washing it three times with PBS. The secondary antibody with fluorescent labeling is then added and incubated at 4 • C for 1 h, followed by three washes with PBS. Finally, flow cytometry analysis or sorting is performed. Required reagents: anti-MSLN antibody (Abcam, cat#ab196235, 1:400 dilution), CoraLite488-conjugated Affinipure Goat-anti-Rabbit IgG (H + L) (Proteintech, cat#SA00013-2, 1:400 dilution), PE anti-mouse CD19 (BioLegend, cat#115507, 1:400 dilution), CoraLite488-conjugated Goat-anti-Mouse IgG (H + L) (Proteintech, cat#SA00013-1, 1:400 dilution), Goat-anti-Human IgG Fc Cross-Adsorbed Secondary Antibody, DyLightTM 650 (Thermo Fisher Scientific, cat#SA5-10137, 1:400 dilution).
## Cell viability assay
OVCAR3 MSLN KO and SKOV3 cells were seeded in triplicate wells (2 × 10 4 cells/well) onto 96-well plates and cultured in complete medium overnight, followed by treatment with MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, Fc (100 nM), or PBS for 0, 24, and 48 h. Cell viability was then determined using a CCK8 assay according to the manufacturer's instructions.
## Wound-healing assay
The migratory ability of the cells was assessed using a wound-healing assay. OVCAR3, OVCAR3 MSLN KO , SKOV3, and PANC-1 cells were seeded and cultured in 12-well plates to form compact monolayers following overnight incubation. Cells were starved in serum-free medium for 8 h, and scratches were made in the wells using a pipette tip. Therefore, a medium containing 2% FBS with MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, Fc (100 nM), W1 mIgG1 or A12H cIgG1 (300 nM), MMPi (1,000 nM, MCE, cat#HY-12169), or PBS was added. Cells were imaged using an inverted microscope at 0, 24, 48, and 72 h, and the migration area was calculated using ImageJ.
## Transwell invasion assay
OVCAR3, OVCAR3 MSLN KO , SKOV3, and PANC-1 cells treated with MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, Fc (100 nM), W1 mIgG1 or A12H cIgG1 (300 nM), MMPi (1,000 nM), or PBS were seeded on 24-well Transwell inserts (8 mm pore size) pre-coated with matrix gel (Beyotime, cat#C0383-5mL) in serum-free medium, whereas the bottom chamber contained medium with 10% FBS. After 24 or 48 h of incubation, cells on the upper side of the inserts were removed with cotton-tipped swabs, and cells that had invaded through the matrix gel were fixed with 4% paraformaldehyde and stained with crystal violet. Images of the invaded cells were captured, and 3 fields/wells were counted using an inverted microscope.
## RT-qPCR
RNA was extracted from different groups of cells using a total RNA kit I (Omega, cat#R6834-01), and cDNA was obtained through reverse transcription using HiScript III All-in-one RT SuperMix Perfect for qPCR (Vazyme, cat#R333-01). Subsequently, qPCR was conducted using Taq Pro Universal SYBR qPCR Master Mix (Vazyme, cat#Q712-02). The following primers were used: Gapdh,
$$For: 5 ′ -CCATGTTCGTCATGGGTGTGAACCA-3 ′ Rev: 5 ′ -GCCAGTAGAGGCAGGGATGATGTTC-3 ′ ; Mmp7, For: 5 ′ -GTCTCTGGACGGCAGCTATG-3 ′ Rev: 5 ′ -GATAGTCCTGAGCCTGTTCCC-3 ′ ; Mmp9, For: 5 ′ -CCGGACCAAGGATACAGTT-3 ′ Rev: 5 ′ -CGGCACTGAGGAATGATCTA-3 ′ .$$
## Western blotting
After washing with PBS, RIPA buffer containing 1% PMSF (Beyotime, cat#ST506) was added to OVCAR3, OVCAR3 MSLN KO , and SKOV3 cells to extract total protein. Following immunoblotting, polyvinylidene fluoride (PVDF) was incubated with primary antibodies at 4 • C overnight and with secondary antibodies at room temperature for 1 h. The required antibodies are as follows: Horseradish peroxidase (HRP)-conjugated β-actin Rabbit mAb (ABclonal, cat#AC028, 1:6000 dilution), Rb mAb to MMP7 (Abcam, cat# AB176325, 1:1000 dilution), Phospho-AKT (Ser473) Monoclonal antibody (Proteintech, cat# 66444-1-Ig, 1:6000), AKT Monoclonal antibody (Proteintech, cat# 60203-2-Ig, 1:6000), E-cadherin Monoclonal antibody (Proteintech, cat# 60335-1-Ig, 1:6000), Goat anti-Rabbit IgG (H + L) Secondary Antibody, HRP (Invitrogen, cat#C31460100, 1:15000 dilution).
## Enzyme-linked immunosorbent assay (ELISA)
High-binding 96-well plates (Corning) were coated with MSLN-Fc, D1-Fc, D2-Fc, D3-Fc, and Fc at 4 μg/mL and incubated overnight at 4 • C. The plates were then blocked with PBS containing 3% skim milk (w/v, Bio-Rad) at 37 • C for 1 h. After washing three times with PBST, serially diluted antibodies were added, and the plates were incubated at 37 • C for 1.5 h. The plates were washed five times with PBST and HRP-conjugated Affinipure Goat Anti-Mouse IgG (H + L) (Proteintech, cat#SA00001-1, 1:5000 dilution), and Anti-Human IgG (Fab specific)-Peroxidase antibody (Sigma-Aldrich, cat#A0293, 1:3000 dilution) was used as a secondary antibody. After 1 h of incubation, the plates were washed five times with PBST. Binding was measured with the subsequent addition of substrate diammonium 2,2-azinobis (3-ethylbenzothiazoline-6-sulfonate) (ABTS, Invitrogen), and absorbance at 405 nm was recorded using a Spark Multimode microplate reader (Tecan).
## SPR
SPR experiments were performed at 25 • C on a Biacore 1K instrument (Cytiva) using PBS as the running buffer. MSLN was immobilized, and serial 2-fold dilutions of W1 mIgG1 and A12H cIgG1 were prepared in running buffer at concentrations ranging from 10 mM to 0 nM, along with a corresponding PBS buffer volume-matched dilution series. Double-referenced data were analyzed using Biacore Evolution software, applying the closest blank for referencing and a steady-state 1:1 interaction model.
## Cell-cell adhesion
The A431-MSLN + -luciferase (1 × 10 5 cells/well) suspension was coincubated with OVCAR3 cells (6 × 10 4 cells/well) affixed to a 96-well plate at 37 • C for 1.5 h. A series of dilutions of W1 mIgG1 and A12H cIgG1 were added to the suspension to block binding between the two cell lines. The detection of Luminescence in the Spark Multimode microplate reader (Tecan) was conducted using a Firefly Luciferase Reporter Gene Assay Kit (Beyotime, cat#RG005).
## Establishment of i.p. xenograft models in nude mice
To generate i.p. xenografts, 4 × 10 6 OVCAR3-luciferase cells in 200 μL of PBS were i.p. injected into female BABL/c-nu/nu mice at 4 weeks of age. The mice were randomly divided into 3 groups: PBS, W1 mIgG1 10 mg/kg, and W1 mIgG1 25 mg/kg. The cells were incubated with OVCAR3-luciferase for 30 min at 4 • C. Following this, the cells were injected i.p. into mice. Bioluminescence signals were measured using the IVIS Lumina X5 in vivo imaging system (PerkinElmer) in a blinded manner, 10 min after i.p. injection of 100 μL of D-luciferin (Beyotime, cat#ST198) at 20 mg/mL, at different time points. At the end of the experimental period, mice were humanely euthanized and dissected to obtain peritoneal organs. All animal studies were approved by the Life Science Ethics Committee of the Wuhan Institute of Virology, Chinese Academy of Sciences. (approval no. WIVA34202204).
## Immunohistochemistry
Formalin-fixed, paraffin-embedded normal intestinal tissue and tumor tissue were obtained from the peritoneal cavity of mice in the PBS group. Tissue sections were incubated for 1 h at room temperature in a humidity chamber with primary antibodies, including MMP7 Rabbit pAb (Abclonal, cat# A0695, 1:100) and E-cadherin monoclonal antibody (Proteintech, cat# 60335-1-Ig, 1:2000). Sections were briefly stained with H&E.
## Alanine scanning mutagenesis
First, the structures of the antigen and antibodies were predicted using AlphaFold3.The predicted structures for MSLN, W1, and A12H were accessed at https://alphafoldserver.com. Alanine mutants of MSLN-Fc were designed based on the binding of truncated antigens to the receptor or antibodies. PCR-mediated site-directed mutagenesis was used to introduce mutations individually into DNAs encoding MSLN-Fc. The plasmids encoding the mutants were transfected into HEK293F cells, and proteins were purified using protein A resin as described earlier. The capacity of these mutations to bind to antibodies and the receptor was evaluated using ELISA and flow cytometry, respectively.
## Statistical analysis
All data were analyzed using Origin and are presented as mean ± SD. Statistical significance between groups was determined by ANOVA, followed by Tukey's multiple comparison test, wherever appropriate. In all figures, *p < 0.05; **p < 0.01; ***p < 0.001; n.s., no significance.
## References
1. Hassan, Bera, Pastan (2004) "Mesothelin: a new target for immunotherapy" *Clin. Cancer Res*
2. Liu, Chan, Tai et al. (2020) "Multiple proteases are involved in mesothelin shedding by cancer cells" *Commun. Biol*
3. Hagerty, Takabe (2023) "Biology of Mesothelin and Clinical Implications: A Review of Existing Literature" *World J. Oncol*
4. Lv, Li (2019) "Mesothelin as a biomarker for targeted therapy" *Biomark. Res*
5. He, Wang, Riedel et al. (2017) "Mesothelin promotes epithelial-to-mesenchymal transition and tumorigenicity of human lung cancer and mesothelioma cells" *Mol. Cancer*
6. Wang, Gong (2022) "Immunotherapy targeting mesothelin in acute myeloid leukemia" *J. Leukoc. Biol*
7. Gerstberger, Jiang, Ganesh (2023) *Metastasis. Cell*
8. Das, Bhattacharya, Chikkaputtaiah et al. (2019) "The basics of epithelial-mesenchymal transition (EMT): A study from a structure, dynamics, and functional perspective" *J. Cell. Physiol*
9. Xia, Duan, Xu et al. (2024) "Mesothelin promotes brain metastasis of non-small cell lung cancer by activating MET" *J. Exp. Clin. Cancer Res*
10. Kaneko, Gong, Zhang et al. (2009) "A Binding Domain on Mesothelin for CA125/MUC16" *J. Biol. Chem*
11. Zhan, Lin, Watson et al. (2023) "Structures of Cancer Antigen Mesothelin and Its Complexes with Therapeutic Antibodies" *Cancer Res. Commun*
12. Rump, Morikawa, Tanaka et al. (2004) "Binding of ovarian cancer antigen CA125/MUC16 to mesothelin mediates cell adhesion" *J. Biol. Chem*
13. Rao, Tian, Ma et al. (2015) "Expression of the Carboxy-Terminal Portion of MUC16/CA125 Induces Transformation and Tumor Invasion" *PLoS One*
14. Rupert, Buerger, Friend et al. (2024) "Structural elucidation of the mesothelin-mucin-16/CA125 interaction" *Structure*
15. Montemagno, Cassim, Pouyssegur et al. (2020) "From Malignant Progression to Therapeutic Targeting: Current Insights of Mesothelin in Pancreatic Ductal Adenocarcinoma" *Int. J. Mol. Sci*
16. Chang, Chen, Chen et al. (2012) "Mesothelin enhances invasion of ovarian cancer by inducing MMP-7 through MAPK/ERK and JNK pathways" *Biochem. J*
17. Chen, Hung, Wang et al. (2013) "Mesothelin binding to CA125/MUC16 promotes pancreatic cancer cell motility and invasion via MMP-7 activation" *Sci. Rep*
18. Servais, Colovos, Rodriguez et al. (2012) "Mesothelin overexpression promotes mesothelioma cell invasion and MMP-9 secretion in an orthotopic mouse model and in epithelioid pleural mesothelioma patients" *Clin. Cancer Res*
19. Koyama, Wang, Liang et al. (2017) "Mesothelin/mucin 16 signaling in activated portal fibroblasts regulates cholestatic liver fibrosis" *J. Clin. Investig*
20. Feng, Xiao, Zhu et al. (2009) "A novel human monoclonal antibody that binds with high affinity to mesothelinexpressing cells and kills them by antibody-dependent cell-mediated cytotoxicity" *Mol. Cancer Ther*
21. Xu, Zhou, Webb et al. (2024) "M9657 Is a Bispecific Tumor-Targeted Anti-CD137 Agonist That Induces MSLN-Dependent Antitumor Immunity without Liver Inflammation" *Cancer Immunol. Res*
22. Chakraborty, Onda, O'shea et al. (2024) "A bispecific antibody that targets the membrane-proximal region of mesothelin and retains high anticancer activity in the presence of shed mesothelin" *Mol. Cancer Ther*
23. Sun, Chu, Adams et al. (2023) "Preclinical assessment of a novel human antibody VH domain targeting mesothelin as an antibody-drug conjugate" *Mol. Ther. Oncolytics*
24. Nicolaides, Kline, Grasso et al. (2023) "NAV-001, a highefficacy antibody-drug conjugate targeting mesothelin with improved delivery of a potent payload by counteracting MUC16/CA125 inhibitory effects" *PLoS One*
25. Schoutrop, Poiret, El-Serafi et al. (2023) "Tuned activation of MSLN-CAR T cells induces superior antitumor responses in ovarian cancer models" *J. Immunother. Cancer*
26. Liu, Onda, Schlomer et al. (2024) "Tumor resistance to anti-mesothelin CAR-T cells caused by binding to shed mesothelin is overcome by targeting a juxtamembrane epitope" *Proc. Natl. Acad. Sci. USA*
27. Han, Zhou, Qu et al. (2018) "FOXC1-induced non-canonical WNT5A-MMP7 signaling regulates invasiveness in triple-negative breast cancer" *Oncogene*
28. Coelho, Ricardo, Amaral et al. (2020) "Regulation of invasion and peritoneal dissemination of ovarian cancer by mesothelin manipulation" *Oncogenesis*
29. Hassan, Ebel, Routhier et al. (2007) "Preclinical evaluation of MORAb-009, a chimeric antibody targeting tumor-associated mesothelin" *Cancer Immun*
30. Ho, Hassan, Zhang et al. (2005) "Humoral Immune Response to Mesothelin in Mesothelioma and Ovarian Cancer Patients" *Clin. Cancer Res*
31. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
32. Chang, Chen, Hsieh et al. (2009) "Mesothelin inhibits paclitaxel-induced apoptosis through the PI3K pathway" *Biochem. J*
33. Zheng, Jia, Tang et al. (2012) "Mesothelin regulates growth and apoptosis in pancreatic cancer cells through p53-dependent and -independent signal pathway" *J. Exp. Clin. Cancer Res*
34. Bharadwaj, Marin-Muller, Li et al. (2011) "Mesothelin confers pancreatic cancer cell resistance to TNF-α-induced apoptosis through Akt/ PI3K/NF-κB activation and IL-6/Mcl-1 overexpression" *Mol. Cancer*
35. Maadi, Soheilifar, Choi et al. (2021) "Trastuzumab Mechanism of Action; 20 Years of Research to Unravel a Dilemma"
36. Fujisaka, Kurata, Tanaka et al. (2015) "Phase I study of amatuximab, a novel monoclonal antibody to mesothelin, in Japanese patients with advanced solid tumors" *Invest. New Drugs*
37. Matsuzawa, Kamachi, Mizukami et al. (0200) "Mesothelin blockage by Amatuximab suppresses cell invasiveness, enhances gemcitabine sensitivity and regulates cancer cell stemness in mesothelin-positive pancreatic cancer cells" *BMC Cancer*
38. Wu, Xu, Li et al. (2023) "A HER2-targeting antibody-MMAE conjugate RC48 sensitizes immunotherapy in HER2-positive colon cancer by triggering the cGAS-STING pathway" *Cell Death Dis*
39. Avula, Rudloff, El-Behaedi et al. (2020) "Mesothelin Enhances Tumor Vascularity in Newly Forming Pancreatic Peritoneal Metastases" *Mol. Cancer Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12670952&blobtype=pdf | # Recurrent vaccine-strain varicella zoster virus reactivation in a child with acute lymphatic leukemia
Antonia Leutert, Aida Zeckanovic, Michael Huber, Patrick Meyer Sauteur, Raphael Morscher
## Abstract
This case illustrates recurrent herpes zoster (HZ) in a child with acute lymphatic leukemia. Interestingly, vaccinestrain HZ was confirmed by identifying the live-attenuated Oka vaccine strain (vOka) using metagenomic sequencing and sequence comparison at three loci that distinguish vOka from wild-type varicella zoster virus (VZV). Although vaccine-strain HZ is generally milder than HZ caused by wild-type VZV, prompt recognition and initiation of antiviral treatment is essential in immunocompromised patients, as fatal varicella due to disseminated vaccine-strain VZV has been reported in this high risk group.
An 11-month-old child presented in good general condition with a rapidly progressing vesiculopapular rash on her right thigh. One month prior to this episode the patient had been diagnosed with infant B-cell precursor acute lymphoblastic leukemia (B-ALL). Intravenous acyclovir was initiated (1.5 g/m2/d) with the differential diagnosis of herpes simplex virus or varicella zoster virus (VZV) infection. The rash expanded over the next four days to form plaques involving the whole upper leg and the knee (Fig. 1).
VZV was detected by specific polymerase chain reaction (PCR) from vesical fluid and specific IgM and IgG antibodies in serum. A VZVspecific PCR in blood was negative. Upon continuous formation of new lesions, intravenous immunoglobulin (0.5 g/kg) was administered on day six. No new vesicles appeared after day seven. Intravenous acyclovir was administered for a total of ten days. Chemotherapy was reconvened on day 12.
Interestingly, the medical history revealed that the patient had been vaccinated with VZV at the site of the initial appearance of vesicles at the age of 9 months, according to the Swiss vaccination schedule. In Switzerland, vaccination against varicella (chickenpox) is routinely recommended for all infants, with two doses administered at 9 and 12 months of age [1]. The parents had also retrospectively reported that a generalized rash developed 10 days after vaccination.
Sequencing of the VZV strain revealed the presence of the liveattenuated Oka vaccine strain (vOka), confirming the diagnosis of vaccine-strain herpes zoster (HZ). Over the following two years, a total of seven HZ reactivations occurred, varying in severity.
## Discussion
VZV is an exclusively human herpes virus that causes chickenpox (varicella), becomes latent in cranial-nerve and dorsal-root ganglia, and frequently reactivates decades later to produce HZ [2]. HZ typically manifests as unilateral radicular pain and a vesicular rash that is generally limited to the dermatome innervated by a single dorsal root or cranial nerve ganglion [3].
HZ in children is relatively rare. Hope-Simpson estimated an incidence rate of 74 per 100,000 persons per year among children younger than 10 years of age. This rate increases substantially with age, reaching 1010 per 100,000 persons per year among those aged 80-89 years [4]. Overall, HZ affects approximately one in three immunocompetent people during their lifetime [3]. Immunocompromised children, however, have a 5-6 times higher risk of HZ [5]. In fact, recovery from varicella has been described to be associated with VZV-specific T cell-mediated immunity, which is also essential for limiting reactivation and replication of latent VZV, and thus for preventing HZ [3].
Viral genomics has identified five VZV clades and their geographical distribution: clades 1, 3, and 5 are of European origin; clade 2 includes Asian strains, such as vOka; and clade 4 contains African strains [6]. vOka was developed by Takahashi and colleagues in 1974 in Japan using an empirical attenuation approach involving the passage of P-Oka, a clinical isolate recovered from the skin lesions of a child with varicella, in cultured human and guinea pig cells [7]. Currently, three loci within open reading frame (ORF) 62 (nucleotide positions 106262, 107252, and 108111) are used to distinguish vOka from wild-type VZV [8]. In our patient, vOka was identified using metagenomic sequencing and sequence comparison at these three loci.
VZV is the only human herpes virus for which highly effective vaccines are available [6], and two doses of varicella vaccine are part of the national vaccination recommendations for children in many countries. Vaccine-strain HZ is extremely rare. The first documented case complicated by multiple recurrence was reported in 2008 [9]. vOka establishes latency in the dorsal root ganglion in the same manner as wild-type VZV, and reactivation can cause vaccine-strain HZ in both immunocompromised and immunocompetent individuals [6]. As was observed in our patient, vaccine-strain HZ rash in children tends to appear predominantly in the lumbar and sacral dermatome areas [10]. Moodley et al. predicted that, with respect to dermatomal localization of the viral eruption, HZ of the lumbar dermatomes in children is likely to be caused by the vaccine-strain, because HZ in those dermatomes is rare in children after wild-type VZV infection [11]. In contrast, HZ in children younger than 10 years caused by wild-type VZV most often involves thoracic dermatomes, which might be related to the intensity of the centrally distributed rash observed during varicella [10].
Bryant et al. showed that, although vOka can cause HZ in immunocompetent children, adolescents, and adults, this occurs at a much lower rate than HZ caused by wild-type VZV [12]. This has also been corroborated in immunocompromised patients: Hardy and colleagues showed that in children with leukemia who received the live-attenuated VZV vaccine subsequently experienced a lower incidence of HZ than those with wild-type VZV infection [13]. Importantly, it has been shown that immunization against varicella does not increase the incidence of HZ in high-risk groups [13]. Symptoms caused by vaccine-strain HZ were generally mild and resolved completely, suggesting thatalthough it is capable of causing central nervous system diseasethis is a rare, self-limiting occurrence with typically milder symptoms in immunocompetent individuals [12]. However, single cases of fatal varicella due to disseminated vaccine-strain VZV have been reported in immunocompetent and immunocompromised children [14,15].
Intravenous acyclovir treatment is recommended for immunocompromised patients with HZ, including patients being treated with highdose corticosteroid treatment for more than 14 days [16]. Vaccine-strain VZV has been shown to have similar susceptibility to antivirals as wild-type VZV [17]. Although VZV immunoglobulin administered shortly after exposure can prevent or modify the course of the disease, it is not effective once the disease has developed [16].
## Conclusion
To our knowledge, this is the first case of recurrent HZ caused by vaccine-strain VZV in an infant with B-ALL. VZV vaccine is generally well tolerated. However, immunocompromised patients are at risk for developing dissemination or HZ following VZV vaccination. Although vaccine-strain HZ is generally milder compared to HZ caused by wildtype VZV, it is important to early recognize this adverse event and to initiate prompt antiviral treatment in order to prevent dissemination and sequelae in these high-risk situations.
## CRediT authorship contribution statement
## Leutert
## Ethical approval
N/A.
## Consent
Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.
## Declaration of Competing Interest
There are no competing interests to declare.
## References
1. (2025) "Federal Commission for Vaccination (EKIF)"
2. Gilden, Kleinschmidt-Demasters, Laguardia et al. (2000) "Neurologic complications of the reactivation of varicella-zoster virus" *N Engl J Med*
3. Harbecke, Cohen, Oxman (2021) "Herpes zoster vaccines" *J Infect Dis*
4. Re (1965) "The nature of Herpes zoster: a long-term study and a new hypothesis" *Proc R Soc Med*
5. Zhang, Kim, Grunebaum (2025) "Pediatric herpes zoster: should I be concerned for immunodeficiency? A review" *Front Pediatr*
6. Gershon (2015) "Varicella zoster virus infection" *Nat Rev Dis Prim*
7. Takahashi, Otsuka, Okuno et al. "Live vaccine used to prevent the spread of varicella in children in hospital" *Lancet*
8. Quinlivan, Jensen, Radford et al. (2012) "Novel genetic variation identified at fixed loci in ORF62 of the Oka varicella vaccine and in a case of vaccine-associated herpes zoster" *J Clin Microbiol*
9. Ota (2008) "Vaccine-strain varicella zoster virus causing recurrent herpes zoster in an immunocompetent 2-year-old" *Pediatr Infect Dis J*
10. Civen (2009) "The incidence and clinical characteristics of herpes zoster among children and adolescents after implementation of varicella vaccination" *Pediatr Infect Dis J*
11. Moodley, Swanson, Grose et al. (2019) "Severe herpes Zoster following varicella vaccination in immunocompetent young children" *J Child Neurol*
12. Bryant (2004) "Vaccine strain and wild-type clades of Varicella-zoster virus in central nervous system and non-CNS disease" *J Clin Microbiol*
13. Hardy, Gershon, Steinberg et al. (1991) "The incidence of zoster after immunization with live attenuated varicella vaccine. A study in children with leukemia. Varicella Vaccine Collaborative Study Group" *N Engl J Med*
14. Leung (2014) "Fatal varicella due to the vaccine-strain varicella-zoster virus" *Hum Vaccin Immunother*
15. Schrauder (2007) "Varicella vaccination in a child with acute lymphoblastic leukaemia" *Lancet*
16. Aaop, Kimberlin, Barnett et al. (2021) "Red book: 2021-2024 report of the committee on infectious diseases"
17. Shiraki, Matsui, Aiba (1993) "Susceptibility of Oka varicella vaccine strain to antiviral drugs" *Vaccine* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12751841&blobtype=pdf | Seong Chan
## Abstract
Okutama tick virus (OKTV) is a novel tick-borne RNA virus that has been reported in Japan and China. In the present study, an OKTV was detected in Haemaphysalis flava that had bitten a raccoon dog in South Korea by reverse transcription polymerase chain reaction using viral family-specific primers. Whole-genome sequencing revealed that the South Korean OKTV strain contains L and S segments with lengths of 6,529 and 1,890 nucleotides, respectively. Phylogenetic analyses revealed that OKTV strains formed two clusters based on the L segment and three clusters based on the S segment, with the South Korean strain forming a common cluster with three Chinese strains (SDQDH01, SDQDH04, and SDQDR04). Sequence comparisons showed high conservation among OKTV strains, with nucleotide identities of at least 97.74% and amino acid identities of at least 98.53% for both the L and N genes. Notably, the South Korean strain exhibited the highest amino acid similarity with the Chinese strain SDQDH04 (99.86% similarity in RdRP and 100% similarity in N protein). Selection pressure analyses revealed low dN/dS ratios for the L (0.0326) and N (0.0927) genes, with no sites detected under positive selection. Collectively, this study provides the first genomic characterization of OKTV in South Korea, expanding its geographical distribution and contributing to our genetic understanding of this virus. Although infectivity in animal hosts has not been established, further studies are needed to assess the zoonotic potential of OKTV.
First genetic characterization and phylogenetic analysis of Okutama tick virus in a tick collected from a raccoon dog (Nyctereutes procyonoides) in South Korea Junho Yoon 1 , Minjoo Yeom 1 , Hai Quynh Do 1 , Kyungmoon Lee 1 , Jong-Woo Lim 1 , Young Deok Suh 2,3 , Do Na Lee 2 , So-Eun Ryu 2,3 , Jang-Hee Han 2,3 , Dae Gwin Jeong 4 , Seong Chan Yeon 2,3* and Daesub Song 1*
## Background
Owing to their remarkable ecological diversity and widespread distribution, ticks serve as vectors for numerous viral pathogens [1]. Tick-borne viruses represent a significant global threat as emerging and reemerging zoonotic pathogens with substantial implications for human and animal health [2]. With advancements in metagenome sequencing, approximately 870 tick-borne viruses belonging to 28 orders, 55 families, and 66 genera have been identified [3]. Among these, tick-borne phenuiviruses and flaviviruses have emerged as prominent viral groups with substantial implications for human and animal health [2]. Representative emerging tick-borne flaviviruses include the Powassan virus (POWV) in North America, Kyasanur Forest disease virus (KFDV) in India, and Omsk hemorrhagic fever virus (OHFV) in Russia, each associated with severe human and animal diseases [4]. Severe fever with thrombocytopenia syndrome virus (SFTSV), a member of the phenuiviruses, was first identified in China in 2009 and has since spread across East Asia, causing recurrent human outbreaks with high fatality rates [5]. Furthermore, numerous novel tick-borne phenuiviruses have been identified in various tick species, particularly in Asian countries [6][7][8].
The novel Okutama tick virus (OKTV), a novel uukuvirus of the family Phenuiviridae, was initially discovered in H. flava in Tokyo, Japan, in 2018 [6]. Since its initial discovery, OKTV has been detected in other locations in Japan, including Cape Toi [9] and the Hokuriku District [10], as well as in China [11][12][13], South Korea [14] and Indonesia [15]. Genetically, OKTV has a two-segment RNA genome; the L segment encodes the RNAdependent RNA polymerase (RdRP) protein, whereas the S segment encodes the nucleocapsid (N) protein [10]. Compared with other viral species belonging to Phenuiviridae, the M segment encoding the envelope glycoprotein is absent in OKTV [10]. Phylogenetic analyses have indicated that OKTV belongs to the same groups as Yongjia tick virus 1 and Dabieshan tick virus but is distantly related to SFTSV [11].
Despite previous reports on OKTV in East Asia, its genetic diversity remains poorly understood due to limited sequence data, and little is known about its infectivity and pathogenicity in humans and animals. As a peninsular country located between China and Japan, South Korea represents a critical region for the surveillance of tick-borne viruses. To the best of our knowledge, OKTV sequences have not yet been reported in South Korea. The aim of this study was to investigate the presence of unrecognized viruses in ticks collected from wild animals and to provide genetic evidence and phylogenetic characterization of OKTV in South Korea.
## Sample preparation and tick species classification
Between June and October 2024, a total of 45 ticks were collected from wild mammals rescued in Seoul, South Korea. Tick species and developmental stages were identified based on morphology [16,17]. For engorged ticks that were difficult to classify morphologically, species identification was conducted using the mitochondrial 16 S rRNA gene [18]. Each tick was homogenized in 400 µL of PBS using a Precellys 24 Touch homogenizer (Bertin Technologies, Montigny-le-Bretonneux, France). Nucleic acids were extracted from the homogenate using the WizPrep Viral DNA/RNA Mini Kit (WizBiosolution, Gyeonggi-do, Republic of Korea). The targeted sequence was amplified using the HS Prime Mix (Genetbio, Daejeon, Republic of Korea) with primers targeting 16S rRNA, following the protocol (Table S1). Polymerase chain reaction (PCR) products were subsequently sequenced using Sanger sequencing at BIONICS Co., Ltd. (Republic of Korea), and the obtained sequences were compared with public data to identify the species. Of the 45 ticks collected, 17 ticks were Haemaphysalis longicornis, 23 were H. flava, and the remaining were Ixodes nipponensis (Table 1 and Table S2).
## Virus detection
Universal primers targeting the L gene of phenuiviruses [19] and NS5 gene of flaviviruses [20] were used to detect the circulation of tick-borne viruses. Reverse transcriptase (RT)-PCR was performed using the SuperScript™ III One-Step RT-PCR Kit (Invitrogen, Carlsbad, CA, USA) with the incubation program (Table S1). The seminested PCR to detect flaviviruses was performed using HS Prime Mix (Genetbio, Daejeon, Republic of Korea), with 1.5 µL of the first RT-PCR product and the incubation program (Table S1). SFTSV and dengue virus RNAs were used as positive controls. Only one sample (H. flava) tested positive for phenuivirus, whereas all samples The asterisk (*) indicates ticks in which Okutama tick virus was confirmed
were negative for flavivirus. The PCR product was subsequently sequenced by Sanger sequencing at BIONICS Co., Ltd. (Seoul, Republic of Korea), and the obtained sequence was identified as OKTV using the Nucleotide BLAST program.
To clarify the detection sensitivity of OKTV, RT-qPCR was conducted using the Luna ® Universal Probe One-Step RT-qPCR Kit (New England Biolabs, Ipswich, MA, USA) with primers and probe targeting the N gene of OKTV, which were designed in Geneious Prime v2025.0.3. (Table S1). The RT-qPCR assay demonstrated a limit of detection (LOD) of 167 copies/µL (Ct = 38.96), with an amplification efficiency of 92.0% (Fig. S1). Additionally, a blood sample collected during a routine veterinary check-up from a raccoon dog bitten by an OKTV-positive tick was also tested. Of all the samples tested, only one tick sample was positive for OKTV (Ct = 25.84), which was the same specimen previously identified as positive by conventional PCR, whereas all the other ticks and the raccoon dog blood sample tested negative (Table S2).
## Whole genome amplification by next-generation sequencing (NGS)
Viral RNA was extracted from an OKTV-positive tick using the Viral RNA Mini Kit (QIAGEN, Hilden, Germany), and a transcriptome library was prepared using the TruSeq Stranded Total RNA Library Prep Human/ Mouse/Rat Kit (Illumina, San Diego, CA, USA). NGS was conducted at Macrogen Inc. (Seoul, Republic of Korea) using the Illumina NovaSeq 6000 system, producing paired-end reads (2 × 101 bp). The raw reads were subjected to preprocessing using BBDuk adapter/quality trimming tool version 38.84 to remove adapters and low-quality sequences. The trimmed reads were mapped to the L and S segment sequences of OKTV (accession numbers: LC483653 and LC483654) [10] using Geneious Prime 2025.0.3.
Two complete viral sequences were obtained from an OKTV-positive H. flava sample. The L segment sequence of OKTV was 6,529 nucleotides (nt) in length, with 143,200 mapped reads and an average coverage of 2,061.7 ± 647.1. The S segment was 1,890 nt in length, with 6,814 mapped reads and an average coverage of 329.6 ± 133.1. Each segment contained a single open reading frame: the L segment, corresponding to the L gene (6,417 nt), encoded the RdRP protein consisting of 2,138 amino acids (aa), and the S segment, containing the N gene (822 nt), encoded a N protein of 273 aa. The identified virus was designated as SNU-W12_Korea_2024, and the sequences were deposited in the NCBI GenBank database under the accession numbers PV567690 and PV567691.
## Phylogenetic tree and pairwise identity analysis
Phylogenetic analyses and pairwise sequence identity comparisons were performed using the viral sequences obtained in this study, together with the complete sequences of OKTV strains and coding RNA sequences of other tick-borne phenuiviruses, retrieved from the NCBI database. Multiple sequence alignments were conducted using MAFFT version 7 with the default parameters [21]. Phylogenetic trees were constructed using the maximum likelihood method in MEGA 12.0 software [22]. Pairwise sequence identities were calculated and visualized as a scatter plot using GraphPad Prism 9 software and a heatmap using RStudio software.
Phylogenetic analysis revealed that OKTV clustered with other two-segmented tick-borne viruses of the genus Uukuvirus, including Dabieshan tick virus (DBTV) and Yongjia tick virus 1 (YJTV-1) (Fig. 1a,b). Pairwise sequence identity analysis showed that the OKTV L gene shared 70.83% nucleotide identity with YJTV-1 and 65.14% with DBTV, while the N gene shared 75.25% with YJTV-1 and 61.66% with DBTV. Compared with other phenuiviruses (excluding DBTV and YJTV-1), OKTV shared only 39.11-47.51% nucleotide identity in the L gene and 40.99-49.21% in the N gene, highlighting its genetic distinction from related viruses (Fig. S2).
For the OKTV phylogenetic trees, strains were divided into two clusters based on the L segment and three clusters based on the S segment. The South Korean strain (SNU-W12) consistently clustered with three strains detected in Shandong, China (SDQDH01, SDQDH04, and SDQDR04) in both L and S segments (Fig. 2a,b). Comparison of the coding RNA sequences among the OKTV strains revealed that the nucleotide identity of the L gene ranged from 97.74 to 99.69%, and the amino acid identity of RdRP ranged from 99.25 to 100% (Table S3a). Similarly, the nucleotide identity of the N gene ranged from 98.18 to 100%, and the amino acid identity of the N protein ranged from 98.53% to 100%, indicating high genetic conservation among the OKTV strains (Table S3b). Furthermore, the South Korean strain (SNU-W12) and Chinese strain (SDQDH04) exhibited the highest amino acid sequence identities, with 99.86% in the RdRP and 100% in the N protein, suggesting a close genetic relationship between the South Korean and Chinese strain (Table S3a, b).
## Estimation of selection pressure
Selection pressures for the RdRp and N protein coding regions were analyzed using the Datamonkey web server (http://www.datamonkey.org). The n o n -s y n o n y m o u s / s y n o n y m o u s (dN/dS) ratio was estimated by applying three complementary methods: Single-Likelihood Ancestor Counting (SLAC) and Fixed Effects Likelihood (FEL) to identify site-specific selection [23], and Fast Unconstrained Bayesian AppRoximation (FUBAR) for Bayesian inference of selection [24]. Sites under positive (dN/dS >1) or negative (dN/dS < 1) selection were determined using a p-value of 0.01 (SLAC and FEL) and a posterior probability of 0.9 (FUBAR), respectively. The mean dN/dS ratios estimated by the SLAC method were 0.0326 for the L gene and 0.0927 for the N gene, indicating predominant negative (purifying) selection in both coding regions (Table 2). Consistent with these results, no sites under positive (diversifying) selection were detected by either the FEL or FUBAR analyses.
## Discussion
Although OKTV has been detected in several Asian countries, knowledge of its pathogenicity, host range, and molecular evolution remains limited. In this study, we present the first complete coding sequences of OKTV from South Korea and compare its genetic features with those of other tick-borne viruses. Phylogenetic analyses and sequence comparisons with other congeneric viruses revealed that OKTV shows clear genetic divergence in the L and N genes, supporting its classification as a distinct species. Since its first identification from ticks collected in 2017, the L and N genes of OKTV strains obtained from different regions and years exhibited consistently high sequence identity, with the L gene showing ≥ 97.74% and the N gene showing ≥ 98.18% nucleotide identity, suggesting that the virus has remained evolutionarily conserved. Compared to DBTV, another two-segmented uukuvirus, OKTV showed a higher sequence identity. Among DBTV strains, the L gene exhibited ≥ 95.8% and the N gene ≥ 96.6% nucleotide identity [25]. Selection analyses revealed low dN/dS ratios with predominant purifying selection in both the L and N genes, underscoring evolutionary constraints that preserve protein functions essential for viral survival and replication [26,27]. This selection pattern is consistent with the findings for other phleboviruses, which likewise showed a low dN/dS ratio in L and N genes [28]. In contrast to SFTSV, in which specific RdRP sites have been reported to be under positive selection [29], no positively selected sites have been identified in OKTV. Taken together, these results suggest that OKTV evolution may have been constrained by negative selection, contributing to genetic conservation. However, the dataset used in this study was small and temporally limited, highlighting the need for additional complete sequences to allow for a more robust evolutionary assessment.
Despite the high genetic similarity among the sequences, phylogenetic analyses revealed that OKTV strains from the three countries separated into two clusters based on the L segment and three clusters based on the S segment, suggesting potential geographic structuring of OKTV strains in East Asia. Notably, the South Korean OKTV strain clustered more closely with the Chinese strains than with the Japanese strains and showed the highest amino acid identity with the Chinese strain. This close genetic relatedness suggests a possible cross-border movement between South Korea and China. The long-distance spread of OKTV is plausibly mediated by the dispersal of ticks via migratory birds [14]. Indeed, it has been suggested that H. flava, known as a primary vector for OKTV, is found in migratory birds along East Asian flyways [30,31] and OKTV RNA has been identified in ticks that fed on migratory birds on Daecheong Island in South Korea [14]. Furthermore, our current detection of OKTV in a tick collected from a raccoon dog in the inland areas of Seoul further suggests the possibility of wider dissemination throughout South Korea. However, because this study was based on a small number of specimens from a single region, we could not infer the extent of OKTV circulation. OKTV may have a broad vector range, as it has been detected in H. flava, H. longicornis, H. campanulata, Rhipicephalus sanguineus, and Dermacentor sinicus in China [11][12][13]. In South Korea, H. longicornis and H. flava are abundant and widely distributed [32], raising the possibility that OKTV is more widespread than currently recognized. Therefore, comprehensive surveillance is required to determine the geographical extent and transmission ecology of OKTV.
OKTV has been detected in both blood-fed and unfed ticks [9][10][11][12][13]. However, to date, no such infections have been confirmed in animals or humans. Although OKTV has been detected in ticks feeding on hedgehogs and hares, there is no direct evidence of infection in these hosts [11]. In our study, the virus was detected in a tick from a raccoon dog, whereas the animal's blood tested negative; thus, host infection could not be established. However, because the blood sample was insufficient, serological tests could not be performed, preventing the confirmation of exposure. Although vertebrate OKTV infections remain unconfirmed, emerging evidence of zoonotic potential among related Uukuvirus species warrants attention. The detection of Dabieshan tick virus RNA in ovine blood in China [33] and febrile human infection with Tacheng tick virus 2 [34] collectively underscore the zoonotic potential of previously unrecognized tick-borne viruses. In this study, virus isolation was not performed because of limited sample material. Therefore, further research is needed to prepare for potential OKTV infections, including viral isolation, experimental animal infection studies, and seroprevalence surveys in wild hosts.
## Conclusion
In conclusion, this study reports the first complete coding sequence of an OKTV identified in South Korea and offers a comparative genetic analysis with strains previously reported in Japan and China. These findings enhance our understanding of the geographic distribution and genetic diversity of OKTV in East Asia. However, information about the ecology and pathogenicity of this virus remains limited. Therefore, continued surveillance and research on OKTV are essential to assess its potential zoonotic risks to both animals and humans.
## References
1. Estrada-Peña (2015) "Ticks as vectors: taxonomy, biology and ecology" *Rev Sci Tech*
2. Mansfield, Jizhou, Phipps et al. (2017) "Emerging tick-borne viruses in the twenty-first century" *Front Cell Infect Microbiol*
3. Moming, Bai, Wang et al. (2024) "The known and unknown of global tick-borne viruses" *Viruses*
4. Dobler (2010) "Zoonotic tick-borne flaviviruses" *Vet Microbiol*
5. Li, Li, Yang et al. (2021) "Severe fever with thrombocytopenia syndrome virus: a highly lethal bunyavirus" *Crit Rev Microbiol*
6. Matsumoto, Masuoka, Hirayama et al. (2018) "Detection and phylogenetic analysis of phlebovirus, including severe fever with thrombocytopenia syndrome virus" *J Vet Med Sci*
7. Li, Shi, Tian et al. (2015) "Unprecedented genomic diversity of RNA viruses in arthropods reveals the ancestry of negative-sense RNA viruses" *eLife*
8. Shen, Duan, Wang et al. (2018) "A novel tick-borne phlebovirus, closely related to severe fever with thrombocytopenia syndrome virus and heartland virus, is a potential pathogen" *Emerg Microbes Infect*
10. Mekata, Kobayashi, Okabayashi (1016) "Detection and phylogenetic analysis of Dabieshan tick virus and Okutama tick virus in ticks collected from cape Toi" *Japan. Ticks Tick Borne Dis*
11. Kobayashi, Murota, Itokawa et al. (2020) "RNA virome analysis of questing ticks from Hokuriku District, Japan, and the evolutionary dynamics of tick-borne phleboviruses" *Ticks Tick Borne Dis*
12. Hu, Jiang, Luo et al. (1922) "Diversity analysis of tick-borne viruses from hedgehogs and hares in Qingdao" *China. Microbiol Spectr*
13. Xiao, Yao, Guan et al. (1016) "Viromes of haemaphysalis longicornis reveal different viral abundance and diversity in free and engorged ticks" *Virol Sin*
14. Ye, Li, Xu et al. (0175) "Virome diversity shaped by genetic evolution and ecological landscape of Haemaphysalis longicornis" *Microbiome*
15. Byun, Rieu, Nam et al. (2025) "Analysis of tickborne pathogens using next-generation sequencing in ticks from wild animals in the Republic of Korea" *J Vet Sci*
16. Byun, Rieu, Nam et al. (2025) "Detection of tickborne pathogens in blood-fed ticks from animals across nine Asian countries" *Microbiol Spectr*
17. Yamaguti, Tipton, Keegan et al. (1971) "Ticks of Japan, Korea, and the Ryukyu Islands" *Brigham Young University Science Bulletin, Biological Series*
18. Barker, Walker (2014) "Ticks of Australia. The species that infest domestic animals and humans" *Zootaxa*
19. Black, Piesman (1994) "Phylogeny of hard-and soft-tick taxa (Acari: Ixodida) based on mitochondrial 16S rDNA sequences" *Proc Natl Acad Sci USA*
20. Matsuno, Weisend, Kajihara et al. (2015) "Comprehensive molecular detection of tick-borne phleboviruses leads to the retrospective identification of taxonomically unassigned bunyaviruses and the discovery of a novel member of the genus phlebovirus" *J Virol*
21. Shimoda, Hayasaka, Yoshii et al. (1016) "Detection of a novel tick-borne flavivirus and its serological surveillance" *Ticks Tick Borne Dis*
22. Katoh, Rozewicki, Yamada (2019) "MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization" *Brief Bioinform*
23. Kumar, Stecher, Suleski et al. (2024) "MEGA12: molecular evolutionary genetic analysis version 12 for adaptive and green computing" *Mol Biol Evol*
24. Pond, Frost (2005) "Not so different after all: a comparison of methods for detecting amino acid sites under selection" *Mol Biol Evol*
25. Murrell, Moola, Mabona et al. (2013) "FUBAR: a fast, unconstrained bayesian approximation for inferring selection" *Mol Biol Evol*
26. Liu, Ma, Liu et al. (1016) "Molecular evolution and spatial transmission of Dabieshan tick virus in Jilin Province, Northeastern China"
27. Velthuis (2014) "Common and unique features of viral RNA-dependent polymerases" *Cell Mol Life Sci*
29. Reguera, Cusack, Kolakofsky (2014) "Segmented negative strand RNA virus nucleoprotein structure" *Curr Opin Virol*
30. Lam, Liu, Bowden et al. (2013) "Evolutionary and molecular analysis of the emergent severe fever with thrombocytopenia syndrome virus" *Epidemics*
31. Park, Kim, Kim et al. (2024) "Deciphering the evolutionary landscape of severe fever with thrombocytopenia syndrome virus across East Asia" *Virus Evol*
32. Choi, Kang, Kim et al. (2014) "Ticks collected from migratory birds, including a new record of Haemaphysalis formosensis, on Jeju Island, Korea" *Exp Appl Acarol*
33. (1993)
34. Byun, Rieu, Han et al. (2024) "Ixodid ticks from wild and domestic animals in East and central Asian flyways" *Acta Trop*
35. Chae, Cho, Cho et al. (2019) "Epidemiological investigation of tick species from near domestic animal farms and cattle, goat, and wild boar in Korea" *Korean J Parasitol*
36. Shao, Chang, Liu et al. (0170) "Detection and phylogenetic analysis of a novel tick-borne virus in Haemaphysalis longicornis ticks and sheep from Shandong, China" *Virol J*
37. Dong, Yang, Wang et al. (2019) "Human Tacheng tick virus 2 infection" *Emerg Infect Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12581384&blobtype=pdf | # Molecular diagnosis of viral hepatitis in the WHO African region: progress toward 2030 elimination goals
Patrick Aluora
## Abstract
The WHO African Region bears a disproportionately high burden of hepatitis B and C virus infections, yet molecular diagnostic capacity remains limited, impeding accurate detection, treatment monitoring, and surveillance. This Letter highlights the critical role of molecular diagnostics, including nucleic acid amplification tests, viral load quantification, and genotypic characterization, in achieving the United Nations' 2030 viral hepatitis elimination targets. I have discussed the infrastructural, financial, and logistical challenges constraining molecular testing in the region, and underscore emerging innovations such as point-of-care viral load assays and portable sequencing technologies. Prioritizing accessible molecular diagnostics and genomic surveillance within national hepatitis control programs is essential to bridge existing gaps and advance toward global elimination goals.
Molecular testing, including nucleic acid amplification tests (NAATs), quantitative HBV DNA and HCV RNA assays, and genotypic characterization, provides unparalleled precision in diagnosing active infection, predicting therapeutic response (While nucleos(t)ide analogue (NA) therapy is effective in most HBV genotypes, growing evidence indicates that response patterns may vary among genotypes), and detecting drug resistance [4,5]. Yet, the WHO African Region faces significant infrastructural, logistical, and financial constraints that hinder their scale-up. Most diagnostic algorithms rely on routine serology, which, although useful for initial screening, provides limited insight into the full spectrum of HBV infection. In such cases, HBV DNA testing offers the necessary molecular confirmation to guide accurate classification and clinical decision-making.
In the WHO African Region, HBV diagnosis still largely depends on serologic screening, with molecular testing seldom integrated into routine practice. Expanding access to HBV DNA testing is therefore essential to uncover occult infections, monitor treatment response,
## Dear editor,
The United Nations (UN) Sustainable Development Goals and the World Health Organization (WHO) Global Health Sector Strategy set an ambitious target to eliminate viral hepatitis as a public health threat by 2030, aiming for a 90% reduction in new infections and a 65% reduction in related mortality [1]. The WHO African Region, home to more than 65 million people chronically infected with hepatitis B virus (HBV) and approximately 8 million with hepatitis C virus (HCV), remains central to achieving this objective [2,3]. Despite the high burden, molecular diagnosis, the cornerstone of effective surveillance, treatment monitoring, and elimination strategies, remains inaccessible in the region.
and identify individuals at risk of viral reactivation. HBV DNA quantification is therefore crucial not only for confirming active infection but also for identifying patients with high viral loads who are at elevated risk of hepatocellular carcinoma, even in the absence of Alanine Aminotransferase (ALT) elevation. It also remains indispensable for monitoring treatment efficacy and detecting viral breakthrough or resistance [6]. Nevertheless, the 2024 WHO guidelines for Low-and Middle-Income Countries (LMICs) provide pathways to initiate treatment without HBV DNA testing when access is limited, an important step to expand care, though it highlights the ongoing gaps in molecular diagnostic capacity across the region. This diagnostic gap perpetuates under-diagnosis and mismanagement for HBV while relying on HCV antibody testing alone risks over-diagnosis, as seropositivity does not confirm active infection, thereby undermining elimination goals [7,8]. Confirmatory HCV RNA and HBV DNA testing is therefore essential.
Several challenges persist. First, the limited availability of molecular platforms in resource-constrained settings leads to reliance on centralized laboratories, often resulting in delayed results and patient loss to followup [9]. Second, prohibitive costs of reagents and lack of subsidized testing programs make molecular diagnosis inaccessible for most patients [10]. Third, there is insufficient investment in local genomic surveillance, despite evidence of diverse HBV and HCV genotypes and recombinants circulating in the WHO African Region, some associated with increased oncogenic potential and altered treatment response [11][12][13][14]. Integrating molecular testing and genomic surveillance can inform vaccine strategies by identifying breakthrough infections, vaccine escape variants, and residual transmission within vaccinated populations. This neglect threatens both clinical management and vaccine policy optimization.
Encouragingly, innovations such as dried blood spot sampling for HBV DNA quantification, portable sequencing technologies (e.g., Oxford Nanopore) and near pointof-care (e.g., GeneXpert) allow decentralized HBV DNA and HCV RNA testing, offering scalable solutions for resource-limited settings, showing promise in overcoming existing barriers [15][16][17]. However, translation into routine practice requires political will, sustainable financing, and regional capacity building. Without prioritizing molecular diagnostics within national hepatitis control programs, the WHO African Region risks lagging, jeopardizing global elimination targets.
In conclusion, achieving the UN 2030 target needs not only scaling up hepatitis vaccination and treatment but also embedding molecular diagnostic capacity as a public health priority in the WHO African Region. This communication seeks to urge policymakers, funders, and the scientific community to align investments towards accessible molecular platforms, robust genomic surveillance, and integration into routine healthcare. Only then can elimination of viral hepatitis transition from aspiration to reality.
## References
1. (2016) "Global health sector strategy on viral hepatitis 2016-2021. Towards ending viral hepatitis"
2. Who, Hepatitis (2021)
3. Who, Hepatitis (2023)
4. Valsamakis (2007) "Molecular Testing in the Diagnosis and Management of Chronic Hepatitis B" *Clin Microbiol Rev*
6. Wong, Lemoine (2025) "The 2024 updated WHO guidelines for the prevention and management of chronic hepatitis B: Main changes and potential implications for the next major liver society clinical practice guidelines" *J Hepatol*
7. Lumley, Mokaya, Maponga et al. (2025) "Hepatitis B virus resistance to nucleos(t)ide analogue therapy: WHO consultation on questions, challenges, and a roadmap for the field" *Lancet Microbe*
8. Lemoine, Thursz (2017) "Battlefield against hepatitis B infection and HCC in Africa" *J Hepatol*
9. Sonderup, Afihene, Ally et al. (2017) "Hepatitis C in sub-Saharan Africa: the current status and recommendations for achieving elimination by 2030" *Lancet Gastroenterol Hepatol*
10. O'hara, Mcnaughton, Maponga et al. (2017) "Elimination of hepatitis B requires recognition of catastrophic costs for patients and their families" *Lancet Gastroenterol Hepatol*
11. Aluora "Genotypic Characterization of Hepatitis B Virus among Voluntary Blood Donors in Nairobi Regional Blood Transfusion Centre"
12. Nairobi (2021)
13. Ochwoto, Kimotho, Oyugi et al. (2016) "Hepatitis B virus genotypes and unique recombinants circulating among outpatients in selected hospitals in Kenya" *International Journal of Infectious Diseases*
14. Matlou, Gaelejwe, Musyoki et al. (2019) "A novel hepatitis B virus recombinant genotype D4/E identified in a South African population"
15. Aluora, Muturi, Gachara (2020) "Seroprevalence and genotypic characterization of HBV among low risk voluntary blood donors in Nairobi, Kenya" *Virol J*
16. Quick, Grubaugh, Pullan et al. (2017) "Multiplex PCR method for MinION and Illumina sequencing of Zika and other virus genomes directly from clinical samples" *Nature Protocols*
17. Saayman, Hechter, Kayuni et al. (2023) "A simplified point-of-service model for hepatitis C in people who inject drugs in South Africa" *Harm Reduct J*
18. Lumley, Kent, Jennings et al. (2025) "Whole genome sequencing of hepatitis B virus using tiled amplicon (HEPTILE) and probe based enrichment on Illumina and Nanopore platforms" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724346&blobtype=pdf | # Can plitidepsin be used as an antiviral against RSV?
Charlotte Estampes, Jenna Fix, Julien Sourimant, Priscila Sutto-Ortiz, Charles-Adrien Richard, Etienne Decroly, Marie Galloux, Jean-François Eléouët
## Abstract
Human respiratory syncytial virus (HRSV) is a main cause of acute lower respiratory tract infections in infants, the elderly, and immunocompromised patients. Although vaccines have recently been approved for the elderly and for pregnant women, there is no curative treatment for HRSV. HRSV replicates in the cytoplasm of infected cells, and transcription and replication of the viral genome depend on the viral RNA polymerase complex, which recruits cellular factors for RNA synthesis. Among them, the eukaryotic translation elongation factor 1A (eEF1A) was previously shown to be critical for HRSV replication. eEF1A activity can be inhibited by plitidepsin (Aplidin), a cyclopeptide extracted from the ascidian Aplidium albicans, which was shown to be highly potent against SARS-CoV-2, with a 50% inhibitory concentration (IC 90 ) of 0.70 to 1.62 nM depending on the cell line. Here, we investigated whether plitidepsin could also inhibit HRSV replication. We found that plitidepsin inhibited HRSV replication with an IC 50 of ≈3 nM in cell cultures. However, further investigation revealed that plitidepsin has pleiotropic effects, affecting the translation of both cellular and viral proteins in a similar manner. Overall, our results show that plitidepsin blocks cellular translation and indicate that plitidepsin can induce a proteasome-mediated degradation of eEF1A, depending on the cell line, also showing the dependence of HRSV replication on cellular factors, such as eEF1A. These results thus highlight an original mechanism of action of plitidepsin on eEF1A, which renders the use of this compound for antiviral therapy very risky. IMPORTANCE Respiratory syncytial virus (RSV) is the main cause of bronchiolitis in infants and the elderly. Although some recent advances have been made, in partic ular vaccines for pregnant women and the elderly, or a new and efficient monoclo nal prophylactic antibody for newborns, there is no curative treatment for human respiratory syncytial virus (HRSV). Previous works suggested that a natural compound extracted from a marine organism, plitidepsin, was capable of inhibiting virus replica tion, in particular SARS-CoV-2. Because the target of plitidepsin has been identified as the cellular protein eukaryotic translation elongation factor 1A (eEF1A) that brings tRNA-aa to the ribosome, and because it was published that RSV needs eEF1A, we tested plitidepsin against RSV. During this work, by using a non-radioactive pulse-chase labeling of protein synthesis, we found that plitidepsin blocks cellular translation with no specificity for the virus. We also observed that eEF1A was degraded after plitidepsin treatment in the BHK21-derived BSRT7 cell line, and that this degradation was inhibited by a proteasome inhibitor. However, this was not observed with Human HEp-2 or simian Vero E6 cell lines. So, we think that our results are new and original and that this information should be useful for the community working either with plitidepsin or eEF1A, with viruses, or other topics. We think that, in contrast to what is suggested by previous studies, it is risky to use plitidepsin as an antiviral in humans.
H uman respiratory syncytial virus (HRSV) is the commonest cause of lower respi ratory tract infection in young children worldwide and the first cause of their hospitalization (1). HRSV is estimated to infect about 33 million children, leading to more than 3 million hospitalizations and 26,000-152,000 deaths in children under 5 years each year (2). The global healthcare costs of HRSV-associated infections in young children in 2017 were estimated to be US$5.45 billion (3). In a systemic multisite study, HRSV was shown to be the first etiological agent responsible for severe pneumonia (more than 30%) in hospitalized children in Asia and Africa (4). HRSV infections are also associated with significant morbidity and mortality in the elderly and immunocompromised people (5,6). The true burden of disease in adults is likely significantly under-recognized, and recent studies indicate that the HRSV impact is similar to that of seasonal influenza in adults older than 65 years (7)(8)(9)(10)(11)(12).
HRSV vaccine research has been ongoing for nearly 60 years without success, and only in 2023, two vaccines based on stabilized HRSV prefusion F protein, Arexvy (GSK) and Abrysvo (Pfizer), were approved for medical use for adults aged 60 or older, both in the USA and Europe. In September 2023, Abrysvo was also approved in the USA for pregnant women at 32-36 weeks' gestation to prevent HRSV infection in newborn children. Finally, Moderna received U.S. FDA and E.U. approval for RSV Vaccine mRESVIA (13). The only pharmaceutical intervention since 1998 has been passive prophylaxis with Palivizumab, a monoclonal antibody targeting the fusion protein, thereby limiting the HRSV entry. However, its use was limited to high-risk infants because of the elevated cost and moderate efficacy (14). But recently, nirsevimab (Beyfortus), a long-acting monoclonal antibody, was approved by several regulatory agencies around the world for the prevention of HRSV infections in newborns and infants (15). A single intramuscular injection of this antibody should protect infants for an entire season compared with monthly doses and reduce costs (vaccine-like pricing expected), allowing for administra tion to all infants. However, there is still no specific curative treatment against HRSV.
HRSV is a non-segmented single-stranded negative-sense RNA virus of the Mono negavirales order, Pneumoviridae family, and Orthopneumovirus genus (16). The HRSV genome is approximately 15.2 kb long and contains 10 genes encoding 11 proteins (17). Replication and transcription rely on four of these proteins: the nucleoprotein N involved in genome and antigenome encapsidation, forming the ribonucleoprotein complex (NC), the RNA-dependent RNA polymerase L which exhibits all the enzymatic activities required for viral replication, transcription, and RNA capping, its cofactor the phosphoprotein P responsible for recruitment of L on the NC template, and the transcription factor M2-1 that has been described as an "antiterminating" factor during transcription and interacts with P and viral mRNAs (18). All viral RNA synthesis takes place in viral factories also called cytoplasmic inclusion bodies (19), which concentrate the viral proteins L, N, P, and M2-1, but also cellular proteins, such as the protein phosphatase 1 (PP1), HSP70, or the human eukaryotic translation elongation factor 1A (eEF1A), all involved in HRVS replication (20)(21)(22). Interestingly, it was found that many RNA viruses utilize eEF1A for replication, although the mechanisms by which they do this differ (23). In a previous study in which eEF1A was knocked down or inhibited with didemnin B, it was suggested that eEF1A plays a key role in the regulation of F-actin stress fiber formation required for HRSV assembly and release, but with no effect on HRSV genome replication (24). Furthermore, eEF1A was found to be associated with the N and P proteins in infected cells by using a proximity ligation assay and co-immunoprecipitation (22). Recently, plitidepsin (Aplidin), an analog of didemnin B and a potent anti-cancer agent targeting eEF1A2 (K D = 80 nM) (25), was shown to be highly potent against SARS-CoV-2 by targeting eEF1A, with a 90% inhibitory concentration (IC 90 ) of 0.88 nM (26). As suggested by the authors, since HRSV also uses eEF1A for viral replication, inhibition of eEF1A could be a new strategy to limit HRSV propagation/infection. In this work, we thus investigated the effect of plitidepsin on HRSV replication. We show that plitidepsin inhibits HRSV replication in infected cells, as well as a minigenome reporter system with an IC 50 of ≈3 nM in cultured cells. Further mechanistic investigation revealed that plitidepsin can induce the degradation of eEF1A in the BHK21-derived cell line, inhibiting the translation of both viral and cellular proteins in a similar range of concentration.
## MATERIALS AND METHODS
## Cells
BSRT7/5 cells (BHK-21 cells that constitutively express the T7 RNA polymerase) (27) and HEp-2 cells (ATCC: CCL-23) were maintained, respectively, in DMEM and MEM supplemen ted with 10% heat-inactivated fetal calf serum (FCS), with 2 mM glutamine, 100 µg/mL penicillin, and 100 U/mL streptomycin. Vero E6 cells were cultivated as BSRT7/5 cells. Cells were grown in an incubator at 37°C in 5% CO 2 . Transfections were performed with 2.5 µL of Lipofectamine 2000 (Thermofisher) per 1 µg of DNA according to the manufacturer's instructions.
## Viruses
Recombinant human RSV rHRSV-mCherry corresponding to HRSV Long strain expressing the mCherry protein was amplified on HEp-2 cells and titrated using a plaque assay procedure as previously described (28,29).
## Plitidepsin antiviral activity
Plitidepsin (MedChemExpress) and Carfilzomib (Cell Signaling #15022) were solubilized in DMSO at 1 and 5 mM as stock solutions, respectively. HEp-2 cells in 96-well plates were infected for 2 h with rHRSV-mCherry at an MOI of 0.2, in the absence of FCS. The medium was then changed to the same medium with 2% FCS and containing serial dilution of plitidepsin, with a 1% final concentration of DMSO in the culture medium. At 48 h post-infection, the red fluorescence intensity of mCherry was quantified using a Tecan infinite M200Pro spectrofluorometer with excitation and emission wavelengths of 580 and 620 nm, respectively. Values obtained for non-treated infected and non-infected cells were used for standardization and normalization. In parallel, the toxicity of the treatment was assessed on non-infected cells using the CellTiter-Glo Luminescent cell viability assay (Promega). The half-maximal inhibitory concentration (IC 50 ) and cytotoxic concentration (CC 50 ) were determined by fitting the data to the dose-response curve implemented in GraphPad version 8 software.
## Fluorescence microscopy
HEp-2 cells infected with rHRSV-mCherry were fixed with PBS-paraformaldehyde 4% for 20 min at room temperature, rinsed with PBS, and permeabilized with PBS-BSA 1%-Triton X-100 0.1% for 10 min. Nuclei were stained with Hoechst 33342 (1 µg/mL) for 5 min, washed with PBS, examined under a Nikon TE200 microscope equipped with a CoolSNAP ES2 (Photometrics) camera, and images were processed with Meta-Vue software (Molecular Devices).
## Minigenome assay
BSRT7/5 cells at 90% confluence in 96-well dishes were transfected with a plasmid mixture containing 125 ng of pM/Luc, 125 ng of pN, 125 ng of pP, 62.5 ng of pL, and 31 ng of pM2-1, as well as 31 ng of pRSV-β-Gal (Promega) to normalize transfection efficiencies as described previously (30,31). After 6 h, the transfection mix was removed and serial dilutions of plitidepsin were added for 14 h. Transfections were done in triplicate, and each independent transfection was performed three times. Cells were harvested 24 h post-transfection, then lysed in luciferase lysis buffer (30 mM Tris, pH 7.9, 10 mM MgCl 2 , 1 mM DTT, 1% Triton X-100, and 15% glycerol). The luciferase and β-galactosidase (β-Gal) activities were determined for each cell lysate with an Infinite 200 Pro (Tecan, Männedorf, Switzerland) and normalized based on the values obtained for cells treated with DMSO only.
## Fluorescence-based nucleotide-incorporation assay
The analysis of the effect of Plitidepsin on the RSV polymerase activity was performed according to the method described by Sutto et al. (32). The RdRp assay reaction contained 0.2 µM recombinant RSV L-P, 2 µM of oligonucleotide template, and 2 µM of a 5′-6-FAM primer, mixed in a buffer containing 20 mM Tris-HCl, pH 7.5, 10 mM KCl, 2 mM DTT, 0.01% Triton X-100, 5% DMSO, 0.2 U/mL RNasin (Ambion), and 6 mM MgCl 2 . Reactions were started by adding specific NTPs at 100 µM and incubated for 2 h at 30°C. Reactions were quenched by adding an equal volume of gel-loading buffer. Samples were denatured at 70°C for 10 min and run on a 20% polyacrylamide urea sequencing gel for 2.5 h at 45 W. The gel was scanned on a Typhoon imager (GE Healthcare).
## Western blotting
Cells were lysed in 2× Laemmli buffer, run on a 12% polyacrylamide gel, and transferred to a nitrocellulose membrane using the Trans-Blot Turbo system (Bio-Rad). The blots were blocked with 5% nonfat milk in PBS Tween20 0.2%, and probed with either a polyclonal rabbit anti-N serum (33), a rabbit anti-GFP (Invitrogen, Waltham, MA, USA), a mouse monoclonal anti-EF1A antibody (Santa Cruz G-8, sc377439), a mouse monoclonal anti-GAPDH antibody (Sigma-Aldrich MAB374), or a mouse anti-alpha-tubulin (DM1A, Sigma) and further incubated with HRP-coupled anti-mouse or anti-rabbit antibodies (Life Science). Membranes were revealed with the Clarity Western ECL kit (Bio-Rad) and analyzed with a ChemiDoc Touch Imaging System and Image Lab software (Bio-Rad).
## Pulse labeling of newly synthesized proteins with L-azidohomoalanine
HEp-2 cells were infected with rHRSV-mCherry at MOI = 1 in 6-well plates (35 mm in diameter) and incubated in MEM supplemented with 2% heat-inactivated FCS, 2 mM glutamine, 100 µg/mL penicillin, and 100 U/mL streptomycin. Twenty-four hours later, serial dilutions of plitidepsin or DMSO were added to the cells for 2 h. Cells were washed twice with methionine-free (Met-) MEM (Gibco), then live-labeled (pulsed) at 37°C for 4 h with 50 µM Click-iT L-azidohomoalanine (AHA; Jena Bioscience) in Met-MEM supplemented with 5%-10% dialyzed FBS (Gibco) and plitidepsin. Cells were lysed in 1 mL of RIPA buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1% Triton X-100, 2 mM DTT, 10 mM of freshly prepared iodoacetamide [Merck], and complete protease inhibitor cocktail [Roche]). After centrifugation for 10 min at 10,000 rpm at 4°C, the supernatants were mixed with 5 µM AFDye 488-DBCO (Click Chemistry Tools, Jena Bioscience) and incubated for 2 h at 21°C. Proteins were immunoprecipitated with specific antibodies coupled to protein A sepharose beads (GE Healthcare) and analyzed by SDS-PAGE, together with the supernatants. Fluorescence was analyzed with a Fuji FLA-3000 scanner and quantified using the AIDA or Fiji software.
## RESULTS
## Antiviral potency of plitidepsin against HRSV
By using an siRNA approach, it was demonstrated 10 years ago that downregulation of eEF1A correlates with a reduction in the amount of infectious HRSV release, accom panied by a reduction in viral genome expression, but not mRNA transcription or protein expression (22). We therefore tested whether plitidepsin, known to inhibit the eEF1A activity, displays antiviral activity against HRSV. For this purpose, HEp-2 cells were infected with rHRSV-mCherry at MOI = 0.2 in a culture medium containing increasing concentrations of plitidepsin. Viral replication was quantified by mCherry fluorescence measurement on live cells 48 h post-infection using the Tecan spectrofluorometer. Values were standardized and normalized by those obtained for non-treated infected and non-infected cells. The IC 50 was determined by dose-response curve fitting. In parallel, the toxicity of plitidepsin was evaluated on non-infected HEp-2 cells. As shown in Fig. 1A, plitidepsin inhibited rHRSV-mCherry replication in a dose-dependent manner, with an IC 50 of 3.5 nM. These data were confirmed by direct observation of mCherry fluores cence and nuclei staining on fixed rHRSV-mCherry-infected cells (Fig. 1B). Noteworthy, a cytotoxic activity of plitidepsin was observed, with a calculated half-maximal cytotoxic concentration (CC 50 ) close to 145 nM (Fig. 1A). With a calculated selective index >30, plitidepsin thus appeared to be a potent inhibitor of HRSV replication in HEp-2 cells.
In a recent paper, using Vero E6 cells, Molina Molina et al. (34) found an inhibi tory effect of plitidepsin against HRSV with an IC 50 of 27 nM, thus five times higher than the one we found using HEp-2 or BSRT7 cells and the IC 50 they measured for SARS-CoV-2 inhibition. The authors hypothesized that plitidepsin promotes a shift in mRNA translation from a cap-dependent mechanism to a cap-independent pathway promoted by m6A mRNA methylation. Consistent with this hypothesis, they concluded that the stronger inhibitory effect of plitidepsin on SARS-CoV-2 replication results from the suppression of cap-dependent mRNA translation, leading to a more pronounced reduction in viral mRNA abundance and viral protein synthesis compared to host mRNA translation and cellular protein production. Conversely, the weaker effect observed on RSV was attributed to the virus's ability to exploit m6A-dependent translation pathways, as previously reported for RSV (34). To further validate this hypothesis, we evaluated the antiviral activity of plitidepsin against RSV in Vero E6 cells using the same experimental conditions as those employed for HEp-2 cells. However, as shown in Fig. S1, in our hands, plitidepsin also inhibited rHRSV-mCherry replication in a dose-dependent manner, but with an IC 50 of ~5 nM, thus in the same range as the one measured when using HEp-2 cells.
## Effect of plitidepsin on an HRSV minigenome
To determine whether plitidepsin could affect the RNA polymerase replication/transcrip tion machinery of HRSV, we used a well-established HRSVspecific minigenome assay system (30). The pM/Luc plasmid, which contains the authentic M/SH gene junction and the Luc reporter gene downstream of the gene start sequence inserted in this gene junction, was co-transfected into BSR-T7/5 cells together with p-β-gal, pL, pP, PN, and pM2-1. Luciferase activity was determined and normalized based on the signal obtained for cells incubated in a similar medium with 1% DMSO. In parallel, β-galactosidase (β-Gal) activity was determined for each cell lysate in order to normalize transfection efficiency. Figure 2A shows that plitidepsin inhibits the function of the HRSV polymerase complex in a dose-dependent manner with an IC 50 of ~5 nM while showing no toxicity toward BSR-T7/5 cells. However, plitidepsin also inhibited, to a lesser extent, the β-Gal activity, the effect being visible from 3.3 nM of plitidepsin.
Since plitidepsin targets the eEF1A cellular translation factor, the inhibition of HRSV RNA polymerase activity could result from either the direct inhibition of polymerase activity of the HRSV L protein or from the inhibition of translation of the viral mRNAs coding for the viral proteins. To assess whether the plitidepsin treatment had a specific or global impact on protein expression, cells were co-transfected either with the minige nome system or with a plasmid encoding EGFP, in the presence of increasing concentra tions of plitidepsin. The expression of proteins was analyzed by western blot. As shown in Fig. 2B, a similar decrease in N and EGFP expression was observed in the presence of plitidepsin, indicating that plitidepsin had a general inhibitory effect on viral or non-viral protein expression. In contrast, the signal corresponding to alpha-tubulin did not decrease significantly, which can easily be explained by the half-life of this cellular protein (~8 days) (35).
## Effect of plitidepsin on in vitro HRSV RdRp activity
To test whether plitidepsin could also directly affect the HRSV RNA polymerase activity, we performed an in vitro RNA polymerase assay using a recombinant L-P complex (36)(37)(38) in the presence of increasing plitidepsin concentrations. As shown in Fig. 3, no inhibitory effect of plitidepsin on RNA synthesis was observed, even for a concentration of 100 µM of plitidepsin. These results strongly suggest that the effect of plitidepsin on were transfected with the complete minigenome system (see above) or with the pEGFP plasmid coding for EGFP. Six hours later, the transfection mix was removed, and medium containing plitidepsin at different concentrations was added. Then, 24 h post-transfection, cells were lysed and analyzed by western blotting using anti-N, anti-GFP, or anti-alpha tubulin antibodies. viral replication was not due to a direct effect of the RdRp complex but mainly due to its impact on protein expression.
## Effect of plitidepsin on cellular and viral protein neosynthesis
To further investigate the impact of plitidepsin on viral and cellular RNA translation, we used a complementary approach, consisting of a pulse labeling of newly synthesized proteins with the methionine analog AHA, revealed by the fluorescent dye AFDye 488-DBCO based on Click chemistry (39). At 24 h post-infection, HEp-2 cells were incubated with serial dilutions of plitidepsin and then labeled with AHA for 4 h. Then, cells were lysed, and the HRSV N protein or the cellular α-tubulin protein was immu noprecipitated. Immunoprecipitated proteins and total cell lysates were analyzed by SDS-PAGE, and the fluorescence was quantified. As shown in Fig. 4A andB, a similar drop in protein synthesis was observed for N and α-tubulin, but also for the total proteins present in the cell lysates for concentrations ≥15 nM. These results correlate with the observations made on transfected BRST7 cells and highlight that plitidepsin has a similar negative impact on the translation of both viral and cellular mRNA.
## Plitidepsin induces proteasome-mediated degradation of eEF1A in BHK21derived cell line
The mechanism by which plitidepsin inhibits the activity of the host factor eEF1A is still debated. It was shown that overexpression of a negative dominant mutant, an Ala399 → Val (A399V) of eEF1A, reduced the sensitivity of cancer cells to didemnin B (40), as well as SARS-CoV-2 to plitidepsin by a factor >10 (26). Interestingly, this A399V substitution also confers resistance to the structurally unrelated ternatin-4 (40), suggesting that these molecules interact with the same binding site on eEF1A (26). However, the mechanism of action of ternatin-4 was recently studied in more detail and revealed that it was mediated by ubiquitination and proteasome degradation of eEF1A in HeLa cells (41). We thus wondered whether plitidepsin could also induce degradation of eEF1A by a similar mechanism. The three cell lines used in this study, BSRT7, HEp-2, and Vero cells, were treated with serial dilutions of plitidepsin in the presence or absence of the proteasome inhibitor carfilzomib, and the expression of eEF1A or the control GAPDH was analyzed by western blot. As shown in Fig. 5A, for concentrations of plitidepsin inhibiting HRSV replication, minigenome activity or translation inhibition, eEF1A expression was impaired in BSRT7 cells, and expression of eEF1A was restored upon treatment in the presence of the proteasome inhibitor Carfilzomib. Of note, this effect was clearly visible as soon as 4-5 h post-treatment (Fig. S2). By contrast, no degradation of eEF1A was observed in HEp-2 and Vero cells (Fig. 5B). These results suggest that plitidepsin, in addition to inhibiting
## DISCUSSION
The mammalian translation elongation factor eEF1A is an essential GTPase and the second most abundant intracellular protein after actin (3% of the total cellular protein). It is localized extensively in the cytoplasm and nucleus (23,42,43). The canonical function of eEF1A is to deliver aminoacyl tRNAs to the ribosomal A site during the elongation stage of protein synthesis. In addition to its canonical functions in transport ing aa-tRNA to the ribosome, eEF1A is found to be involved in cellular mechanisms, such as regulation of cytoskeleton organization by interacting with actin and tubulin, protein degradation mediated by the proteasome, nuclear aa-tRNAs protein export, signaling transduction pathway concerning apoptosis and oncogenesis, or binding to viral RNA (43)(44)(45). There are two variants of eEF1A, eEF1A1 and eEF1A2, that share 92% amino acid identity (46). In contrast to the ubiquitous expression of eEF1A1 in many cell types, eEF1A2 expression is limited to the terminally differentiated cells of the brain, heart, and skeletal muscle (47). The two eEF1A variants have similar translation activity but may differ with respect to their secondary, "moonlighting" functions. eEF1A1 also plays an important role in the process of heat shock stress response. eEF1A2 activates Akt in a PI3K-dependent fashion, stimulating cell migration, actin remodeling, and invasion, and inhibiting apoptosis (23).
Like for many viruses (23,43), eEF1A has been identified as an important cellular factor for HRSV replication (22). As an abundant, multifunctional protein, it is not surprising that many viruses have adapted to use eEF1A as a cofactor for viral transcrip tion, translation, assembly, and pathogenesis. So, eEF1A was shown to bind to some viral RNA structures, some viral structural and non-structural proteins, or to interact with viral polymerase complexes, such as those of vesicular stomatitis virus (VSV) (43,48). Like for HRSV, VSV forms two different viral RNA polymerase complexes in infected cells: the transcriptase and the replicase. The transcriptase complex synthesizes capped mRNAs, whereas the replicase complex initiates genomic minus-strand RNA synthesis at the precise 3′ ends of the plus-strand antigenomic and negative-strand genomic RNAs. The transcriptase was described as a protein complex containing the L and P proteins, the cellular eEF1A and heat shock protein 60, and a submolar amount of cellular mRNA cap guanylyltransferase (49). The replicase complex was described as containing the viral proteins L, P, and nucleocapsid (N) but not eEF1A, heat shock protein 60, or the guanylyltransferase. Hence, it was proposed that eEF1A is important for transcription of viral mRNAs but not for genomic RNA replication in VSV.
Disentangling the canonical and non-canonical roles of eEF1A and the extent to which each contributes to viral function is essential if eEF1A is to be targeted therapeut ically. It was previously shown that didemnin B treatment, a drug targeting eEF1A, protected cells from HRSV-induced cell death (22). Didemnin B did not significantly affect HRSV transcription and replication, especially at 24 h post-infection, but significantly reduced infectious virus production and release, possibly as a consequence of changes in actin stress fiber formation (24). More recently, plitidepsin (Aplidin), an analog of didemnin B that was approved in Australia for the treatment of myeloma (50), was shown to be very efficient against SARS-CoV-2 in cultured cells, as well as in a mice model (26).
Since it has entered clinical trials as an anti-SARS-CoV-2 drug and was shown to have a favorable long-term safety profile in adult patients hospitalized for COVID-19 (51)(52)(53)(54), we wondered whether this compound could also be efficient against HRSV. Plitidepsin is a cyclic depsipeptide that was first isolated from a Mediterranean marine tunicate (Aplidium albicans) and, at present, is manufactured by total synthesis and commercial ized by PharmaMar, S.A., as Aplidin (55). Plitidepsin has antitumoral and immunosuppres sive activities (56). Although not fully clarified, the molecular mechanisms of action of plitidepsin against tumor cells and SARS-CoV-2 have been investigated. Plitidepsin induces cell-cycle arrest and apoptosis (57). These effects rely on the induction of early oxidative stress, the rapid activation of Rac1 GTPase, and activation of c-Jun N-termi nal kinase (JNK), ERK, and p38 mitogen-activated protein kinases (p38/MAPK), which finally result in caspase-dependent apoptosis (58,59). It was determined that eEF1A is the primary target of plitidepsin, which can bind to eEF1A at the interface between domains 1 and 2 of this protein in the GTP conformation with a measured KD of 80 nM (25). However, it was proposed that the antitumoral effect of plitidepsin is not due to translation inhibition but inhibition of eEF1A binding to double-stranded RNA-depend ent protein kinase (PKR). In the presence of plitidepsin, PKR would disengage from eEF1A, thereby regaining its kinase activity to initiate extrinsic apoptosis through activation of MAPK and NF-κB signaling cascades (60). Plitidepsin was also reported as an ER stress inducer by activating the unfolded protein response (61). In parallel, plitidepsin was also shown to induce the phosphorylation of eIF2a, resulting in the arrest of protein synthesis at the initiation step (61,62). For SARS-CoV-2, it has been determined that the effect of plitidepsin is also mediated by eEF1A by using a mutated (A399V) version of eEF1A1 in 293T cells (26). It is likely that this effect is mediated by the inhibition of translation of the viral proteins (26,57,63). Very recently, the mechanism of SARS-CoV-2 inhibition was revisited (34). In this paper, using Vero E6 cells, the authors found that plitidepsin at 50 nM reduced the translation of RNAs, including cellular RNAs, but with a higher impact on viral mRNA translation and without affecting cellular viability. The molecular mechanisms involved in the cell's proteostatic response to eEF1A blockade, namely a shift from cap or internal ribosome entry sites-mediated translation toward an N-6-methyladenosine (m6A)-dependent translation, that could explain cell survival. They also tested plitidepsin against several other viruses, including HRSV, and found an IC 50 of 27 nM in Vero E6 cells, about ten times more than what we found using HEp-2 cells. For in vivo effects, it was also shown that treatment of a monocyte-derived macrophage cell line by plitidepsin in the same range as ours (2.5-10 nM) reduced the production of the proinflammatory cytokine IL-6 in the presence of SARS-CoV-2 virions, and that was associated with a reduction of NF-κB p65 subunit phosphorylation and of its transcription activity in the inflammatory cascade (64). This effect was also observed in vivo in SARS-CoV-2 or influenza virus-infected mice.
In our experiments, we observed that plitidepsin had an inhibitory effect on viral replication and minigenome activity with an IEC 50 ≈ 3-5 nM, and a toxic effect on cells with an IC 50 ≈ 150 nM. However, further investigation revealed that this inhibitory effect was due to a general inhibition of translation in the host cell, with no significant difference between cellular and viral protein expression. We then wondered whether eEF1A, which is essential for cellular translation, could be actively degraded by plitidep sin treatment, as it was observed recently with ternatin-4, another eEF1A-targeting drug (41). We found that, in the BHK21-derived cell line BSRT7, eEF1A was degraded after plitidepsin treatment for 18H at concentrations as low as 1 nM, which was inhibited by the proteasome inhibitor carfilzomib, indicating that an eEF1A ubiquitination mecha nism can occur. However, this effect on eEF1A was inhibited in the presence of the proteasome inhibitor carfilzomib, indicating that an eEF1A ubiquitination mechanism can occur. Strikingly, this degradation was not observed when using Vero or HEp-2 cells. During this study, it was published by Molina Molina et al. that cellular cap-dependent translation was inhibited by plitidepsin in Vero cells (34), which is in accordance with our results. They also determined that plitidepsin inhibits SARS-CoV-2 with an IC 50 varying from 4.6 to 17.9 nM depending on the viral strains, by decreasing de novo cap-dependent translation of SARS-CoV-2 and non-viral RNAs but affecting less than 13% of the host proteome, thus preserving cellular viability. They compared the effect of plitidepsin on other viruses, including HRSV, and found an IC 50 of ~27 nM. Although in the nanomolar range, their hypothesis to explain this reduced sensitivity to plitidepsin was that RNA viruses having m 6 A methylation in their mRNAs, a property described for RSV (65), would use as an alternative cap-independent translation mechanism via m 6 A reading. They also found that at 50 nM plitidepsin, cellular translation was inhibited, but the host's protein levels were poorly affected. Using pulse-chase experiments, we found that cellular translation was mainly affected in the presence of 30 nM of plitidepsin after 6 h of treatment (see Fig. 4). A simple explanation for cell survival despite translation stoppage would be to consider the persistence of cellular functions thanks to half-lives of cellular proteins which are generally long (35,66).
In conclusion, we show here that treatment of cells with plitidepsin induces the arrest of translation in treated cells and the possible degradation of eEF1A by the proteasome pathway, depending on the cell line, which is correlated with a global extinction of translation. These experiments also show the dependence of HRSV replication on the cellular factor eEF1A. Although effective concentrations of plitidepsin against HRSV are in the same range as those found for SARS-CoV-2, the side effects of plitidepsin on cells and its potential toxicity raise the question of its use in vivo for antiviral assays against HRSV.
## References
1. Langedijk, Bont (2023) "Respiratory syncytial virus infection and novel interventions" *Nat Rev Microbiol*
2. Li, Wang, Blau et al. (2022) "Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in children younger than 5 years in 2019: a systematic analysis" *Lancet*
3. Zhang, Akmar, Bailey et al. (2020) "Cost of respiratory syncytial virus-associated acute lower respiratory infection management in young children at the regional and global level: a systematic review and metaanalysis" *J Infect Dis*
4. O'brien, Baggett, Brooks et al. (2019) "Causes of severe pneumonia requiring hospital admission in children without HIV infection from Africa and Asia: the PERCH multi-country case-control study" *Lancet*
5. Falsey, Hennessey, Formica et al. (2005) "Respiratory syncytial virus infection in elderly and high-risk adults" *N Engl J Med*
6. Munting, Manuel (2021) "Viral infections in lung transplantation" *J Thorac Dis*
7. Ackerson, Tseng, Sy et al. (2019) "Severe morbidity and mortality associated with respiratory syncytial virus versus influenza infection in hospitalized older adults" *Clin Infect Dis*
8. Korsten, Adriaenssens, Coenen et al. (2021) "Burden of respiratory syncytial virus infection in community-dwelling older adults in Europe (RESCEU): an international prospective cohort study" *Eur Respir J*
9. Falsey, Walsh, House et al. (2021) "Risk factors and medical resource utilization of respiratory syncytial virus, human metapneumovirus, and influenzarelated hospitalizations in adults-a global study during the 2017-2019 epidemic seasons (hospitalized acute respiratory tract infection [HARTI] Study)" *Open Forum Infect Dis*
10. Heppe-Montero, Gil-Prieto, Del et al. (2012) "Impact of respiratory syncytial virus and influenza virus infection in the adult population in Spain between"
11. *Int J Environ Res Public Health*
12. Busack, Shorr (2022) "Going viral-RSV as the neglected adult respiratory virus" *Pathogens*
13. Maggi, Veronese, Burgio et al.
14. Grazia, Costantino, Vitale et al. (2022) "Rate of hospitaliza tions and mortality of respiratory syncytial virus infection compared to influenza in older people: a systematic review and meta-analysis" *Vaccines (Basel)*
15. Wilson, Goswami, Baqui et al. (2023) "Efficacy and safety of an mRNA-Based RSV PreF vaccine in older adults" *N Engl J Med*
16. Mazur, Terstappen, Baral et al. (2023) "Respiratory syncytial virus prevention within reach: the vaccine and monoclonal antibody landscape" *Lancet Infect Dis*
17. Ruckwardt (2023) "The road to approved vaccines for respiratory syncytial virus" *NPJ Vaccines*
18. Afonso, Amarasinghe, Bányai et al. (2016) "Taxonomy of the order Mononegavirales: update"
19. Collins, Crowe (2007) "Respiratory syncytial virus and metapneumo virus"
20. Cao, Gao, Liang (2021) "Structural insights into the respiratory syncytial virus RNA synthesis complexes" *Viruses*
21. Garcı́a, Garcı́a-Barreno, Vivo et al. (1993) "Cytoplasmic inclusions of respiratory syncytial virus-infected cells: formation of inclusion bodies in transfected cells that coexpress the nucleoprotein, the phosphoprotein, and the 22K protein" *Virology (Auckl)*
22. Munday, Wu, Smith et al. (2015) "Interactome analysis of the human respiratory syncytial virus RNA polymerase complex identifies protein chaperones as important cofactors that promote L-protein stability and RNA synthesis" *J Virol*
23. Richard, Rincheval, Lassoued et al. (2018) "RSV hijacks cellular protein phosphatase 1 to regulate M2-1 phosphorylation and viral transcription" *PLoS Pathog*
24. Wei, Li, Marcial et al. (2014) "The eukaryotic elongation factor 1A is critical for genome replication of the paramyxovirus respiratory syncytial virus" *PLoS One*
25. Abbas, Kumar, Herbein (2015) "The eEF1A proteins: at the crossroads of oncogenesis, apoptosis, and viral infections" *Front Oncol*
26. Snape, Li, Wei et al. (2018) "The eukaryotic translation elongation factor 1A regulation of actin stress fibers is important for infectious RSV production" *Virol J*
27. Losada, Muñoz-Alonso, García et al. (2016) "Translation elongation factor eEF1A2 is a novel anticancer target for the marine natural product plitidepsin" *Sci Rep*
28. White, Rosales, Yildiz et al. (2021) "Plitidepsin has potent preclinical efficacy against SARS-CoV-2 by targeting the host protein eEF1A" *Science*
29. Buchholz, Finke, Conzelmann (1999) "Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter" *J Virol*
30. Rameix-Welti, Goffic, Hervé et al. (2014) "Visualizing the replication of respiratory syncytial virus in cells and in living mice" *Nat Commun*
31. Bouillier, Rincheval, Sitterlin et al. (2019) "Generation, amplification, and titration of recombinant respiratory syncytial viruses" *J Vis Exp*
32. Blondot, Dubosclard, Fix et al. (2012) "Structure and functional analysis of the RNA-and viral phosphoprotein-binding domain of respiratory syncytial virus M2-1 protein" *PLoS Pathog*
33. Sourimant, Rameix-Welti, Gaillard et al. (2015) "Fine mapping and characterization of the Lpolymerase-binding domain of the respiratory syncytial virus phosphoprotein" *J Virol*
34. Sutto-Ortiz, Selisko, Ferron et al. (2025) "Mapping the impact of 1'-, 2'-and 4'nucleotide modifications on the respiratory syncytial virus RNAdependent RNA polymerase" *Antiviral Res*
35. Castagné, Barbier, Bernard et al. (2004) "Biochemical characterization of the respiratory syncytial virus P-P and P-N protein complexes and localization of the P protein oligomerization domain" *J Gen Virol*
36. Molina, Bech-Serra, Franco-Trepat et al. (2025) "Targeting eEF1A reprograms translation and uncovers broad-spectrum antivirals against cap or m 6 A protein synthesis routes" *Nat Commun*
37. Rolfs, Frey, Shi et al. (2021) "An atlas of protein turnover rates in mouse tissues" *Nat Commun*
38. Gilman, Liu, Fung et al. (2019) "Structure of the respiratory syncytial virus polymerase complex" *Cell*
39. Sutto-Ortiz, Tcherniuk, Ysebaert et al. (2021) "The methyltransferase domain of the respiratory syncytial virus L protein catalyzes cap N7 and 2'-Omethylation" *PLoS Pathog*
40. Yu, Abeywickrema, Bonneux et al. (2023) "Structural and mechanis tic insights into the inhibition of respiratory syncytial virus polymerase by a non-nucleoside inhibitor" *Commun Biol*
41. Morey, Esmaeili, Duennwald et al. (2021) "SPAAC pulsechase: a novel click chemistry-based method to determine the half-life of cellular proteins" *Front Cell Dev Biol*
42. Carelli, Sethofer, Smith et al. (2015) "Ternatin and improved synthetic variants Research Article mSphere December"
43. "kill cancer cells by targeting the elongation factor-1A ternary complex"
44. Oltion, Carelli, Yang et al. (2023) "An E3 ligase network engages GCN1 to promote the degradation of translation factors on stalled ribosomes" *Cell*
45. Sasikumar, Perez, Kinzy (2012) "The many roles of the eukaryotic elongation factor 1 complex" *Wiley Interdiscip Rev RNA*
46. Li, Wei, Abbott et al. (2013) "The unexpected roles of eukaryotic translation elongation factors in RNA virus replication and pathogenesis" *Microbiol Mol Biol Rev*
47. Mateyak, Kinzy (2010) "eEF1A: thinking outside the ribosome" *J Biol Chem*
48. Xu, Liu, Song (2021) "Functions and regulation of translation elongation factors" *Front Mol Biosci*
49. Negrutskii, Shalak, Novosylna et al. "El'skaya AV. 2023. The eEF1 family of mammalian translation elongation factors" *BBA Adv*
50. Lee, Francoeur, Liu (1992) "Tissuespecific expression in mammalian brain, heart, and muscle of S1, a member of the elongation factor-1 gene family" *J Biol Chem*
51. Das, Mathur, Gupta et al. (1998) "RNA polymerase of vesicular stomatitis virus specifically associates with translation elongation factor-1 αβγ for its activity" *Proc Natl Acad Sci*
52. Qanungo, Shaji, Mathur et al. (2004) "Two RNA polymer ase complexes from vesicular stomatitis virus-infected cells that carry out transcription and replication of genome RNA" *Proc Natl Acad Sci*
53. Jimenez, Wilke, Branco et al. (2020) "Enriching cancer pharmacology with drugs of marine origin" *British J Pharmacol*
54. Varona, Landete, Lopez-Martin et al. (2022) "Preclinical and randomized phase I studies of plitidepsin in adults hospitalized with COVID-19" *Life Sci Alliance*
55. Aguareles, Fernández, Carralón-González et al. (2023) "Outcomes and clinical characteristics of the compassionate use of plitidepsin for immunocompromised adult patients with COVID-19" *Int J Infect Dis*
56. Varona, Landete, Paredes et al. (2023) "Plitidepsin in adult patients with COVID-19 requiring hospital admission: a long-term follow-up analysis" *Front Cell Infect Microbiol*
57. Landete, Caliman-Sturdza, Lopez-Martin et al. (2024) "A phase iii randomized controlled trial of plitidepsin, a marine-derived compound, in hospitalized adults with moderate COVID-19" *Clin Infect Dis*
58. Alonso-Álvarez, Pardal, Sánchez-Nieto et al. (2017) "Plitidepsin: design, development, and potential place in therapy" *Drug Des Devel Ther*
59. Zhang, Cai, Yu et al. (2023) "Anticancer small-molecule agents targeting eukaryotic elongation factor 1A: state of the art" *Int J Mol Sci*
60. Papapanou, Papoutsi, Giannakas et al. (2021) "Plitidepsin: mechanisms and clinical profile of a promising antiviral agent against COVID-19" *J Pers Med*
61. García-Fernández, Losada, Alcaide et al. (2002) "Aplidin induces the mitochondrial apoptotic pathway via oxidative stress-mediated JNK and p38 activation and protein kinase C δ" *Oncogene*
62. Cuadrado, Garcia-Fernandez, Gonzalez et al. (2003) "Aplidin induces apoptosis in human cancer cells via glutathione depletion and sustained activation of the epidermal growth factor receptor, Src, JNK, and p38 MAPK" *J Biol Chem*
63. Losada, Muñoz-Alonso, Martínez-Díez et al. (2018) "Binding of eEF1A2 to the RNAdependent protein kinase PKR modulates its activity and promotes tumour cell survival" *Br J Cancer*
64. Losada, Berlanga, Molina-Guijarro et al. (2020) "Generation of endoplasmic reticulum stress and inhibition of autophagy by plitidepsin induces proteotoxic apoptosis in cancer cells" *Biochem Pharmacol*
65. Gordiyenko, Llácer, Ramakrishnan (2019) "Structural basis for the inhibition of translation through eIF2α phosphorylation" *Nat Commun*
66. Wong, Damania (2021) "SARS-CoV-2 dependence on host pathways" *Science*
67. Losada, Izquierdo-Useros, Aviles et al. (2024) "Plitidepsin as an immunomodulator against respiratory viral infections" *J Immunol*
68. Xue, Zhao, Zhang et al. (2019) "Viral N6methyladenosine upregulates replication and pathogenesis of human respiratory syncytial virus" *Nat Commun*
69. Boisvert, Ahmad, Gierliński et al. (2012) "A quantitative spatial proteomics analysis of proteome turnover in human cells" *Mol Cell Proteomics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12451639&blobtype=pdf | # Erratum: Summary of taxonomy changes ratified by the International Committee on Taxonomy of Viruses (ICTV) from the Fungal and Protist Viruses Subcommittee, 2025
Sead Sabanadzovic, Chantal Abergel, Marıá Ayllón, Leticia Botella, Marta Canuti, Yuto Chiba, Jeanmichel Claverie, Robert Coutts, Stefania Daghino, Livia Donaire, Marco Forgia, Ondřej Hejna, Jichun Jia, Daohong Jiang, Ioly Kotta-Loizou, Mart Krupovic, Andrew Lang, Matthieu Legendre, Shin- Yi Lee Marzano, Fan Mu, Uri Neri, Luca Nerva, Judit Pénzes, Anna Poimala, Sofia Rigou, Yukiyo Sato, Wajeeha Shamsi, Suvi Sutela, Nobuhiro Suzuki, Massimo Turina, Syun-Ichi Urayama, Eeva Vainio, Jiatao Xie, Ictv Taxonomy, Summary Consortium, Shin-Yi Lee Marzano
of Biology, Institute for Plant Sciences, University of Cologne, Cologne, Germany; 23 Department of Molecular Biology and Genetics, Aarhus University, 8000 Aarhus C, Denmark; 24 Institute of Plant Science and Resources, Okayama University, Kurashiki, Japan;
## References
1. "Information Génomique & Structurale"
2. *CNR*
4. Luke, Helsinki |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645950&blobtype=pdf | # A HiBiT-tagged pseudovirus-like particle platform for safe, rapid quantification of virus neutralization and antibody-dependent enhancement
Jonathan Mitchell, Vincent Mastrodomenico, Jim Hartnett, William Heelan, Denise Garvin, Mei Cong, Jamison Grailer
## Abstract
Accurate quantification of virus neutralization is essential for evaluating the efficacy of vaccines and antibody-based therapies. However, conventional neutraliza tion assays require strict biocontainment or utilize surrogates of infection that increase biosafety but reduce biological relevance. These limitations create critical bottlenecks for the timely development of antiviral immunotherapeutics. Here, we describe a neutralization platform using non-replicating, HiBiT-tagged pseudovirus-like particles (HiBiT-PsVLPs) for safe yet biologically relevant assessment of neutralization. These HiBiT-PsVLPs package the HiBiT protein tag internally and incorporate glycoproteins from pathogenic viruses to recapitulate entry at reduced biosafety levels. Unlike traditional pseudoviruses, HiBiT-PsVLPs lack reporter genes and employ the NanoBiT split luciferase system for rapid, complementation-based luminescent readout of entry and neutraliza tion. Using the SARS-CoV-2 Spike protein as proof-of-concept, we demonstrate efficient pseudotyping, entry, and neutralization of HiBiT-PsVLPs. HiBiT-PsVLP neutralization is specific, reproducible, and reflective of antibody potency and stability. Assay results align closely with an established surrogate virus neutralization test (sVNT), with HiBiT-PsVLPs capturing more diverse mechanisms-of-action (MoAs). Moving beyond SARS-CoV-2, we adapt the HiBiT-PsVLP platform for other clinically relevant viruses, including HIV-1 and emerging pathogens, such as Ebola, Marburg, Lassa, and Nipah viruses. By pairing HiBiT-PsVLPs with Fc gamma receptor (FcγR)-expressing cells, we also demonstrate the capability of this system to detect antibody-dependent enhancement (ADE) of virus entry as an important safety consideration for immunotherapies. Combined, these data establish the utility of the HiBiT-PsVLP platform for safely and rapidly measuring critical antibody activities across diverse viruses and drug development stages. IMPORTANCE Standard neutralization assays are often slow, labor-intensive, and restricted to high-containment facilities, thus complicating and delaying the develop ment of vaccines and antibody-based treatments. Here, we present a novel neutralization assay system using HiBiT-tagged pseudovirus-like particles (HiBiT-PsVLPs). These particles incorporate entry proteins from diverse pathogenic viruses but are non-replicating and lack viral nucleic acids, thus mitigating the biosafety risks of conventional assays. The particles encapsulate the HiBiT peptide, enabling rapid, luminescent quantitation of entry and neutralization. We demonstrate that this platform accurately measures neutralizing activity of monoclonal antibodies across development stages and sensitively detects antibody-dependent enhancement, a critical safety consideration. Altogether, HiBiT-PsVLPs offer a safe, rapid, and scalable platform to accelerate the development of vaccines and antibody therapeutics targeting a broad range of viruses.
V iral disease outbreaks are increasing in frequency alongside rising global connectiv ity, urbanization, climate change, and other socio-economic, environmental, and ecological factors (1). Over the past decade alone, the world has experienced multiple epidemics, including the 2014-2016 Ebola epidemic in Western Africa, the Zika epidemic in the Americas (2015-2016), the COVID-19 pandemic, and the Mpox global health emergency. Amidst these emerging virus outbreaks, established pathogens, such as human immunodeficiency virus (HIV), hepatitis viruses, and influenza viruses, remain major causes of global morbidity and mortality (2)(3)(4). New countermeasures are thus needed to combat both endemic and emerging viral threats.
Monoclonal antibodies (mAbs) have arisen as powerful, dual-modality interventions with niche roles in antiviral defense (5). By virtue of their high affinity and specific ity, mAbs can serve as potent prophylactics and therapeutics with minimal offtarget effects. As prophylactics, mAbs confer immediate protection against infection, provid ing an important stopgap for vaccine-induced immunity in the nascent stages of an outbreak (6). Such protection is especially critical for populations with high exposure risk, including healthcare professionals and close contacts of infected individuals. MAbs may also serve as supplements or alternatives to vaccination for immunocompromised individuals unlikely to mount a protective vaccine response (7). As therapeutics, mAbs can decrease infectious viral loads, reduce the risk of hospitalization and death, and potentially lower the likelihood of transmission (6).
Like antibodies elicited by vaccination or infection, mAbs exert their antiviral activity via immune effector functions (e.g., antibody-dependent cellular cytotoxicity) and/or virus neutralization (8). Neutralizing antibodies (nAbs) bind specifically to viral epito pes essential for receptor binding, membrane fusion, endosomal escape, or genome uncoating, thereby inhibiting virus entry. This ability to block infection underpins the prophylactic and therapeutic efficacy of neutralizing monoclonal antibodies (NMAbs). Despite their antiviral properties, however, virusspecific antibodies can paradoxically exacerbate disease in certain contexts. In its best-understood form, this antibodydependent enhancement (ADE) occurs when non-or sub-neutralizing antibodies bind virus particles and engage Fc gamma receptors (FcγRs) on phagocytes, increasing viral uptake, entry, and replication (9). Because neutralization and ADE are crucial to NMAb efficacy and safety, robust, mechanism-of-action (MoA)reflective methods for measuring these processes are vital for NMAb development.
A variety of methods currently exist for assessing neutralization and ADE, each with distinct advantages and limitations (10). Live virus assays, such as the plaque reduction neutralization test, are typically considered the "gold standard" for their biological relevance. However, these assays are intrinsically low-throughput and require high biocontainment for pathogenic viruses and multiple days for readout. Surrogate methods are therefore commonly deployed, including enzyme-linked immunosorb ent assay (ELISA)-based surrogate virus neutralization tests (sVNTs) and pseudovirus neutralization assays (PNAs). sVNTs measure interactions between purified viral receptor binding domains (RBDs) and their cognate receptors (11). Although fast and free of biosafety constraints, these assays feature several key drawbacks: (i) they do not reflect the full complexity of virus entry, (ii) they capture only those nAbs that disrupt RBD-receptor binding, and (iii) they cannot detect ADE. By comparison, PNAs offer a balance between biosafety and biological relevance by using replication-incompetent viruses bearing the entry protein(s) from the virus of interest (12). These pseudoviruses recapitulate virus entry into host cells, allowing for the detection of more diverse antibody MoAs than sVNTs. Reporter genes, such as those encoding fluorescent proteins or luciferases, may be packaged within pseudoviruses for quantitative assessment of neutralization and ADE. Despite these advantages, PNAs nonetheless carry biosafety risks, including the potential for replication-competent virus production and insertional mutagenesis (i.e., for lentivirus-based pseudoviruses) (13,14). Additionally, PNAs require one or more days for readout (15).
To overcome the limitations of existing assays, we utilized the HiBiT protein tagging system to develop a safe, biologically relevant method to rapidly quantify neutraliza tion and ADE. HiBiT is an 11 amino acid peptide tag that binds to its complementary polypeptide LgBiT with high affinity to reconstitute functional NanoBiT luciferase (16). By fusing HiBiT to the HIV-1 Pr55 Gag polyprotein, we derive enveloped VLPs that package the HiBiT peptide internally. This highly modular VLP platform can be pseudotyped with diverse viral glycoproteins to permit cell entry. Pairing these HiBiT-tagged pseudoviruslike particles (HiBiT-PsVLPs) with LgBiT-expressing cells yields a luminescent assay system with improved biosafety and reduced turnaround times relative to conventional PNAs. Unlike sVNTs, HiBiT-PsVLP bioassays capture MoAs beyond inhibition of RBD-receptor binding and enable detection of both neutralization and ADE. Here, we establish the feasibility and robustness of the HiBiT-PsVLP platform and demonstrate its broad applicability for NMAb development against endemic and emerging viruses.
## RESULTS
## Production and characterization of SARS-CoV-2 S HiBiT-PsVLPs
To generate HiBiT-tagged VLPs, we fused HiBiT to the C-terminus of HIV-1 Pr55 Gag (Fig. 1a). Transient expression of this Gag HiBiT fusion protein yielded electron-dense struc tures in culture supernatants consistent in size (~100 nm) and morphology with HIV-1 Gag VLPs (Fig. 1b) (17,18). HiBiT detection assays indicated successful packaging of Gag HiBiT within enveloped VLPs, in contrast to a secreted, non-enveloped control protein (PCSK9 HiBiT ) (Fig. 1c).
We next sought to pseudotype Gag HiBiT VLPs with the SARS-CoV-2 Spike protein. To this end, we co-expressed Gag HiBiT with truncated Spike protein lacking the cytoplas mic tail, as removal of this endodomain has been shown to improve pseudotyping (19). Immunoblotting confirmed Spike expression and processing, which were unaffec ted by Gag HiBiT co-expression (Fig. 1d). Minimal Spike secretion was observed in the absence of Gag HiBiT , whereas secretion, primarily of processed Spike, was enhanced by Gag HiBiT co-expression. Electron microscopy revealed VLPs with distinctive crown-like halos characteristic of coronavirus Spike proteins (Fig. 1e). HiBiT content per particle was similar across three independent batches of SARS-CoV-2 S HiBiT-PsVLPs and was increased relative to non-pseudotyped, Bald HiBiT-VLPs, indicating that Spike incorpora tion did not reduce HiBiT packaging efficiency (Fig. 1f). Combined, these data demon strate successful production of enveloped Gag HiBiT VLPs pseudotyped with SARS-CoV-2 Spike protein (Fig. 1g).
## Cell entry by SARS-CoV-2 S HiBiT-PsVLPs
To measure entry and neutralization of SARS-CoV-2 S HiBiT-PsVLPs, we generated a clonal 293T target cell line stably expressing LgBiT, SARS-CoV-2 receptor ACE2, and the Spike-activating protease TMPRSS2 (Fig. S1) (20). SARS-CoV-2 S HiBiT-PsVLP entry into these target cells is predicted to culminate in NanoBiT luciferase complementa tion and luminescence (Fig. 2a). Indeed, the addition of SARS-CoV-2 S HiBiT-PsVLPs to target cells resulted in a luminescent signal that increased rapidly from 0 to 2 h and plateaued by 3 to 4 h (Fig. 2b). HiBiT-PsVLP entry was Spike-dependent, as Bald HiBiT-VLPs yielded similar luminescence to target cells alone. SARS-CoV-2 S HiBiT-PsVLP entry was ACE2-dependent and enhanced by TMPRSS2 expression (Fig. 2c). Conversely, TMPRSS2 inhibition by camostat mesylate reduced SARS-CoV-2 S HiBiT-PsVLP entry into ACE2 + /TMPRSS2 + target cells but had no effect on entry of HiBiT-PsVLPs pseudotyped with Vesicular Stomatitis Virus glycoprotein (VSV-G) (Fig. 2d). Collectively, these results demonstrate rapid luminescent detection of SARS-CoV-2 Spike-mediated cell entry using the HiBiT-PsVLP system.
## Neutralization of SARS-CoV-2 S HiBiT-PsVLPs by anti-Spike NMAbs
Antibody neutralization of SARS-CoV-2 S HiBiT-PsVLPs was first assessed using freshly cultured target cells and a biosimilar of anti-Spike NMab bamlanivimab (21). Given the entry kinetics observed in Fig. 2b, a neutralization assay endpoint of 3 h was chosen. Bamlanivimab biosimilar reduced assay luminescence in a concentration-dependent manner, fully neutralizing SARS-CoV-2 S HiBiT-PsVLPs at the highest concentrations tested (Fig. 3a andb). Similar results were obtained using cryopreserved, thaw-anduse target cells, which expressed ACE2 and TMPRSS2 at levels comparable to freshly cultured cells (Fig. 3c; Fig. S1c). Thaw-and-use cells (commonly referred to as ready-to-use or assay-ready cells) offer distinct advantages over continuous cell culture, including greater assay reproducibility, cost/time savings, and flexibility in scheduling assays (22). As such, thaw-and-use target cells were utilized for neutralization assays throughout the remainder of this study.
Specificity of SARS-CoV-2 S HiBiT-PsVLP neutralization was evaluated using anti-Spike mAb clones previously isolated from convalescent patients (23). We selected a HiBiT-PsVLP input of ~1-2 × 10 5 RLUs per well, which yields >5-fold signal-to-background and falls within a robust operating range for neutralization (Fig. S2). This input was used for all subsequent neutralization experiments. NMAb clone 414-1 neutralized SARS-CoV-2 S HiBiT-PsVLPs, whereas the non-neutralizing anti-Spike clone 415-6 had no effect (Fig. 3d). KZ52, an NMAb targeting the Ebola virus glycoprotein (EBOV GP), also failed to neutralize SARS-CoV-2 S HiBiT-PsVLPs, further demonstrating specificity for anti-Spike NMAbs (24).
To assess compatibility with early-stage NMAb discovery workflows, we compared assay performance in 96-vs 384-well plate formats. Raw luminescence values decreased in 384-well format, as expected given reduced assay volumes (data not shown). Nonetheless, HiBiT-PsVLP neutralization aligned closely between formats, confirming the assay can be miniaturized for high-throughput workflows (Fig. 3e). During NMAb discovery, candidate NMAbs are routinely expressed in mammalian cells, and mAb-containing supernatants are screened for neutralizing activity (25). To simulate this approach, we generated human IgG1 expression constructs bearing the variable regions of anti-Spike NMAb imdevimab (IMD-huIgG1) or the anti-EBOV GP NMAb mAb114 (mAb114-huIgG1) (26,27). Culture supernatants from cells transfected with these constructs were assayed for neutralization of SARS-CoV-2 S and EBOV GP HiBiT-PsVLPs (Fig. 3f). Supernatant from IMD-huIgG1 expressing cells neutralized SARS-CoV-2 S but not EBOV GP HiBiT-PsVLPs. Conversely, supernatant from mAb114-huIgG1 expressing cells neutralized EBOV GP but not SARS-CoV-2 S HiBiT-PsVLPs. These results confirm viral glycoproteinspecific neutralization of HiBiT-PsVLPs by transiently expressed NMAbs in culture supernatants.
## Qualification of the SARS-CoV-2 HiBiT-PsVLP bioassay
The SARS-CoV-2 S HiBiT-PsVLP neutralization bioassay was qualified in accordance with International Conference on Harmonization of Technical Requirements for Pharmaceuti cals for Human Use Q2(R2) guidelines (28). Results are summarized in Table 1 and Fig. 4. Assay specificity was confirmed in Fig. 3d. The assay was linear (R 2 = 0.996) across a range of 50% to 200% relative potency (Fig. 4a). Figure 4b shows dose-response curves from a representative assay comparing NMAb samples at 50%, 70%, 100%, 150%, and 200% relative potency. Repeatability (n = 6) was 3.9% (Fig. 4c). Intermediate precision was 11.2%. Assay accuracy across the different relative potencies ranged from 96.8% to 105.6% (Table 1). Stability-indicating properties of the assay were tested using heat-stressed NMAb samples. The assay detected a loss of NMAb potency with increasing incubation time at 65°C relative to an unstressed sample maintained at 4°C (Fig. 4d). Relative potencies decreased by ~83% and ~91% for samples incubated at 65°C for 24 or 48 h, respectively, confirming that the assay is stability-indicating. Together, these results demonstrate that the analytical performance of this bioassay is appropriate for quality testing of NMAbs in accordance with Good Manufacturing Practices (GMP).
## Benchmarking against a surrogate virus neutralization test
sVNTs are an established method commonly used to assess antibody neutralization without the need for virus or pseudovirus (10,11). Like HiBiT-PsVLPs, sVNTs largely alleviate biosafety concerns and provide a rapid and quantitative measurement of NMAb activity. Given these similarities, we sought to compare the SARS-CoV-2 HiBiT-PsVLP bioassay with a commercially available sVNT using a panel of anti-Spike mAbs (11). Non-neutralizing anti-Spike clone 415-6 was used as a negative control and showed minimal activity in both tests (Fig. 5a). Conversely, NMAbs targeting the Spike RBD exhibited concentration-dependent neutralizing activity in both assays, yielding sigmoidal, 4-parameter logistic dose-response curves with R 2 values ≥0.95 (Fig. 5b through k). Neutralization potencies, expressed as half-maximal inhibitory concentra tions (IC 50 values), were determined for each of these NMAbs and compared between assays. A strong positive correlation (Pearson r = 0.79) was observed between the HiBiT-PsVLP bioassay and the sVNT (Fig. 5m).
Notably, NMAb clone 4A8 targeting the Spike N-terminal domain (NTD) exhibited neutralizing activity exclusively in the HiBiT-PsVLP bioassay (Fig. 5l). This result was expected, as the sVNT includes only the purified Spike RBD, whereas HiBiT-PsVLPs incorporate the full ectodomain and transmembrane region of Spike. Thus, while highly correlative to the sVNT for RBDspecific NMAbs, the HiBiT-PsVLP bioassay offers broader epitope coverage and detection of more diverse MoAs.
## HiBiT-PsVLPs for anti-HIV NMAb development
Given the inherent flexibility of pseudotyping, we predicted that the HiBiT-PsVLP platform could be adapted for other clinically relevant viruses. Despite a sharp decline in HIV-related deaths since their peak in 2004, more than 500,000 such deaths still occur annually (2). Thus, there is a continued need for new and improved countermeasures against HIV, including NMAbs and other prophylactics capable of reducing virus transmission.
HIV-1 neutralization is commonly assessed using the TZM-bl assay, a specialized PNA involving transactivation of cellular reporter genes by pseudovirus-encoded HIV Tat protein (29). Although this assay offers improved biosafety compared to replication-com petent HIV-1, it nonetheless requires multiple days for readout. We therefore reasoned that HiBiT-PsVLPs could provide a faster method for assaying HIV-1 neutralization. To test this prediction, we first produced a 293T target cell line stably expressing LgBiT along with the HIV-1 receptor CD4 and co-receptors CCR5 and CXCR4 (Fig. S3). We next generated HiBiT-PsVLPs bearing the HIV-1 envelope (Env) protein from two different CCR5-tropic isolates, BaL and JRFL, and one CXCR4-tropic strain, NL4-3. The endodomain of each HIV Env protein was truncated to improve pseudotyping (30). Addition of HIV Env HiBiT-PsVLPs to target cells resulted in increased luminescence over time, with peak entry observed by 4 to 5 h (Fig. 6a).
For neutralization studies, we utilized HiBiT-PsVLPs bearing Env protein from the JRFL isolate. This isolate exhibits a tier 2 neutralization phenotype considered typical of most circulating HIV-1 strains (31,32). HIV Env(JRFL) HiBiT-PsVLPs were challenged with VRC01, a potent anti-HIV gp120 NMAb that targets the CD4 binding site and neutralizes ~90% of HIV-1 isolates (33). Based on results from Fig. 6a, a 4 h assay endpoint was selected. HIV Env(JRFL) HiBiT-PsVLPs were neutralized by VRC01, whereas a nonspecific control antibody had no impact on entry (Fig. 6b). These data demonstrate the feasibility of using HiBiT-PsVLPs to probe HIV-1 neutralization with dramatically reduced turnaround times relative to the TZM-bl assay.
## HiBiT-PsVLPs for emerging viral pathogens
The World Health Organization and Coalition for Epidemic Preparedness Innovations have identified panels of key emerging viruses to guide global pandemic preparedness efforts (34,35). These priority pathogens were selected based on their epidemic and pandemic potential, their high rates of morbidity and mortality, and the lack of specific countermeasures against them. As many of these viruses are Risk Group four agents, we predicted that the HiBiT-PsVLP platform could provide a safe yet biologically relevant surrogate for assessing their neutralization. To address this hypothesis, we first generated HiBiT-PsVLPs bearing the glycoprotein (GP) from three distinct filoviruses: Zaire Ebola virus (EBOV), Sudan Ebola virus (SEBOV), or Marburg virus (MARV), each of which causes severe hemorrhagic fever in humans. Despite licensed vaccines and NMAbs for EBOV, no specific countermeasures currently exist for MARV or SEBOV.
Entry of filovirus GP HiBiT-PsVLPs was confirmed using LgBiT-expressing 293T cells (Fig. 7a). Filovirus GP-mediated entry was notably slower than that driven by SARS-CoV-2 S or HIV Env, consistent with differences in viral entry pathways (Fig. S4). Whereas SARS-CoV-2 S and HIV Env can mediate entry at the plasma membrane, filovirus GPs require trafficking to late endosomes and/or lysosomes for binding to the intracellular receptor NPC-1 and subsequent membrane fusion (36)(37)(38). This additional trafficking step likely delays luminescence onset in the HiBiT-PsVLP assay. Nonetheless, each filovirus HiBiT-PsVLP was neutralized by an NMAb targeting its GP (Fig. 7b through d), demonstrating specific neutralization despite pathway-dependent differences in entry kinetics.
We further extended the HiBiT-PsVLP platform to two other priority pathogens: (i) Lassa virus (LASV), an arenavirus causing acute hemorrhagic illness, and (ii) Nipah virus (NiV), a paramyxovirus causing acute respiratory disease and fatal encephalitis. LgBiT-expressing 293T cells were selected as the target cell line, as they endogenously express the primary LASV receptor, α-dystroglycan, and NiV receptors, EphrinB2 and EphrinB3 (39)(40)(41)(42). HiBiT-PsVLPs bearing the LASV glycoprotein complex (GPC) entered target cells and were neutralized specifically by an anti-LASV GPC NMAb (Fig. 7e andf). NiV HiBiT-PsVLPs were generated using the viral attachment glycoprotein (G) and an endodomain-truncated viral fusion protein (F) (43). These HiBiT-PsVLPs entered target cells and were neutralized by an anti-NiV F NMAb (Fig. 7g andh). Together, these studies demonstrate the flexibility of the HiBiT-PsVLP platform and its potential to facilitate NMAb development across a variety of emerging viral pathogens.
## Measuring antibody-dependent enhancement with HiBiT-PsVLPs
ADE assays typically rely on live viruses or pseudoviruses and are subject to many of the same pitfalls as conventional neutralization assays. We therefore sought to deter mine whether HiBiT-PsVLPs could provide a safe and rapid surrogate for measuring ADE. To this end, we engineered LgBiT expression into THP-1 human monocyte cells that endogenously express FcγRI and FcγRIIa (Fig. S5) (44,45). We then assayed entry of EBOV GP HiBiT-PsVLPs into this target cell line in the presence of different mAbs. EBOV GP HiBiT-PsVLP entry was enhanced approximately sevenfold by cosfroviximab, a non-neutralizing anti-EBOV-GP mAb previously shown to induce ADE in vitro (Fig. 8a) (46). Isotype-matched control mAb lacking specificity for EBOV GP had no effect on HiBiT-PsVLP entry. In contrast to cosfroviximab, anti-EBOV-GP NMAb KZ52 yielded a bi-phasic response, enhancing entry at low concentrations and neutralizing HiBiT-PsVLPs at higher concentrations (Fig. 8b). The Fc domain of KZ52 was required for enhancement but was dispensable for neutralization, consistent with a previous report (Fig. 8c) (24). To confirm the role of FcγRs, we generated THP-1/HaloTag-LgBiT CRISPR knockout lines lacking FcγRI, FcγRII, or both receptors (Fig. S5). ADE by cosfroviximab and KZ52 was reduced or eliminated by knockout of FcγRI, whereas FcγRII knockout had no significant effect on ADE (Fig. 8d ande). Neutralizing activity of KZ52 was preserved across the different knockout lines (Fig. 8d). Taken together, these results are consistent with the detection of classical, FcγR-associated ADE driven primarily by mAb Fc interaction with FcγRI.
## DISCUSSION
NMAbs play unique roles in antiviral defense, with potent and, ideally, broadly act ing NMAbs serving as both therapeutics and prophylactics. Robust, MoAreflective neutralization assays are essential to identify the most promising NMAbs and to shepherd these candidates through the development pipeline. In this study, we established the HiBiT-PsVLP system as a safe, fast, and versatile bioassay platform with applications spanning multiple stages of NMAb development. This platform is compati ble with standard discovery workflows, enables rapid characterization of NMAb potency and stability, and demonstrates performance characteristics suitable for quality release testing. Additionally, this system can be adapted to measure ADE and thereby identify potential safety concerns at early stages of NMAb development.
Surrogate neutralization assays provide safer, higher-throughput, and faster alternatives to live virus neutralization tests and are thus better suited to many research and GMP environments. Nonetheless, each surrogate assay features distinct advantages and limitations. ELISA-based methods (e.g., sVNTs) are rapid and safe but fail to capture the full complexity of virus entry, typically measuring only the interaction between a viral RBD and its cognate receptor. Conversely, pseudoviruses incorporate the full ectodomains of viral glycoproteins and recapitulate native virus entry, thus offering greater biological relevance than sVNTs. However, PNAs are dramatically slower than sVNTs, with assay times typically ranging from 1 to 4 days. By comparison, HiBiT-PsVLPs deliver the same benefits as traditional PNAs with reduced assay times approaching those of sVNTs. Biosafety of HiBiT-PsVLPs is also improved over HIV-based pseudoviruses, as HiBiT-PsVLPs lack enzymes encoded by the HIV pol gene and do not rely on packaging of viral nucleic acids. These changes mitigate the risks of insertional mutagenesis and Unlike conventional PNAs, HiBiT-PsVLP bioassays require the use of LgBiT-expressing target cells for complementation-based readout. For many widely used, "workhorse" cell lines, such as the HEK293T cells used in this study, LgBiT expression can be easily introduced via stable plasmid transfection. For more specialized models (e.g., primary cell cultures), alternative gene delivery methods, such as LgBiT mRNA transfection or viral transduction, may be necessary. In each case, the time required for LgBiT gene delivery would be at least partially offset by the reduced turnaround times of HiBiT-PsVLP bioassays relative to PNAs. Also, unlike pseudoviruses, HiBiT-PsVLPs do not provide intrinsic signal amplification through reporter gene expression, resulting in lower signal-to-background ratios. Nonetheless, luminescence and signal-to-background are sufficient to quantitatively assess neutralization and ADE, as demonstrated in this study. Like pseudoviruses, HiBiT-PsVLPs can be rapidly adapted for different viral patho gens by exchanging the glycoprotein(s) on their surface. In our study, HiBiT:p24 ratios were elevated for SARS-CoV-2 S HiBiT-PsVLPs relative to Bald controls. Because each Gag molecule carries a single HiBiT tag, this increase is unlikely to reflect changes in HiBiT stoichiometry per Gag protein but may instead arise from differences in particle composition or detection efficiency. Importantly, the elevated ratios were similar across independent preparations, supporting the reproducibility of HiBiT incorporation. Beyond SARS-CoV-2 and HIV, we anticipate the HiBiT-PsVLP platform could be applied to other endemic viruses of global public health concern. As demonstrated here, HiBiT-PsVLPs can also be adapted for many priority viruses of epidemic/pandemic potential, highlight ing the utility of this platform for pandemic preparedness. HiBiT-PsVLPs could also be leveraged for mutagenesis studies of viral glycoproteins to identify mutations that alter NMAb sensitivity. Similarly, these assays could be utilized in the discovery and develop ment of NMAb cocktails designed to reduce the likelihood of mutational escape.
As with PNAs, the current HiBiT-PsVLP platform is limited to enveloped viruses. Additionally, glycoproteins from some enveloped viruses, such as Dengue virus, are difficult to incorporate into heterologous virions and may likewise prove incompatible with our HIV-1 Gag-based system (47). In such cases, the HiBiT tag can be appended to structural proteins derived from the same virus or related family members to generate HiBiT-tagged VLPs capable of incorporating these glycoprotein(s) (48). The HiBiT-VLP approach was also recently extended to non-enveloped viruses by fusing the HiBiT tag to structural proteins required for VLP assembly (49)(50)(51).
Although the present study focused primarily on applications for NMAb develop ment, we anticipate that HiBiT-PsVLP bioassays could be implemented for other antiviral modalities. NAb titers are considered an important correlate of protection against many viruses, and a similar HiBiT-tagged VLP approach was recently described for screening post-vaccination sera against SARS-CoV-2 (52,53). In contrast, our HiBiT-PsVLP system omits the HIV-1 pol gene for improved biosafety and employs a streamlined, homoge nous workflow well suited to high-throughput screening. The proven flexibility of our platform and its ability to assess ADE provide additional advantages for vaccine research. HiBiT-PsVLP bioassays could also aid in developing small-molecule entry inhibitors targeting viral glycoproteins or essential host factors.
Beyond antiviral drug development, we envision key applications for HiBiT-PsVLP technology in basic virology research. In our assays, luminescence kinetics generally varied in ways consistent with known entry pathways-for example, signals appeared more rapidly for viruses that fuse at the plasma membrane compared to those requiring endocytosis and intracellular trafficking. Such pathway-dependent differences suggest that HiBiT-PsVLPs could provide a biologically relevant system for probing entry mechanisms (e.g., receptor usage, endocytic dependence, pH sensitivity) in a manner similar to traditional pseudoviruses. Given their non-replicative nature, HiBiT-PsVLPs could also serve as a safe platform for screening glycoprotein mutations that alter receptor usage and host range, without many of the gain-of-function concerns associated with replication-competent systems.
## MATERIALS AND METHODS
## Cell lines and cell culture
293T (CRL-3216) and THP-1 (TIB-202) cell lines were purchased from the American Type Culture Collection (ATCC). 293T cells were cultured in Dulbecco's Modified Eagle Medium (DMEM; Gibco, 11995) supplemented with 10% fetal bovine serum (FBS; Avantor Seradigm, 89510-194), 1 mM sodium pyruvate (Gibco, 11360070), and 1× MEM nonessential amino acids (Gibco, 11140050). THP-1 cells were cultured in RPMI 1640 medium (Gibco, 22400) supplemented with 10% FBS.
## Plasmids
Sequence encoding the HiBiT peptide (VSGWRLFKKIS) was appended to the 3′ end of an HIV-1 Gag ORF codon-optimized for Rev-independent expression (54). This Gag HiBiT ORF was inserted into a mammalian expression vector under control of the Cytomegalovirus (CMV) immediate enhancer/chicken β-actin (CAG) promoter. LgBiT and HaloTag-LgBiT were encoded by LgBiT Expression Vector (Promega, N268A) and CMV HaloTag-LgBiT Vector (Addgene, 236917), respectively. PCSK9 HiBiT , human ACE2, and human CD4 ORFs were inserted into CAG or CMV promoter-driven vectors. A single ORF encoding human CCR5 and human CXCR4 separated by a P2A self-cleaving peptide linker was cloned into a CMV promoter-driven vector. Human TMPRSS2 FLAG ORF was inserted into an EF1α promoter-driven vector. Human codon-optimized ORFs encoding viral entry proteins were cloned into CAG or CMV promoter-driven vectors. The endodomains of several viral glycoproteins were truncated to improve pseudotyping based on previous reports (Table 2). Antibody expression constructs were generated via insertion of the heavy and light chain variable sequences of imdevimab or mAb114 into vectors encoding the human IgG1 heavy and lambda (imdevimab) or kappa (mAb114) light chain constant regions.
## Antibodies and reagents
Antibodies used for immunoblotting were as follows: anti-HiBiT monoclonal antibody (Promega, N7200), anti-SARS Spike protein antibody (Novus Biologicals, NB100-56578), anti-GAPDH antibody (Santa Cruz, SC-47724), anti-Mouse IgG (H+L) HRP conjugate antibody (Promega, W4021), and anti-Rabbit IgG (H+L) HRP conjugate antibody (Promega, W4011).
The following antibodies were purchased from Biolegend and used for flow cytometric and fluorescenceactivated cell sorting (FACS) analyses: PE anti-DYKDDDDK (FLAG) (637309), PE anti-human CD4 (317410), Alexa Fluor 488 anti-human CD195 (CCR5) (359103), APC anti-human CD184 (CXCR4) (306509), Brilliant Violet 421 anti-human CD64 (305020), and FITC anti-human CD32 (303204). For ACE2 detection, Human ACE2 antibody (R&D Systems, AF933) was used in combination with Alexa Fluor 647 AffiniPure Donkey Anti-Goat IgG secondary (Jackson ImmunoResearch, 705-605-147).
Antibodies used in neutralization assays are listed in Table 3. F(ab′) 2 fragment of anti-Ebola clone KZ52 was generated using the Pierce (Fab′) 2 Micro Preparation Kit (Thermo Scientific, 44688). Camostat mesylate was purchased from Sigma (SML0057).
## Transfection and stable cell line generation
293T-derived target cell lines were generated via plasmid transfections using ViaFect Reagent (Promega, E4981), followed by antibiotic selections and clonal isolation via FACS. LgBiT expression was confirmed by flow cytometry and/or by detecting LgBiT activity using the Nano-Glo HiBiT Lytic Detection System (Promega, N3040) with recombinant HaloTag-HiBiT protein (Aldevron). Expression of all other transgenes was confirmed via flow cytometry. THP-1 cells (ATCC TIB-202) were electroporated with HaloTag-LgBiT expression vector using Ingenio Electroporation Reagent (Mirus, MIR50114) with a GenePulser Xcell (Bio-Rad). Following antibiotic selection, cells were sorted for LgBiT expression via HaloTag labeling and FACS. CRISPR RNAs (crRNAs) targeting CD64 and CD32 (Integrated DNA Technologies; IDT) were duplexed with tracrRNA (IDT, 1072533) and incubated with Cas9 nuclease (IDT, 1081059) to yield ribonucleoprotein (RNP) complexes. RNPs were electroporated into the THP-1/HaloTag-LgBiT pool using Ingenio Reagent with a Nucleofector system (Lonza). Cell pools with complete knockout of FcγRI, FcγRII, or both were obtained via FACS. Final knockout phenotypes were confirmed by flow cytometry.
## Flow cytometry
Cells were washed with phosphatebuffered saline supplemented with 2% FBS, labeled with primary antibodies for 30 min on ice, and washed twice with PBS + 2% FBS. Where necessary, cells were labeled with secondary antibody for 30 min on ice, followed by two additional rounds of washing. For detection of HaloTag-LgBiT, cells were labeled with Janelia Fluor 646 HaloTag Ligand (Promega GA1121) for 30 min at 37°C and washed once with PBS + 2% FBS. Samples were analyzed on a Fortessa X-20 flow cytometer (BD), and data were analyzed using FlowJo Software (BD).
## Production and characterization of HiBiT-PsVLPs
HiBiT-PsVLPs were generated via transient transfection of 293T producer cells using ViaFect Reagent with Gag HiBiT expression vector and plasmid(s) encoding the specified viral glycoprotein(s). Transfection Carrier DNA (Promega, E4881) was substituted for plasmids encoding viral glycoproteins to generate Bald HiBiT-VLP controls. Culture supernatants were collected 48 to 72 h post-transfection, centrifuged at 1,000 × g for 10 min at 4°C, and passaged through 0.45 µm PVDF filters (Merck Millipore, SLHVR33RS). HiBiT-PsVLP preps were used immediately or stored at -80°C. HiBiT packaging was assessed using the Nano-Glo HiBiT Lytic Detection System and the Nano-Glo HiBiT Extracellular (i.e., non-lytic) Detection System (Promega, N2421). HiBiT content per particle, expressed as the ratio of HiBiT to HIV-1 p24 protein, was determined using the Nano-Glo HiBiT Lytic Detection System in parallel with the Lumit p24 Immunoassay (Promega, CS2039B25). Ratios were normalized to a reference batch of Bald HiBiT-VLPs. For immunoblotting, culture supernatants and 293T producer cells were harvested following HiBiT-PsVLP production and treated with mammalian lysis buffer (Promega, G9381) containing protease inhibitors (Promega, G6521). Protein concentrations in cell lysates were quantified using a BCA assay (ThermoFisher, 23227). Samples were separated via SDS-PAGE on a precast 4%-20% Criterion TGX protein gel (Bio-Rad, 5671094) and transferred onto nitrocellulose membranes using the iBlot 2 Dry Blotting System (Invitrogen). Membranes were blocked with trisbuffered saline containing 0.1% Tween-20 (TBST) and 5% non-fat dry milk for 1 h at room temperature, washed with TBST, and incubated overnight with primary antibodies at 4°C. After washing with TBST, membranes were incubated with secondary antibodies for 1 h at room temperature, washed again with TBST, and incubated for 1 min with ECL Western Blotting Substrate (Promega, W1001). Blots were imaged on a ChemiDoc MP Imaging System with Image Lab software (Bio-Rad).
For electron microscopy, HiBiT-PsVLPs were produced as described above and concentrated using Amicon Ultra-15 Centrifugal Filters with a 100 kDa molecular weight cutoff (Merck Millipore, UFC9100008). Concentrates were loaded onto a 20% sucrose cushion, centrifuged at 112,000 × g for 70 min at 10°C using an Optima TLX Ultracentri fuge (Beckman Coulter), and pellets were resuspended in PBS. Samples were nega tively stained with Nano-W (Nanoprobes) on a formvar-coated 300 mesh Cu Thin-Bar grid (EMS), coating side down using a two-step method. Samples were viewed on a Philips CM120 transmission electron microscope at 80 kV and documented with an AMT BioSprint12 digital camera at the Electron Microscope Core of the University of Wisconsin.
## HiBiT-PsVLP entry assays
For time-course analyses, the indicated target cells were incubated for 1 h at 37°C, 5% CO 2 in DMEM + 0.5% FBS containing Nano-Glo Vivazine Substrate (Promega, N2581) in 96-well white flatbottom assay plates (Corning). Bio-Glo-NB DrkBiT Peptide (Promega), a membrane-impermeable peptide that binds free LgBiT to prevent extracellular NanoBiT complementation, was added during this pre-incubation step to eliminate background luminescence unrelated to HiBiT-PsVLP entry. HiBiT-PsVLPs were overlaid onto target cells, and luminescence was measured at the indicated time points using a GloMax Discover Microplate Reader (Promega). Specificity of SARS-CoV-2 S HiBiT-PsVLP entry was assayed by overlaying HiBiT-PsVLPs onto 293T cells expressing LgBiT with or without ACE2 and TMPRSS2. Assay plates were incubated at 37°C, 5% CO 2 for 4 h, followed by the addition of Bio-Glo-NB Live Cell Reagent (Promega), further incubation at 37°C for 15 min, and luminescence measurement. Data were analyzed using GraphPad Prism Software.
## Neutralization assays
Frozen stocks of HiBiT-PsVLPs were thawed at 37°C and incubated with serial mAb titrations in DMEM + 0.5% FBS for 30 min at 37°C, 5% CO 2 in 96-well white flatbottom assay plates. Target cells were incubated with Bio-Glo-NB DrkBiT Peptide in DMEM + 0.5% FBS for 15 min at 37°C, 5% CO 2 , and added to assay wells containing the HiBiT-PsVLPs and mAbs. Mock control wells containing target cells alone were included on each plate. Assay plates were incubated at 37°C, 5% CO 2 . Incubation times were selected based on glycoproteinspecific entry kinetics of HiBiT-PsVLPs: 3 h for SARS-CoV-2 S, 4 h for HIV Env, 5 h for LASV GPC and NiV-F/G, 22 h for MARV GP and SEBOV GP, and 24 h for EBOV GP. Luminescence was determined by adding the Bio-Glo-NB Live Cell Reagent at assay endpoints, as described above. Baseline-corrected luminescence was determined by subtracting the average RLU value measured in Mock control wells.
Percent neutralization was calculated as 100 × (1 -[RLU +inhibitor /RLU no inhibitor ]). 4-PL curves were plotted using GraphPad Prism Software. Neutralization assays in 384-well white flatbottom assay plates (Corning) were performed as above, albeit with reduced volumes of each assay component.
To assess neutralization by transiently expressed mAbs, 293T cells were transfec ted with IMD-or mAb114-huIgG1 expression plasmids. Supernatants were collected 72 h post-transfection, and neutralizing activity was assayed as above. For TMPRSS2 inhibition, a serial titration of camostat mesylate was prepared in DMEM + 0.5% FBS and incubated with target cells and Bio-Glo-NB DrkBiT Peptide for 25 min at 37°C, 5% CO 2 . HiBiT-PsVLPs were then added, and the remainder of the assay was performed as above.
The SARS-CoV-2 Surrogate Virus Neutralization Test (Genscript, L00847-A) was performed according to the manufacturer's protocol. NMAb IC 50 values derived from the sVNT were plotted against those from HiBiT-PsVLP bioassays, and a linear correlation analysis was performed using GraphPad Prism Software.
## SARS-CoV-2 HiBiT-PsVLP bioassay qualification
Assay linearity was determined by two separate analysts across three independent assays each. For each assay, titrations of cilgavimab biosimilar were prepared at 50%, 70%, 100%, 150%, or 200% potency relative to a reference titration. JMP software (JMP) was used to plot 4-PL curves, test parallelism (F test), and determine relative potency values. Measured relative potency values were plotted against expected values using GraphPad Prism Software, and data were fitted with a linear regression. The mean and standard deviation of measured potency values were used to calculate accuracy (i.e., %recov ery) and intermediate precision. Repeatability was determined from six independent titrations of cilgavimab biosimilar analyzed in a single experiment. For stability-indicating studies, cilgavimab biosimilar was incubated at 4°C or at 65°C for the indicated lengths of time before performing the HiBiT-PsVLP neutralization bioassay.
## Antibody-dependent enhancement assays
MAbs or F(ab′) 2 fragments were prepared in RPMI + 0.5% FBS and added to EBOV GP HiBiT-PsVLPs in 96-well white flatbottom assay plates. THP-1/HaloTag-LgBiT target cell lines were incubated with Bio-Glo-NB DrkBiT Peptide in RPMI + 0.5% FBS for 15 min at 37°C, 5% CO 2 , then added to assay wells. Assay plates were incubated for 24 h at 37°C, 5% CO 2 , and luminescence was determined by adding the Bio-Glo-NB Live Cell Reagent at the assay endpoint, as described above. Fold changes in luminescence were calculated as RLU +antibody /RLU no antibody . Data were analyzed using GraphPad Prism Software.
## References
1. Baker, Mahmud, Miller et al. (2022) "Infectious disease in an era of global change" *Nat Rev Microbiol*
2. Carter, Zhang, Tram et al. (2024) "Global, regional, and national burden of HIV/AIDS, 1990-2021, and forecasts to 2050" *Lancet HIV*
3. (2022) "Global hepatitis report 2024: action for access in low-and middle-income countries. Web Annex. Method for global reporting on disease burden and service coverage data for viral hepatitis B and C"
4. Uyeki, Hui, Zambon et al. (2022) *Influenza. Lancet*
5. Pantaleo, Correia, Fenwick et al. (2022) "Antibodies to combat viral infections: development strategies and progress" *Nat Rev Drug Discov*
6. Gupta, Purcell, Corti et al. (2023) "Pandemic preparedness strategies must go beyond vaccines" *Sci Transl Med*
7. Cowan, Amson, Christofides et al. (2023) "Monoclonal antibodies as COVID-19 prophylaxis therapy in immunocompromised patient populations" *Int J Infect Dis*
8. Burton (2023) "Antiviral neutralizing antibodies: from in vitro to in vivo activity" *Nat Rev Immunol*
9. Wells, Esposito, Henderson et al. (2025) "Mechanisms of antibody-dependent enhancement of infectious disease" *Nat Rev Immunol*
10. Rocha, Quadros, Fernandes et al. (2023) "An overview of the conventional and novel methods employed for SARS-CoV-2 neutralizing antibody measurement" *Viruses*
11. Tan, Chia, Qin et al. (2020) "A SARS-CoV-2 surrogate virus neutralization test based on antibodymediated blockage of ACE2-spike protein-protein interaction" *Nat Biotechnol*
12. Xiang, Li, Wu et al. (2022) "Application of pseudovirus system in the development of vaccine, antiviral-drugs, and neutralizing antibodies" *Microbiol Res*
13. Schlimgen, Howard, Wooley et al. (2016) "Risks associated with lentiviral vector exposures and prevention strategies" *J Occup Environ Med*
14. Bilska, Tang, Montefiori (2017) "Short communication: potential risk of replication-competent virus in HIV-1 Env-pseudotyped virus preparations" *AIDS Res Hum Retroviruses*
15. Nie, Li, Wu et al. (2020) "Quantification of SARS-CoV-2 neutralizing antibody by a pseudotyped virus-based assay" *Full-Length Text Journal of Virology*
16. Schwinn, Machleidt, Zimmerman et al. (2018) "CRISPR-mediated tagging of endogenous proteins with a luminescent peptide" *ACS Chem Biol*
17. Gheysen, Jacobs, De Foresta et al. (1989) "Assembly and release of HIV-1 precursor Pr55 gag virus-like particles from recombinant baculovirus-infected insect cells" *Cell*
18. Wagner, Fliessbach, Wanner et al. (1992) "Studies on processing, particle formation, and immunogenicity of the HIV-1 gag gene product: a possible component of a HIV vaccine" *Arch Virol*
19. Schmidt, Weisblum, Muecksch et al. (2020) "Measuring SARS-CoV-2 neutralizing antibody activity using pseudotyped and chimeric viruses" *J Exp Med*
20. Hoffmann, Kleine-Weber, Schroeder et al. (2020) "SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor" *Cell*
21. Jones, Brown-Augsburger, Corbett et al. (1906) "The neutralizing antibody, LY-CoV555, protects against SARS-CoV-2 infection in nonhuman primates" *Sci Transl Med*
22. Riss, Moravec, Duellman et al. (2021) "Treating cells as reagents to design reproducible assays"
23. Wan, Xing, Ding et al. (2020) "Human-IgG-neutralizing monoclonal antibodies block the SARS-CoV-2 infection" *Cell Rep*
24. Maruyama, Rodriguez, Jahrling et al. (1999) "Ebola virus can be effectively neutralized by antibody produced in natural human infection" *J Virol*
25. Rogers, Zhao, Huang et al. (2020) "Isolation of potent SARS-CoV-2 neutralizing antibodies and protection from disease in a small animal model" *Science*
26. Corti, Misasi, Mulangu et al. (2016) "Protective monother apy against lethal Ebola virus infection by a potently neutralizing antibody" *Science*
27. Hansen, Baum, Pascal et al. (2020) "Studies in humanized mice and convalescent humans yield a SARS-CoV-2 antibody cocktail" *Science*
28. (2023) "International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use"
29. Montefiori (2009) "Measuring HIV neutralization in a luciferase reporter gene assay" *Methods Mol Biol*
30. Zhang, Deng, Fang et al. (2022) "Endodomain truncation of the HIV-1 envelope protein improves the packaging efficiency of pseudovi ruses" *Virology (Auckl)*
31. Seaman, Janes, Hawkins et al. (2010) "Tiered categorization of a diverse panel of HIV-1 Env pseudovi ruses for assessment of neutralizing antibodies" *J Virol*
32. Montefiori, Roederer, Morris et al. (2018) "Neutralization tiers of HIV-1" *Curr Opin HIV AIDS*
33. Wu, Yang, Li et al. (2010) "Rational design of envelope identifies broadly neutralizing human monoclonal antibodies to HIV-1" *Science*
34. (2024) "Pathogens prioritization: a scientific framework for epidemic and pandemic research preparedness"
35. Gouglas, Christodoulou, Plotkin et al. (2019) "CEPI: driving progress toward epidemic preparedness and response" *Epidemiol Rev*
36. Carette, Raaben, Wong et al. (2011) "Ebola virus entry requires the cholesterol transporter Niemann-Pick C1" *Nature*
37. Côté, Misasi, Ren et al. (2011) "Small molecule inhibitors reveal Niemann-Pick C1 is essential for Ebola virus infection" *Nature*
38. Chandran, Sullivan, Felbor et al. (2005) "Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection" *Science*
39. Cao, Henry, Borrow et al. (1998) "Identification of αdystroglycan as a receptor for lymphocytic choriomeningitis virus and Lassa fever virus" *Science*
40. Negrete, Levroney, Aguilar et al. (2005) "EphrinB2 is the entry receptor for Nipah virus, an emergent deadly paramyxovirus" *Nature*
41. Negrete, Wolf, Aguilar et al. (2006) "Two key residues in ephrinB3 are critical for its use as an alternative receptor for Nipah virus" *PLoS Pathog*
42. Bonaparte, Dimitrov, Bossart et al. (2005) "Ephrin-B2 ligand is a functional receptor for Hendra virus and Nipah virus" *Proc Natl Acad Sci*
43. Palomares, Vigant, Handel et al. (2013) "Nipah virus envelope-pseudotyped lentiviruses efficiently target ephrinB2-positive stem cell populations in vitro and bypass the liver sink when administered in vivo" *J Virol*
44. Fleit, Kobasiuk (1991) "The human monocyte-like cell line THP-1 expresses FCγRI and FCγRII" *J Leukoc Biol*
45. Mackay, Stanevsky, Wang et al. (2006) "Selective dysregulation of the FcγIIB receptor on memory B cells in SLE" *J Exp Med*
46. Kuzmina, Younan, Gilchuk et al. (2018) "Antibodydependent enhancement of Ebola virus infection by human antibodies isolated from survivors" *Cell Rep*
47. Hu, Hsieh, King et al. (2007) "Characterization of retrovirus-based reporter viruses pseudotyped with the precursor membrane and envelope glycoproteins of four serotypes of dengue viruses" *Virology (Auckl)*
48. Sasaki, Anindita, Phongphaew et al. (2018) "Development of a rapid and quantitative method for the analysis of viral entry and release using a NanoLuc luciferase comple mentation assay" *Virus Res*
49. Qiao, Wang, Zhang et al. (2022) "A split NanoLuc complemen tation-based human norovirus-like particle entry assay facilitates evaluation of anti-norovirus antibodies in live cells" *Antiviral Res*
50. (2025) *Full-Length Text Journal of Virology*
51. Kimura, Sekiguchi, Okitsu et al. (2022) "A highly quantitative detection system for cell entry of human norovirus-like particles based on the complementation of NanoLuc luciferase" *Virology (Auckl)*
52. Ishida, Noguchi, Kimura et al. (2023) "Tracking of human parvovirus B19 virus-like particles using short peptide tags reveals a membrane-associated extracellular release of these particles" *J Virol*
53. Miyakawa, Jeremiah, Kato et al. (2022) "Rapid detection of neutralizing antibodies to SARS-CoV-2 variants in post-vaccination sera" *J Mol Cell Biol*
54. Miyakawa, Jeremiah, Ohtake et al. (2020) "Rapid quantitative screening assay for SARS-CoV-2 neutralizing antibodies using HiBiT-tagged virus-like particles" *J Mol Cell Biol*
55. Sabo, De, Santos et al. (2020) "IQGAP1 negatively regulates HIV-1 gag trafficking and virion production" *Cell Rep* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12585652&blobtype=pdf | # Characterization of Small Molecule Inhibition of Avian Influenza Hemagglutinin
Amir Shimon
## Abstract
Influenza virus causes over 300,000 respiratory deaths per year. Moreover, the 1918 H1N1 influenza pandemic resulted in over 50 million deaths worldwide. Of particular concern are influenza strains H5N1 and H7N9 that currently circulate in avians and occasionally cross over to humans. In the case of humans infected with avian influenza, the mortality rate can rise to >50%. Influenza entry is mediated by the membrane protein hemagglutinin (HA). The Caffrey laboratory is interested in characterizing the HA mechanism of action and exploit this information for the design of novel antiviral therapeutics. In this poster we present the characterization of HA and its interactions by a combination of disciplines including virology, biochemistry, x-ray crystallography and Cryo EM. Specifically, potential inhibitors are discovered by screening libraries of small molecules using a virus-like-particle (VLP) system. Binding of the small molecule inhibitor is validated by NMR, Thermal Shift Assay (TSA), and Surface Plasmon Resonance (SPR). X-ray crystallography and Cryo EM are then used to give high resolution structural information of the binding site, thereby giving insight into the mechanism of drug inhibition and guidance for chemical optimization. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724223&blobtype=pdf | # Ancestral reconstruction supports that loss of Nef-mediated T cell modulation coincided with the emergence of pathogenic lentiviruses
Angelina Baldino, Mitchell Mumby, Cassandra Edgar, Abayomi Olabode, Art Poon, Jimmy Dikeakos
## Abstract
Human immunodeficiency virus type 1 (HIV-1) originated following cross-species transmission of simian immunodeficiency virus (SIV) infecting chimpanzees (SIVcpz). While SIV infection of its natural primate hosts is typically non-pathogenic, SIVcpz/HIV-1 is highly pathogenic. SIV/HIV pathogenesis is influenced by the viral accessory protein Nef, which manipulates several immune factors to facilitate disease progression. Specifically, Nef can modulate T cell activation by downregulating the T cell receptor subunit CD3ζ and the co-stimulatory receptor CD28. Several studies have demonstrated that while Nef downregulates CD28 from the T cell surface, this ability is not completely conserved throughout the SIV/HIV phylogeny. In comparison, Nef-mediated CD3 downregulation is highly conserved in SIV derived from lower-order primates yet completely absent from SIVcpz/HIV-1. Here, we used Nef proteins previously generated via an ancestral reconstruction pipeline, corresponding to common ances tors in the SIV/HIV phylogeny leading to HIV-1 group M, and evaluated their ability to downregulate CD3ζ and CD28 in both expression and transduction systems. We observed that only the ancestral Nef of all primate lentiviruses efficiently downregulated both CD3ζ and CD28, while ancestors along the SIVcpz/HIV-1 lineage failed to down regulate CD3ζ and minimally recovered the ability to downregulate CD28 over time. Furthermore, we showed that the loss of CD3ζ and CD28 downregulation was associated with increased expression of cell surface T cell activation markers and inflammatory cytokines, consistent with a pathogenic infection. Altogether, this study illustrates the evolutionary dynamics of Nef modulation of T cell activation and its impact on patho genesis in the SIV/HIV lineage.
IMPORTANCEWe characterized ancestral Nef proteins derived from the lineage that gave rise to HIV-1 group M. We show that the loss of Nef-mediated CD3ζ and CD28 downregulation occurred in the common ancestor of the SIVcpz/HIV-1 lineage and coincided with increased T cell activation, a hallmark of pathogenicity. Moreover, this suggests that evolutionary changes to Nef-mediated immune modulation contributed to the emergence of pathogenic viruses like HIV-1. Together, these findings support the application of ancestral reconstruction to better understand the functional evolution of viral proteins, particularly in the context of cross-species transmission. KEYWORDS ancestral reconstruction, Nef, simian immunodeficiency virus, human immunodeficiency virus H uman immunodeficiency virus type 1 (HIV-1) was introduced to humans following multiple zoonotic transmission events, leading to four HIV-1 subgroups (M, N, O, and P), where group M (HIV-1/M) is the major causal agent of acquired immunodefi ciency syndrome (AIDS) (1). HIV-1 arose following cross-species transmission events of
simian immunodeficiency virus (SIV) infecting the common chimpanzee, Pan troglodytes (SIVcpz) (2)(3)(4)(5). The full-length proviral sequence of SIVcpz is a genomic mosaic generated from viral recombination events. Notably, the 5′ portion, nef gene, and 3′ long terminal repeat resemble SIV infecting red-capped mangabeys (Cercocebus torquatus; SIVrcm), while the vpu, tat, rev, and env genes are most related to SIV infecting greater spot-nosed (Cercopithecus nictitans; SIVgsn), mustached (Cercopithecus nictitans; SIVmus), and mona monkeys (Cercopithecus mona; SIVmon) (4,6,7).
Serological evidence has reported endemic SIV infections in over 40 primate species (4,8,9). Surprisingly, early studies of SIV in sooty mangabeys (Cercocebus atys; SIVsmm), African green monkeys (Chlorocebus sabaeus; SIVagm), and mandrills (Mandrillus sphinx; SIVmnd) found SIV infections of these hosts to be largely non-pathogenic, characterized by low levels of inflammation and T cell activation, despite high levels of viral replica tion (9)(10)(11)(12). In contrast, pathogenic infections, such as SIVcpz or HIV-1, display chronic immune activation, leading to increased rates of T cell activation, proliferation, and cell turnover within their respective hosts (13)(14)(15). These differences in pathogenicity may be explained by variations in T cell activation and inflammation. Conventional T cell activation occurs through the T cell receptor (TCR)-CD3 complex (16). The CD3ζ subunits contain intracellular immunoreceptor tyrosine-based activation motifs (ITAMs), enabling tyrosine phosphorylation and subsequent signal transduction through the TCR (16). In addition, complete T cell activation is achieved upon co-stimulatory signaling through engagement of co-stimulatory receptors like CD28 (16). Co-stimulation is required for T cell activation, as weak TCR signals may result in suppression of the immune response and anergy (16). Together, T cell activation requires specific signaling cascades via motifs on the CD3ζ subunit of the TCR and co-stimulatory signals from CD28 (16).
To facilitate successful infection, SIV and HIV rely on accessory proteins to coun teract host immune responses in vivo. In particular, the Nef accessory protein has been implicated in viral pathogenicity by manipulating the host immune responses to promote immune evasion (17,18). For example, Nef downregulates CD4 and major histocompatibility complex class I (MHC-I) from the cell surface to evade virus detection by the adaptive immune response (19,20). Nef can also counter aspects of the host's intrinsic innate immune response by downregulating serine incorporator 5 (SERINC5) and, to a lesser extent, SERINC3, to enhance viral infectivity (21). Numerous studies have also shown that Nef can alter T cell responses, which may explain differences observed in the pathogenicity of SIV and HIV infections. Specifically, Nef downregulates the CD28 co-stimulatory receptor from the T cell surface, which can be inhibited upon mutation of the Nef dileucine (LL 165 ) and diacidic (DD 175 ) motifs (22)(23)(24). Moreover, SIV Nefs from the lineage that gave rise to HIV-2 downregulate CD28 more efficiently than Nefs from SIVcpz/HIV-1 and SIVmus/mon/gsn (22). Therefore, Nef-mediated CD28 downregulation is not completely conserved and functionally varies throughout the SIV/HIV phylogeny.
Nef also downregulates CD3 via interactions with the CD3ζ subunit, hereafter referred to as CD3 (22,(25)(26)(27)(28)(29)(30). While this function is highly conserved in SIV derived from lower-order primates, Nef proteins derived from the HIV-1 lineage-HIV-1, SIVcpz, SIVmus/mon/gsn-fail to downregulate CD3 altogether (22,25,26,29,30). The Nefs that downregulate CD3 efficiently are derived from SIV infections that are non-pathogenic in their natural hosts (22). Since pathogenicity may be related to T cell activation, it is unsurprising that pathogenic infections correlate with the inability of Nef to downregu late CD3. Furthermore, Nefs derived from SIVcpz/HIV-1 and SIVmus/mon/gsn-which do not downregulate CD3-display increased expression of T cell activation markers (22,26,30). Overall, the ability of SIV Nef to modulate T cell activation by downregulating CD3 and CD28 is consistent with a non-pathogenic phenotype, while retaining these receptors on the T cell surface leads to increased T cell activation and inflammation, consistent with a pathogenic infection. Overall, the virus is not the only essential component for determining pathogenicity; instead, it is a summation of both virus-and hostspecific adaptations (22,31).
Prior studies seeking to understand the evolution of Nef as a pathogenic factor have traditionally utilized modern-day-or extant-nef sequences (reviewed in reference 32). This approach can limit information regarding the evolutionary dynamics of Nef within a phylogeny. We previously employed an ancestral reconstruction pipeline to generate an accurate time-scaled phylogeny relating Nef sequences from both SIV and HIV lineages. This approach predicted the ancestral Nef sequences at each common ancestor, or node, in the phylogenetic lineage leading to HIV-1 group M and generated corresponding nucleotide sequences (33). Thus far, the reconstructed ancestral Nefs have exclusively been tested for their abilities to downregulate SERINC5 and CD4 (33). The investigation of Nef functions pertaining to T cell activation, such as CD3 and CD28 downregulation, has not been assessed.
Herein, we further investigated the reconstructed phylogeny by assessing the ability of these predicted Nefs to downregulate CD3 and CD28-two functions that likely contribute to pathogenicity by modulating T cell responses. We found that the Nef corresponding to the ancestor of all primate lentiviruses efficiently downregulated CD3 and CD28, while the ancestors derived from SIVcpz/HIV-1 failed to do so. Further more, we observed that robust CD3 and CD28 downregulation led to low levels of T cell activation, consistent with a non-pathogenic infection. In comparison, expression of Nefs incapable of CD3 and CD28 downregulation led to high levels of T cell acti vation, consistent with the pathogenic nature of SIVcpz/HIV-1. This characterization of Nef ancestors provides insight into how Nef-mediated CD3 and CD28 downregu lation may have evolved through time and helps create predictions for how these changes contributed to the pathogenic nature of SIVcpz and HIV-1 infection within their respective hosts.
## RESULTS
## Ancestral Node 35 Nef efficiently downregulates cell-surface CD3, independ ent of host species
The reconstructed phylogeny contains six internal nodes in the path from the ancestral primate lentivirus to the common ancestor of HIV-1/M (33). The ancestral nodes (Fig. S1) correspond to the following lineages: Node 35 (all primate lentiviruses, including SIVcpz, SIVrcm, SIVsun, SIVmus/mon/gsn, HIV-1, and HIV-2), Node 51 (HIV-1/SIVsun), Node 52 (HIV-1/SIVcpz), Node 55 (HIV-1/SIVcpzptt), Node 56 (HIV-1/M,N/SIVcpzptt), and Node 59 (HIV-1/M). We first tested the ability of the reconstructed Nefs to downregu late cell surface CD3. To do so, we expressed the various Nefs in HEK293T cells and used the huCD8α-CD3ζ fusion constructs (Fig. 1A) to evaluate the downregulation of CD3 sequences derived from different primate hosts, including human/chimpanzee (huCD3), which has been evaluated previously (30). We also included a Nef isolate derived from a well-characterized strain of SIV-SIVmac239-which has been previously shown to robustly downregulate both huCD3 and rhesus macaque CD3 (rhCD3) (22,25,30). As a functional negative control, we included a mutated form of SIVmac239 Nef containing two mutations in its core domain, I 123 L L 146 F, that selectively impair CD3 downregulation (25). We also included a lab-adapted HIV-1 Nef (NL4.3 Nef ), which has also been shown to be incapable of CD3 downregulation (22,25,30). Finally, we included a plasmid that expresses eGFP only (ΔNef ) as our overall negative control. Accordingly, we assessed the ability of the various Nefs to downregulate huCD3. As expected, no huCD3 downregulation was observed in the ΔNef and NL4.3 Nef negative controls (Fig. 1C through E; 34.5% and 46.4% of SIVmac239 WT, respectively, P < 0.0001). Additionally, SIVmac239 WT Nef demonstrated robust huCD3 downregulation, while its mutant form was significantly impaired, as expected (Fig. 1C through E; 53.2% of SIVmac239 WT, P < 0.0001). We observed that Node 35 Nef, which represents the ancestral primate lentivirus, displayed a significant increase in huCD3 downregulation compared to the ΔNef negative control (Fig. 1C through E; 84.4% of SIVmac239 WT, P < 0.0001). Furthermore, Node 35 Nef downregulated huCD3 to a significantly greater extent when compared to the rest of the ancestral Nefs (Fig. 1C through E; P < 0.0001), which displayed a similar level of downregulation as ΔNef (Fig. 1C through E; ranging from 37.5% to 46.9% of SIVmac239 WT). To compare the relative expression levels of the ancestral Nef proteins, the codon-optimized nef sequences were cloned into an expression vector enabling C-terminal eGFP fusion, as described previously (33). The resulting constructs were transfected into HEK293T cells, and Nef-eGFP expression levels were evaluated by Western blot analysis (Fig. S2). Although the expression levels varied slightly among the reconstructed Nef-eGFP variants, all fusion proteins were expressed at high levels (Fig. S2). Together, these findings provide evidence that the ancestral reconstruction accurately generated Nef ancestors, given that all SIVcpz/HIV-1 Nefs were incapable of CD3 downregulation, as expected.
We next tested if Nef-mediated CD3 downregulation differs within various species contexts, as some Nef functions are species-dependent. For instance, SIV Nefs can antagonize their autologous tetherin, a host factor that interferes with virion release, but not human tetherin (34)(35)(36). Contrastingly, some Nef functions, such as SERINC5 downregulation, are independent of the host species, with SIV and HIV Nefs retaining the ability to downregulate SERINC5 homologs across different primates (30,33,37,38). Thus, we generated additional huCD8α-CD3ζ fusions encoding the CD3ζ cytoplasmic tail of rhesus macaques (rhCD3; Fig. 2A), sooty mangabeys (smmCD3; Fig. 2B), African green monkeys (agmCD3; Fig. 2C), and Ma's night monkeys (masCD3; Fig. 2D). Given that sooty mangabeys and African green monkeys are naturally infected with SIV, we expected to see a similar trend to huCD3 downregulation, where HIV-1 Nefs are incapable of downregulating CD3. On the other hand, Ma's night monkey is a new world primate and is not naturally infected with SIV. However, recent evidence suggests that these primates contain compatible CD4 receptors for HIV-1 and may serve as a novel non-human primate model for studying pathogenic HIV-1 infections (39,40). Therefore, we expected the pattern of masCD3 downregulation to emulate our previous experiments where only Node 35 Nef could downregulate CD3.
The evaluation of rhCD3, smmCD3, agmCD3, and masCD3 downregulation demon strated that Node 35 Nef displayed significant CD3 downregulation (Fig. 2A through D; 61.9%, 98.8%, 101.5%, and 132.5% of SIVmac239 WT, respectively; ranging from P < 0.01 to P < 0.0001), but Node 51-59 Nefs did not. In these experiments, we observed no significant difference in CD3 downregulation between Node 35 Nef and SIVmac239 WT Nef (Fig. 2B andC; P > 0.05), except in the context of rhCD3 downregulation, where SIVmac239 WT Nef demonstrated significantly greater CD3 downregulation than Node 35 Nef (Fig. 2A; P < 0.0001). Interestingly, when evaluating masCD3 downregulation, Node 35 Nef displayed significantly greater masCD3 downregulation than SIVmac239 WT Nef (Fig. 2D; P < 0.0001). This was contrary to all previous experiments in which Node 35 Nef exhibited significantly less (rhCD3) or similar (smmCD3 agmCD3) CD3 downregu lation to SIVmac239 WT Nef. Taken together, the Nef representing the ancestor of all primate lentiviruses (Node 35) displayed significant CD3 downregulation, regardless of species context. In comparison, Nefs derived from the SIVcpz/HIV-1 lineage (Nodes 51-59) failed to downregulate CD3 across different species contexts. Therefore, this function appears independent of the host species, suggesting that downstream experiments in primary human cells can be generalizable across most primates. orange), huCD3ζ (middle; blue), and huCD8α-huCD3ζ fusion construct (bottom). (B) Gating strategy used to isolate live (Zombie NIR -) and transfected (eGFP + ) cells. (C) Representative pseudocolor plots illustrating huCD8α-huCD3ζ (PE; Y-axis) and eGFP (X-axis) levels gated on live (Zombie NIR -) cells. (D) Representative histogram of cell surface levels of huCD8α-huCD3ζ (X-axis) within live, transfected (Zombie NIR -, eGFP + ) cells. The count is normalized to the mode of the set.
(E) Summary of fold huCD3 downregulation (±SE) from three independent experiments (n = 3), calculated as the percentage of SIVmac239 WT Nef (positive control). SP, signal peptide; EC, extracellular domain; TM, transmembrane domain; SSC-A, side scatter area; FSC-A, forward scatter area; FSC-H, forward scatter height; FMO, fluorescence minus one; NIR, near-infrared; eGFP, enhanced green fluorescent protein; PE, phycoerythrin; and SE, standard error. Significance: ****P < 0.0001; ns, not significant. Schematic created in BioRender. https://BioRender.com/p20gixl. We next tested the function of the various Nef proteins in primary human CD4 + (huCD4 + ) T cells transduced with pseudoviruses expressing each ancestral Nef. In parallel, we also assessed the ability of these Nefs to downregulate CD28. Of note, the CD28 cytoplasmic amino acid sequences are identical between humans, chimpanzees, rhesus macaques, sooty mangabeys, and African green monkeys, but not with Ma's night monkeys. Given the highly conserved nature of the CD28 sequences, we predicted this Nef function to, like CD3 downregulation, also be species independent. We included the same controls described previously, with the addition of a mutated NL4.3 Nef (NL4.3 LL 164-165 AA Nef ), containing a mutated dileucine motif, which prevents the interaction between Nef and AP-2 to impair CD28 downregulation (24). While trends of CD3 downregulation observed in transduced primary huCD4 + T cells were consistent with Fig. 1C through E, the magnitude of Nef-mediated CD3 downregulation was considerably more pronounced upon transduction (Fig. 3A andC). Specifically, Node 35 Nef demonstrated significantly greater CD3 downregulation than ΔNef (Fig. 3C; 17.6-fold, P < 0.0001), while NL4.3 and Nodes 51-59 Nefs did not differ from ΔNef (Fig. 3C; ranging from 0.95-to 1.1-fold). Moreover, SIVmac239 WT Nef significantly downregulated CD3 compared to ΔNef (Fig. 3C; 13.8-fold, P < 0.0001), while the mutant SIVmac239 I 123 L L 146 F Nef was significantly impaired compared to SIVmac239 WT Nef (Fig. 3C; 1.2-fold, P < 0.0001). Finally, within transduced primary cells, there was no significant difference in huCD3 downregulation between Node 35 Nef and SIVmac239 WT Nef (Fig. 3C; P > 0.05).
The pattern of Nef-mediated CD28 downregulation differed slightly from CD3 downregulation (Fig. 3B andD). Both SIVmac239 WT Nef and SIVmac239 I 123 L L 146 F Nef downregulated CD28 (Fig. 3D; 3.7-and 4.8-fold, respectively), confirming that the I 123 L L 146 F mutation selectively disrupts CD3 downregulation, although this mutation led to a significant increase in CD28 downregulation compared to SIVmac239 WT Nef (Fig. 3D; P < 0.01). As expected, Node 35 Nef significantly downregulated CD28 compared to ΔNef (Fig. 3D; 3.5-fold, P < 0.0001). However, unlike CD3 downregulation, we observed differential CD28 downregulation with Nodes 51-59. For instance, we found Nodes 51 and 52 Nef downregulated CD28 significantly less than Node 35 Nef (Fig. 3D; 1.3-and 1.1-fold, respectively, P < 0.0001). Interestingly, we observed that CD28 downregulation gradually improved in Nodes 55 (1.2-fold), 56 (1.5-fold), and 59 Nefs (1.6-fold). Thus, CD28 downregulation is lost from Node 35 to Node 52 and then gradually recovered in Nodes 55-59. These findings are contrary to CD3 downregulation, which is completely lost in the SIVcpz/HIV-1 lineage (Nodes 51-59), suggesting the evolutionary dynamics of CD28 downregulation mediated by Nef differs from CD3 downregulation.
## Efficient CD3 and CD28 downregulation significantly impairs T cell activation
We next evaluated the downstream effects of Nef-mediated CD3 and/or CD28 downre gulation on T cell activation following transduction. To do this, we transduced pri mary huCD4 + T cells with the pseudoviruses expressing each ancestral Nef. Following transduction, T cells were re-stimulated via the TCR using a soluble anti-CD3/CD28/CD2 antibody complex. We then measured the expression of surface activation markers CD69 and CD25 and the intracellular production of proinflammatory cytokines IL-2 and IFNγ (Fig. 4A). We hypothesized that the expression of Nefs capable of robust CD3 and/or CD28 downregulation would lead to lower T cell activation, as efficient downregulation of CD3 and/or CD28 would render transduced cells less able to become activated upon the addition of a secondary stimulus. As expected, in the presence of Node 35 Nef, we observed a significant decrease in CD69 expression compared to all ancestral and control Nefs (Fig. 4B; P < 0.0001), consistent with its ability to efficiently downregulate both CD3 and CD28. This was also not significantly different than the unactivated control, which was transduced with ΔNef (Fig. 4B; P > 0.05). SIVmac239 WT Nef, which also robustly downregulates CD3 and CD28, displayed an intermediate activation profile, with CD69 expression levels significantly greater than Node 35, but significantly less than HIV-1 Nefs such as NL4.3 WT Nef (Fig. 4B; P < 0.05). In the context of CD25, we observed cells expressing Node 35 Nef retained less CD25 on the cell surface when compared to Nodes 51-59, although these were not significantly different (Fig. 4C; P > 0.05). Similarly, when we assessed the expression of IL-2 and IFNγ, we observed Node 35 Nef to result in significantly less intracellular IL-2 and IFNγ production than all the ancestral Nefs (Fig. 4D andE; ranging from P < 0.05 to P < 0.0001). This did not differ from the cytokine expression levels in the presence of SIVmac239 Nef (WT or mutant), which were also significantly less than NL4.3 WT Nef (Fig. 4D andE; P < 0.05). In conclusion, T cell activation in the presence of the reconstructed Nefs generally aligns with the magnitude by which each viral protein downregulates CD3 and/or CD28 from the cell surface.
## Mapping regions involved in CD3 and CD28 downregulation by Node 35 Nef
To elucidate the motifs involved in CD3 and CD28 downregulation by Node 35 Nef, we first aligned its amino acid sequence to SIVmac239 Nef, which shares these func tional capabilities, and identified putative motifs involved in CD3 or CD28 downregu lation using sequence homology (Fig. S3A). We then generated the corresponding mutations in Node 5B andC) and CD28 downregulation (5.1-and 4.3-fold, respectively; Fig. 5B andC). In comparison, the I 123 L L 146 F mutation in SIVmac239 and the correspond ing A 130 F mutation in Node 35 Nef significantly impaired CD3 downregulation (2.7and 1.2-fold, respectively, P < 0.0001), but not CD28 downregulation (Fig. 5B andC). As expected, both the W 203 S and LM 195 AA mutations in SIVmac239 Nef significantly impaired CD28 downregulation (1.4-and 0.95-fold, respectively, P < 0.0001), but not CD3 downregulation (Fig. 5B andC). However, while the corresponding dileucine motif
## DISCUSSION
In the current report, we assessed the ability of reconstructed ancestral Nef proteins to downregulate cell-surface CD3 and CD28. Functional outputs were also used to further validate the previously employed bioinformatics pipeline (33). Overall, these analyses allowed us to characterize the evolutionary changes of Nef functions and determine the contribution of these Nef-mediated functions to pathogenicity. We determined that only the ancestor of all primate lentiviruses, Node 35 Nef, efficiently downregulated cell surface CD3. This function was subsequently lost in the SIVcpz/HIV-1 lineage and never recovered, regardless of the host primate context. These findings align with previous reports suggesting that Nef-mediated CD3 downregulation is only retained in the majority of SIV and HIV-2 Nefs, while absent in the SIVcpz/HIV-1 lineage (22,26,30). Although we determined that Nef-mediated CD3 downregulation is independent of the host species, we observed some divergence in its hostspecific magnitude. For instance, SIVmac239 WT Nef was less efficient than Node 35 Nef at downregulating masCD3, which may implicate the role of host-virus co-evolution. We posit that SIVmac239 Nef evolved robust rhCD3 downregulation following extensive co-evolution, which came at the consequence of modestly impairing its ability to downregulate CD3 from an unrelated primate species, such as Ma's night monkeys. On the contrary, smmCD3 and agmCD3 are more genetically similar to rhCD3 than to masCD3, explaining why we observed a similar pattern of CD3 downregulation between these experiments. Additionally, these findings may provide further insight into the potential for Ma's night monkeys to serve as a novel model for studying HIV-1 infections. The results from the evaluation of masCD3 downregulation align with what would be expected for a non-human primate model of HIV-1 infection, where CD3 downregulation is not observed in the presence of HIV-1 Nefs-like NL4.3 Nef-but does occur in the presence of other SIV Nefs such as SIVmac239 Nef (Fig. 2D).
In addition, Node 35 Nef was the only ancestral Nef capable of robust CD28 downregulation. Similar to Nef-mediated CD3 downregulation, this function was subsequently lost in Nodes 51 and 52. The magnitude of CD28 downregulation by Node 59 Nef, which represents the ancestor of HIV-1 group M, is similar to the lab-adapted HIV-1 NL4.3 Nef, which aligns with reports suggesting that HIV-1 Nefs exhibit weak CD28 downregulation that is not as efficient as SIV and HIV-2 Nefs (22, 24, 30, 42).
We then examined the downstream consequences of Nef-mediated CD3 and CD28 downregulation on T cell activation within transduced cells. Given the magnitude of CD3 and CD28 downregulation we observed previously, we expected transduced cells expressing SIVmac239 WT Nef to display similar levels of T cell activation to Node 35. The finding that SIVmac239 WT Nef led to an intermediate activation profile in primary human cells may suggest that a specific host factor in rhesus macaques is required for maintaining lower immune activation in the presence of a Nef isolate that downregulates CD3 and CD28. In addition, other studies that observed low levels of T cell activation (i.e., CD69 expression) in the presence of SIVmac239 WT Nef used replication-competent virus and infected peripheral blood mononuclear cells (PBMCs) (22,30). For our study, we utilized a replication-incompetent virus and transduced purified primary CD4 + T cells to limit the potential confounding effects of other immune cells in the population. Altogether, this may explain why we did not observe the same activation profile between Node 35 and SIVmac239 Nef. Our finding that efficient Nef-mediated CD3 downregulation by Node 35 Nef led to lower levels of immune activation is consistent with the non-pathogenic phenotype of SIV infections within their respective natural hosts. When this function is lost, we observed greater T cell responses, consistent with a pathogenic infection. However, it remains unclear why Nef would lose its ability to downregulate CD3, given its role in reducing the magnitude of the host immune response. It has been hypothesized that retaining CD3 on the surface of infected T cells may provide a selective advantage for the virus, as greater T cell signaling increases T cell recruitment and generates larger pools of T cells susceptible to viral infection (25). An increase in T cell responsiveness can also boost viral gene expression and replication (26). Additionally, cell surface CD3 expression enhances Env incorporation into virions, thereby increasing viral infectivity and spread (26). This aligns with our findings that the loss of Nef-mediated CD3 downregulation in the SIVcpz/HIV-1 group M lineage is associated with features of a pathogenic infection, such as the increased expression of T cell activation markers and inflammatory cytokines.
We have also provided insight into how motifs in Node 35 Nef may facilitate CD3 and CD28 downregulation by mapping the corresponding residues and functional motifs observed with SIVmac239 Nef using sequence homology. The finding that the SIVmac239 I 123 L L 146 F and Node 35 A 130 F mutants selectively impaired CD3 down regulation but not CD28 downregulation suggests that these Nefs rely on similar residues within the core Nef domain to associate with and downregulate CD3. Similarly, mutation of the Nef dileucine motif (SIVmac239 LM 194-195 AA and Node 35 LL 178-179 AA) significantly impaired CD28 downregulation, while not affecting CD3 downregulation, suggesting that the dileucine motif is required for CD28 downregulation to facilitate Nef and AP-2 binding (41). Interestingly, while the W 203 S mutation in SIVmac239 Nef demonstrated a similar outcome as the LM 195 AA mutant, the corresponding mutation in Node 35 Nef (M 187 S) did not. Previously, the W 203 S mutation in SIVsmm Nef was shown to disrupt CD28 downregulation-but not CD3 downregulation-by impacting Nef complex formation with AP-2 subunits (41). We speculate that Node 35 Nef and SIVmac239 Nef exhibit similar, but distinct, mechanisms to downregulate CD28 via AP-2 engagement. Both Nef proteins require binding of their dileucine motif to AP-2, but rely on alternate residues to stabilize the interactions within the Nef ternary complex in the context of CD28 downregulation.
The loss of Nef-mediated CD3 downregulation in SIVcpz/HIV-1 coincides with the acquisition of another viral protein, Vpu (22). Specifically, HIV-1, SIVcpz, and the SIVmus/mon/gsn clade encode vpu, while other SIVs and HIV-2 encode a vpx gene (44,45). This suggests that Vpu originated from the common ancestor of SIVmus/mon/gsn and was then transmitted to SIVcpz following its recombination with SIVrcm (6,45). The acquisition of vpu by SIVcpz through recombination was considered necessary for its cross-species transmission to humans, as the Nef protein in SIVcpz/HIV-1 lost the ability to downregulate human tetherin, a function that was regained by Vpu (46). This is especially interesting because Nef proteins derived from the SIVmus/mon/gsn clade were previously found to lack the ability to downregulate CD3, despite representing an SIV lineage (22). However, Vpu does not directly impact T cell activation via the downregulation of CD3. More specifically, Nef-mediated CD3 downregulation prevents the induction of NF-kB, while HIV-1 Vpu inhibits NF-kB signaling at later steps in the pathway to prevent antiviral gene expression (47)(48)(49). This suggests that Vpu-mediated suppression of T cell activation downstream in the NF-kB pathway may have reduced the selective pressure for Nef to suppress T cell activation by downregulating CD3 (47)(48)(49). For the current study, we used a proviral backbone lacking Vpu to ensure that our findings were not the result of overlapping functions between the two viral proteins. Nonetheless, the evolution of Nef-mediated CD3 downregulation coincided with the acquisition of Vpu, thereby providing the virus with an alternative method of immune suppression while maintaining the benefits of increased T cell signaling through the TCR.
While the magnitude of Nef-mediated CD3 downregulation likely drives the downstream impacts on T cell activation, we also observed robust CD28 downregulation by Nefs that efficiently downregulated CD3 (such as Node 35 and SIVmac239 WT Nef ). These Nefs may robustly downregulate both CD3 and CD28 to ensure maximal suppression of immune activation. Interestingly, CD28 downregulation was lost in the first node of the SIVcpz/HIV-1 lineage (Node 51), coinciding with the loss of CD3 downregulation and the acquisition of Vpu. Previous work from our group has shown that both HIV-1 Nef and Vpu independently downregulate CD28 within infected cells (23). While it is unknown whether SIVcpz Vpu downregulates CD28, using the same rationale as with CD3 downregulation, we speculate that Vpu-mediated CD28 downre gulation in the SIVcpz/HIV-1 lineage likely reduced the selective pressure on Nef to downregulate CD28 to the same extent as its SIV counterparts that lack Vpu. This could explain why the loss of Nef-mediated CD3 and CD28 downregulation occurred at the same time, which happened to coincide with the acquisition of Vpu.
Our findings show that only the Nef corresponding to the ancestor of all pri mate lentiviruses, Node 35, could efficiently downregulate CD3 and CD28, while Nefs corresponding to the intermediate nodes (Nodes 51-59) completely lost the ability to downregulate CD3 and displayed only modest improvements in CD28 downregulation over time. These observations strongly align with previous literature on the functional evolution of lentiviral Nef proteins, which assessed the function of extant Nef sequences from different lineages to infer the respective evolutionary history. Combined with our earlier work demonstrating consistent patterns of SERINC5 and CD4 downregulation across the same ancestral Nefs, these findings further validate the accuracy of our bioinformatics pipeline in reconstructing functionally relevant ancestral proteins within a given viral lineage (33). Taken together, our results strongly support the broader application of this pipeline to reconstruct ancestral Nef proteins in other SIV/HIV lineages beyond the SIVcpz/HIV-1 group M lineage to better understand the functional evolu tion of Nef. More broadly, we propose that this workflow can be extended to other HIV-1 proteins, and even to proteins from unrelated viruses, as a general strategy to explore how viral protein functions evolved-particularly in the context of cross-species transmission events.
## Conclusions
To conclude, our study provides functional evidence that the ability of Nef to downregu late CD3 and CD28 was present in the ancestral primate lentivirus but lost in the lineage giving rise to HIV-1. The corresponding effect on T cell activation further demonstrates the important functional consequences of these Nef functions to the pathogenicity of infection. Together, these results shed light on the evolutionary dynamics of Nef modulation of T cell activation and its impact on the pathogenesis of modern infections.
## MATERIALS AND METHODS
## Ancestral sequence reconstruction
The reconstructed ancestral Nef sequences were generated as described previously (33). Briefly, a representative selection of Nef amino acid sequences from different primate lentiviruses was obtained from the Los Alamos National Laboratory (http:// www.hiv.lanl.gov/) and GenBank (https://www.ncbi.nlm.nih.gov/genbank) databases. A multiple sequence alignment was generated using MAFFT (version 7.271) (50), and a maximum-likelihood tree was reconstructed using IQ-TREE (version 1.3.11.1) (51) to serve as a starting tree for Bayesian analysis. BEAST (version 1.10.4) (52) was then used to sample time-scaled phylogenies from the posterior distribution. From the resulting posterior sample of trees, we identified the most frequent trajectory of internal nodes from the root to the common ancestor of HIV-1/M, comprising six internal nodes. Among the n = 4,535 trees that followed this maximum support trajectory, we randomly selected 1,000 trees. For each of these trees, we used Historian (version 0.1) (53) to reconstruct the ancestral amino acid sequences at the internal nodes and then generated a consensus sequence across the 1,000 trees. The consensus amino acid sequence at each node was converted to nucleotides using codon distributions from extant sequences, from which the most frequent codon was assigned to each residue (see https://doi.org/10.5281/ zenodo.8010261).
## Cell culture
HEK293T cells (ATCC, Manassas, VA, USA) were maintained in Dulbecco's modified Eagle's medium containing 4 mM L-glutamine (Cytiva Life Sciences, Vancouver, BC, Canada), 4.5 g/L glucose (Cytiva Life Sciences), and supplemented with 1% 100 μg/mL penicillinstreptomycin (HyClone, Logan, UT, USA) and 10% fetal bovine serum (FBS-Wisent, St. Bruno, QC, Canada). Cell lines were grown at 37°C in the presence of 5% CO 2 and sub-cultured according to the manufacturer's instructions.
Primary human PBMCs were isolated from healthy donors, and CD4 + T cells were purified using the RosetteSep Human CD4 + T cell Enrichment Cocktail (StemCell, Vancouver, BC, Canada) and cryopreserved. Upon thawing, cells were activated with Immunocult Human CD3/CD28/CD2 T cell activator (StemCell) and maintained in Immunocult-XF T cell Expansion Medium (StemCell) supplemented with recombinant human IL-2 (10 ng/mL, ThermoFisher, Whitby, ON, Canada). Primary cells were grown at 37°C in the presence of 5% CO 2 and sub-cultured for 3 days post-thaw.
## DNA constructs
Previously, the ancestral nef sequences were commercially synthesized (GeneArt Gene Synthesis, ThermoFisher) and cloned into pN1 expression vectors (Clontech) expressing Nef C-terminally fused to eGFP (33). For this study, the ancestral nef sequences were cloned into replication-incompetent pNL4.3 ΔGag/Pol eGFP ΔVpu Nef proviral plasmids using the pN1 Nef-eGFP expression vectors as templates for PCR amplification (33). Additionally, controls, including NL4.3 Nef, NL4.3 LL 164-165 AA Nef, SIVmac239 Nef, and SIVmac239 I 123 L L 146 F Nef, were also cloned into the pNL4.3 backbone for downstream pseudovirus generation. To improve viral production, the reconstructed Nef amino acid sequences were analyzed using BLASTp to identify related primary Nef sequences (54). For each reconstructed Nef sequence, individual residues and codons were compared to those in the related primary isolates. If the amino acid residue at a given position was conserved, the codon from the primary isolate was substituted into the reconstruc ted nef sequence. This manual codon optimization ensured that codons typical of naturally circulating viral strains were incorporated, with the aim of enhancing viral gene expression and pseudovirion production. This process was performed manually for all reconstructed nef sequences. Importantly, the resulting codon-optimized sequences encoded the same amino acid sequences as the original reconstructed Nef sequences (33). The manually optimized nef sequences were synthesized using the GeneArt Gene Synthesis service. The proviral plasmid was engineered to readily clone nef using flanking 5′ XmaI and 3′ NotI restriction sites as previously described (23). Forward and reverse primers were designed to clone each respective nef gene into the proviral backbone by harboring XmaI and NotI restriction sites, respectively. The pCMV-DR8.2 plasmid (Addgene, Waterdown, MA, USA; Cat#12263, gift from Didier Trono) encoding Gag/Pol and the pMD2.G plasmid encoding vesicular stomatitis virus glycoprotein (VSV-G; Addgene, Cat#12259, gift from Didier Trono) were also used for transfections where indicated.
To exogenously express CD3ζ in the absence of the TCR, we used a human CD8α-human CD3ζ (huCD8α-huCD3ζ) fusion construct encoding the extracellular and transmembrane domains of huCD8α (accession: M12824; residues 1-208) fused to the cytoplasmic tail of huCD3ζ (accession: NM_000734; residues . This was a kind gift from Dr. Frank Kirchhoff (University Clinic of Ulm, Ulm, Germany) and has been previously described (30). To generate fusion vectors expressing CD3ζ from different primates, the nucleotide sequences corresponding to the CD3ζ cytoplasmic domain of rhesus macaques (Macacca mulatta; accession: DQ437670), sooty mangabeys (accession: XM012091107), African green monkeys (accession: XM007989658), and Ma's night monkeys (Aotus nancymaae; accession: NM001308517) were first synthesized using the GeneArt Gene Synthesis service. Overlapping PCR mutagenesis was subsequently used to replace the huCD3ζ nucleotide sequence for the CD3ζ cytoplasmic domain of the previously mentioned primate species. In the first PCR reaction, the nucleotide sequence corresponding to the extracellular and transmembrane domains of huCD8α was amplified using a forward primer bearing the XbaI restriction site and a reverse primer with a tail complementary to the 5′ region of the CD3ζ cytoplasmic sequence of the aforementioned primate species. In the second PCR reaction, the nucleotide sequence corresponding to the primate CD3ζ cytoplasmic domain was amplified with a forward primer complementary to the 3′ region of the extracellular/transmembrane domain of huCD8α and a reverse primer harboring the MluI restriction site. In the final PCR reaction, both purified amplicons were mixed 1:1 and amplified using the 5′ forward primer bearing the XbaI restriction site and the 3′ reverse primer bearing the MluI site to generate the huCD8α-CD3ζ construct. This fusion was then cloned into the pCG huCD8α-CD3ζ plasmid accordingly.
To induce the relevant mutations into Node 35 and SIVmac239 Nef, the Q5 Site-Directed Mutagenesis Kit was used according to the manufacturer's instructions (New England Biolabs, Ipswich, MA, USA). For protein expression analysis, the reconstructed nef sequences, along with SIVmac239 nef and all corresponding nef mutants, were PCR amplified using sequencespecific forward primers harboring a 5′ EcoRI restriction site and sequencespecific reverse primers harboring a 3′ BamHI restriction site. To generate a plasmid expressing a Nef-eGFP fusion protein, the reverse primer was additionally designed to mutate the nef stop codon to a TGT cysteine codon. PCR products were subsequently purified and cloned into the pN1 expression vector (Clontech), which contains the 3′ egfp ORF. To clone the Node 35 and SIVmac239 nef mutants into the proviral backbone, sequencespecific forward primers harboring a 5′ XmaI restriction site and sequencespecific reverse primers harboring a 3′ NotI restriction site were designed, followed by PCR amplification, purification, and subsequent cloning into the pNL4.3 ΔGagPol eGFP ΔVpu Nef proviral plasmid. All DNA constructs used in this study were verified by Sanger Sequencing at the London Regional Genomics Centre prior to functional analysis.
## Cell surface CD3 downregulation in HEK293T cells
For transfection assays to assess cell surface CD3 downregulation, HEK293T cells were seeded in 6-well plates 24 hours prior to transfection. HEK293T cells were then triply transfected with 1 μg of the cloned pNL4.3 ΔGag/Pol eGFP ΔVpu Nef plasmids, 1 μg of the pCMV-DR8.2 plasmid, and 0.2 μg of the pCG huCD8α-CD3ζ plasmids using PolyJet (FroggaBio, North York, ON, Canada). Forty-eight hours post-transfection, HEK293T cells were washed in 1× phosphatebuffered saline (PBS) without calcium and magnesium (Wisent), lifted with 0.25% Trypsin-EDTA (ThermoFisher), and collected in a 96-well U-bottom plate. Cells were then washed with 1× PBS containing calcium and magnesium (Wisent), stained with Zombie NIR viability dye (BioLegend, San Diego, CA, USA), and subsequently fixed with 1.5% paraformaldehyde (PFA; Electron Microscopy Sciences, Hatfield, PA, USA). Cells were then washed with fluorescenceactivated cell sorting (FACS) buffer (3% FBS and 5 mM EDTA in 1× PBS). For cell surface staining of the huCD8α as a proxy for cell surface CD3 levels, cells were stained with 1:25 anti-human CD8α phycoerythrin (PE; clone HITα; BioLegend). The isotype control sample was stained with equivalent amounts of the PE mouse IgG1, κ-Isotype Control Antibody (clone MOPC-21; BioLegend). Finally, cells were washed again in FACS and resuspended in 1× PBS prior to flow cytometry analysis.
## Virus generation
For VSV-G pseudotyped lentivirus production, HEK293T cells were seeded in 6-well plates 24 hours prior to transfection. HEK293T cells were then triply transfected with 0.83 μg of the cloned pNL4.3 ΔGag/Pol eGFP ΔVpu Nef plasmids, 0.83 μg of the pCMV-DR8.2 plasmid, and 0.4 μg of the pMD2.G plasmid encoding VSV-G using PolyJet. Virus-contain ing supernatants were collected 72 hours post-transfection, supplemented with 20% FBS, filtered using a 0.45 μm filter (ThermoFisher), and stored at -80°C for downstream experiments.
## Viral transduction
For transduction of primary cells, huCD4 + T cells 6 days post-thaw were pelleted and resuspended with each VSV-G-pseudotyped virus, 8 μg/mL Polybrene (Sigma-Aldrich, St. Louis, MO, USA), and complete Roswell Park Memorial Institute medium (RPMI) 1640 media-containing L-glutamine and supplemented with 100 μg/mL penicillin-strepto mycin, 1% sodium-pyruvate (HyClone), 1% non-essential amino acids (HyClone), and 10% FBS. Subsequently, cells were spinoculated at 2,880 × g for 2 hours at room temperature. After spinoculation, cells were resuspended in complete RPMI and IL-2 (10 ng/mL) for an additional 48 hours.
## Cell surface CD3 and CD28 downregulation in primary cells
For cell surface staining of huCD4 + T cells 48 hours post-transduction, cells were collected and washed with 1× PBS and stained with Zombie NIR viability dye. Cells were then fixed with 1% PFA and washed in FACS buffer. For cell surface staining of CD3 and CD28, 1:80 anti-human CD3 PE (clone UCHT1; BioLegend) and 1:25 anti-human CD28 APC (clone CD28.2; BioLegend) were used. Isotype control samples were stained with PE mouse IgG1, κ-Isotype Control Antibody (clone MOPC-21; BioLegend) and APC mouse IgG1, κ-Isotype Control Antibody (clone MOPC-21; BioLegend). Finally, cells were washed again in FACS and resuspended in 1× PBS prior to analysis.
## T cell activation and intracellular cytokine analysis
To assess T cell activation, huCD4 + T cells were cultured and transduced as described above. Forty-eight hours post-transduction, they were re-activated with Immunocult Human CD3/CD28/CD2 T cell activator. Cells were then incubated for an additional 24 hours before being stained with Zombie NIR viability dye, fixed with 1% PFA, and stained with 1:20 anti-human CD69 BrilliantViolet421 (BV421; clone FN50; BioLegend) and 1:200 anti-human CD25 APC (clone BC96; BioLegend) for flow cytometry analysis. Isotype control samples were stained with BV421 mouse IgG1, κ Isotype Ctrl Antibody (clone MOPC-21; BioLegend) and APC mouse IgG1, κ-Isotype Control Antibody (clone MOPC-21; BioLegend). For intracellular cytokine staining, cells were transduced and re-stimulated as described previously, with the addition of 1:1,000 Brefeldin A (BioLe gend) 8 hours post-re-activation to inhibit cytokine secretion. Cells were stained with Zombie NIR viability dye, fixed with 1% PFA, and washed with FACS buffer. Cells were then permeabilized with Perm/Wash Buffer (BD Biosciences, Franklin Lakes, NJ, USA) and incubated with Human TruStain FcX blocking solution (BioLegend) to prevent unwanted antibody binding. Cells were then stained with 1:20 anti-human IL-2 PE (clone MQ1-17H12; BioLegend) and 1:20 anti-human IFNγ AlexaFluor647 (AF647; clone 4S.B3; BioLegend). Isotype control samples were stained with PE rat IgG2α, κ Isotype Ctrl Antibody (clone RTK2758; BioLegend) and AF647 Mouse IgG1, κ Isotype Ctrl (ICFC) Antibody (clone MOPC-21; BioLegend). Cells were then washed with Perm/Wash buffer and FACS and resuspended in 1× PBS prior to analysis.
## Flow cytometry analysis
For flow cytometry analysis of the HEK293T transfection experiments, cells were analyzed with the BD Biosciences LSR-II cytometer. For flow cytometry analysis of all other experiments, cells were analyzed with the Beckman Coulter CytoFLEX S (Brea, CA, USA). Both cytometers are located at the London Regional Flow Cytometry Facility (London, ON, Canada). Flow cytometry data were analyzed using FlowJo software (version 10.10.0, FlowJo LLC, Ashland, OR, USA).
## Data and statistical analysis
Relative levels of cell surface receptors and intracellular expression of cytokines were determined by quantifying geometric mean fluorescence intensity (gMFI) of the respective fluorophores after gating on single, live (Zombie NIR -), and transfected/trans duced cells (eGFP + ), as illustrated in Fig. 1B. For the analysis of cell surface huCD8α-CD3ζ levels, Nef-mediated CD3 downregulation was calculated as a percentage of SIVmac239 wild-type Nef by dividing the gMFI of PE for SIVmac239 WT Nef by the gMFI of PE for each Nef sample, and then multiplying by 100. In primary cells, fold CD3/CD28 downregulation was calculated by dividing the gMFI of PE/APC for ΔNef by the gMFI of PE/APC for each Nef sample, such that fold downregu lation by ΔNef was equal to 1.0. One-way ANOVA was used to test for significant variation among groups and Tukey's honestly significant difference test to identify significant differences between specific pairs of groups while adjusting for multiple comparisons. For all statistical tests, P values less than or equal to 0.05 were considered statistically significant. Statistical tests and figures were generated using GraphPad Prism (Version 8, Boston, MA, USA).
## Western blot analysis
To determine Nef-eGFP fusion protein expression levels, HEK293T cells were seeded in 6-well plates 24 hours prior to transfection. HEK293T cells were then transfected with 1 μg of the cloned pN1 Nef-eGFP plasmids using PolyJet. Twenty-four hours post-trans fection, the cells were washed once with cold 1× PBS, then lysed by rocking in lysis buffer (0.5 M HEPES, 1.25 M NaCl, 1 M MgCl 2 , 0.25 M EDTA, 0.1% Triton X-100, and 1× complete Protease Inhibitor Tablets [Roche, Indianapolis, IN, USA]) for 1 hour at 4°C. Supernatants were then clarified by spinning at 20,000 × g for 30 minutes at 4°C, mixed with 5× SDS-PAGE sample buffer (0.3 M Tris, pH 6.8, 2.8 M β-mercaptoethanol, 0.5% bromophenol blue, 50% glycerol, and 10% SDS), and boiled at 98°C for 8 minutes. Samples were run on 12% SDS-PAGE gels, followed by transferring to nitrocellulose membranes (Cytiva; Cat# 10600002). Membranes were blocked with 5% milk in TBST (50 mM Tris, 150 mM NaCl, and 0.1% Tween 20) for 1 hour at room temperature, followed by overnight incubation at 4°C with the appropriate primary antibody in 5% milk in TBST. The following primary antibodies were used: 1:5,000 mouse anti-GFP monoclonal antibody (ThermoFisher; Cat# MA5-15256) and 1:20,000 mouse anti-GAPDH monoclonal antibody (ThermoFisher; Cat#AM4300). Membranes were washed three times with TBST and incubated for 2 hours at room temperature with 1:20,000 (for GAPDH detection) and 1:5,000 (Nef-eGFP detection) of the HRP-conjugated goat anti-mouse IgG (H + L) secondary antibody (ThermoFisher; Cat#31430) in 5% milk in TBST. Blots were subse quently washed in TBST and developed using Immobilon Classico Western HRP Substrate (Millipore; Cat#WBLUC0500) and imaged using a C-DiGit chemiluminescence Western blot scanner (LI-COR Biosciences, Lincoln, NE, USA).
## References
1. Nastri, Pagliano, Zannella et al. (2023) "HIV and drug-resistant subtypes" *Microorgan isms*
2. Van Heuverswyn, Li, Bailes et al. (2007) "Genetic diversity and phylogeographic clustering of SIVcpzPtt in wild chimpanzees in Cameroon" *Virology (Auckl)*
3. Keele, Van Heuverswyn, Li et al. (2025) *Peeters Full-Length Text Journal of Virology*
4. Hahn (2006) "Chimpanzee reservoirs of pandemic and nonpan demic HIV-1" *Science*
5. Sharp, Hahn (2011) "Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med 1:a006841"
6. Worobey, Santiago, Keele et al. (2004) "Origin of AIDS: contaminated polio vaccine theory refuted" *Nature*
7. Bailes, Gao, Bibollet-Ruche et al. (2003) "Hybrid origin of SIV in chimpanzees" *Science*
8. Sharp, Shaw, Hahn (2005) "Simian immunodeficiency virus infection of chimpanzees" *J Virol*
9. Pandrea, Sodora, Silvestri et al. (2008) "Into the wild: simian immunodeficiency virus (SIV) infection in natural hosts" *Trends Immunol*
10. Vandewoude, Apetrei (2006) "Going wild: lessons from naturally occurring T-lymphotropic lentiviruses" *Clin Microbiol Rev*
11. Apetrei, Sumpter, Souquiere et al. (2011) "Immunovirological analyses of chronically simian immunodeficiency virus SIVmnd-1-and SIVmnd-2-infected mandrills (Mandrillus sphinx)" *J Virol*
12. Pandrea, Apetrei, Dufour et al. (2006) "Simian immunodeficiency virus SIVagm.sab infection of Caribbean African green monkeys: a new model for the study of SIV pathogenesis in natural hosts" *J Virol*
13. Silvestri, Sodora, Koup et al. (2003) "Nonpathogenic SIV infection of sooty mangabeys is characterized by limited bystander immunopathology despite chronic high-level viremia" *Immunity*
14. Santiago, Lukasik, Kamenya et al. (2003) "Foci of endemic simian immunodeficiency virus infection in wild-living eastern chimpanzees (Pan troglodytes schweinfurthii)" *J Virol*
15. Keele, Jones, Terio et al. (2009) "Increased mortality and AIDS-like immunopathology in wild chimpanzees infected with SIVcpz" *Nature*
16. Klatt, Silvestri, Hirsch (2012) "Nonpathogenic simian immunodefi ciency virus infections" *Cold Spring Harb Perspect Med*
17. Shah, Al-Haidari, Sun et al. (2021) "T cell receptor (TCR) signaling in health and disease" *Sig Transduct Target Ther*
19. Basmaciogullari, Pizzato (2014) "The activity of Nef on HIV-1 infectivity" *Front Microbiol*
20. Arold, Franken, Strub et al. (1997) "The crystal structure of HIV-1 Nef protein bound to the Fyn kinase SH3 domain suggests a role for this complex in altered T cell receptor signaling" *Structure*
21. Dirk, Pawlak, Johnson et al. (2016) "HIV-1 Nef sequesters MHC-I intracellularly by targeting early stages of endocytosis and recycling" *Sci Rep*
22. Kwon, Kaake, Echeverria et al. (2020) "Structural basis of CD4 downregulation by HIV-1 Nef" *Nat Struct Mol Biol*
23. Rosa, Chande, Ziglio et al. (2015) "HIV-1 Nef promotes infection by excluding SERINC5 from virion incorporation" *Nature*
25. Schindler, Münch, Kutsch et al. (2006) "Nef-mediated suppression of T cell activation was lost in a lentiviral lineage that gave rise to HIV-1" *Cell*
26. Pawlak, Dirk, Jacob et al. (2018) "The HIV-1 accessory proteins Nef and Vpu downregulate total and cell surface CD28 in CD4 + T cells" *Retrovirology (Auckl)*
27. Swigut, Shohdy, Skowronski (2001) "Mechanism for downregulation of CD28 by Nef" *EMBO J*
28. Joas, Sauermann, Roshani et al. (2020) "Nef-mediated CD3-TCR downmodulation dampens acute inflammation and promotes SIV immune evasion" *Cell Rep*
29. Mesner, Hotter, Kirchhoff et al. (2020) "Loss of Nef-mediated CD3 down-regulation in the HIV-1 lineage increases viral infectivity and spread" *Proc Natl Acad Sci*
30. Howe, Jung, Desrosiers (1998) "Zeta chain of the T-cell receptor interacts with nef of simian immunodeficiency virus and human immunodeficiency virus type 2" *J Virol*
31. Schindler, Schmökel, Specht et al. (2008) "Inefficient Nef-mediated downmodulation of CD3 and MHC-I correlates with loss of CD4+T cells in natural SIV infection" *PLoS Pathog*
32. Arhel, Lehmann, Clauss et al. (2009) "The inability to disrupt the immunological synapse between infected human T cells and APCs distinguishes HIV-1 from most other primate lentiviruses" *J Clin Invest*
33. Schmökel, Sauter, Schindler et al. (2011) "The presence of a vpu gene and the lack of Nef-mediated downmodulation of T cell receptor-CD3 are not always linked in primate lentiviruses" *J Virol*
34. Alexander, Du, Howe et al. (1999) "Induction of AIDS in rhesus monkeys by a recombinant simian immunodeficiency virus expressing nef of human immunodeficiency virus type 1" *J Virol*
35. Kirchhoff, Schindler, Specht et al. (2008) "Role of Nef in primate lentiviral immunopathogenesis" *Cell Mol Life Sci*
36. Olabode, Mumby, Wild et al. (2023) "Phylogenetic reconstruction and functional characterization of the ancestral Nef protein of primate lentiviruses" *Mol Biol Evol*
37. Zhang, Wilson, Landford et al. (2009) "Nef proteins from simian immunodeficiency viruses are tetherin antagonists" *Cell Host Microbe*
38. Sauter, Schindler, Specht et al. (2009) "Tetherin-driven adaptation of Vpu and Nef function and the evolution of pandemic and nonpandemic HIV-1 strains" *Cell Host Microbe*
39. Jia, Serra-Moreno, Neidermyer et al. (2009) "Speciesspecific activity of SIV Nef and HIV-1 Vpu in overcoming restriction by tetherin/BST2" *PLoS Pathog*
40. (2025) *Full-Length Text Journal of Virology*
41. Nakano, Matsuda, Yoshikawa et al. (2015) "Down-modulation of primate lentiviral receptors by Nef proteins of simian immunodeficiency virus (SIV) of chimpanzees (SIVcpz) and related SIVs: implication for the evolutionary event at the emergence of SIVcpz" *J Gen Virol*
42. Heigele, Kmiec, Regensburger et al. (2016) "The potency of Nef-mediated SERINC5 antagonism correlates with the prevalence of primate lentiviruses in the wild" *Cell Host Microbe*
43. Meyerson, Sharma, Wilkerson et al. (2015) "Identification of owl monkey CD4 receptors broadly compatible with early-stage HIV-1 isolates"
44. Warren, Meyerson, Dirasantha et al. (2019) "Selective use of primate CD4 receptors by HIV-1" *PLoS Biol*
45. Buffalo, Stürzel, Heusinger et al. (2019) "Structural basis for tetherin antagonism as a barrier to zoonotic lentiviral transmission" *Cell Host Microbe*
46. Bell, Schaefer, Trible et al. (2001) "Downmodulation of the costimulatory molecule, CD28, is a conserved activity of multiple SIV Nefs and is dependent on histidine 196 of Nef" *Virology (Auckl)*
47. Schaefer, Bell, Pfeifer et al. (2002) "The conserved process of TCR/CD3 complex downmodulation by SIV Nef is mediated by the central core, not endocytic motifs" *Virology (Auckl)*
48. Naruki, Saito, Nomaguchi et al. (2025) "Possible acquisition and molecular evolution of vpu genes inferred from comprehensive sequence analysis of human and simian immunodeficiency viruses" *J Mol Evol*
49. Bibollet-Ruche, Bailes, Gao et al. (2004) "New simian immunodeficiency virus infecting De Brazza's monkeys (Cercopithecus neglectus): evidence for a cercopithecus monkey virus clade" *J Virol*
50. Neil, Sandrin, Sundquist et al. (2007) "An interferon-αinduced tethering mechanism inhibits HIV-1 and Ebola virus particle release but is counteracted by the HIV-1 Vpu protein" *Cell Host Microbe*
51. Langer, Hammer, Hopfensperger et al. (2019) "HIV-1 Vpu is a potent transcriptional suppressor of NF-κB-elicited antiviral immune responses"
52. Hotter, Krabbe, Reith et al. (2017) "Primate lentiviruses use at least three alternative strategies to suppress NF-κBmediated immune activation" *PLoS Pathog*
53. Sauter, Hotter, Van Driessche et al. (2015) "Differential regulation of NF-κB-mediated proviral and antiviral host gene expression by primate lentiviral Nef and Vpu proteins" *Cell Rep*
54. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
55. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol Biol Evol*
56. Drummond, Rambaut (2007) "BEAST: Bayesian evolutionary analysis by sampling trees" *BMC Evol Biol*
57. Holmes (2017) "Historian: accurate reconstruction of ancestral sequences and evolutionary rates" *Bioinformatics*
58. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12646003&blobtype=pdf | # MicroRNA response in insect salivary glands to plant virus infection
Yan Xiao, Guohua Liang, Jiaming Zhu, Feng Cui, Wan Zhao
## Abstract
Most arboviruses rely on insect vectors for transmission, with salivary glands serving as a critical gateway for viral spread to new hosts. MicroRNAs (miRNAs) are key regulators of gene expression, yet their roles in salivary glands during virus infection remain poorly understood. Using the small brown planthopper (SBPH)-rice stripe virus (RSV) system, we sequenced small RNAs from salivary glands of nonvirulifer ous and viruliferous insects and identified 5,909 known miRNAs from 201 families. Of these, 1,143 miRNAs were differentially expressed upon RSV infection, including 1,090 upregulated from 36 families and 53 downregulated from 11 families. These differen tially expressed miRNAs were predicted to target 2,876 genes. Gene Ontology analysis showed that targets of upregulated miRNA were enriched in "protein binding, " while those of downregulated miRNAs were associated with "cytoskeleton" and "regulation of dephosphorylation. " The neurotrophin signaling pathway was the top-enriched KEGG pathway for the upregulated miRNA targets. Two miRNAs, miR-276-5p and miR-13a-3p, were specifically modulated by RSV in the salivary glands but not in the guts. miR-276-5p was found to enhance RSV secretion from the salivary glands into rice without altering viral load in insects, whereas miR-13a-3p could play a negative role in viral accumula tion in insects. Taken together, our findings highlight tissuespecific miRNA responses in vector-virus interactions and uncover distinct roles of salivary gland miRNAs in regulating viral transmission.IMPORTANCE Most plant viruses depend on insect vectors for transmission. The salivary glands of insect vectors are the last barrier for these viruses to overcome before being transmitted to plant hosts. In this work, we dissected the microRNA (miRNA) response to plant virus infection in insect salivary glands using the model of the small brown planthopper and rice stripe virus (RSV). The abundance of hundreds of miRNAs changes in the salivary glands after RSV infection. Two specific miRNAs play distinct roles. One enhances the release of RSV from salivary glands into rice plants, and the other regulates viral accumulation within the insects. These findings deepen our understanding of small RNA reactions and potential functions of miRNAs to viral infection in the salivary glands of insect vectors. KEYWORDS plant virus, rice stripe virus, salivary glands, microRNA, small brown planthopper, viral transmission A rboviral diseases account for over 17% of all human infectious diseases and place approximately 80% of the global population at risk (1). More than 70% of plant viruses are transmitted by insects, resulting in annual agricultural losses of up to $60 billion (2). Insects are the primary vectors for arboviruses, which can establish long-term, persistent infections in insect vectors, such as mosquitoes for certain human viruses or planthoppers for rice viruses, often with little to no apparent impact on insects' physiology (3, 4). Arboviruses typically initiate infection by crossing the midgut
epithelium, then spread via the hemolymph to the salivary glands, where they are transmitted to hosts through saliva during feeding or passed vertically via the ovaries (5). The midgut and salivary glands serve as key barriers for viral infection and transmission (6). Unlike the midgut, which supports viral replication, the salivary glands represent the final tissue barrier for cross-kingdom transmission. The ability to cross the salivary gland barrier and the viral load in this tissue are critical for viral transmission.
Traditionally, research on salivary glands has focused on protein effectors, which modulate viral infection by influencing viral persistence, replication, release, and host interactions (7)(8)(9)(10). However, the discovery of microRNAs (miRNAs) in saliva and salivary glands of arthropod vectors has revealed a new class of potential regulators in virusvector interactions (11)(12)(13)(14). Unlike proteins that primarily function through protein-pro tein interactions (8)(9)(10), miRNAs exert broad regulatory effects by base-pairing with target mRNAs, influencing gene expression in both vectors and hosts (14)(15)(16). Although small RNA sequencing has confirmed the presence of miRNAs in the salivary glands and saliva of mosquitoes, ticks, and planthoppers, their functional roles in viral infec tion and transmission remain poorly characterized (11)(12)(13)(14). Nonetheless, emerging evidence highlights their potential significance. In the small brown planthopper (SBPH), salivary gland-enriched miR-315-5p promotes rice black-streaked dwarf virus infection by suppressing a melatonin receptor in the insect (17). Another SBPH miRNA, miR-263a, enhances rice stripe virus (RSV) replication by directly targeting the 3′ extended terminal region of viral genomic RNA1 and the 5′ UTR of insect cathepsin B-like gene (14,18). In ticks, infection with Powassan virus (POWV) alters salivary miRNA expression, with 35 miRNAs upregulated and 17 downregulated. Functional assays revealed that inhibiting specific upregulated miRNAs-such as isc-miR-315, isc-miR-5307, and several novel candidates-leads to increased POWV levels in host cells, suggesting these miRNAs suppress viral replication (19). Furthermore, bioinformatic analyses indicate that vector salivary miRNAs may target host mRNAs involved in immune and inflamma tion pathways, potentially shaping the local environment to influence viral infection or secretion (11)(12)(13). Notably, one study demonstrated that SBPH saliva delivers miR-263a into rice plants, where it silences GATA19 to activate jasmonate signaling and enhance plant resistance to RSV-providing direct evidence of salivary miRNA activity in host tissues (14). Despite these advances, the specific role of salivary gland miRNAs in regulating viral secretion during feeding remains poorly understood.
RSV is one of the most destructive rice viruses in East Asia, especially China, Japan, and Korea (20). RSV, a member of the Tenuivirus genus, has a genome consisting of four single-stranded RNA segments, encoding a nucleocapsid protein (NP), an RNA-depend ent RNA polymerase, and five nonstructural proteins (21). NP gene RNA or NP protein levels are commonly used as reliable indicators of viral load and are standard metrics for detecting and quantifying RSV in both insect vectors and plants (4). RSV is efficiently transmitted between rice plants by SBPH (Laodelphax striatellus) in a persistent and circulative-propagative manner (20). The salivary glands play a crucial role in viral transmission, with proteins, such as Importin α2, LssaMP, and Exportin 6, identified as key viral receptors and effectors facilitating viral infection and secretion (7,8,22). However, the response of salivary gland miRNAs to RSV infection and their functional significance has not yet been systematically investigated.
In this study, we constructed and compared small RNA libraries from the salivary glands of nonviruliferous and viruliferous SBPH to characterize the overall miRNA response to RSV infection. By analyzing differentially expressed miRNAs and their potential targets, we revealed distinct regulatory patterns for upregulated and downre gulated miRNAs. We further investigated two miRNAs specifically modulated by RSV in the salivary glands for their roles in viral accumulation and release.
## RESULTS
## Remodeling of miRNA expression by RSV infection in the salivary glands of small brown planthoppers
To elucidate the profiles of miRNAs in the salivary glands of the SBPH during RSV infection, sRNA libraries were constructed and sequenced from salivary gland samples of nonviruliferous (N) and viruliferous (V) fourth-instar nymphs. Three biological replicates for each group were sequenced. Principal component analysis revealed clear separation between N and V samples along principal component 1. Clustering was tighter among N samples, suggesting greater consistency in the uninfected state (Fig. S1). Following the exclusion of adapter sequences and low-quality reads, an average of 21,366,618 and 21,627,593 clean reads were obtained for N and V groups (NCBI number PRJNA1307885) (Table 1). The dominant sizes of sRNAs were 22 and 27 nt, and the size distributions of sRNAs were essentially identical for the two groups (Fig. 1A). A total of 6,277 miRNAs were identified, including 5,909 known miRNAs (Table S1) and 368 novel miRNAs (Table S2). A significantly higher number of miRNAs was detected in V samples compared to N samples (Fig. 1B). Based on sequence similarity of their mature forms, the 5,909 known miRNAs were classified into 201 established miRNA families. Among them, 1,186 miRNAs (|log₂FC| > 1, P-adjust < 0.05) were differentially expressed miRNAs (DEMs), including 1,143 known miRNAs. Of these, 1,090 miRNAs were upregulated, representing 36 families, while 53 miRNAs were downregulated, belonging to 11 families (Table S3).
Comparison of the top 20 most highly expressed known miRNAs with DEMs revealed that miR-276 and miR-9 families were both upregulated and highly abundant, whereas miR-281, miR-8, miR-13, and miR-124 families were among the downregulated miRNAs yet ranked highly in expression (Fig. 1C andD), suggesting that downregulated miRNAs tend to be more abundantly expressed in salivary glands under RSV infection.
## Potential regulatory networks of miRNA target genes
To investigate the regulatory roles of the known DEMs, target gene prediction was performed, yielding 2,876 targets in total-2,623 for upregulated miRNAs and 464 for downregulated miRNAs, with 211 targets shared between both groups (Fig. 1E). Gene ontology (GO) annotation of these targets revealed that "binding, " "cell part, " and "cellular process" were the most highly represented categories in molecular function, cellular component, and biological process (Fig. S2). In addition, "transcription regulator activity" was specifically obtained in the targets of upregulated miRNAs (Fig. S2A), and "extracellular region" was uniquely associated with the targets of downregulated miRNAs (Fig. S2B). In the GO enrichment analyses (P value < 0.05), "protein binding" was the most significantly enriched GO term among the targets of upregulated miRNAs. The remaining 19 enriched GO terms were relatively evenly distributed, with "protein modification process, " "regulation of DNA-templated transcription, " "regulation of nucleic acidtemplated transcription, " "regulation of RNA metabolic process, " and "regulation of RNA biosynthetic process" containing the highest numbers of target genes (Fig. 2). For the targets of downregulated miRNAs, the strongest enrichment was observed in "cytoskele ton" and "regulation of dephosphorylation. " The other 18 enriched GO terms were broadly distributed, with "molecular function regulator activity, " "phosphatase activity, " "phosphoric ester hydrolase activity, " and "dephosphorylation" encompassing the largest numbers of target genes (Fig. 2). When P-adjust <0.05 was applied for the enrichment analysis, only "protein binding" was significantly enriched for the targets of upregulated miRNAs. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation revealed that the targets of upregulated and downregulated miRNAs shared six common pathways, representing the majority of the assigned pathways in both groups (Fig. S3). How ever, some pathways were unique to each group: folding, sorting, and degradation (under genetic information processing) and lipid metabolism (under metabolism) were specifically to the targets of upregulated miRNAs (Fig. S3A), while circulatory system (under organismal systems) and carbohydrate metabolism (under metabolism) were unique to the targets of downregulated miRNAs (Fig. S3B). In addition, KEGG enrichment analysis (P value < 0.05) revealed that the neurotrophin signaling pathway was the most significantly enriched among the targets of upregulated miRNAs, while the remaining 19 enriched pathways were relatively evenly distributed (Fig. 3). In contrast, for the targets of downregulated miRNAs, the 20 enriched KEGG pathways showed a uniform distribu tion (Fig. 3). When P-adjust <0.05 was applied for the enrichment analysis, only the neurotrophin signaling pathway was significantly enriched for the targets of upregulated miRNAs.
## RSV specifically regulates miR-276-5p and miR-13a-3p in salivary glands
Among the 1,143 known DEMs, 775 DEMs with an average read count greater than 1 in the V samples, representing 23 miRNA families, were selected for experimental validation. The most abundantly expressed miRNAs from each family were chosen for cloning. Reverse transcriptase PCR (RT-PCR), followed by Sanger sequencing, confirmed the presence of 12 miRNAs, and seven of these showed significant expression changes after RSV acquisition (Fig. 4A; Table S4). Among them, the expression trends of miR-1, miR-92a-3p, miR-276-5p, and miR-13a-3p were consistent with sRNA sequencing results. miR-1, miR-92a-3p, and miR-276-5p were upregulated upon RSV infection, while miR-13a-3p was downregulated.
The expression of the four miRNAs was further compared in the salivary glands and guts with and without RSV infection using quantitative real-time PCR (qPCR). As shown in Fig. 4B, miR-1 and miR-92a-3p were upregulated in both the salivary glands and guts upon RSV infection, suggesting they may function in both tissues during RSV infection. In contrast, miR-276-5p and miR-13a-3p responded to RSV uniquely in the salivary glands but not in the guts. miR-276-5p was specifically upregulated, and miR-13a-3p was specifically downregulated by RSV in the salivary glands. Therefore, we selected these two miRNAs for functional study.
## Effects of miR-276-5p and miR-13a-3p on RSV accumulation and secretion
To investigate whether miR-276-5p and miR-13a-3p regulate RSV infection, the agomirs (chemically modified double-stranded miRNA mimics) and antagomirs (single-stranded inhibitors) for each miRNA were synthesized and injected into viruliferous SBPH nymphs. The treated insects were allowed to feed on healthy rice seedlings for 1 day. Injection of miR-276-5p agomir significantly increased miR-276-5p amount in the whole body and salivary glands of SBPH (Fig. 5A). No significant changes in viral accumulation in terms of NP RNA levels were observed in either whole insects or salivary glands (Fig. 5B). However, rice leaves fed upon by insects treated with miR-276-5p agomir showed significantly higher RNA levels of NP compared to those fed upon by agomir-NC-treated insects, indicating the enhanced viral secretion (Fig. 5C). Injection of miR-276-5p antagomir did not affect viral accumulation either (Fig. 5D andE), but miR-276-5p antagomir treatment significantly decreased viral secretion (Fig. 5F). In addition, no significant influence on insect survival was observed after injection with agomir or antagomir of miR-276-5p. These results suggest that miR-276-5p promotes RSV secretion from the salivary glands into host plants without affecting viral replication.
For miR-13a-3p, injection of miR-13a-3p agomir significantly reduced NP RNA levels in whole insects but did not affect viral loads in the salivary glands or viral transmission to rice leaves (Fig. 6A through C). miR-13a-3p antagomir treatment only led to a slight but not significant increase in NP RNA levels in whole insects with no impact on viral accumulation in the salivary glands (Fig. 6D andE) or virus secretion into rice leaves (Fig. 6F). Additionally, no significant impact on insect survival was observed following injection of either the agomir or antagomir of miR-13a-3p. These findings indicate that miR-13a-3p may play a negative role in viral accumulation in insects.
## Potential regulatory networks of the targets of miR-276-5p
A total of 33 transcripts, corresponding to 30 genes, were predicted as candidate targets of miR-276-5p. Among these, three target genes-zinc finger protein 2-like, guanylate kinase-associated protein Mars, and Protein phosphatase 1B-were predicted by both miRanda and RNAhybrid, lending higher confidence to these interactions; the remaining targets were predicted by only one of the two algorithms (Table S5). GO annotation of these targets revealed that the most highly represented functional categories among these targets were "binding" and "membrane part" (Fig. S4A). Notably, in comparison to the GO profiles of targets for upregulated miRNAs, the "membrane part" category was specifically and prominently represented among miR-276-5p targets, suggesting a specialized role for this miRNA in membrane-associated processes. In the GO enrich ment analyses using a P value threshold of <0.05, the top three most significantly enriched terms were "N-acetylglucosaminylphosphatidylinositol deacetylase activity, " "RNA polymerase II general transcription initiation factor binding, " and "1-phosphatidyli nositol-4-phosphate 3-kinase activity. ". The remaining enriched GO terms were relatively evenly distributed, with "transferase activity, transferring phosphorus-containing groups" encompassing the largest number of target genes (Fig. S4B). However, when the more stringent P-adjust <0.05 was applied to correct for multiple testing, no GO terms remained significantly enriched. KEGG pathway annotation showed that signal transduction (under environmental information processing) was the most frequently represented pathway (Fig. S5A), a trend also observed in the KEGG analysis of targets of both up-and downregulated miRNAs. KEGG enrichment analysis at P value <0.05 indicated significant enrichment in pathways, such as FoxO signaling pathway, human immunodeficiency virus 1 infection, oocyte meiosis, and cell cycle (Fig. S5B). Nevertheless, when the P-adjust <0.05 criterion was applied, no KEGG pathways were found to be significantly enriched.
In conclusion, the predicted targets of miR-276-5p are enriched in membrane-asso ciated components and key signaling functions-such as 1-phosphatidylinositol-4-phos phate 3-kinase activity and the FoxO signaling pathway-many of which are linked to cytoskeletal organization and membrane trafficking. This suggests a biologically plausible mechanism by which miR-276-5p could influence viral stability or secretion in the salivary glands through the regulation of membrane dynamics and intracellular signaling pathways. However, these enrichments did not reach statistical significance after correction for multiple testing (P-adjust <0.05), indicating that the pathway-level evidence remains suggestive rather than conclusive. Therefore, while the predicted regulatory network provides valuable hypotheses for how miR-276-5p may modulate host processes relevant to viral transmission, it should be viewed as a starting point for further investigation. Experimental validation of individual target genes will be essential to fully understand the functional role of miR-276-5p in the virus-insect-host interaction.
## DISCUSSION
Salivary glands are critical determinants of arbovirus transmission. In this study, we present a comprehensive profile of the miRNA response to RSV in the salivary glands of SBPH, revealing distinct regulatory patterns between upregulated and downregula ted miRNAs following viral infection. We identified two salivary glandspecific miRNAs that respond to RSV and play divergent roles in viral accumulation and transmission, potentially finetuning different stages of the viral transmission cycle. Our findings offer new insights into the molecular interplay between insect vectors and plant arboviruses and highlight promising targets for disrupting the transmission of insect-borne plant viruses.
While both the salivary glands and guts are essential for viral replication and transmission, they often display distinct molecular and functional responses across various vector-virus systems. For example, in Aedes aegypti mosquitoes infected with dengue virus, 190 genes showed differential expression in the salivary glands, whereas 13 genes were differentially expressed in the guts. Only the tubulin beta chain was differentially regulated in both tissues (23). In Ae. aegypti infected with Zika virus, the gut mounts a robust antiviral response, primarily through activation of the siRNA and piRNA pathways, whereas the salivary glands create a more permissive environment, potentially facilitating transmission via immune suppression. There, immune-related pathways, such as Toll, IMD, and JAK/STAT, show more pronounced transcriptional changes (24). Our previous transcriptomic analysis of the SBPH revealed a stronger immune response to RSV in the salivary glands than in the guts. Specifically, RSV infection in the guts upregulates genes related to lysosomal function, digestion, and detoxification, while suppressing those involved in DNA replication and repair. In the salivary glands, RSV promotes RNA transport but suppresses key signaling pathways, such as MAPK, mTOR, Wnt, and TGF-β (25). In this study, we characterized the miRNA response to RSV infection in the salivary glands and identified potential differences in regulatory mechanisms between the two tissues. We found that miR-1 and miR-92a-3p are responsive to RSV infection in both tissues, whereas miR-276-5p and miR-13a-3p are specifically regulated in the salivary glands upon viral infection. This suggests that miR-1 and miR-92a-3p may indirectly affect viral accumulation or secretion in the salivary glands by modulat ing viral replication in the guts. In contrast, miR-276-5p and miR-13a-3p likely play direct roles in regulating viral stability or secretion within the salivary glands, making them promising targets for strategies aimed at blocking the horizontal transmission of RSV. These findings highlight the tissuespecific nature of host-virus interactions and underscore the unique role of the salivary glands in modulating the balance between viral persistence and transmission.
Emerging studies have uncovered the dynamics of salivary miRNAs in vector-borne virus transmission. For example, sRNA-seq analysis of Ae. aegypti and Aedes albopictus saliva identified 103 exogenous miRNAs. Notably, 59 and 30 known miRNAs were upregulated in the saliva of chikungunya virus (CHIKV)-infected Ae. aegypti and Ae. albopictus, respectively, suggesting these miRNAs may play functional roles in CHIKV dissemination and transmission (11). However, a more recent study comparing miRNA expression in both saliva and salivary glands in non-infected and CHIKV-infected Ae. aegypti only identified two upregulated and three downregulated miRNAs in saliva, while no significant changes in miRNAs were observed in the salivary glands. Instead, CHIKV infection primarily activated the siRNA and piRNA pathways (26). In female Ixodes scapularis ticks infected with POWV, sRNA sequencing revealed 379 salivary gland miRNAs with 35 upregulated and 17 downregulated. Functional assays showed that inhibiting specific upregulated miRNAs, such as isc-miR-315, isc-miR-5307, and several novel candidates, increased viral levels in host cells, indicating these miRNAs act to suppress viral replication (19). In our study, RSV infection upregulated 1,090 miRNAs and downregulated 53 miRNAs in the salivary glands of SBPH. These results collectively indicate that miRNA response to arboviruses may be tissue-and context-spe cific and highlight the complexity of vector-virus interactions across different systems.
Investigating miRNA dynamics in the salivary glands would offer valuable insights into the complex interactions between viruses and their insect vectors.
We found that miR-276-5p enhanced RSV transmission by promoting viral secre tion into rice plants. miR-276 is a highly conserved miRNA family and has been reported in many insects, including Ae. aegypti, Drosophila melanogaster, Bactrocera dorsalis, Tribolium castaneum, Locusta migratoria, and Aphis gossypii (27). This miRNA family consists of two mature forms-miR-276-5p and miR-276-3p. They are primarily known to regulate key biological processes, such as insect development, metabolism, and reproduction. For instance, in D. melanogaster, miR-276-3p suppresses the insu lin signaling pathway by repressing InR expression, ultimately resulting in decreased developmental growth (28). In L. migratoria, miR-276-3p promotes the egg-hatching synchrony by upregulating brm expression in the ovaries (27). In Anopheles coluzzii, blood meal-induced miR-276-5p finetunes the expression of branched-chain amino acid transferase, contributing to the termination of the reproductive cycle (29). Additionally, in A. gossypii, miR-276-5p enhances susceptibility to the insecticide spirotetramat by downregulating acetyl-CoA carboxylase in resistant strains (30). We demonstrated that miR-276-5p enhances RSV transmission by facilitating viral secretion into rice plants without affecting viral load within insects. Further studies are required to explore the potential mechanisms for this specific phenomenon in the future.
## MATERIALS AND METHODS
## Insect preparation
Nonviruliferous (N) and viruliferous (V) SBPH populations were maintained separately on rice seedlings (Oryza sativa subsp. japonica) Wuyujing 3 in glass incubators at 24°C under a 16 h light/8 h dark photoperiod. The RSV-carrying rates of the viruliferous SBPH populations were monitored every 3 months using a dot enzyme-linked immunosorbent assay with homemade monoclonal anti-NP antibodies (4).
## RNA extraction and sRNA-seq library construction
Total RNA was separately extracted from the salivary glands of nonviruliferous and viruliferous fourth-instar SBPH nymphs. For each condition, three biological replicates were prepared, with 30 salivary glands per replicate. RNA extraction was performed using TRIzol reagent (Invitrogen, Carlsbad, CA, USA; 15596026) following the manufac turer's protocol. The quality and concentration of each RNA sample were assessed by 1.2% agarose gel electrophoresis and by measuring the OD260/OD280 ratio using a NanoDrop One spectrophotometer (Thermo Scientific, Waltham, MA, USA; 840-317400). One microgram of prepared total RNA with a concentration ≥50 ng/µL, RQN >7, OD260/280 between 1.8 and 2.2 was ligated to the 5′ and 3′ adaptors and reverse transcribed into cDNAs using the QIAseq miRNA Library Kit (Qiagen, Germany). The adapter-ligated cDNA was enriched by PCR, then purified and size-selected on a 6% Novex TBE PAGE gel to isolate small RNA fragments of 18-30 nt. Final libraries were sequenced using 140-160 bp single-end reads on the NovaSeq X Plus platform at Majorbio Bio-Pharm Technology Co., Ltd., Shanghai, China.
## Analysis of sRNA sequencing data
sRNA sequencing data were analyzed using the online platform of Majorbio Cloud Platform. Briefly, clean reads mapped to the SBPH reference transcript database (31) were aligned against miRBase (32) and Rfam (33) databases to annotate known miRNAs and other non-coding RNAs. Unannotated reads were analyzed using miRDeep2 to predict novel miRNAs (34). Expression levels of known and novel miRNAs were quantified and normalized using transcripts per million. DEMs between N and V samples were identified using DESeq2 with the following thresholds: a minimum twofold change in expression (log₂ fold change >1 or <-1) and an adjusted P value <0.05 (35).
## Target prediction and functional enrichment analysis
Target genes of known DEMs were predicted using miRanda (36) and RNAhybrid (37) algorithms on the Majorbio Cloud Platform, based on the SBPH genome databases (31). Prediction relied on the "seed region" rule, requiring base pairing between positions 2-8 of the mature miRNA and the target mRNA. For miRanda, the parameters were set as follows: score cutoff ≥160, energy cutoff ≤-18 kcal/mol, and strict mode enabled. For RNAhybrid, the energy cutoff was set to ≤-18 kcal/mol and the P value cutoff to ≤0.05. A gene was considered a candidate target if predicted by at least one of the two algorithms.
Predicted target genes were functionally annotated using public databases, including GO, KEGG, EggNOG, NR, Swiss-Prot, Pfam, and Rfam. GO enrichment analysis was performed using the Python package Goatools (https://github.com/tanghaibao/GOat ools), with significance defined by both P value and adjusted P value <0.05. KEGG pathway enrichment was conducted in R using Fisher's exact test, and pathways with a P value and adjusted P value <0.05 were considered statistically enriched.
## cDNA synthesis and quantitative PCR
Total RNA was extracted from 10 salivary glands, eight guts, five whole insect bodies of nonviruliferous and viruliferous fourth-instar SBPH nymphs, or two rice leaves using TRIzol reagent (Invitrogen). For miRNA quantification, 2 µg of total RNA was reverse transcribed into cDNA using the miRcute Plus miRNA First-Strand cDNA Synthesis Kit (Tiangen, Beijing, China; KR211) according to the manufacturer's instructions (14). For viral RNA detection, 2 µg of RNA was reverse transcribed using the SuperScript III First-Strand Synthesis System (Thermo Scientific; 18080051) with random hexamer primers (Promega, Madison, WI; C1181).
qPCR was performed on a LightCycler 480 Instrument II (Roche, Basel, Switzerland). miRNA expression levels were measured using the miRcute miRNA qPCR Detection Kit (Tiangen; FP411), while mRNA and viral RNA transcripts were amplified using LightCycler 480 SYBR Green I Master (Roche; 04887352001). The qPCR reaction system was a 20 µL volume, consisting of 10 µL of 2× SYBR Green I Master Mix (Roche) or 2× miRcute miRNA premix (Tiangen), 4 µL of cDNA template, and 0.5 µL of each primer (10 µM). For the evaluation of miRNA expression, the forward primer for each target miRNA was designed based on its mature sequence, while the reverse primer was provided by the miRcute miRNA qPCR Detection Kit (Tiangen). All qPCR products were sequenced to confirm primer specificity. In SBPH, EF2 (Contig0.299) was used as the reference gene for normalizing viral NP levels, and U6 snRNA was used as the endogenous control for miRNA expression (18). In rice, UBQ10 (LOC4328390) served as the internal control for normalization of viral NP level (14). Each experiment included six to eight biological replicates for insect samples and at least 12 replicates for rice samples. All primer sequences are provided in Table S6.
## Injection of miRNA agomir or antagomir
Agomirs and antagomirs targeting miR-276-5p and miR-13a-3p were designed and synthesized by GenePharma (Shanghai, China). Agomirs, which function as miRNA mimics, were delivered as chemically modified dsRNAs with the following sequences: for miR-276-5p (sense: 5′-AGCGAGGUAUAGAGUUCCUACG-3′, antisense: 5′-UAGGAACUCUAUACCUCGCUUU-3′), for miR-13a-3p (sense: 5′-UAUCACAGCCACUUU GAUGUGGU-3′, antisense: 5′-CACAUCAAAGUGGCUGUGAUAUU-3′), and a negative control (NC) agomir (sense: 5′-UUCUCCGAACGUGUCACGUTT-3′, antisense: 5′-ACGUGA CACGUUCGGAGAATT-3′). Antagomirs, designed to inhibit endogenous miRNA function, were chemically modified ssRNAs complementary to the mature miRNA sequen ces: miR-276-5p (5′-CGUAGGAACUCUAUACCUCGCU-3′), miR-13a-3p (5′-ACCACAUCAAA GUGGCUGUGAUA-3′), with a scrambled sequence (5′-CAGUACUUUUGUGUAGUACAA-3′) used as the NC antagomir. A volume of 13.8 nL of each oligonucleotide (250 µM) was microinjected into viruliferous fourth-instar SBPH nymphs using a Nanoliter 2000 microinjection system (World Precision Instruments). Four days after injection, whole insects or dissected salivary glands were collected for RNA extraction to assess the impact of miRNA upregulation or inhibition on RSV accumulation.
## Determination of RSV secretion by salivary glands
To assess the effect of miRNA modulation on RSV secretion, groups of 15 virulifer ous fourth-instar SBPH nymphs were microinjected with agomir, antagomir, or their respective NC. After 3 days of incubation to allow for miRNA modulation, the treated insects were transferred onto 3-week-old healthy rice seedlings (15 insects per plant) and allowed to feed for 1 day. The insects were then carefully removed, and the rice leaves were collected for viral detection. Two leaves from each plant were used as one biological replicate, and at least 12 replicates were prepared per treatment group. RSV accumulation in the rice leaves was quantified by qPCR to determine the relative RNA level of NP.
## Statistical analysis
All bar charts were created using GraphPad Prism 9 (GraphPad Software, San Diego, CA) based on original experimental data. Statistical significance between two groups was assessed using Student's t-test.
## References
1. Oliveira, Bahia, Vale (2020) "How are arbovirus vectors able to tolerate infection?" *Dev Comp Immunol*
2. Selvarajan, Balasubramanian (2016) "Cutting-edge technologies for detection of plant viruses in vegetatively propagated crop plants"
3. Liu, Swevers, Kolliopoulou et al. (2019) "Arboviruses and the challenge to establish systemic and persistent infections in competent mosquito vectors: the interaction with the RNAi mechanism" *Front Physiol*
4. Zhao, Yang, Kang et al. (2016) "Different pathogenicities of Rice stripe virus from the insect vector and from viruliferous plants" *New Phytol*
5. Whitfield, Falk, Rotenberg (2015) "Insect vector-mediated transmission of plant viruses" *Virology (Auckl)*
6. Franz, Kantor, Passarelli et al. (2015) "Tissue barriers to arbovirus infection in mosquitoes" *Viruses*
7. Ma, Lu, Wang et al. (2021) "Membrane association of importin α facilitates viral entry into salivary gland cells of vector insects" *Proc Natl Acad Sci*
8. Lu, Zhu, Yu et al. (2022) "Key role of exportin 6 in exosome-mediated viral transmission from insect vectors to plants" *Proc Natl Acad Sci*
9. Sri-In, Weng, Chen et al. (2019) "A salivary protein of Aedes aegypti promotes dengue-2 virus replication and transmission" *Insect Biochem Mol Biol*
10. (1016)
11. Wang, Lu, Guo et al. (2024) "Leafhopper salivary vitellogenin mediates virus transmission to plant phloem" *Nat Commun*
12. Maharaj, Widen, Huang et al. (2015) "Discovery of mosquito saliva microRNAs during CHIKV infection" *PLoS Negl Trop Dis*
13. Hackenberg, Langenberger, Schwarz et al. (2017) "In silico target network analysis of de novo-discovered, tick salivaspecific microRNAs reveals important combinatorial effects in their interference with vertebrate host physiology" *RNA*
14. Arcà, Colantoni, Fiorillo et al. (2019) "MicroRNAs from saliva of anopheline mosquitoes mimic human endogenous miRNAs and may contribute to vector-hostpathogen interactions" *Sci Rep*
15. Zhao, Lu, Zhu et al. (2025) "A double-agent microRNA regulates viral cross-kingdom infection in animals and plants" *EMBO J*
16. Zhou, Liu, Yan et al. (2014) "miR-281, an abundant midgutspecific miRNA of the vector mosquito Aedes albopictus enhances dengue virus replication" *Parasites Vectors*
17. Slonchak, Hussain, Torres et al. (2014) "Expression of mosquito microRNA Aae-miR-2940-5p is downregulated in response to West Nile virus infection to restrict viral replication" *J Virol*
18. Zhang, Dong, Wang et al. (2021) "MicroRNA-315-5p promotes rice black-streaked dwarf virus infection by targeting a melatonin receptor in the small brown planthopper" *Pest Manag Sci*
19. Zhao, Yu, Jiang et al. (2021) "Coordination between terminal variation of the viral genome and insect microRNAs regulates rice stripe virus replication in insect vectors" *PLoS Pathog*
20. Hermance, Widen, Wood et al. (2019) "Ixodes scapularis salivary gland microRNAs are differentially expressed during Powassan virus transmission" *Sci Rep*
21. Cho, Lian, Kim et al. (2013) "Current insights into research on rice stripe virus" *Plant Pathol J*
22. Zhao, Wang, Xu et al. (2019) "Distinct replication and gene expression strategies of the Rice Stripe virus in vector insects and host plants" *J Gen Virol*
23. Huo, Zhao, Meng et al. (2022) "Laodelphax striatellus saliva mucin enables the formation of stylet Full-Length Text Journal of Virology November"
24. "sheathes to facilitate its feeding and rice stripe virus transmission" *Pest Manag Sci*
25. Bonizzoni, Dunn, Campbell et al. (2012) "Complex modulation of the Aedes aegypti transcriptome in response to dengue virus infection" *PLoS One*
26. Zhu, Jiang, Zhang et al. (2025) "Transcriptome analysis of Aedes aegypti midgut and salivary gland post-Zika virus infection" *Curr Res Parasitol Vector Borne Dis*
27. Zhao, Lu, Yang et al. (2015) "Organspecific transcriptome response of the small brown planthopper toward rice stripe virus" *Insect Biochem Mol Biol*
28. Fiorillo, Yen, Colantoni et al. (2022) "MicroRNAs and other small RNAs in Aedes aegypti saliva and salivary glands following chikungunya virus infection" *Sci Rep*
29. He, Wei, Jiang et al. (2016) "MicroRNA-276 promotes egg-hatching synchrony by up-regulating brm in locusts" *Proc Natl Acad Sci*
30. Lee, Kim, Jang et al. (2023) "Ecdysoneinduced microRNA miR-276a-3p controls developmental growth by targeting the insulin-like receptor in Drosophila" *Insect Mol Biol*
31. Lampe, Jentzsch, Kierszniowska et al. (2019) "Metabolic balancing by miR-276 shapes the mosquito reproductive cycle and Plasmodium falciparum development" *Nat Commun*
32. Wei, Zheng, Peng et al. (2016) "miR-276 and miR-3016-modulated expression of acetyl-CoA carboxylase accounts for spirotetramat resistance in Aphis gossypii Glover" *Insect Biochem Mol Biol*
33. Tong, Chen, Xiao et al. (2022) "Alternative splicing landscape of small brown planthopper and different response of JNK2 isoforms to Rice stripe virus infection" *J Virol*
34. Kozomara, Griffithsjones (2014) "miRBase: annotating high confidence microRNAs using deep sequencing data" *Nucleic Acids Res*
35. Nawrocki, Burge, Bateman et al. (2015) "Rfam 12.0: updates to the RNA families database" *Nucleic Acids Res*
36. Friedländer, Mackowiak, Li et al. (2012) "miRDeep2 accurately identifies known and hundreds of novel microRNA genes in seven animal clades" *Nucleic Acids Res*
37. Love, Huber, Anders (2014) "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2" *Genome Biol*
38. John, Enright, Aravin et al. (2004) "Prediction of human microRNA targets" *PLoS Biol*
39. Rehmsmeier, Steffen, Hochsmann et al. (2004) "Fast and effective prediction of microRNA/target duplexes" *RNA* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12791788&blobtype=pdf | Sara Kim, Linda Sircy, Yun Lim, Khaleel Yahya, Terry Stevens-Ayers, Nina Ozbek, Rachel Blazevic, Larry Mose, Louise Kimball, Michael Boeckh, Alpana Waghmare
## Abstract
citation ID: ofaf695.1982 P-1813. The natural history of respiratory viral antibodies during low circulation of respiratory viruses during the COVID-19 pandemic
Methods. VirScan, an immunosurvey that can identify antibodies against viruses with human tropism, was performed on samples from a cohort of patients in a SARS-CoV-2 surveillance study from May 2020 to June 2021. The median study period was 229 days. We selected 4 RVs (human metapneumovirus [HMPV], influenza A and B, and respiratory syncytial virus. The enrollment samples were ranked into quartiles (Q1-Q4) based on the geometric mean (gMean) epitope binding signal (EBS), a VirScan surrogate for an antibody titer. Q1 represented the highest gMean EBS, likely due to recent RV exposure. We used linear mixed effect (LME) models to model longitudinal changes in gMean EBS for each RV and calculated fold change (ratio of gMean EBS at end of study to enrollment). Fold change values were log₂-transformed for visualization and statistical analysis.
Results. A total of 204 immunocompetent adults were included [median age 43.5 (IQR: 32.3 -54.8)]. LME models showed a consistent decline in gMean EBS over time for Q1 compared to Q4, which demonstrated a modest increase or stability in antibody levels (Figure 1A-D). This trend was most pronounced for HMPV, which also exhibited greater variability in response compared to other RVs (Figure 1A). The log 2 --transformed fold change in gMean EBS between timepoints was compared for each RV between Q1 and Q4. Q1 fold changes were significantly greater than Q4 for all RVs, indicating greater waning (Fig. 2A-D, P< 0.0001).
Conclusion. We demonstrate that VirScan can assess population level changes in antibody repertoires. Our results show waning humoral immunity across 4 RV during a period of low circulation in patients with the highest EBS (antibody titers) at enrollment. We were able to assess antibody levels for multiple viruses using a single assay, thus allowing for broad immune profiling compared to traditional methods of assessing antibody kinetics. Future studies will utilize VirScan to assess changes at an epitope level.
Disclosures. Michael J. Boeckh, MD PhD, Allovir: Advisor/Consultant| Ansun Biopharma: Grant/Research Support|AstraZeneka: Advisor/Consultant| AstraZeneka: Grant/Research Support|GSK: Grant/Research Support|Merck: Advisor/Consultant|Merck: Grant/Research Support|Moderna: Advisor/ Consultant|Moderna: Grant/Research Support|Symbio: Advisor/Consultant|Vir Biotechnology: Grant/Research Support Alpana Waghmare, MD, Ansun Biopharma: Clinical trial site|AstraZeneca: Advisor/Consultant|GSK: Advisor/ Consultant|GSK: Grant/Research Support|Merck: Advisor/Consultant|Merck: Grant/Research Support|Pfizer: Clinical trial site|Shionogi: Clinical trial site|Vir Biotechnology: Advisor/Consultant |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793649&blobtype=pdf | Jeffrey Pearson, Lindsey Baden, Lisa Cosimi, Benjamin Gewurz, Eric Gillett, Simran Gupta, Nicolas Issa, David Kubiak, Alexis Liakos, Jessica Little, Pritha Sen, Ann Woolley
Background. Cytomegalovirus (CMV) infection causes significant morbidity and mortality in immunocompromised patients. Despite novel antiviral agents like maribavir and letermovir, resistance remains a concern. This study describes CMV resistance rates at a large academic medical center.
Methods. Patients tested for CMV antiviral resistance from January 2022 through March 2025 were included, while tests with insufficient sample volume were excluded. The study's primary outcome was the prevalence of CMV resistance to various antivirals in patients who had testing completed. Secondary outcomes included the most encountered resistance mutations. All data was analyzed using descriptive statistics. This study was deemed exempt by the Mass General Brigham Institutional Review Board (protocol 2025P001067).
Results. From January 2022 through March 2025, CMV resistance testing was performed 123 times in 76 unique patients. Insufficient serum quantities were collected in 47 tests ordered, so the analysis included 76 tests from 53 unique patients. Resistance was detected overall in 29/76 tests (38.2%). UL97 phosphotransferase mutations conferring resistance to ganciclovir were identified in 23/66 tests (34.8%), most commonly A594V and C603W. UL97 phosphotransferase mutations conferring resistance to maribavir were identified in 9/61 tests (14.8%), most commonly H411Y and T409M. UL54 DNA polymerase mutations potentially conferring resistance to ganciclovir, foscarnet, and cidofovir were identified in 2/69 tests (2.9%). UL56 terminase mutations conferring resistance to letermovir were identified in 2/9 tests (22.2%). When UL97 mutations conferring resistance to ganciclovir were identified, maribavir resistance was identified in 5 of 21 tests (23.8%), and all nine tests that resulted in maribavir resistance occurred in patients previously exposed to maribavir.
Conclusion. In our cohort, CMV resistance to ganciclovir was most frequently observed. Though less common, resistance to novel agents like maribavir and letermovir was identified in patients previously exposed to these antivirals. The development of resistance is correlated with increased morbidity and mortality, so timely resistance testing is paramount to help guide management and improve outcomes. foot_0 . The efficacy of RSV antivirals used for post-exposure prophylaxis (PEP) is unknown. 10 2 on Day (D)0. RSV RT-PCR was performed on nasal washes collected twice daily on D2-12. If PCR-confirmed RSV infection had not occurred by D5am after RSV exposure, participants were randomized to receive daily oral EDP-323 high dose (600mg), low dose (200mg with 600mg loading dose) or PBO for 5D. PEP efficacy was evaluated in this pre-specified population using Fisher's Exact test (two-tailed) of RSV-uninfected vs infected (pre-defined as PCR-positive on 2 consecutive specimens). |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671126&blobtype=pdf | # A laboratory-adapted and a clinical isolate of dengue virus serotype 4 differently impact Aedes aegypti life-history traits relevant to vectorial capacity
Mariana Maraschin, Diego Novak, Valdorion Klein Junior, Lucilene Granella, Luiza Hubner, Athina Medeiros, Tiago Gräf, Guilherme Toledo-Silva, Daniel Mansur, José Oliveira, Peer Reviewer, Wei Huang
## Abstract
Dengue virus cases are on the rise globally, and strategies to control its primary vector, the mosquito Aedes aegypti, represent a promising approach. However, the interaction between virus serotypes and genotypes with Ae. aegypti is poorly characterized in terms of life-history traits related to vector capacity and mosquito disease tolerance. Here, we infected Ae. aegypti with two phylogenetically distant strains of dengue virus serotype 4 genotype II: a laboratory-adapted strain, DENV4-TVP/360, and a recent clinical strain isolated from southern Brazil, DENV4-LRV 13/422. These strains exhibit 26 amino acid differences in their sequences. We assessed various life-history traits of Ae. aegypti, including mortality, fecundity, fertility, and induced flight capacity, as well as vector competence-related parameters, such as infection intensity and prevalence, following exposure to different viral concentrations. We found that while neither strain significantly reduced mosquito lifespan, infection prevalence was highly influenced by the initial dose of DENV4-TVP/360. In contrast, the infectious prevalence of DENV4-LRV 13/422 was smaller (~40% at 14 days post-infection [DPI]), regardless of the initial virus titers. The DENV4-TVP/360 strain also enhanced mosquito-induced flight capacity at early (1 day post-infection) and late (21 DPI) time points, when dispersion is critical for vector competence. On the other hand, DENV4-LRV13/422 reduced Ae. aegypti fertility. A better understanding of how arbovirus strains influence mosquito life-history traits connected to disease spread is critical in public health efforts to mitigate arbovirus outbreaks that are focused on the mosquito vector. IMPORTANCE As dengue virus (DENV) cases rise globally, vector control strategies targeting Aedes aegypti remain essential to public health. The effectiveness of these interventions depends on understanding how viral strains interact with mosquito biology. We show that two phylogenetically distinct DENV serotype 4 strains-one laboratory-adapted, one recently isolated-differentially affect Ae. aegypti traits, such as fertility, flight capacity, and infection dynamics. These traits are linked to mosquito fitness and vector competence, and our results show that sequence variation can shape mosquito-virus interactions. The lab strain enhanced flight capacity at key time points, potentially aiding transmission, while the clinical isolate reduced fertility independently of viral dose. Infection prevalence and sensitivity to viral dose also differed between strains. These findings highlight how viral genotype influences mosquito performance and transmission potential. Incorporating viral genetic diversity into arbovirus studies can improve disease spread models and inform mosquito-based control strategies.
T he ability of hosts to sustain infection without significant impact on their health and fitness is called disease tolerance, a key feature of vector competence and pathogen transmission by mosquitoes (1). Aedes aegypti is predominantly tolerant to the arboviruses they transmit, such as dengue and Zika (2,3). Several life-history traits of mosquitoes, such as survival, reproductive output, and dispersal, influence arbovirus transmission (4) and can be used to quantify mosquito disease tolerance (5). Well-defined mosquito health metrics are critical for a better understanding of the molecular mechanisms that enable Ae. aegypti to sustain a chronic arbovirus infection during its life cycle. Identifying and characterizing such mechanisms and how their manipulation changes transmission is fundamental to developing strategies to block vector competence and arbovirus epidemics (6,7).
We evaluated the impact of two dengue virus serotype 4 genotype II strains (DENV4) on different life-history traits of Ae. aegypti. The strains are a laboratory-adap ted reference strain DENV4-TVP/360 (GenBank: KU513442), originally isolated in the Dominican Republic in 1981 (8,9) and a clinical isolate from a patient presenting high serum viral load in southern Brazil, DENV4-LRV 13/422 (GenBank: KU513441) (10,11). They were chosen because they exhibit different in vitro infectivity rates and immuno modulatory capacities in vertebrate cells (11). Specifically, the LRV 13/422 strain is able to infect C6/36 Ae. albopictus cells and Huh7.5 hepatocyte-derived human cells more efficiently than the TVP/360 strain. Also, LRV 13/422 is less sensitive to interferon-alpha (IFN-α) inhibition in Huh7.5 cells, and preliminary analyses revealed an extensive genetic divergence between the two strains (11).
Based on differences in in vitro infection efficiency and immune responses in vertebrate cells, we hypothesized that the two strains would differentially affect Ae. aegypti life-history traits under laboratory conditions. Initially, we analyzed in more detail the virus genome of both strains, finding that they were phylogenetically distant with 26 differences in their amino acid sequences. Then, we measured Ae. aegypti life-history traits after infection with the DENV strains, focusing on parameters that are directly linked to mosquito fitness, such as survival, fecundity, and induced-flight activity (INFLATE), a proxy for dispersion (12). We found strain-specific phenotypes in the infection prevalence, fertility, and INFLATE at certain time points following infec tion. To modulate mosquito disease tolerance, we need to improve our ability to quantify mosquito fitness and homeostatic status during infection and to understand the molecular basis of strain-specific effects on vector life-history traits (13)(14)(15).
## MATERIALS AND METHODS
## Mosquitoes
Ae. aegypti (Red Eye strain) were reared and maintained under standard conditions as described previously (16,17) at the Federal University of Santa Catarina, Brazil, with a 12 h light/dark cycle at 28°C and 70%-80% relative humidity. Larvae were maintained in filtered, dechlorinated tap water at a density of approximately 100 larvae per liter. They were fed Purina powdered dog chow, with a total of 500 mg provided in 2-3 additions over the course of the larval stage, without changing the water. Adults were provided a 10% sucrose solution ad libitum while housed in 7.3 L plastic cages (21 cm diameter × 25 cm height) with ~200-300 mosquitoes per cage. Infections were performed in females aged between 3 and 7 days post-adult eclosion, which were used in all assays and maintained under the same light-dark cycle at 28°C and 80% relative humidity. Additionally, for egg production to maintain the insectarium, females were fed artificially using a water-jacketed glass artificial feeder and a parafilm membrane containing peripheral human blood (collected into 5 mL tubes containing 7 mg of K 2 -ethylenediaminetetraacetic acid (EDTA) and 1 mM adenosine triphosphate (ATP) (pH 7) as a phagostimulant. Informed consent was obtained from all blood donors. This protocol was approved by the Federal University of Santa Catarina (UFSC)-CAAE: 89894417.8.0000.0121.
## Dengue virus serotype 4 stock preparations
We used dengue virus 4 strain TVP/360-GenBank: KU513442, hereafter called DENV4/ TVP, and dengue virus 4 strain LRV13/422-GenBank: KU513441 at P5 with a known history passage, hereafter called DENV4/LRV (11). Virus strains were kindly provided by Dr. Claudia Nunes Duarte dos Santos from Instituto Carlos Cagas-Fiocruz Paraná, Brazil. Ae. albopictus mosquito cells of the C6/36 lineage (ATCC, CRL-1660) were maintained and propagated in 1× L-15 medium with a pH of 7.6, supplemented with 5% SFB, 1% P/S, and 0.26% tryptone, at a temperature of 28°C in a Biochemical Oxygen Demand (B.O.D.) incubator. The cells were seeded at 80% confluency (3 × 10 7 cells) in a 150 cm 2 bottle, which was kept at 28°C overnight. The following day, the cells were infected with a multiplicity of infection of 0.01 for 5 days to produce DENV4/TVP or DENV4/LRV. The cell supernatant, containing viral particles, was collected, centrifuged at 460 × g for 10 min at 4°C, aliquoted, and stored at -80°C. The plaque assay was performed to determine the viral titer.
## Phylogenetic analysis
A comparative phylogenetic analysis between LRV 13/422 and TVP/360 strains was performed by using augur/auspice as available in Nextstrain (18). Initially, all DENV4 genomes with more than 70% coverage compared to the reference strain (NC_002640) were retrieved from NCBI Virus. Sequences were then aligned with MAFFT (19) and visually inspected in Aliview (20). A maximum likelihood phylogenetic tree was constructed with IQ-TREE (21) and the best substitution model was inferred by the model testing function. Branch support was calculated with SH-aLRT in 1,000 pseudoreplicates. The phylogenetic tree was then visualized in FigTree (https://github.com/rambaut/fig-tree), revealing that both LRV 13/422 and TVP360 strains and the DENV4 reference genome belonged to genotype 2. We then reconstructed the ancestral sequences of this genotype and translated the mutations using augur. The sequence AF326573 was used as the root sequence since NC_002640 was derived from an engineered strain (AF326825), and AF326573 was the natural isolation from which AF326825 originated (9). Auspice was used to visualize and stylize the tree and to extract the LRV 13/422 and TVP/360 mutation path from the root sequence.
## Ae. aegypti infection with dengue virus serotype 4
For mosquito infection experiments, human blood was collected from healthy donors (UFSC-CAAE: 89894417.8.0000.0121) using tubes containing EDTA reagent. To isolate red blood cells (RBCs), 1 mL of blood was centrifuged at 6,400 rpm for 4 min at room temperature. Following centrifugation, the serum was discarded, and the cells were gently washed twice with sterile 1× phosphate-buffered saline solution (GIBCO). Ae. aegypti females were fasted for 18-24 h before a blood meal. During this fasting period, mosquitoes had access to water (but not sucrose solution) before being offered a meal containing a 1:1 mixture of RBC cells and L-15 medium with either DENV4 TVP/360 virus or DENV4 LRV13/422. The control group (Mock) received a 1:1 mixture of RBC and C6/36 cell supernatant. All groups included ATP, pH 7.4, at a final concentration of 1 mM as a phagostimulant. These solutions were presented to the females using artificial glass feeders heated by water at 37°C for approximately 1 h. Subsequently, mosquitoes were anesthetized using cold (-20°C), and only fully engorged females were separated into small cages and kept in a B.O.D. incubator at 28°C until the experiment concluded. Each cage housed 20 mosquitoes and was supplied with a 10% sucrose solution soaked in cotton, offered ad libitum.
## Plaque assays
Vero cells (ATCC, CCL-81) were used to quantify DENV4 TVP/360 and Vero E6 cells (ATCC, CRL-1586) were used to quantify DENV4 LRV 13/422 based on plaque optimization for each viral strain. Cells were seeded at a density of 5 × 10 4 cells/well and 1 × 10 5 cells/ well, respectively, in a 24-well plate in Dulbecco's Modified Eagle Medium (DMEM) F-12 supplemented with 5% FBS, 1% penicillin/streptomycin, 1% glutamine (1× complete DMEM) maintained at 37°C and 5% CO 2 . The following day, the mosquito samples were thawed and underwent a decontamination process, which involved immersing each mosquito in 1 × 45″ in 70% alcohol, followed by 1 × 45″ in 1% hypochlorite, another round of 1 × 45″ in 70% alcohol, and finally 1 × 45″ in 0.9% sterile saline. Subsequently, each mosquito was transferred to a 1.5 mL tube containing 200 µL of complete DMEM F-12 medium (as defined above) and kept on ice. The mosquitoes were then macerated individually using a manual vortex with a sterile pestle dedicated to each sample. The homogenate was then centrifuged at 3,200 × g for 5 min at 4°C. Each mosquito homogenate was diluted (ranging from 10 -1 to 10 -5 ) in DMEM F-12-1× complete medium, added to Vero cells, and incubated in 200 µL of each dilution for 60 min at 37°C and 5% CO 2 . The same procedure was performed for the control group (Mock) using 200 µL of complete DMEM F-12 medium. After incubation, the medium containing the viral dilutions was removed, and 700 µL of semi-solid medium containing DMEM 2× supplemented with 1% fetal bovine serum, 1% P/S, and 1.6% carboxymethyl cellulose was added. The plates were then incubated for 5 days at 37°C and 5% CO 2 . Following incubation, the cells were fixed in paraformaldehyde 3% for 20 min, stained in 1% crystal violet, and counted.
## Survival curves
All infections were performed in female mosquitoes between 4 and 6 days following adult emergence. Infected females were cold-anesthetized immediately after feeding and transferred to cardboard cups with a density of 20 fully engorged females per cup (maximum capacity of 470 mL-10 cm height × 9 cm diameter). Ad libitum access to a 10% sucrose solution was provided through cotton pads placed on top of a woven mesh, which were replaced 2-3 times weekly. Survival rates were monitored daily, six times a week, until all insects within the cups died. The survival cages were maintained in the insectary at a temperature of 28°C (±10%) and humidity of 80% (±10%). Survival data presented represent pooled results from a minimum of two independent experiments.
## Ae. aegypti fecundity and fertility
Female mosquitoes were artificially fed with whole blood, mock, or blood supplemented with DENV4-TPV/360 or DENV4-LRV 13/422 as detailed in section "Ae. aegypti infection with dengue virus serotype 4. " Fully engorged females were cold-anesthetized and individually housed in cages containing a dark plastic cup with water and filter paper to allow egg deposition. Fecundity was scored 5 days following the meal by counting the number of eggs deposited per female. Collected eggs were allowed to rest in insectary conditions for 7 days, when they were submerged in water containing filtered and dechlorinated tap water plus powdered dog chow to allow larval development. Fertility was assessed 3 days post-eclosion by counting the percentage of eclosed larvae.
## Ae. aegypti induced-flight activity
The protocol was adapted from reference 12. A rectangular plastic cage measuring 19 cm × 20 cm was divided into four equal quadrants, in addition to the base. Each quadrant was assigned a value ranging from zero, the lowest (base), to four, the highest. Five mosquitoes were quickly anesthetized at a temperature of -20°C and introduced into the cage, where they remained for 10 min, acclimating to the experimental conditions in the insectary temperature and humidity. To begin the INFLATE test, a mechanical stimulus was applied to the cage, consisting of lifting it about 20 cm above the surface and gently tapping it, causing the mosquitoes to detach and fall to the base of the cage. The cage was kept stable for 1 min, during which time the mosquitoes initiated flight activity. The number of mosquitoes that landed in each quadrant of the cage was visually recorded. Mosquitoes still flying after 1 min were assigned the highest value (four). Mosquitoes landing on the line between quadrants received the value of the upper quadrant. Then, a 2 min rest period was observed, and the process was repeated. Each repetition generated a value called the "inflate value, " calculated by multiplying the value of each quadrant by the number of mosquitoes landing in it. These values were then summed and divided by the total number of mosquitoes in the cage (five). This process was repeated 10 times, corresponding to a technical replicate (n = 10). The average of the values from the 10 technical replicates is referred to as the "Inflate Index. " The protocol was conducted with infected and non-infected mosquitoes (control group, mock). Ultimately, a biological replicate was obtained (n = 1). In Fig. 4, each dot represents one biological replicate (consisting of five mosquitoes assayed 10 times).
## Statistical analysis
Infection intensity, day of death, eggs per female, percentage of eclosion, and INFLATE index were tested for normality. For data that did not follow a normal distribution, non-parametric tests, such as the Kruskal-Wallis test, were performed, as indicated in the figure legends, for side-by-side comparisons or Dunn's multiple comparisons test. Infection prevalence was analyzed using a contingency table (Yes-infected vs No-uninfected) and Fisher's exact test. Survival curves were analyzed with a Log-rank (Mantel-Cox) test. Statistical analysis was conducted using GraphPad Prism.
## RESULTS
## Sequence differences between DENV4-TVP/360 and DENV4-LRV 13/422 sequences
We compared the sequences of the laboratory-adapted (DENV4-TVP/360) and the clinical isolate (DENV4-LRV 13/422) strains of DENV4 with a reference strain, DENV4/814669, obtained from the Dominican Republic in 1981 (8,9). Both strains were phylogeneti cally distant (Fig. S2) with the clinical isolate (LRV 13/422) exhibiting 24 unique amino acid (AA) substitutions compared to the reference sequence DENV4/814669, while the laboratory-adapted (TVP/360) strain exhibited two unique AA substitutions, compared to DENV4/814669 (Fig. S1; Table S1). Three mutations were shared by both strains: S2H, in the precursor membrane glycoprotein (prM), K14Q, in the non-structural protein 3 (NS3), and R23K in the non-structural protein 5 (NS5). The two exclusive amino acid substitutions of the DENV4-TVP/360 strain were located in the non-structural protein NS4B, being phylogenetically closer to the reference strain, DENV4/814669, than the virus isolated from southern Brazil in 2013 (DENV4-LRV 13/422). Most of the non-synony mous mutations of the Brazilian strain (LRV 13/422) were present in the envelope protein (E), with six AA substitutions; non-structural protein 1 (NS1) with five AA substitutions; non-structural protein 2A (NS2A), with five AA substitutions, and the non-structural protein 5 (NS5), with six AA substitutions. Interestingly, these proteins are critical for flavivirus infection, replication, and dissemination into mosquito tissues (22)(23)(24)(25)(26)(27).
## Infection dynamics of DENV4 strains differ in Ae. aegypti
To explore how the laboratory-adapted (DENV4-TVP/360) and the clinical isolate (DENV4-LRV 13/422) interact with Ae. aegypti over time, we infected mosquitoes with two doses of each strain and measured virus titers in single whole-body mosquitoes at 0, 7, 14, and 21 days following the infectious blood meal (Fig. 1A andB). Since the initial infectious dose strongly influences DENV infection in Ae. aegypti (28), we used the maximum available dose for each strain and a 1/10 dilution. The infectious particle virus concentration (input in Fig. 1C) is shown in red and expressed as plaque-forming units (PFU) per microliter of blood offered to mosquitoes (PFU/μL of blood). Both strains were able to infect and replicate in Ae. aegypti (Fig. 1B). Using the same infectious dose when comparing infection dynamics between groups is essential, and our experimental design accounted for this. There is no statistical difference in the viral titers in the blood meals provided to mosquitoes between the DENV4 TVP "low dose" and DENV4 LRV "high dose" groups, as shown by the analysis of independent biological replicates (Fig. S3A). Consistently, there is also no statistical difference in the amount of infectious virus ingested by Ae. aegypti immediately after feeding (Fig. S3B).
We measured two readouts: infection prevalence, the percentage of infected mosquitoes, and infection intensity, the amount of infectious particles per mosquito. Overall, infection intensity showed no dose-dependent increase with input dose in all the experimental conditions. For the DENV4 TVP strain, mosquitoes ingested either 2. PFUs (±2.5 × 10² S.E.M., n = 15) at the lowest dose. Here too, despite a 6.6-fold increase in virus input, infection intensity at 14 DPI did not differ (P > 0.999, K-S test). Taken together, input doses ranging across ~4 × 10⁵ PFUs (a 407,781-fold difference between the TVP high dose and the LRV low dose) had little effect on virus titer per mosquito at 14 DPI. The only significant difference was detected at 14 DPI when comparing the highest doses of TVP and LRV. However, this range of input doses differentially influenced infection prevalence depending on the virus strain, with the laboratory-adapted strain (DENV4-TVP/360) being highly influenced by virus input, as expected based on several mosquito-DENV studies (28)(29)(30). However, the clinical isolate (DENV4-LRV 13/422) exhibited a relatively low infection prevalence (~40%), irrespective of the input dose (Fig. 1C). This result suggests that factors determining infection prevalence, also known as the midgut infection barrier, might be more relevant to vector competence than immune resistance factors that decrease virus replication inside mosquito tissues once the infection is already established (31)(32)(33).
## The impact of different DENV4 strains on life-history traits of Ae. aegypti
We challenged mosquitoes with two doses of DENV4/TVP and DENV4/LRV and scored mosquito survival during the full life span of the population. Consistent with our previous results ( 5), the median time to death of Ae. aegypti did not differ between uninfected (mock) and infected mosquitoes, regardless of input viral doses and virus strain (Fig. 2).
The impact of virus infection on parameters directly connected to vector fitness was evaluated following mosquito challenge with both DENV4 strains. Feeding Ae. aegypti with blood supplemented with different concentrations of infectious particles did not alter the total number of eggs laid by each fully engorged female (Fig. 3A). In Fig. 3, "blood" represents whole human blood (RBCs + plasma), which has a higher total protein concentration than mock, as evidenced by a greater fecundity of Ae. aegypti (blood ~70 eggs/female vs mock ~35 eggs/female). Challenging mosquitoes with the clinical strain of DENV4 (LRV) resulted in a significant reduction in the percentage of eclosion (~50%), irrespective of the dose tested (Fig. 3B). This result suggests a lower adaptation of the LRV strain, which is consistent with its recent interaction with our colony mosquitoes (Red Eyes strain) instead of field mosquitoes circulating in southern Brazil (34). On the other hand, we did not observe statistically significant differences in fertility with the laboratory-adapted TVP strain, consistent with the higher adaptability of this virus under laboratory conditions. Interestingly, there is no difference in fertility between blood-fed and mock-fed mosquitoes, averaging around 70% eclosion, suggesting that females can optimize fecundity (egg production) according to the nutritional status of the meal to maximize fertility, similar to what has been described for desiccation stress in Ae. aegypti (35) .
Ae. aegypti dispersal involves its flight activity and is directly connected to vectorial capacity and arbovirus epidemics (36). We took advantage of a recently described method to assess the induced flight activity (INFLATE) (12) and challenged Ae. aegypti females with the highest available doses of both strains of DENV4. The laboratory-adap ted TVP/360 strain significantly enhanced the mosquito flight activity 24 h after feeding when compared to "mock" and DENV4-LRV 13/422 (Fig. 4B), a time point where infection of the midgut epithelium is taking place. At 21 days post-feeding, an epidemiologically relevant time point, where DENV4 has already infected the salivary glands and the mosquito is infectious (37,38), the INFLATE index was also significantly higher, specifically in the laboratory-adapted TVP strain (Fig. 4E). The clinical isolate did not induce alterations in the flight activity during the mosquito lifespan, compared to mock-infected Ae. aegypti (Fig. 4B through E). We observed an overall decline in the INFLATE index 1 day post-feeding (Fig. 4B) as reported by reference 12, compared to all the other time points. This reduction was independent of the DENV4 challenge. After the completion of blood digestion, at days 7, 14, and 21 in our assays, mosquitoes were leaner and lighter, which translated into higher INFLATE values compared to 1 DPI (Fig. 4C through E). At 21 days post-feeding, Ae. aegypti is already experiencing senescence, causing a reduction in the overall INFLATE index compared to 7 and 14 DPI (Fig. 4E).
## DISCUSSION
Mosquitoes can harbor and transmit microbes, such as dengue virus and Plasmodium spp., without exhibiting overt signs of pathology, a phenomenon often described as disease tolerance. While this trait has been relatively well characterized phenotypically, its underlying molecular mechanisms remain poorly understood (1,2,39,40). In our previous work, we explored the impact of different arboviruses on mosquito mortality to establish a dose-response framework to study Ae. aegypti's response to infections (5). A quantitative analysis of disease tolerance can be achieved by scoring how parameters relevant to host physiology vary during a gradient of infection (14,15,41). In this manuscript, we investigated the effect of two distinct DENV4 genotype II strains on mosquito life-history traits useful to describe organismal fitness and disease tolerance. We found that both DENV4 strains differentially impacted infection prevalence, female fertility, and induced flight capacity. Defining the molecular basis that underlines such phenotypes will be critical in efforts to mitigate mosquito-borne diseases based on vector control.
Dengue virus serotype 4 is one of the least studied dengue viruses. Although its presence in Brazil has been relatively limited (42)(43)(44)(45), studies on the dynamics of different serotype circulation and emergence (46), as well as the occurrence of antibodydependent enhancement of disease in humans (47)(48)(49), highlight the importance of examining the interaction between DENV4 and its main vector, the mosquito Ae. aegypti. This is particularly relevant since DENV4 has been shown to circulate undetected in mosquito populations without reported human cases (50,51), be vertically transmitted by mosquitoes (52), and displace DENV1, a common DENV serotype in Brazil (53) when co-infecting Ae. aegypti (54).
Ae. aegypti dispersal is a critical parameter to arbovirus infection in humans (55,56). We found that DENV4-TVP/360, a laboratory-adapted strain, enhanced Ae. aegypti induced-flight capacity (INFLATE), a proxy for mosquito dispersal, specifically at an early (1 DPI) and late (21 DPI) time points (Fig. 4B through E). This modulation was specific to the TVP strain since the recent clinical isolate (LRV) did not show differences in the INFLATE index compared to mock-fed mosquitoes (Fig. 4B through E). It was previously shown that DENV infection enhances mosquito behaviors linked to flight (57)(58)(59). Recently, immune activation by zymosan was shown to reduce Ae. aegypti's induced flight capacity (60). The molecular and physiological basis of immune-induced flight phenotypes in Ae. aegypti, as well as its epidemiological implications, remains to be defined (61). The contribution of different defense strategies, such as antiviral resistance or disease tolerance, during mosquito immune response to DENV is poorly defined (1). While DENV load increases 100-1,000-fold in susceptible mosquito strains during its spread from the midgut to the salivary glands (37), Ae. aegypti does not experience significant fitness costs associated with chronic infection (5), although this may vary depending on different DENV serotypes, genotypes, and mosquito populations (62). The RNAi machinery is considered the main antiviral pathway in Ae. aegypti (63), but the knockout or overexpression of its core gene, Dicer-2, revealed conflicting results regarding its actual contribution to mosquito immune resistance to arbovirus infection and vector competence (64)(65)(66). In Fig. 1C, we show that despite a marked difference in prevalence at 14 DPI between the LRV 13/422 and TVP/360 strain, depending on the initial DENV4 input dose, infection intensity (Fig. 1C-middle panel) exhibited a modest dose-depend ency. This result suggests that immune resistance plays a minor role in the DENV4 replication cycle within the mosquito at the whole-body level, and unknown factors controlling the midgut infection barrier might be more important to mosquito infection and, therefore, disease transmission.
In summary, we described that two different strains of DENV4 genotype II differentially impacted Ae. aegypti infection dynamics, fertility, and flight capacity. The determi nation of how arbovirus infection affects life-history traits of insect vectors is critical for the development of strategies to fight mosquito-borne diseases (67). Some of the limitations of our study include the fact that we have demonstrated viral strain-induced changes in parameters associated with Ae. aegypti vector competence and physiology under laboratory conditions, and not in field settings, with the influence of real-world ecological parameters. The DENV4 TVP/360 strain is laboratory-adapted and reaches high titers in cell culture. It remains possible that the observed infection prevalence (Fig. 1C) and induced-flight activity (Fig. 4) result from artificial adaptation in cell culture. Also, feeding mosquitoes with decreasing doses of arboviruses leads to reduced percentages of infections (Fig. 1C). The observed differences should be interpreted with caution and understood as group or population effects.
## Funder Grant(s)
Author(s)
## References
1. Oliveira, Bahia, Vale (2020) "How are arbovirus vectors able to tolerate infection?" *Dev Comp Immunol*
2. Lambrechts, Saleh (2019) "Manipulating mosquito tolerance for arbovirus control" *Cell Host Microbe*
3. Magistrado, El-Dougdoug, Short (2023) "Sugar restriction and blood ingestion shape divergent immune defense trajectories in the mosquito Aedes aegypti" *Sci Rep*
4. Cator, Johnson, Mordecai et al. (2020) "The role of vector trait variation in vector-borne disease dynamics" *Front Ecol Evol*
5. Maraschin, Talyuli, Rulff Da Costa et al. (2023) "Exploring dose-response relationships in Aedes aegypti survival upon bacteria and arbovirus infection" *J Insect Physiol*
6. Chandrasegaran, Lahondère, Escobar et al. (2020) "Linking mosquito ecology, traits, behavior, and disease transmission" *Trends Parasitol*
7. Wu, Sánchez, Hm et al. (2020) "Vector bionomics and vectorial capacity as emergent properties of mosquito behaviors and ecology" *PLoS Comput Biol*
8. Mackow, Makino, Zhao et al. (1987) "The nucleotide sequence of dengue type 4 virus: analysis of genes coding for nonstructural proteins" *Virology (Auckl)*
9. Durbin, Karron, Sun et al. (2001) "Attenuation and immunogenicity in humans of a live dengue virus type-4 vaccine candidate with a 30 nucleotide deletion in its 3'untranslated region" *Am J Trop Med Hyg*
10. Sarathy, Infante, Li et al. (2015) "Characteri zation of lethal dengue virus type 4 (DENV-4) TVP-376 infection in mice lacking both IFN-α/β and IFN-γ receptors (AG129) and comparison with the DENV-2 AG129 mouse model" *J Gen Virol*
11. Kuczera, Bavia, Mosimann et al. (2016) "Isolation of dengue virus serotype 4 genotype II from a patient with high viral load and a mixed Th1/Th17 inflammatory cytokine profile in South Brazil" *Virol J*
12. Gaviraghi, Oliveira (2020) "A simple and reliable method for longitudinal assessment of untethered mosquito induced flight activity" *J Insect Physiol*
13. Shirasu-Hiza, Schneider (2007) "Confronting physiology: how do infected flies die?" *Cell Microbiol*
14. Torres, Oliveira, Tate et al. (2016) "Tracking resilience to infections by mapping disease space" *PLoS Biol*
15. Louie, Song, Hotson et al. (2016) "How many parameters does it take to describe disease tolerance?" *PLoS Biol*
16. Oliveira, Gonçalves, Lara et al. (2011) "Blood meal-derived heme decreases ROS levels in the midgut of Aedes aegypti and allows proliferation of intestinal microbiota" *PLoS Pathog*
17. Oliveira, Talyuli, Goncalves et al. (2017) "Catalase protects Aedes aegypti from oxidative stress and increases midgut infection prevalence of Dengue but not Zika" *PLoS Negl Trop Dis*
18. Hadfield, Megill, Bell et al. (2018) "Nextstrain: real-time tracking of pathogen evolution" *Bioinformatics*
19. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
20. Larsson (2014) "AliView: a fast and lightweight alignment viewer and editor for large datasets" *Bioinformatics*
22. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol Biol Evol*
23. Erb, Butrapet, Moss et al. (2010) "Domain-III FG loop of the dengue virus type 2 envelope protein is important for infection of mammalian cells and Aedes aegypti mosquitoes" *Virology (Auckl)*
24. Liu, Liu, Nie et al. (2016) "Flavivirus NS1 protein in infected host sera enhances viral acquisition by mosquitoes" *Nat Microbiol*
25. Liu, Liu, Du et al. (2017) "Evolutionary enhancement of Zika virus infectivity in Aedes aegypti mosquitoes" *Nature New Biol*
26. Saraiva, Fang, Kang et al. (2018) "Aminopeptidase secreted by Chromobacterium sp. Panama inhibits dengue virus infection by degrading the E protein" *PLoS Negl Trop Dis*
27. Cheng, Tan, Chan et al. (2021) "In vitro and in vivo stability of P884T, a mutation that relocalizes dengue virus 2 non-structural protein 5" *ACS Infect Dis*
28. Besson, Lezcano, Overheul et al. (2022) "Arbovirus-vector protein interactomics identifies Loquacious as a co-factor for dengue virus replication in Aedes mosquitoes" *PLoS Pathog*
29. Novelo, Hall, Pak et al. (2019) "Intrahost growth kinetics of dengue virus in the mosquito Aedes aegypti" *PLoS Pathog*
30. Nguyet, Duong, Trung et al. (2013) "Host and viral features of human dengue cases shape the population of infected and infectious Aedes aegypti mosquitoes" *Proc Natl Acad Sci U S A*
31. Duong, Lambrechts, Paul et al. (2015) "Asymptomatic humans transmit dengue virus to mosquitoes" *Proc Natl Acad Sci U S A*
32. Bennett, Olson, Muñoz et al. (2002) "Variation in vector competence for dengue 2 virus among 24 collections of Aedes aegypti from Mexico and the United States" *Am J Trop Med Hyg*
33. Souza-Neto, Powell, Bonizzoni (2019) "Aedes aegypti vector competence studies: a review" *Infect Genet Evol*
34. Hodoameda, Ebel, Mukhopadhyay et al. (2024) "Extreme infectious titer variability in individual Aedes aegypti mosquitoes infected with Sindbis virus is associated with both differences in virus population structure and dramatic disparities in specific infectivity" *PLoS Pathog*
35. De Oliveira, Netto, Francisco et al. (2023) "Aedes aegypti in Southern Brazil: spatiotemporal distribution dynamics and association with climate and environmental factors" *Trop Med Infect Dis*
36. Prasad, Sreedharan, Bakthavachalu et al. (2023) "Eggs of the mosquito Aedes aegypti survive desiccation by rewiring their polyamine and lipid metabolism" *PLoS Biol*
37. Kraemer, Sinka, Duda et al. (2015) "The global distribution of the arbovirus vectors Aedes aegypti and Ae"
38. Salazar, Richardson, Sánchez-Vargas et al. (2007) "Dengue virus type 2: replication and tropisms in orally infected Aedes aegypti mosquitoes" *BMC Microbiol*
39. Mayton, Hernandez, Vitek et al. (2021) "A method for repeated, longitudinal sampling of individual Aedes aegypti for transmission potential of arboviruses" *Insects*
40. Shaw, Tate, Schneider et al. (2018) "Vector immunity and evolutionary ecology: the harmonious dissonance" *Trends Immunol*
41. Shaw, Marcenac, Catteruccia (2022) "Plasmodium development in Anopheles: a tale of shared resources" *Trends Parasitol*
42. Gupta, Vale (2017) "Nonlinear disease tolerance curves reveal distinct components of host responses to viral infection" *R Soc Open Sci*
43. Nunes, Faria, Vasconcelos et al. (2012) "Phylogeography of dengue virus serotype 4, Brazil, 2010-2011"
44. Fang, Wu, Wu et al. (2013) "Recent advances in DENV receptors" *Sci World J*
45. Ortiz-Baez, Cunha, Vedovello et al. (2019) "Origin, tempo, and mode of the spread of DENV-4 Genotype IIB across the state of São Paulo, Brazil during the 2012-2013 outbreak" *Mem Inst Oswaldo Cruz*
46. Bezerra, Sousa Sc De, Tauil et al. (2021) "Entry of dengue virus serotypes and their geographic distribution in Brazilian federative units: a systematic review" *Rev Bras Epidemiol*
47. Salles, Da, Sá-Guimarães et al. (2018) "History, epidemiology and diagnostics of dengue in the American and Brazilian contexts: a review" *Parasit Vectors*
48. Dejnirattisai, Jumnainsong, Onsirisakul et al. (2010) "Crossreacting antibodies enhance dengue virus infection in humans" *Science*
49. Beltramello, Williams, Simmons et al. (2025) *Research Article Microbiology Spectrum*
50. Sallusto (2010) "The human immune response to Dengue virus is dominated by highly cross-reactive antibodies endowed with neutralizing and enhancing activity" *Cell Host Microbe*
51. Katzelnick, Gresh, Halloran et al. (2017) "Antibody-dependent enhancement of severe dengue disease in humans" *Science*
52. Boyles, Mavian, Finol et al. (2020) "Under-the-radar dengue virus infections in natural populations of Aedes aegypti mosquitoes"
53. Ayers, Xie, Coatsworth et al. (2021) "Infection kinetics and transmissibility of a reanimated dengue virus serotype 4 identified originally in wild Aedes aegypti from Florida" *Front Microbiol*
54. Torres-Avendaño, Apodaca-Medina, Castillo-Ureta et al. (2021) "Natural vertical transmission of dengue virus serotype 4 in Aedes aegypti larvae from urban areas in Sinaloa" *Mexico. Vector Borne Zoonotic Dis*
55. De Bruycker-Nogueira, Souza, Chouin-Carneiro et al. (2016) "DENV-1 Genotype V in Brazil: spatiotemporal dispersion pattern reveals continuous co-circulation of distinct lineages until" *Dos Santos FB*
56. Vazeille, Gaborit, Mousson et al. (2016) "Competi tive advantage of a dengue 4 virus when co-infecting the mosquito Aedes aegypti with a dengue 1 virus" *BMC Infect Dis*
57. Sedda, Vilela, Aguiar et al. (2018) "The spatial and temporal scales of local dengue virus transmission in natural settings: a retrospective analysis" *Parasit Vectors*
58. Leandro, De Castro, Garey et al. (2024) "Spatial analysis of dengue transmission in an endemic city in Brazil reveals high spatial structuring on local dengue transmission dynamics" *Sci Rep*
59. Lima-Camara, Bruno, Luz et al. (2011) "Dengue infection increases the locomotor activity of Aedes aegypti females" *PLoS One*
60. Tallon, Lorenzo, Moreira et al. (2020) "Dengue infection modulates locomotion and host seeking in Aedes aegypti" *PLoS Negl Trop Dis*
61. Javed, López-Denman, Paradkar et al. (2024) "Flight traits of dengue-infected Aedes aegypti mosquitoes" *Comput Biol Med*
62. Gaviraghi, Barletta, Silva et al. (2024) "Activation of innate immunity selectively compromises mitochondrial complex I, proline oxidation, and flight activity in the major arbovirus vector Aedes aegypti" *Mol Microbiol*
63. Maire, Lambrechts, Hol (2024) "Arbovirus impact on mosquito behavior: the jury is still out" *Trends Parasitol*
64. Keirsebelik, David, Pavan et al. (2024) "Dengue virus serotype 1 effects on mosquito survival differ among geographically distinct Aedes aegypti populations" *Insects*
65. Olmo, Ferreira, Izidoro-Toledo et al. (2018) "Control of dengue virus in the midgut of Aedes aegypti by ectopic expression of the dsRNA-binding protein Loqs2" *Nat Microbiol*
66. Dong, Dong, Dizaji et al. (2022) "The Aedes aegypti siRNA pathway mediates broad-spectrum defense against human pathogenic viruses and modulates antibacterial and antifungal defenses" *PLoS Biol*
67. Merkling, Crist, Henrion-Lacritick et al. (2023) "Multifaceted contributions of Dicer2 to arbovirus transmission by Aedes aegypti" *Cell Rep*
68. Samuel, Pohlenz, Dong et al. (2023) "RNA interference is essential to modulating the pathogenesis of mosquito-borne viruses in the yellow fever mosquito Aedes aegypti" *Proc Natl Acad Sci U S A*
69. Achee, Grieco, Vatandoost et al. (2019) "Alternative strategies for mosquito-borne arbovirus control" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793596&blobtype=pdf | H Davidson, Lisa Han, Kevin Mcconeghy, Yasin Abul, Ivis Perez, Evan Dickerson, Laurel Holland, Tiffany Wallace, Mandi Winkis, Clare Nugent, M Chb, Olajide Olagunju, Debbie Keresztesy, Elise Didion, David Canaday, Stefan Gravenstein
Background. Timely local surveillance data regarding acute respiratory infections (ARI) is crucial for nursing homes (NHs) to proactively manage outbreaks. Centers for Disease Control and Prevention (CDC) national-level viral surveillance data lags by weeks, hindering timely and informed decisions. Our study evaluated feasibility and utility of a pragmatic approach to improve respiratory virus testing.
## Point of Care Test Results in Nov 2023 -Jan 2025
Methods. We enrolled 23 NHs in OH, MO, IN, RI, VA, and PA. We placed a CLIA-waived Cepheid GeneXpert Xpress point-of-care (POC) device, to enable rapid (< 30 minutes) on-site molecular testing for Respiratory Syncytial Virus (RSV), Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), and influenza A and B from long-and short-stay residents' and staff members' nasal swabs. The device enabled reported symptoms to be entered at the time of testing and automatically transmitted results to a web-based dashboard, providing real-time feedback on respiratory viral activity from Nov 2023 -Jan 2025. Monthly surveys of NH staff gathered qualitative feedback on use of the device, impact on testing protocols, and virus awareness.
Results. Enrolled NHs had a mean bed size of 155 (range 50-427) averaging a monthly census of 125 residents. Of 1745 POC tests, 360 (20.6%) identified a virus: 111 (6.4%) influenza A, 4 (0.2%) influenza B, 81 (4.6%) RSV, and 160 (9.2%) SARS-CoV-2; 4 individuals with coinfection were also identified (Table 1).
Barriers for initial setup included: unfamiliar technology, device placement issues, staff training, network access, need for additional data entry by busy NH staff, resistance to change, and staff turnover. Monthly surveys revealed that the presence of the device impacted testing protocols, led to a greater recognition of respiratory virus infections among staff, and reduced time to diagnose these infections.
Conclusion. We demonstrated the feasibility and utility of implementing CLIA-waived POC multiplex molecular testing for viral ARIs within a network of nursing homes. The high test-positivity (20.6%) demonstrated value for clinical management and an approach that could be adapted for regional or national real-time reporting.
Disclosures |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737532&blobtype=pdf | # Development and Validation of a Commercial TaqMan-Based RT-qPCR Kit for Rotavirus and Norovirus Detection in the Brazilian Acute Diarrhea Surveillance Network
Geison Cambri, Thiago Jacomasso, Fernanda Marcicano Burlandy, Fábio Correia Malta, Alexandre Fialho, Audrey Cilli, Simone Guadagnucci, Monteiro Teixeira, Patrícia Santos Lobo, Hugo Reis Resque, Lucia Berto, Alessandro Afornali, Fabricio Klerynton Marchini, Irina Nastassja Riediger, Luana Soares, Rita De Cássia, Compagnoli Carmona, Tulio Machado Fumian
## Abstract
Acute gastroenteritis (AGE) is a major cause of illness and death in children under five, especially in low-and middle-income countries, and rotavirus A (RVA) and norovirus are the leading viral agents. The present study aimed to describe the development of a commercial multiplex TaqMan-based RT-qPCR assay to detect those viruses to enhance surveillance and public health responses in Brazil. The assay validation involved optimizing primers and probes for multiplex RT-qPCR, assessing analytical sensitivity, and confirming specificity. A multicenter pilot study across Brazil's AGE surveillance network assessed the assay's performance. The IBMP NAT assay demonstrated high specificity and sensitivity for detecting RVA and norovirus GI and GII. No cross-reactivity was observed. LoD95 values were low: 18.6 (GI), 71.2 (GII), and 12.3 (RVA) copies/reaction. In 379 clinical samples, diagnostic sensitivity and specificity exceeded 96% for all targets. The assay showed strong reproducibility across operators and instruments. Stability tests confirmed consistent performance under freeze-thaw, transport, and storage conditions. Compared to in-house RT-qPCR, the IBMP NAT test yielded lower Ct values, indicating improved detection of low viral loads. The IBMP NAT Kit significantly advances molecular diagnostics, enabling rapid, sensitive, and reliable detection of RVA and norovirus in fecal specimens. It strengthens public health surveillance and supports timely responses to AGE outbreaks, helping reduce disease burden in vulnerable populations.
## 1. Introduction
Acute gastroenteritis (AGE) is one of the leading causes of morbidity and mortality in children under the age of five, especially in low-and middle-income countries (LMICs), where access to healthcare, hygiene, and education is often limited [1,2]. Studies conducted in LMICs that screened for multiple enteropathogens have identified rotavirus and norovirus as the predominant enteric pathogens responsible for pediatric AGE [3][4][5][6].
Rotavirus alphagastroenteritidis (formerly species A rotavirus; RVA) is the leading cause of AGE in children under 5 years, posing a significant global health threat especially in LMICs [3]. In Brazil, since 2013, more than 5000 severe RVA infection cases have been officially reported to the national RVA surveillance program of the Brazilian Ministry of Health [7]. The outer capsid proteins, VP7 (a capsid glycoprotein) and VP4 (a spike protein), independently elicit neutralizing antibodies and form the basis of the binary classification system for G and P types, respectively [8]. Currently, 42 G and 58 P genotypes are recognized as capable of infecting both humans and animals (https://rega.kuleuven.be/ cev/viralmetagenomics/virus-classification/rcwg, accessed on 18 January 2025); however, some genotypes, such as G1P [8], G2P [4], G3P [8], G4P [8], G9P [8], and G12P [8] are globally significant in clinical settings [9].
Noroviruses belong to the genus Norovirus within the Caliciviridae family. Noroviruses are classified into ten distinct genogroups (GI-GX) and divided into 48 genotypes, however only GI, GII, GIV, GVIII and GIX are known to infect humans [10,11]. Among these, GI and GII noroviruses are responsible for most human infections [12]. Since the introduction of the RVA vaccine in several countries, norovirus has emerged as the predominant cause of AGE, particularly in countries where the RVA vaccine has been publicly available. In those countries, norovirus-related illnesses now account for a substantial proportion of medical visits due to AGE in pediatric populations [13][14][15]. Currently, noroviruses are recognized as the leading cause of AGE across all age groups worldwide, responsible for approximately 20% of AGE episodes and over 200,000 deaths annually in low-and middle-income countries [11,16,17].
Early diagnosis of viral AGE cases plays a pivotal role, especially in preventing the unnecessary and inappropriate use of antimicrobials, as well as to evaluate the ongoing prevention strategies, such as RVA vaccination. Additionally, comprehensive screening and increasing the capacity for RVA and norovirus testing are crucial. A wide range of diagnostic methods have been employed for detecting rotavirus infection in stool samples, including commercial enzyme-linked immunosorbent assay (ELISA) and immunochromatographic tests, particularly in clinical settings. In research laboratories involved in vaccine development and epidemiological studies, molecular assays, both conventional and real-time RT-PCR, are frequently used to complement antigen detection methods. For norovirus, due to the antigenic variability, which poses challenges for high-sensitivity antigen detection, RT-qPCR remains the gold standard for detecting and quantifying the viral RNA in stool samples. In Brazil, laboratory surveillance of AGE cases is initially conducted at Central Public Health Laboratories (LACENs) in each state where RVA infection is detected using ELISA. Reference laboratories perform further diagnostic confirmation using molecular approaches to determine circulating genotypes.
Although several multiplex RT-qPCR assays for RVA and norovirus have been described previously, the IBMP-NAT (Instituto de Biologia Molecular do Paraná-Nucleic Acid Test) Rotavirus and Norovirus Kit represent the first fully validated, ready-to-use registered commercial assay specifically standardized for the Brazilian public health surveillance network. Our assay integrates (i) simultaneous detection of RVA and norovirus GI and GII genogroups with an internal endogenous control, (ii) analytical performance aligned with recommended diagnostic standards, and (iii) validation across three national Rotavirus Reference Laboratories using routine clinical stool samples. This combination ensures high diagnostic sensitivity, reproducibility, and logistical feasibility for large-scale implementation within the Central Public Health Laboratories (LACEN) system.
## 2. Materials and Methods
## 2.1. Ethical Clearance
This study received approval from the Ethics Committees of the Oswaldo Cruz Foundation (FIOCRUZ; in CAAE 76063123.5.0000.5248), the Adolfo Lutz Institute (IAL; in CAAE 91636825.1.0000.0059), and the Evandro Chagas Institute (IEC; in CAAE 92340825.7.0000.0019). The surveillance was conducted through a hierarchical network, in which samples were obtained based on medical requests from hospitals and health centers monitored by the Brazilian Unified Health System (SUS). This study was conducted within the scope of the RRRL/MoH as part of a federal public health policy for viral AGE surveillance in Brazil. Patient-informed consent was waived by the Ethical Committee, and patient data was maintained anonymously and securely.
## 2.2. Stool Samples
This study included 902 stool samples (379 and 523 samples for the analytical performance and inter-laboratory evaluations, respectively) from inpatients and outpatients (children and adults) with AGE symptoms who were treated in sentinel units or any healthcare facility unit, as outlined in Consolidation Ordinances No. 4 and No. 5, 28 September 2017, of Brazilian Ministry of Health. For instance, AGE was defined as ≥three liquid or semi liquid evacuations within a 24 h period. Stool samples were systematically sent along with clinical-epidemiological records to one of the Rotavirus Reference Laboratory that are part of the AGE surveillance network in Brazil. The surveillance network consists of one Rotavirus National Reference Center (Evandro Chagas Institute), two Rotavirus Regional Reference Laboratories (Adolf Lutz Institute and Oswaldo Cruz Institute) and states central laboratories, all overseen by the General Coordination of Public Health Laboratories of the Brazilian Ministry of Health (CGLab/MoH). In addition to RVA surveillance, the reference laboratories also conduct norovirus diagnostic testing for all received samples.
## 2.3. Viral RNA Extraction
Viral nucleic acids were extracted from 140 µL of clarified stool suspension (10% w/v in Tris-calcium buffer, pH 7.2) using the Extracta 32 automated system and the Extracta Fast DNA and RNA Viral extraction kit (Loccus, São Paulo, Brazil). The procedure, based on magnetic bead technology, was performed according to the manufacturer's instructions. Briefly, 140 µL of clarified stool suspension, supplemented with 5 µL of proteinase K (20 mg/mL), was transferred to the extraction plate containing lysis buffer and magnetic silica particles for nucleic acid binding. The system then performed automated lyses, capture and washing steps to remove potential inhibitors, followed by elution of purified RNA in 60 µL of RNase-free elution buffer. Extracted RNA was stored at -80 • C until further analysis.
## 2.4. Development and Design of the Assay
Primers and probes used for detecting RVA and norovirus GI/GII were based on those previously described by Zeng et al. [18], Hill et al. [19] and Kageyama et al. [20], respectively. To combine these primers and probe sets and a set targeting a human gene as internal control into one multiplex reaction, concentrations for each individual set were determined using IBMP's proprietary RT/Taq formulation and a synthetic DNA control molecule, designed to contain the three viral targets (RVA NSP3 region, Norovirus GI and GII ORF1-2 junction region). Optimal primer concentrations were evaluated using EvaGreen intercalating dye, which allowed for the inspection of secondary amplification products that could hinder reaction performance. The selected concentration for each primer pair was the one that offered the best compromise between reaction kinetics (judged by the shape of the amplification curves) and dimer/hairpin formation (based on the melting profiles of PCR products in templated and non-templated reactions). Probe concentrations were selected based on amplification kinetics in a temperature gradient. Primers and probes sequences are not fully disclosed due to commercial confidentiality.
Primers and probes in their optimized concentrations were combined in a single reaction mix. Probes were labeled with dye and quencher pairs that are compatible with most qPCR instruments. The multiplexed reactions were then validated as described below.
## 2.5. Multiplex TaqMan-Based RT-qPCR for RVA and Norovirus Detection
RT-qPCR reactions were performed using 5 µL of the extracted RNA in a final volume of 20 µL, containing 15 µL of a PCR mixture (IBMP MixFit I, IBMP, Curitiba, PR, Brazil), which includes 5 µL of IBMP mastermix 4X, 12 ng of Taq DNA polymerase (IBMP), 75 ng of reverse transcriptase enzyme (IBMP) and primers and probes (IBMP; IDT, Coralville, IA, USA; and Thermo Fisher Scientific, Waltham, MA, USA). Reactions were conducted in the ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA), under the following thermal cycling conditions: 15 min at 50 • C, 10 min at 95 • C, 40 cycles of 15 s at 95 • C, and 1 min at 57 • C. The RT-qPCR analysis parameters were configured in the ABI 7500 software with a baseline range of 3-15 and threshold values set to 5000 for norovirus GI, 10,000 for norovirus GII, 30,000 for RVA, and 5000 for the internal control. The results were considered positive for stool clarified samples that exhibited a characteristic sigmoid curve, and when the Ct values of both the endogenous control and rotavirus or norovirus genes were ≤40. Samples were considered negative when no amplification was detected for norovirus and rotavirus, and the internal control showed amplification with a Ct value ≤ 35. All runs included both positive and negative controls using synthetic molecules. The positive controls exhibited amplification for all targets, whereas the negative controls showed amplification exclusively for the internal control.
## 2.6. Analytical Sensitivity and Specificity of Multiplex TaqMan-Based RT-qPCR Assay
The analytical sensitivity of the assay was determined using six serial five-fold dilutions of rotavirus and norovirus (GI and GII) reference samples from Vircell (catalog number: MBC026-R; Vircell, Santa Fe, Spain) and ATCC (catalog number: VR-3234SD and VR-3235SD; American Type Culture Collection, Manassas, VA, USA), respectively, with known concentrations. A minimum of 63 and maximum of 72 replicates for each dilution was tested and the limit of detection with 95% confidence (LoD95) was determined using a probit model with interpolation of detection rates and analyte concentrations in copies per reaction. The initial concentration for each virus was 40, 180 and 280 copies/reaction for norovirus GI, norovirus GII and RVA, respectively. RNA reference samples were prepared in TE buffer (pH 8.0) containing 5 µg/mL of salmon sperm DNA, 10 µg/mL of poly(A) RNA, with addition of gBlock molecule at a concentration of 10 copies/µL; gBlock was added as it contains the DNA sequence for the internal control of the multiplex reaction.
We also compared the new TaqMan-based commercial RT-qPCR assay using 379 clarified stool samples which had previously tested positive or negative. RVA-and noroviruspositive stool samples (with a wide range of Ct values) and negative samples were selected by the three RRRLs and sent to IBMP to be extracted and tested using the new multiplex protocol. Additionally, positive samples for other enteric organisms were also tested to verify the specificity.
## 2.7. Assay Repeatability, Reproducibility and Storage Stability
To assess the robustness of the TaqMan-based RT-qPCR assay, repeatability (interoperator variability) and reproducibility (inter-thermocycler variability) were evaluated using the same dilution series from the sensitivity analysis. The dilution series consisted of six serial fivefold dilutions, with initial concentrations of 40, 180, and 280 copies/reaction for norovirus GI, norovirus GII, and rotavirus, respectively. For each concentration, a minimum of five replicates were performed per PCR run. The experiments were conducted by three independent operators using three distinct ABI 7500 Real-Time PCR Systems (Applied Biosystems, Foster City, CA, USA), with each operator performing one experiment on each thermocycler. From both assays, the mean, standard deviation (SD) and coefficient of variation (CV) were calculated independently for each target. A CV value of less than 20% for both assays was regarded as acceptable.
The stability of the new multiplex protocol was assessed through freeze-thaw cycles, transport simulation, and shelf-accelerated stability. Freeze-thaw cycles were tested by incubation of reagents at -30 to -15 • C until freezing and thawing them at room temperature. The simulation of transport was performed by incubation of reagents at -70 to -80 • C for 96 h. Finally, assessment of shelf stability was performed by an accelerated test [21,22]. Accelerated stability was calculated through a stability projection based on incubation at a stress temperature storage condition. This stability projection assumed that the product's degradation rate decreased by a constant factor (Q10) when the storage temperature was reduced by 10 • C. The Q10 value was commonly defined in the in vitro reagent industry with acceleration factors of 2, 3, or 4, due to its activation energy under stress conditions. A Q10 value of 2, for example, provided a conservative estimate of the product's degradation rate and was chosen in our analysis [21,22].
The formula for calculating this stability projection was defined as follows: Q10 = X ∆T/10 X = Acceleration factor. ∆T = Temperature difference ( • C) between stress conditions and ideal storage conditions (-20 • C in this case).
Thus, incubations at 4 • C for periods of 69 days were approximately equivalent to one year of incubation at -20 • C. Accordingly, 3 batches of the IBMP NAT Rotavirus and Norovirus new protocol were incubated under this condition to assess the equivalence of storage at -20 • C for 365 days.
## 2.8. Pilot Multicenter Study
A multicenter pilot study was conducted across Brazil's national rotavirus surveillance network to evaluate the performance of the newly developed IBMP Rotavirus and Norovirus NAT test. The clinical validation involved parallel testing of 523 stool samples from patients with AGE between 2023 and 2024, using both the NAT test and established in-house RT-qPCR methods or commercial ELISA for rotavirus at three reference laboratories: IEC, National Reference Laboratory (n = 92); IAL, Regional Reference Laboratory (n = 188); and IOC, Regional Reference Laboratory (n = 243).
All samples were obtained through the Brazilian AGE Surveillance network from patients referred for routine diagnostic testing. The study design enabled evaluation of interlaboratory sensitivity, methodological concordance, and assay performance under standardized yet operationally diverse conditions.
## 2.9. Statistical Analysis
The Ct value for the amplification of each target gene was analyzed. The statistical significance of the differences in the different groups of data was analyzed by one-way analysis of variance (ANOVA). p < 0.05 was considered statistically significant. Stability comparison was performed using Repeated Measures ANOVA and t-tests on the Log10 (copies/reaction) values corresponding to the limit of detection for each assay, using GraphPad Prism 9.5.1 software (San Diego, CA, USA).
The Kappa coefficient was calculated to assess the level of agreement between diagnostic methods, using the reference laboratory results as the standard and including 95% confidence intervals (CI). Interpretation of Kappa values followed the criteria proposed by [23]: values between 0.81 and 1.00 indicated almost perfect agreement, 0.61-0.80 substantial, 0.41-0.60 moderate, 0.21-0.40 fair, and ≤0.20 slight agreement. Fisher's exact test and the Chi-square test were performed.
## 3. Results
## 3.1. Multiplex Hydrolysis Probe-Based RT-qPCR Design and Analytical Sensitivity
While adapting the reactions into a multiplex setup, we identified a cross reaction where RING1c probe (norovirus GI target) is hydrolyzed in the presence of norovirus GII RNA, resulting in an unspecific amplification signal. To investigate the causes, we evaluated the alignment of the binding regions of primers and probes to norovirus GI and GII RNAs (Figure 1). This analysis revealed high homology between the two strains at this site. Moreover, because the probes for the norovirus GI and GII are in different strands, they are not displaced by the probe with more homology, thus allowing for both probes to be hydrolyzed (although with different efficiencies) regardless of the species in the sample. To confirm this hypothesis, we synthesized RING1, the complement to RING1c that binds to the same strand as RING2. Using both probes in the same strand improved the specificity of the signal, providing clean amplification profiles for both norovirus targets and causing no visible impact on the RVA targets (Figure 2).
After defining the optimal primer concentrations and annealing temperature (Ta), the qPCR assays were evaluated in a multiplex format, incorporating primers and a probe for the detection of a human gene as an internal control. Serial dilutions of RVA RNA (1000 to 10 copies/µL) and norovirus RNA (100 to 0.1 copies/µL) were tested, confirming specific and appropriate amplification. These results provided the basis to proceed with experiments aimed at assessing both the analytical and diagnostic sensitivity and specificity of the assay.
## 3.2. Specificity and Sensitivity
The specificity of the IBMP NAT multiplex RT-qPCR was evaluated against a panel of microorganisms, including Clostridium difficile (Vircell, 21MBC04001-R), Campylobacter jejuni (Vircell, 20MBC088001-R), Yersinia enterocolitica (Vircell, 20MBC027001-R), Shigella flexneri (Vircell, 21MBC89001-R), Escherichia coli EAEC (Vircell, 21MBC121001-R), Salmonella enteritidis (Vircell, 20MBC003001-R), Cryptosporidium parvum Tyzzer (ATCC, PRA-67DQ), Entamoeba histolytica HM-1:IMSS (ATCC, 30459D and 20459DQ), Blastocystis hominis strain BT1 (ATCC, 50608D), Giardia intestinalis WB clone C6 (ATCC, 50803D), and also clinical samples positive for sapovirus, enterovirus species A (coxsackievirus A6), enterovirus species B (echovirus 18), and adenovirus types 40 (AdVF40) and 41 (AdVF41). No cross-reactivity was detected with any of the tested organisms.
Sensitivity analysis was performed based on the results of the robustness assays. To determine the Limit of Detection with 95% confidence (LoD95), six serial fivefold dilutions were prepared, starting from initial concentrations of 40, 180, and 280 copies per reaction for norovirus GI, norovirus GII, and RVA, respectively. The estimated LoD95 values were 18.6 copies/reaction for norovirus GI, 71.2 copies/reaction for norovirus GII, and 12.3 copies/reaction for RVA (Figure 3).
## 3.3. Performance Evaluation in Clinical Samples
A panel of 379 fecal suspension samples was used for the clinical performance evaluation of the IBMP NAT Rotavirus and Norovirus Kit. This panel consisted of samples positive for RVA, norovirus (GI and GII), and negative samples that tested negative for all three viruses but presented with similar clinical symptoms. The samples were provided by the three Brazilian Reference Laboratories. In total, the panel included 137 negative samples, 80 positive samples for RVA, 71 positive samples for norovirus GI, and 96 positive samples for norovirus GII. Also, five samples that were co-detected with more than one of the target viruses.
For the RVA target, the IBMP NAT Rotavirus and Norovirus Kit demonstrated a diagnostic sensitivity and specificity of 98.75% and 96.32%, respectively. The Positive Predictive Value (PPV) was 87.78%, while the Negative Predictive Value (NPV) was 99.65%. The detection of the norovirus GI demonstrated a diagnostic sensitivity and specificity of 98.59% and 98.7%, respectively. The PPV was 94.59%, and the NPV was 99.67%. For the norovirus GII target, the IBMP NAT Rotavirus and Norovirus Kit showed a diagnostic sensitivity and specificity of 97.92% and 99.29%, respectively, and the PPV was 97.92%, and the NPV was 99.29% (Table 1).
Table 1. Performance of the molecular assay developed by IBMP for the detection of RVA, norovirus genogroup GI, and norovirus genogroup GII, compared to the reference method. The table presents absolute concordance values (detected/not detected), total number of samples tested, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), along with their respective 95% confidence intervals (95% CI).
## Rotavirus (RVA)
## Reference
## 3.4. Robustness and Stability of the RT-qPCR Diagnostic Kit
To evaluate inter-operator reproducibility, three independent operators performed amplification assays targeting COG1, COG2, and NSP3 using serial dilutions of known input concentrations. For each target and concentration, the mean Ct value, standard deviation, and relative standard deviation (RSD%) were calculated. The results demonstrated high reproducibility among operators, with RSD values up to 5.79%, even at lower concentrations (Supplementary Table S1). These findings support the analytical consistency of the assay when performed by different users under the same conditions.
Intra-assay reproducibility of the IBMP NAT assay was evaluated using serial dilutions of known input concentrations for the targets COG1, COG2, and NSP3. Each condition was tested in replicates by three independent operators, each using three different real-time PCR thermal cyclers. For every target and concentration, mean Ct values, standard deviations, and relative standard deviations (RSD%) were calculated. For most tested concentrations, RSD values remained below 3%, confirming strong reproducibility. The highest variability was observed at the lowest input level for COG1 (1.6 copies/µL), with an RSD of 5.23%, which is expected due to increased stochastic variation at low template concentrations (Supplementary Table S2). These findings demonstrate the robustness and precision of the assay across different instruments and operators, reinforcing its suitability for routine diagnostic applications.
Stability of the IBMP NAT assay was evaluated through three experimental approaches: freeze-thaw cycling, simulated transport, and accelerated stability at 4 • C. For each condition, the 95% limit of detection (LoD95%) was determined in Log10 copies per reaction for the three assay targets (COG1, COG2, and NSP3) across three different production lots (EXT 012/24, EXT 015/24, and EXT 023/24). Statistical analyses were conducted using repeated-measures ANOVA and Student's t-tests to assess intra-and inter-lot consistency and the impact of stress condition. In the freeze-thaw test, no significant differences were observed in LoD95% values between the first and fifth freeze-thaw cycles for any of the targets (COG1, COG2, NSP3), with p-values ranging from 0.55 to 0.96. The consistency across lots was confirmed by high inter-lot p-values (COG1 = 0.87; COG2 = 0.87; NSP3 = 0.42), and low t-values indicated no statistically meaningful shift (Supplementary Table S3). Similarly, in the simulated transport test, no significant changes were detected in LoD95% before and after simulated transport for any of the targets. p-values for COG1, COG2, and NSP3 were 0.97, 0.82, and 0.51, respectively. This indicates that the assay maintains its analytical sensitivity under physical stress conditions mimicking transportation scenarios (Supplementary Table S3).
The accelerated stability test, simulating 12 months of storage at 4 • C, also showed no statistically significant degradation in performance. For all targets, LoD95% value remained stable, with p-values above the significance threshold (COG1 = 0.57; COG2 = 0.25; NSP3 = 0.16). Although a slight increase in variability was observed for the NSP3 target (t = 2.10), it did not reach statistical significance (Supplementary Table S3). Overall, these results support the robustness of the IBMP NAT assay under routine handling and storage conditions. The assay exhibited consistent performance across production lots and maintained its sensitivity following environmental stress simulations.
## 3.5. Multicenter Evaluation the IBMP NAT TaqMan-Based RT-qPCR Assay in Clinical Samples
To evaluate the performance of the IBMP NAT test TaqMan-based RT-qPCR protocol at the three reference laboratories, a total of 523 clinical samples, previously tested positive for RVA, norovirus GI or GII during routine diagnosis, were used for the analysis. Previous negative samples were also included as negative controls. The IBMP NAT test was compared to the diagnostic methodologies used at each site. Both IEC and IOC laboratories employed in-house TaqMan-based RT-qPCR assays to detect RVA and norovirus GI and GII. At IAL, RVA was detected using the Ridascreen ® Rotavirus ELISA (R-Biopharm, Darmstadt, Germany), while norovirus GI and GII was detected using the same in house TaqMan-based RT-qPCR assay.
At IAL, the NAT test demonstrated high concordance with ELISA for RVA detection. Among ELISA-positive samples (n = 115), the NAT assay detected RVA in 95.6% (n = 110). Of 73 ELISA-negative samples, 61 (83.5%) were also negative by NAT assay. Notably, the IBMP NAT test identified RVA in 12 ELISA-negative samples (16.4%), all of which showed high Ct values (Ct > 31), including six samples with Ct between 31 and 33, and five with Ct > 33. For norovirus GI detection, the IBMP NAT assay detected 22 of 29 positive samples (75.9%) identified by in-house RT-qPCR. Among the seven undetected samples, six had Ct values > 33. Both assays showed 100% concordance for the 159 norovirus GI-negative samples. Regarding norovirus GII, the IBMP NAT assay detected 37 of 42 positive samples (88.1%) identified by in-house RT-qPCR. The five undetected samples included three with Ct > 38. Among 146 negative samples, three were positive by IBMP NAT test (2%), though two of which had Ct values > 34.
At IEC and IOC laboratories, the performance of the IBMP NAT test was compared against in-house RT-qPCR protocols. At IEC, complete concordance (100%) was observed for RVA detection in 42 samples using both methods. The NAT assay additionally identified 10 RVA-positive samples among those previously negative by in-house RT-qPCR, all exhibiting high Ct values (>33), suggesting low viral loads. For norovirus GI, a 100% concordance was achieved between methods across all tested samples (4 positive and 88 negative). Regarding norovirus GII, the NAT assay detected 28 of 33 samples previously identified as positive by in-house RT-qPCR, corresponding to a sensitivity of 84.8%. The five undetected samples all exhibited high Ct values (>36) in the in-house assay. Both assays demonstrated 100% concordance for GII-negative samples (n = 59).
At IOC, 100% concordance was observed for RVA detection in 85 samples using both methods. Among 158 negative samples, the IBMP NAT test detected RVA in six samples, all with high Ct values (Ct > 32), including two samples with Ct values between 32 and 33, and four with Ct > 33. For norovirus GI, 100% concordance was achieved between methods across all tested samples (23 positive and 202 negative). Regarding norovirus GII, 100% concordance was observed in 61 samples using both methods. Among 164 negative samples, the IBMP NAT assay yielded six positive results (3.6%), with Ct values ranging from 15 to 31.
Overall, the IBMP NAT test demonstrated high sensitivity, detecting RVA in 98% (238/242) of samples previously identified as positive by in-house RT-qPCR. Additionally, the assay identified RVA in 28 samples that were negative by ELISA or RT-qPCR (10.1% of 278), though those discordant samples predominantly exhibited high Ct values (>33), suggesting lower viral loads. The overall concordance between methods for RVA detection was 90% (kappa value of 0.9 (95% CI, 0.86-0.93) (Table 2). The IBMP NAT test demonstrated 87.5% sensitivity (49/56) for norovirus GI detection, with six of seven undetected samples exhibiting high Ct values (>33). The assay showed perfect specificity (100%, 449/449) for norovirus GI-negative samples. For norovirus GII, the IBMP NAT test achieved 92.6% sensitivity (126/136), with 80% of undetected samples (8/10) showing high Ct values (Ct > 36). Among the 360 negative samples, the IBMP NAT assay yielded nine positive results (2.5%). The overall kappa values of norovirus GI and GII for the IBMP NAT test compared to the in-house RT-qPCR were 0.93 (95% CI, 0.88-0.98) and 0.92 (95% CI, 0.88-0.96), respectively (Table 2). Concordance values between the IBMP NAT and in-house RT-qPCR assays, as well as sensitivity and sensibility values obtained in each reference laboratory for each RVA, norovirus GI and GII are presented in Tables 2 andS4.
We also compared the Ct values for RVA, norovirus GI and GII obtained using both methodologies across the three reference laboratories. Overall, the IBMP NAT test detected all three viruses with lower Ct values compared to the in-house RT-qPCR. For RVA, among 126 positive samples tested, Ct values ranged from 12.6 to 36.7 and from 16.1 to 30.7 with the in-house RT-qPCR and the IBMP NAT test, respectively. The IBMP NAT test yielded significantly lower Ct values (median of 21.3) compared to the in-house RT-qPCR (median of 24.9) (p < 0.0001). Regarding norovirus GI, 49 samples were detected by both assays across the three reference laboratories. Ct values ranged from 19.1 to 37 with the in-house RT-qPCR and from 14.4 to 33 with the IBMP NAT test. Again, the IBMP NAT test produced significantly lower Ct values (median of 24.1) compared to the in-house assay (median of 28) (p = 0.0005). For norovirus GII, a similar downward trend was observed, with the IBMP NAT test yielding lower Ct values. Among the positive samples, Ct values ranged from 14.6 to 37 with the in-house RT-qPCR and from 13.5 to 34 with the IBMP NAT test. The median Ct values were 24.3 and 19.6, respectively (p < 0.0001) (Figure 4).
## 4. Discussion
The global implementation of RVA vaccines from 2006 onward, and their subsequent widespread use have significantly reduced RVA disease burden, as evidenced by numerous impact assessments and vaccine effectiveness studies worldwide [24][25][26]. Despite these advances, RVA remains a major cause of AGE, even with substantial declines in RVA-associated AGE cases globally [3,27,28]. For instance, a recent vaccine effectiveness study demonstrated that RVA was the leading cause of severe AGE in both vaccinated and unvaccinated children in India [29], aligning with findings from a prior ROTAVAC efficacy trial reanalysis [30]. Similarly, a Tanzanian study on ROTARIX introduction observed reduced diarrhea admissions, yet RVA persisted as the primary pathogen driving hospitalizations in children under five [25]. Noroviruses also figure as a major cause of AGE in people of all ages worldwide. In countries where the RVA vaccine has been implemented, cases of norovirus-related illnesses frequently rank as the primary reason for medical visits due to AGE in pediatric populations [13,17,26]. Together, these pathogens are responsible for millions of cases annually, imposing a significant economic and social impact on healthcare systems.
In the present study, we describe the development and evaluation of a NAT kit able to detect RVA, norovirus GI and GII, and an internal control in stool samples from AGE patients. The PCR primers were selected from a highly conserved region of the RVA nonstructural protein 3 (NSP3) sequence, while for norovirus, primers and probes target a highly conserved junction region between ORFs1 and 2. Molecular diagnostics play a critical role in addressing these challenges, as conventional methods such as antigen-based assays or electron microscopy lack the sensitivity, specificity, and throughput required for reliable detection and large-scale surveillance. For instance, some studies have demonstrated that molecular methods, especially real time-based methods, are more sensitive than antigen-detection EIA assays and conventional RT-PCR [31,32]. Real-time PCR, especially in multiplex formats, provides rapid, highly sensitive, and specific results, allowing simultaneous detection of multiple targets in a single reaction. By developing a standardized kit, rather than relying on in-house protocols, issues related to inter-laboratory variability and assay validation can be minimized, thereby facilitating scalability and incorporation of molecular testing into national surveillance programs.
From a public health perspective, the availability of a national validated RT-qPCR kit contributes not only to clinical diagnosis but also to the early detection of outbreaks, and emergence epidemic virus genotypes, which is crucial for the implementation of timely control measures. Furthermore, routine surveillance supported by molecular tools enables the monitoring of viral circulation patterns, seasonality, and genotype distribution, all of which are essential to inform vaccination strategies and evaluate their longterm effectiveness. In addition, harmonized detection systems strengthen international data comparability, enhancing the capacity of global networks to respond to emerging threats. Multiple surveillance studies conducted across Brazilian regions have consistently identified RVA and norovirus as major etiological agents of AGE. For instance, in Belém, Northern Brazil, norovirus showed high prevalence (~25% positivity) among children hospitalized with AGE [33]. Similarly, a study in Rio Grande do Sul, Southern Brazil, attributed approximately 50% of AGE outbreaks to norovirus infections [34]. More recently, Sarmento et al. [35] reported a high norovirus prevalence (37.2%) among both outpatients and inpatients with AGE, during a four-year study (2019-2022) with samples from several Brazilian states. Moreover, multiple viral AGE outbreaks have been documented in Brazil, revealing diverse transmission sources. Early investigations analyzed foodborne transmission [36] and clinical cases [37], and both sample types of samples to definitively confirm outbreak sources [38]. More recent reports described outbreaks linked to commercial ice pops [39], outbreak at hospital setting with norovirus and RVA co-circulation [40], and outbreaks linked to emergent strains, including a rare GII.10[P16] recombinant [41,42]. In 2024, a major AGE outbreak involving over 217,000 cases across 180 municipalities in Goiás, Brazil, has been primarily attributed to RVA and norovirus infections, with approximately 57,000 cases reported in August alone (https://g1.globo.com/go/goias/noticia/2024/09/18/surto-de-diarreiaaguda-atinge-mais-de-180-cidades-em-goias-diz-secretaria-de-saude.ghtml, accessed on 7 October 2025). More recently, during the summer and rainy season in the Southern Hemisphere, between 29 December 2024 and 6 March 2025, a large AGE outbreak attributed to norovirus occurred in the coastal region of São Paulo State, Brazil. According to the Notifiable Diseases Information System (SINAN) of the Brazilian Ministry of Health, 54 outbreaks were reported, involving more than 76,000 cases requiring medical care, underscoring the substantial public health impact of norovirus-associated gastroenteritis in the region (https://g1.globo.com/sp/sao-paulo/noticia/2025/01/08/institutoadolfo-lutz-confirma-presenca-de-norovirus-em-amostras-de-fezes-coletadas-na-baixadasantista-apos-casos-de-viroses.ghtml, accessed on October 7 2025).
The establishment of a real-time PCR kit targeting RVA and norovirus provides a valuable platform for both diagnostic and epidemiological applications. Its integration into disease surveillance systems, supported by the strategic use of nationally produced reagents, can significantly enhance the ability to monitor disease dynamics, improve outbreaks response, and ultimately reduce the burden associated with viral gastroenteritis, while simultaneously reinforcing the autonomy and sustainability of Brazil's public health infrastructure.
## 5. Conclusions
The IBMP NAT Rotavirus and Norovirus Kit represents a significant advancement in molecular diagnostics, offering high sensitivity, specificity, and reproducibility for detecting RVA and norovirus G I and GII. Its low limits of detection enable accurate identification of infections even at minimal viral loads, which is crucial for early diagnosis and outbreak control. The use of these highly conserved target regions provides strong robustness and long-term sustainability for molecular surveillance of different genotypes.
Additionally, its superior analytical sensitivity compared to conventional RT-qPCR methods enhances diagnostic confidence, reduces false negatives, and supports more effective clinical decision-making and public health surveillance. The kit development strengthens diagnostic capacity, enables rapid outbreak response, and will enhance nationwide surveillance of RVA and norovirus within the Brazilian Surveillance Network. In conclusion, the development of a national kit marks a strategic advance for Brazil's health autonomy, reducing dependence on imported reagents and ensuring a sustainable, locally validated diagnostic supply. By guaranteeing continuous testing capacity even amid global disruptions, this initiative reinforces diagnostic excellence and strengthens autonomous and resilient national surveillance for RVA and norovirus.
## References
1. Troeger, Blacker, Khalil et al. (2016) "Estimates of the Global, Regional, and National Morbidity, Mortality, and Aetiologies of Diarrhoea in 195 Countries: A Systematic Analysis for the Global Burden of Disease Study" *Lancet Infect. Dis*
2. Wardlaw, Salama, Brocklehurst et al. (2010) "Why Children Are Still Dying and What Can Be Done" *Lancet*
3. Cohen, Platts-Mills, Nakamura et al. (2022) "Aetiology and Incidence of Diarrhoea Requiring Hospitalisation in Children under 5 Years of Age in 28 Low-Income and Middle-Income Countries: Findings from the Global Pediatric Diarrhea Surveillance Network" *BMJ Glob. Health*
4. Kotloff (2017) "The Burden and Etiology of Diarrheal Illness in Developing Countries" *Pediatr. Clin. N. Am*
5. Operario, Platts-Mills, Nadan et al. (2017) "Etiology of Severe Acute Watery Diarrhea in Children in the Global Rotavirus Surveillance Network Using Quantitative Polymerase Chain Reaction" *J. Infect. Dis*
6. Platts-Mills, Liu, Rogawski et al. (2018) "Use of Quantitative Molecular Diagnostic Methods to Assess the Aetiology, Burden, and Clinical Characteristics of Diarrhoea in Children in Low-Resource Settings: A Reanalysis of the MAL-ED Cohort Study"
7. Gutierrez, Arantes, Bello et al. (2015) "Emergence and Dissemination of Equine-like G3P[8] Rotavirus A in Brazil between" *Microbiol. Spectr*
8. Greenberg, Estes, Rotaviruses (2009) "From Pathogenesis to Vaccination" *Gastroenterology*
9. Matthijnssens, Ciarlet, Heiman et al. (2008) "Full Genome-Based Classification of Rotaviruses Reveals a Common Origin between Human Wa-Like and Porcine Rotavirus Strains and Human DS-1-Like and Bovine Rotavirus Strains" *J. Virol*
10. Chhabra, De Graaf, Parra et al. (2019) "Updated Classification of Norovirus Genogroups and Genotypes" *J. Gen. Virol*
11. Hoa Tran, Trainor, Nakagomi et al. (2013) "Molecular Epidemiology of Noroviruses Associated with Acute Sporadic Gastroenteritis in Children: Global Distribution of Genogroups, Genotypes and GII.4 Variants" *J. Clin. Virol*
12. Siebenga, Vennema, Zheng et al. (2009) "Norovirus Illness Is a Global Problem: Emergence and Spread of Norovirus GII.4 Variants, 2001-2007" *J. Infect. Dis*
13. Payne, Vinjé, Szilagyi et al. (2013) *Norovirus and Medically Attended Gastroenteritis in U.S. Children. N. Engl. J. Med*
14. Bucardo, Reyes, Svensson et al. (2014) "Predominance of Norovirus and Sapovirus in Nicaragua after Implementation of Universal Rotavirus Vaccination" *PLoS ONE*
15. Hemming, Räsänen, Huhti et al. (2013) "Major Reduction of Rotavirus, but Not Norovirus, Gastroenteritis in Children Seen in Hospital after the Introduction of RotaTeq Vaccine into the National Immunization Programme in Finland" *Eur. J. Pediatr*
16. Bartsch, Lopman, Ozawa et al. (2016) "Global Economic Burden of Norovirus Gastroenteritis" *PLoS ONE*
17. Tarr, Pang, Zhuo et al. (2021) "Attribution of Pediatric Acute Gastroenteritis Episodes and Emergency Department Visits to Norovirus Genogroups I and II" *J. Infect. Dis*
18. Zeng, Halkosalo, Salminen et al. (2008) "One-Step Quantitative RT-PCR for the Detection of Rotavirus in Acute Gastroenteritis" *J. Virol. Methods*
19. Hill, Mull, Jothikumar et al. (2010) "Detection of GI and GII Noroviruses in Ground Water Using Ultrafiltration and TaqMan Real-Time RT-PCR" *Food Environ. Virol*
20. Kageyama, Kojima, Shinohara et al. (2003) "Broadly Reactive and Highly Sensitive Assay for Norwalk-Like Viruses Based on Real-Time Quantitative Reverse Transcription-PCR" *J. Clin. Microbiol*
21. Magari (2003) "Assessing Shelf Life Using Real-Time and Accelerated Stability Tests" *BioPharm Int*
22. Anderson, Scott (1991) "Determination of Product Shelf Life and Activation Energy for Five Drugs of Abuse" *Clin. Chem*
23. Landis, Koch (1977) "The Measurement of Observer Agreement for Categorical Data" *Biometrics*
24. Patel, Steele, Gentsch et al. (2011) "Real-World Impact of Rotavirus Vaccination" *Pediatr. Infect. Dis. J*
25. Platts-Mills, Amour, Gratz et al. (2017) "Impact of Rotavirus Vaccine Introduction and Postintroduction Etiology of Diarrhea Requiring Hospital Admission in Haydom, Tanzania, a Rural African Setting" *Clin. Infect. Dis*
26. Burnett, Parashar, Tate (2020) "Real-World Effectiveness of Rotavirus Vaccines, 2006-2019: A Literature Review and Meta-Analysis" *Lancet Glob. Health*
27. Varghese, Alokit Khakha, Giri et al. (2021) "Rotavirus Strain Distribution before and after Introducing Rotavirus Vaccine in India"
28. Kyu, Vongpradith, Dominguez et al. (2025) "Global, Regional, and National Age-Sex-Specific Burden of Diarrhoeal Diseases, Their Risk Factors, and Aetiologies, 1990-2021, for 204 Countries and Territories: A Systematic Analysis for the Global Burden of Disease Study 2021" *Lancet Infect. Dis*
29. Abraham, Premkumar, Platts-Mills et al. (2024) "Effect of Non-Rotavirus Enteric Infections on Vaccine Efficacy in a ROTASIIL Clinical Trial" *Am. J. Trop. Med. Hyg*
30. Praharaj, Platts-Mills, Taneja et al. (2019) "Diarrheal Etiology and Impact of Coinfections on Rotavirus Vaccine Efficacy Estimates in a Clinical Trial of a Monovalent Human-Bovine (116E) Oral Rotavirus Vaccine" *Clin. Infect. Dis*
31. Pang, Joensuu, Hoshino et al. (1999) "Rotaviruses Detected by Reverse Transcription Polymerase Chain Reaction in Acute Gastroenteritis during a Trial of Rhesus-Human Reassortant Rotavirus Tetravalent Vaccine: Implications for Vaccine Efficacy Analysis" *J. Clin. Virol*
32. Pang, Lee, Boroumand et al. (2004) "Increased Detection of Rotavirus Using a Real Time Reverse Transcription-polymerase Chain Reaction (RT-PCR) Assay in Stool Specimens from Children with Diarrhea" *J. Med. Virol*
33. Reymão, Fumian, Justino et al. (2018) "Norovirus RNA in Serum Associated with Increased Fecal Viral Load in Children: Detection, Quantification and Molecular Analysis" *PLoS ONE*
34. De Andrade, Rocha, Carvalho-Costa et al. (2004) "Noroviruses Associated with Outbreaks of Acute Gastroenteritis in the State of" *J. Clin. Virol*
35. Sarmento, De Andrade, Malta et al. (2019) "Norovirus Epidemiology and Genotype Circulation during the COVID-19 Pandemic in Brazil"
36. Morillo, Luchs, Cilli et al. (2010) "Rapid Detection of Norovirus in Naturally Contaminated Food: Foodborne Gastroenteritis Outbreak on a Cruise Ship in Brazil" *Food Environ. Virol*
37. Silva, De Andrade, Fumian et al. (2016) "17 Associated with a Foodborne Acute Gastroenteritis Outbreak in Brazil" *Food Environ. Virol*
38. Morillo, Luchs, Cilli et al. (2014) *Pe Genotype: Tracking a Foodborne Outbreak on a Cruise Ship Through Molecular Epidemiology*
39. Fumian, Ferreira, De Andrade et al. (2020) "Norovirus Foodborne Outbreak Associated With the Consumption of Ice Pop, Southern Brazil" *Food Environ. Virol*
40. Gutierrez, De Figueiredo, Fialho et al. (2019) "Nosocomial Acute Gastroenteritis Outbreak Caused by an Equine-like G3P[8] DS-1-like Rotavirus and GII.4 Sydney[P16] Norovirus at a Pediatric Hospital in Rio de Janeiro"
41. Fumian, Malta, Sarmento et al. "Acute Gastroenteritis Outbreak Associated with Multiple and Rare Norovirus Genotypes after Storm Events in Santa Catarina" *Brazil. J. Med*
42. Burlandy, Malta, Mello et al. "Foodborne Acute Gastroenteritis Outbreak Associated with a Rare Norovirus Recombinant GII"
43. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548434&blobtype=pdf | # Identification of clinical and virological correlates associated with influenza A candidate vaccine virus (CVV) attenuation in a ferret model
Claudia Pappas, Nicole Brock, Jessica Belser, Troy Kieran, Joanna Pulit-Penaloza, Xiangjie Sun, Hui Zeng, Li Wang, Bin Zhou, Terrence Tumpey, Taronna Maines
## Abstract
Influenza A viruses continuously circulate among avian and swine species, posing a persistent threat to public health. The development of influenza candidate vaccine viruses (CVVs) plays a pivotal role in the global strategy for influenza pandemic preparedness. Safety-testing of CVVs for attenuation in ferrets represents a critical step that takes place prior to making these viruses available to vaccine manufacturers. Development of pathogenicity standards is needed to establish acceptable thresholds of disease so that CVV safety can be assessed without the need for comparison to the parental virus. To assess the capacity of diverse CVVs to cause pathogenesis in mamma lian hosts, clinical and virological parameters were compiled from CVV assessments in ferrets conducted using consistent methods over approximately 20 years to identify disease parameters most reflective of attenuation compared to wild-type strains. These analyses revealed an overall reduction in ferret weight loss and fever relative to wild-type controls. Viral titers in nasal washes were reduced with limited spread to tissues beyond the respiratory tract. Regression models further support the significance of clinical signs in distinguishing the virulence of wild-type viruses and CVVs. These findings provide support for the development of standardized parameters for assessing pathogenicity of CVVs and their suitability for manufacturers.
IMPORTANCEThe development and safety testing of candidate vaccine viruses (CVVs) against emerging zoonotic influenza strains prior to sharing with vaccine manufacturers is a critical component of influenza pandemic preparedness. The extensive data set reported here provides critical information that will drastically streamline the safety testing process, thereby enabling more efficient CVV assessments and improving public health in the event of an influenza pandemic.
KEYWORDS zoonotic influenza virus, ferrets, candidate vaccine virusI nfluenza A viruses (IAV) continuously circulate among a diversity of mammalian and avian species, with occasional spillover to humans, posing a persistent threat to human health (1). Risk assessment rubrics, such as the Influenza Risk Assessment Tool (IRAT) developed by the Centers for Disease Control and Prevention (CDC), aid in the identifica tion of zoonotic viruses of public health concern (2, 3) and inform resource allocation for public health countermeasures. Vaccines are widely recognized as the most effective defense against IAV but can be challenging to rapidly deploy in the event of novel virus emergence in a population due to lengthy development and production consid erations (4). As such, the development of influenza candidate vaccine viruses (CVVs) plays a pivotal role in the global strategy for influenza pandemic preparedness (5). Once CVVs are meticulously chosen, developed, and characterized (6), they become available
to national authorities and vaccine manufacturers for stockpiling, further studies, and clinical trials.
Testing CVV attenuation relative to parental wild-type (WT) strains is a component of development and production practices to reduce risk to humans and animals during vaccine manufacturing. There are several tests to evaluate pathogenicity of CVVs in vitro and in vivo, including trypsin-independent replication in cell culture, gene sequencing, genetic stability, pathogenicity in chickens, and attenuation in ferrets (7). Ferrets are the preferred mammalian model for assessing attenuated phenotypes compared to WT viruses, as they exhibit many clinical signs of IAV infection similar to those observed in humans and are, therefore, valuable for studying the pathogenicity of zoonotic IAV (8). Due to the ever-increasing diversity of zoonotic IAVs crossing species barriers worldwide, there is substantial phenotypic and genetic variability among WT strains that can affect the relative degree of attenuation observed when safety-testing CVVs (9)(10)(11). While a few studies have reported CVV attenuation relative to WT virus in ferrets (12)(13)(14)(15), there remains a need to identify and standardize attenuation metrics that are broadly applicable when assessing a diverse range of CVVs, thereby supporting the establish ment of an acceptable alternative to direct comparison with the parental WT virus.
Achieving safe development and production of IAV CVVs represents a multi-national effort inclusive of CVV-testing laboratories, vaccine manufacturers, and national regulatory authorities (7,16). In this context, we present aggregated ferret pathogenicity evaluations of 30 CVVs conducted using consistent methods from one laboratory, with CVV attenuation compared to WT strains of similar subtype/clade produced in the same laboratory or elsewhere. Analyses of clinical (weight loss, fever) and virological parame ters (viral detection in nasal washes [NW] and tissues) were performed to identify the metrics most consistently associated with attenuation, independent of strain heteroge neity. Collectively, these findings demonstrate consistent attenuation of CVVs compared to WT viruses and support the development of a standardized protocol with established thresholds for assessing pathogenicity and safety testing of future CVVs.
## RESULTS
## IAV WT and CVVs tested at CDC
CDC has performed IAV CVV testing under a consistent experimental protocol for approximately 20 years, with zoonotic subtypes spanning H1 (including a 2009 H1N1 pandemic virus), H2, H3, H5, H7, H9, and H10, highlighting the diversity of viruses that pose a threat to human health (Tables 1 and2). As IAV CVV safety-testing typically includes demonstrating attenuation in the ferret model (7), multiple WT viruses of public health concern were characterized in ferrets to serve as reference strains to measure relative attenuation of CVVs in this model. With the exception of the LPAI virus A(H7N7), IBCDC-1, which was a CVV generated by classical reassortment (17), all CVVs shown in Table 1 were generated by reverse genetics (15). CVVs typically consist of six internal genes derived from the A/Puerto Rico/8/1934 A(H1N1) (PR8) virus, combined with the hemagglutinin (HA) and neuraminidase (NA) genes from the WT influenza viral target of interest. In the case of H5 or H7 highly pathogenic avian influenza (HPAI) viruses, the sequences corresponding to the multi-basic amino acids (MBAA) in the HA were removed to prevent HA cleavage by ubiquitous cellular proteases.
To support high yield in eggs, which is the typical method for vaccine develop ment, additional mutations or modifications, none of which result in gain-of-function in mammalian cells, were introduced into certain CVVs listed in Table 1. Collectively, the use of reverse genetics with a 2:6 WT to PR8 backbone configuration represents a robust, safe, and consistent approach for the rapid generation of CVVs to support pandemic preparedness (7).
## Clinical signs of infection of IAV CVVs in ferrets
CVVs typically exhibit reduced clinical signs of infection in ferrets relative to parental WT strains, attributed primarily to the presence of a PR8 virus backbone and removal of the MBAA cleavage site in the HA when present (7,15). All ferrets inoculated with the CVVs listed in Table 1 survived the infection as did RG-PR8-inoculated ferrets, whereas lethal phenotypes were reported among many of the HPAI parental WT strains. Furthermore, the frequency of clinical signs of infection (encompassing lethargy, diarrhea, nasal discharge, respiratory and neurological symptoms, and mortality) was reduced when ferrets were inoculated with CVVs compared with their matched WT viruses (Fig. 1). However, no systematic evaluation has been performed to assess the relative degree of attenuation across a diverse panel of CVVs. To this end, we aggregated data from the viruses shown in Table 1 and compared the relative differences in key clinical parameters between WT and CVV strains, with an aim to identify which recorded parameters during CVV standard pathotyping assessments were most substantially reduced compared to WT. Elevated temperatures in ferrets is a common clinical sign associated with zoonotic IAV infection. All except one of the WT viruses included here had a mean peak temper ature of >1°C during days 1-9 post-inoculation (p.i.), with a median rise over baseline across all viruses of 1.7°C (Fig. 2A). CVVs consistently had lower mean peak temperatures than their paired parental WT viruses (median value 0.8°C) with one exception (RG32A), though the degree of reduction varied extensively depending on the parental strain, from 0.1 to 2.0°C. Differences between mean peak rises in temperature were statistically significant between WT and CVV groups (P = 4e-8, Table S1), further supporting that CVVs were consistently associated with detectable and meaningful reductions in peak temperature readings during acute infection.
Weight loss is a frequently measured parameter to assess morbidity in laboratory animals (37). While parental strains demonstrated a wide range of weight loss (approach ing 20% below baseline), CVVs tested all had mean values < 6%, indicating generally mild morbidity (Fig. 2B). Similar to temperature, the range of attenuated weight loss varied on a per-strain basis. Despite the heterogeneity of maximum weight loss reported among WT strains, mean differences were statistically significant between WT and CVV groups (P = 1.3e-7, Table S1). Limiting this analysis to only H5 subtype viruses (inclusive of A(H5N1), A(H5N2), A(H5N6), and A(H5N8) viruses) resulted in comparable results as the entire data Contains two amino acid changes in the antigenic site that reflect those found in the highly pathogenic avian influenza (HPAI) virus. i Weight loss on days 3 and 7 post inoculation reported by van den Brand JM et al. (35). j Number of ferrets that exhibited weight loss (expressed as mean maximum percentage) or temperature rise (expressed as mean rise above baseline in °C) from preinoculation values/total number of ferrets (in parenthesis); 3/3 unless otherwise specified. k Clinical signs observed days 1-14 p.i., defined by the absence (no) or presence (yes) of one or more of the following symptoms: lethargy score of 1 for at least 2 days, or score of 2 for one or more days (36), respiratory symptoms (sneezing, wheezing, dyspnea), nasal (nasal discharge), diarrhea and neurological (incoordination). l For A/chicken/Guangdong-Shantou/481/2022 and A/harbor seal/Germany/PV20762_TS/2014 viruses, ferrets were inoculated with 10 6 TCID 50 intranasal or intratracheally, respectively. Data for A/chicken/Guangdong-Shantou/481/2022 and A/harbor seal/Germany/PV20762_TS/2014 were not obtained from CDC laboratories. nd, not determined. Wild-type viruses that were used as reference are in bold. set (Table S2), supporting that attenuation of these clinical parameters were maintained across different data sets.
## Virus replication of IAV CVVs in ferrets
The PR8 virus backbone is known for causing mild illness in ferrets although it can replicate moderately in the upper respiratory tract of ferrets up to day 5 p.i., with minimal levels of virus detected in the lungs (12,38). We first assessed relative differences between viral titers of paired WT and CVVs in two upper respiratory tract specimens: NW (mean peak titers days 1-7 p.i.) and nasal turbinate tissue (mean titers on day 3 p.i.). Attenuation of peak NW viral titer was observed in all but one pair, though the degree of attenuation varied in a strainspecific manner (Fig. 3A). Mean differences in peak NW titer were statistically significant between WT and CVV groups (P = 1.5e-3, Table S1; Fig. 3A).
Comparison of nasal turbinate viral titers revealed a similar pattern (Fig. 3B), with greater distance between both mean and median titers of groups, with statistical significance of P = 2.7e-4 (Table S1). Statistical significance between differences for both mean peak NW and mean day 3 nasal turbinate titers was maintained when aggregated viruses were limited to the H5 subtype (Table S2).
In contrast to NW and nasal turbinate specimens, a more pronounced attenuative effect was observed among CVV strains compared to WT viruses in tissues more distal from the nares (Fig. 4). For both trachea and lung specimens, median values for CVVs were at the limit of detection (Table S1). Similar to upper respiratory tract specimens, the degree of attenuation varied depending on the parental strain; however, CVVs overall had statistically significant lower day 3 p.i. viral titers compared to WT strains for both trachea (P = 6.3e-3) and lung (P = 2.3e-7) specimens (Table S1). In particular, the mean virus titers in lung tissues were >2.5 logs lower among CVV compared with WT strains, with the largest single mean WT-CVV pair reduction of viral titer in this specimen (>6 logs for IBCDC-1). Extrapulmonary spread of IAV is not uncommon among HPAI viruses that cause severe disease. Accordingly, mean viral titers of CVVs were also significantly reduced in brain (P = 1.1e-2) and olfactory bulb (P = 3.1e-5) tissues (Table S1). Furthermore, frequency of virus detection was reduced in these and additional extrapulmonary tissues sampled, including the intestine and spleen (Fig. 5). Taken together, these results further support that CVV strains can maintain a capacity for replication throughout the mamma lian respiratory tract though the frequency and magnitude of virus replication is reduced When values were identical between CVV and WT groups, only one dot is shown. compared to parental WT strains. These results align with established guidelines, which indicate that the replication of the virus and the associated clinical symptoms should resemble those produced by the attenuated PR8 parent virus, while being less severe than disease caused by the corresponding WT zoonotic virus (7).
## Predictive magnitude of variables for IAV CVV attenuation
The evaluations above identified many independent variables that could potentially predict IAV CVV attenuation. We used generalized linear models (GLM) to assess different combinations of features to identify those elements most strongly associated with prediction of CVV attenuation. To do this, the evaluation of key variables for predictive CVV feature selection was performed using Ridge and Lasso regression models (Table S4 andS5), and the predictive strength of each model was determined by ranking each based on Akaike Information Criteria (AIC) where lower AIC values indicate higher-per forming models. Models were tested with a full panel of 30 independent variables (full list in Table S3), spanning quantifiable virological titer and clinical measurements, and measures of the relative frequency of detection of these virological and clinical measurements (as indicated as _num, giving a number from 0 to 3 out of 3 inocu lated ferrets in the group). Fig. 6 specifies which variables were included in each tested iteration of the GLM shown, with blue squares indicating the model included this variable, and tan squares indicating that the variable was not included. The top performing GLM accounts for 100% of the variation in the response variable (R-Squared) (Fig. 6; Table S5). This model incorporates a mix of quantifiable clinical [mean maximum weight loss (wt_loss), mean maximum rise in body temperature (temp)] and frequency detection metrics [number of animals with temperature change (temp_num), number of animals with virus detected in trachea (Tr_num)] as predictor variables. Isolating single variables reveals a notable impact of wt_loss (R-Squared = 0.51) and temp (0.44), and a comparatively lower influence of temp_num (0.22) and Tr_num (0.20). When we examine the model with only clinical signs (wt_loss and temp, 0.62), or by removing Tr_num only (0.72) or temp_num only (0.62), we see some reduction in explanatory power, but most variability is still explained by wt_loss and temp, further emphasizing their impact. The Bayesian GLM results underscore the significance of these findings, S3;
for the highest-performing model, features were wt_loss (mean maximum weight loss), temp_num and weight_loss_num (number of ferrets with detectable temperature rise or mean maximum weight loss, respectively), and Tr_num (number of ferrets with virus detected in trachea). Full scope of all model metrics for models are reported in Table S5. indicating a 100% probability of significance and a large effect (99.98%) for temp, followed by temp_num (99.98%, 99.98%), wt_loss (99.92%, 77.5%), and Tr_num (97.98%, 95.95%). Interestingly, while absence of clinical signs (clinical_none) was not part of the final models, a GLM employing only this parameter yielded reasonable explanatory power (0.37), showcasing the pivotal role of clinical signs in assessing CVV attenuation (Table S5) and supporting inclusion of these parameters when performing studies of this nature. Taken together, these results indicate that clinical metrics, specifically aggregates of frequency of detection of clinical signs and not the specific experimental values themselves, were consistently associated with the highest predictive utility for CVV attenuation among all variables captured during CVV safety testing.
## DISCUSSION
IAV CVV development represents a critical component of pandemic preparedness, that has garnered support domestically and internationally (6), by facilitating rapid genera tion of high-yield vaccine preparations for use when a novel IAV becomes capable of sustained transmission in the human population. According to international zoonotic CVV manufacturing recommendations (6), CVV attenuation must be demonstrated in the ferret model, as defined by comparison with WT viruses or within pathogenicity standards, for instance, replication should be within a predefined range and restricted to the respiratory tract with an absence of extrapulmonary spread (7,15). The data presented in this analysis supports that CVVs are consistently attenuated relative to WT strains in the ferret model. However, understanding the relative degree of attenu ation achieved across a panoply of clinical and virological parameters, and how this attenuation may vary depending on the heterogeneity of IAV subtypes is key to moving forward in establishing standard thresholds for virulence in CVV safety testing in the future and thereby eliminating the need for a paired parental strain for comparison. Data science approaches permit a robust way to achieve these outcomes. We employed statistical and logistic regression approaches to identify critical variables associated with virus attenuation consistently present across a diversity of CVVs, ultimately identifying the frequency of clinical parameters (notably reductions in peak temperature and peak weight loss) as valuable metrics of virus attenuation independent of WT strain heteroge neity.
Prior studies have reported CVV attenuation phenotypes in ferrets with both seasonal and novel IAV (13,15). However, while these studies support that IAV can lead to varied degrees of attenuation of CVVs, none of them analyzed a data set of this size, inclusive of many IAV subtypes, lineages, and clades. Furthermore, as the diversity of studies captured here have high uniformity across laboratory facilities where the work was conducted, consistent protocols under which the experiments were performed, and sourcing of ferrets from a consistent vendor, the overall consistency across parameters which can vary between institutions (39) makes this data set unique in the field. The use of linear regression to identify specific in vivo correlates with phenotypic outcomes has been used in the past (40)(41)(42), but not previously in the context of CVV characterization.
This study collectively supports that CVVs exhibit attenuated clinical signs and virological titers compared with well-matched WT strains in the ferret model. The variation of attenuation observed is not surprising given the extensive heterogeneity of strains present in the data set; the few instances where attenuation of virological parameters was not detected is likely, in part, reflective of the inherent variances of the outbred animal model, minor variances in collection of specimens for titration, differences in the timing of collection, and other considerations. The restriction of viruses within a single HA subtype (H5, see Table S2) when analyzed by GLM resulted in increased statistical significance across many of the parameters assessed. Subtype-spe cific stratification was not possible for other subtypes in the data set due to limited sample sizes. While the low pathogenicity of CVVs in ferrets is, in part, attributed to the presence of PR8 virus backbone sequences, these viruses are nonetheless capable of replication in ferrets. The PR8 virus primarily replicates in the upper respiratory tract of ferrets with limited systemic dissemination (12,38). This feature, and its ability to replicate to high titer in embryonated eggs, makes it a suitable backbone for CVVs. However, alternative virus backbones should not be ruled out.
Distilling in vivo-generated data to discrete values necessary to conduct statisti cal modeling can be a challenge (43). As such, analyses were conducted with both quantifiable measures (percent weight loss and temperature rise, both of which were normalized prior to analysis (37), or infectious virus titer) and frequency of detection assessments (number of ferrets that displayed the parameter out of the total per group). It is interesting that frequency measures, and not quantifiable values, were often the most strongly associated with statistically meaningful outcomes when assessing attenuation of CVVs relative to WT strains. Group sizes were necessarily small for these experiments though appropriate for the conclusions drawn (44); it is expected that these frequency assessments would still maintain high value if group sizes were increased. Future studies assessing the potential utility of summary metrics not included in this work, such as area under the curve, to capture overall viral shedding in NW specimens during the acute phase of infection, would be of interest.
The current study has limitations of note. The analyses conducted here were restricted to one institution; similar comparative aggregate work would be important to perform at other institutes conducting this work to ensure the metrics most associated with attenuated phenotypes in CVV are independent of experimental inoculation or laboratoryspecific protocols. For clinical parameters, we only examined peak values and did not look at the frequency or duration of detection during the acute phase of infection, nor did we evaluate CVV transmission. There were selected instances (Table 1) where the WT virus used for comparison differed from the CVV (primarily by differ ences in lineages within the same clade, or different donor strains within the same lineage/clade); these decisions were made to adhere to the 3 R's (reduction, refinement, replacement) governing animal research (45) so that additional ferrets were not used solely for attenuation comparison purposes when possible. Both cell-based (PFU) and egg-based (EID 50 ) titrations were used to quantify infectious virus among wild-type and CVV viruses in this study due to strain-and subtypespecific replication capacities. Statistical analyses to assess relative degree of attenuation in viral titer between WT and CVV strains were only conducted for pairs for which the same titration matrix was used across the pair, to eschew any matrixspecific confounding (46). The majority of ferret inoculations were conducted with 10 6 infectious units with few exceptions (footnotes in Table 1); however, all ferrets were inoculated with a consistently high dose of virus permitting data aggregation across this parameter (47). Although gross pathology scores are determined during pathogenicity tests for comparison to WT (7), we did not include them in our analyses due to interpretation variability, descriptive nature, and lack of standardization. Instead, we focused on objective metrics that could be quantified with laboratory tests that provide reliable and replicable data.
Our study supports the findings of Chen et al., which reviewed over 15 years of chicken pathogenicity tests on H5 and H7 CVVs, showing consistent attenuation in this species (48). Although USDA still requires select agent exclusion of H5 and H7 CVVs, chicken lethality testing as an exclusion method was removed by the Division of Agricultural Select Agents and Toxins (DASAT) in 2018, as long as trypsin dependency is demonstrated and genomic sequence confirms consistency with LPAI, thereby reducing the testing timeline for rapid vaccine responses. However, pathogenicity testing in ferrets is still a WHO recommendation for CVVs derived from both high and low pathogenicity viruses (7). The high increase in the number of zoonotic influenza viruses with pandemic potential over the past decade has led to a rise in CVV production and WT virus testing, drawing renewed attention toward the development of standards for ferret pathogenic ity tests to streamline CVV evaluations so that vaccines can proceed in a timelier manner while drastically reducing the number of ferrets required for testing. Our analysis of cumulative data from pathogenicity tests conducted in ferrets has confirmed that these viruses are attenuated in these animals. This finding could potentially reduce the need for WT virus testing and enable more tailored biosafety assessments of CVVs in specific situations.
In vivo experimentation represents a cornerstone of biomedical research, and it is not anticipated that the use of ferrets to support CVV development will be eliminated entirely. Considering the unpredictable nature of IAV, it is likely that future, emerging influenza strains chosen for CVV development will include viruses with alterations in receptor binding specificity or other changes related to the virus backbone or adaptation markers that might modify the degree of attenuation observed. Ferret safety testing will undoubtedly be necessary in these scenarios. By understanding which parameters are most substantially and consistently attenuated in CVVs generated under standard protocols, these analyses represent a critical step in developing attenuation standard thresholds that may be consistently applied to IAV CVV testing, thereby improving the timeliness and efficiency of this critical component of pandemic preparedness.
## MATERIALS AND METHODS
## WT and CVV development
WT IAV used for comparison analyses against their corresponding CVVs are listed in Table 1. Virus stocks were propagated in the allantoic cavities of 10-11 day old embryonated chicken eggs at 35-37°C for 24-48 h, minimizing the number of passages in order to avoid deleterious mutations (18,24). Pooled allantoic fluid was clarified by centrifugation and aliquots were exclusivity tested by real-time reverse transcription (RT)-PCR to rule out the presence of other subtypes of influenza virus and stored at -80°C until use. Viral infectivity was determined by the EID 50 end point inoculation into embryonated eggs or by plaque forming units (PFU) (49). The references for data obtained from outside laboratories are provided in Table 1.
The generation of reassortant viruses was performed according to guidance for development of vaccine reference viruses (7) and described in (15). All studies performed at CDC with live virus were conducted under biosafety level 3 containment, including enhancements required by the U.S. Department of Agriculture and the Federal Select Agent Program (50).
## Ferret pathotyping experiments
Ferret experiments were performed under the guidance of the Centers for Disease Control and Prevention's Institutional Animal Care and Use Committee in an AAALAC International-accredited animal facility. Adult male Fitch ferrets, 6-12 months of age (Triple F Farms, Sayre, PA) and serologically negative for currently circulating influenza viruses were used in this study. Ferrets were intranasally (i.n.) inoculated with 10 6 PFU or EID 50 of virus diluted in PBS in a 1 mL volume unless otherwise specified in Table 1. The details for experiments using WT viruses can be found in references provided in Table 1. Studies using CVVs were performed according to established protocols and guidelines (15,24,51). For each CVV, three inoculated ferrets were monitored for 14 days for clinical signs of infection (including weight loss, fever, lethargy, sneezing, nasal discharge, and neurological dysfunction); any ferret that lost >25% of preinoculation body weight or exhibited signs of neurological involvement was humanely euthanized. Body temper ature was measured once daily, most often in the mornings, using subcutaneous implanted transponders (BMDS). NW specimens were collected every-other-day for 7 days for determination of virus titer. Three additional CVV-inoculated ferrets were euthanized on day 3 p.i. for tissue collection and determination of virus spread (7). All viral titers are reported as log 10 infectious units/mL or g. The titration limit of detection was 10 1.5 EID 50 or 10 PFU.
## References
1. Uyeki, Katz, Jernigan (2017) "Novel influenza A viruses and pandemic threats" *Lancet*
2. Cox, Trock, Burke (2014) "Pandemic preparedness and the influenza risk assessment tool (IRAT)" *Curr Top Microbiol Immunol*
3. Yamaji, Zhang, Kamata et al. (2024) "Pandemic risk characterisation of zoonotic influenza a viruses using the tool for influenza pandemic risk assessment (TIPRA)" *Lancet Microbe*
4. Moore, Ostrowsky, Kraigsley et al. "Influenza Vaccines R, Taskforce DR. 2021. A research and development (R&D) roadmap for influenza vaccines: Looking toward the future" *Vaccine (Auckl)*
5. Gip (2019) "Global influenza strategy 2019-2030"
6. (2005) "WHO guidance on development of influenza vaccine reference viruses by reverse genetics"
7. (2019) "Annex 3: guidelines for the safe development and production of vaccines to human pandemic influenza viruses and influenza viruses with pandemic potential. Guidelines for the safe development and production of vaccines to human pandemic influenza viruses and influenza viruses with pandemic potential"
8. Belser, Katz, Tumpey (2011) "The ferret as a model organism to study influenza A virus infection" *Dis Model Mech*
9. Focosi, Maggi (2024) "Avian Influenza virus A(H5Nx) and prepandemic candidate vaccines: state of the art" *Int J Mol Sci*
10. Naguib, Verhagen, Mostafa et al. (2019) "Global patterns of avian influenza A (H7): virus evolution and zoonotic threats" *FEMS Microbiol Rev*
11. Nuñez, Ross (2019) "A review of H5Nx avian influenza viruses" *Ther Adv Vaccines Immunother*
12. Dong, Matsuoka, Maines et al. (2009) "Development of a new candidate H5N1 avian influenza virus for pre-pandemic vaccine production" *Influenza Other Respir Viruses*
13. Robertson, Nicolson, Harvey et al. (2011) "The development of vaccine viruses against pandemic A(H1N1) influenza" *Vaccine (Auckl)*
14. Webby, Perez, Coleman et al. (2004) "Responsiveness to a pandemic alert: use of reverse genetics for rapid development of influenza vaccines" *Lancet*
15. Belser, Johnson, Pulit-Penaloza et al. (2017) "Pathogenicity testing of influenza candidate vaccine viruses in the ferret model" *Virology (Auckl)*
16. (2018) "Informal consultation on WHO biosafety risk assessment and guidelines for the production and quality control of novel human influenza candidate vaccine viruses and pandemic vaccines"
17. Jadhao, Achenbach, Swayne et al. (2008) "Development of Eurasian H7N7/PR8 high growth reassortant virus for clinical evaluation as an inactivated pandemic influenza vaccine" *Vaccine (Auckl)*
18. Maines, Jayaraman, Belser et al. (2009) "Transmission and pathogen esis of swine-origin 2009 A(H1N1) influenza viruses in ferrets and mice" *Science*
19. Pulit-Penaloza, Belser, Tumpey et al. (2015) "Mammalian pathogenicity and transmissibility of a reassortant Eurasian avian-like A(H1N1v) influenza virus associated with human infection in China" *Virology (Auckl)*
20. Pulit-Penaloza, Jones, Sun et al. (2025) *Barnes J, Full-Length Text Journal of Virology*
21. Tumpey, Wentworth, Maines et al. (2018) "Antigenically diverse swine origin H1N1 variant influenza viruses exhibit differential ferret pathogenesis and transmission phenotypes" *J Virol*
22. Pappas, Yang, Carney et al. (2015) "Assessment of transmission, pathogenesis and adaptation of H2 subtype influenza viruses in ferrets" *Virology (Auckl)*
23. Sun, Belser, Pulit-Penaloza et al. (2024) "Pathogenesis and transmission assessment of 3 swine-origin influenza a(h3n2) viruses with zoonotic risk to humans isolated in the United States, 2017-2020" *J Infect Dis*
24. Chen, Peng, Zheng et al. (2023) "Characteri zation of an emergent chicken H3N8 influenza virus in Southern China: a potential threat to public health" *J Virol*
25. Maines, Lu, Erb et al. (2005) "Avian Influenza (H5N1) viruses isolated from humans in Asia in 2004 exhibit increased virulence in mammals" *J Virol*
26. Maines, Chen, Matsuoka et al. (2006) "Lack of transmission of H5N1 avianhuman reassortant influenza viruses in a ferret model" *Proc Natl Acad Sci*
27. Pearce, Pappas, Gustin et al. (2017) "Enhanced virulence of clade 2.3.2.1 highly pathogenic avian influenza A H5N1 viruses in ferrets" *Virology (Auckl)*
28. Pulit-Penaloza, Brock, Pappas et al. (2020) "Characterization of highly pathogenic avian influenza H5Nx viruses in the ferret model" *Sci Rep*
29. Pulit-Penaloza, Belser, Brock et al. (2022) "Pathogenesis and transmissibility of north american highly pathogenic avian influenza A(H5N1) virus in ferrets" *Emerg Infect Dis*
30. Pulit-Penaloza, Sun, Creager et al. (2015) "Pathogenesis and transmission of novel highly pathogenic avian influenza H5N2 and H5N8 viruses in ferrets and mice" *J Virol*
31. Belser, Lu, Maines et al. (2007) "Pathogenesis of avian influenza (H7) virus infection in mice and ferrets: enhanced virulence of Eurasian H7N7 viruses isolated from humans" *J Virol*
32. Belser, Gustin, Pearce et al. (2013) "Pathogenesis and transmission of avian influenza A (H7N9) virus in ferrets and mice"
33. Ridenour, Hossain, Mateu-Petit et al. (2015) "Development of influenza A(H7N9) candidate vaccine viruses with improved hemagglutinin antigen yield in eggs" *Influenza Other Respir Viruses*
34. Sun, Belser, Pappas et al. (2019) "Risk assessment of fifthwave H7N9 Influenza A viruses in mammalian models"
35. Belser, Sun, Brock et al. (2020) "Genetically and antigenically divergent influenza A(H9N2) viruses exhibit differential replication and transmission phenotypes in mammalian models" *J Virol*
36. Van Den Brand, Wohlsein, Herfst et al. (2016) "Influenza A (H10N7) virus causes respiratory tract disease in harbor seals and ferrets" *PLoS One*
37. Reuman, Keely, Schiff (1989) "Assessment of signs of influenza illness in the ferret model" *J Virol Methods*
38. Belser, Kieran, Mitchell et al. (2024) "Key considerations to improve the normalization, interpretation and reproducibility of morbidity data in mammalian models of viral disease" *Dis Model Mech*
39. Liu, Wood, Ellis et al. (2003) "Preparation of a standardized, efficacious agricultural H5N3 vaccine by reverse genetics" *Virology (Auckl)*
40. Belser, Lau, Barclay et al. (2022) "Working group on the standardization of the ferret model for influenza risk a. 2022. Robustness of the ferret model for influenza risk assessment studies: a cross-laboratory exercise" *MBio*
41. Stark, Long, Ortiz et al. (2013) "Clinical profiles associated with influenza disease in the ferret model" *PLoS One*
42. Van Den Brand, Stittelaar, Van Amerongen et al. (2012) "Comparison of temporal and spatial dynamics of seasonal H3N2, pandemic H1N1 and highly pathogenic avian influenza H5N1 virus infections in ferrets" *PLoS One*
43. Kieran, Sun, Maines et al. (2024) "Optimal thresholds and key parameters for predicting influenza A virus transmission events in ferrets" *Npj Viruses*
44. Kieran, Maines, Belser (2024) "Data alchemy, from lab to insight: transforming in vivo experiments into data science gold" *PLoS Pathog*
45. Belser, Maines, Katz et al. (2013) "Considerations regarding appropriate sample size for conducting ferret transmission experiments" *Future Microbiol*
46. Balls, Straughan (1996) "The three Rs of russell & burch and the testing of biological products" *Dev Biol Stand*
47. Creager, Kieran, Zeng et al. (2023) "Utility of human in vitro data in risk assessments of influenza a virus using the ferret model" *J Virol*
48. Kieran, Maines, Belser (2025) "Eleven quick tips to unlock the power of in vivo data science" *PLoS Comput Biol*
49. Chen, Donis, Suarez et al. (2020) "Biosafety risk assessment for production of candidate vaccine viruses to protect humans from zoonotic highly pathogenic avian influenza viruses" *Influenza Other Respir Viruses*
50. Reed, Muench (1938) "A simple method of estimating fifty per cent endpoints" *Am J Epidemiol*
51. (2020) *Biosafety in Microbiology and Biomedical Laboratories. 6th ed. U.S. Government Printing Office*
52. (2007) "Annex 5: WHO biosafety risk assessment and guidelines for the production and quality control of human influenza pandemic vaccines"
53. Friedman, Hastie, Tibshirani (2010) "Regularization paths for generalized linear models via coordinate descent" *J Stat Soft*
54. Goodrich, Gabry, Ali et al. (2024) "Rstanarm: bayesian applied regression modeling via stan"
55. (2025) *Full-Length Text Journal of Virology*
56. Makowski, Ben-Shachar, Patil et al. (2020) "Automated results reporting as a practical tool to improve reproducibility and methodological best practices adoption" *CRAN*
57. Hadley, Averick, Bryan et al. (2019) "Welcome to the tidyverse"
58. Wickham, Sievert (2009) "ggplot2: elegant graphics for data analysis"
59. Kassambara (2018) "ggpubr:'ggplot2'based publication ready plots"
60. Iannone, Cheng, Schloerke et al. (2022) "Gt: Easily create presentation-ready display tables"
61. Garnier, Ross, Rudis et al. (2021) "Rvision-colorblind-friendly color maps for R" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12829795&blobtype=pdf | # Intra-patient neuraminidase mutations in avian H5N1 influenza virus reduce sialidase activity to complement weaker hemagglutinin binding and facilitate human infection
Yohei Watanabe, Madiha Ibrahim, Yasuha Arai, Daisuke Kuroda, Emad Elgendy, Shin-Ichi Nakakita, Yohei Takeda, Nghia Vuong, Bui, Takao Ono, Shota Ushiba, Tomo Daidoji, Nongluk Sriwilaijaroen, Haruko Ogawa, Kazuhiko Matsumoto, Yasuo Suzuki, Takaaki Nakaya
## Abstract
Clade 2.2 H5N1 influenza viruses have caused an unusually high number of human infections, providing a unique opportunity to investigate early molecular steps associated with host adaptation. Although most work has focused on hemagglutinin (HA), the contribution of neuraminidase (NA) to these early adaptive events has remained unclear. By analyzing publicly available sequences from clade 2.2-infected patients, we identified 20 NA mutations and compared their phenotypes to 20 mutations acquired during diversification in primary human airway cells under drug-free conditions. Most patient-derived NA mutations resulted in modest reductions in sialidase activity, keeping activity within a functional range that supported improved replication in α2,6 sialylglycan (α2,6 Sia)-dominant environments, whereas excessive reduction impaired fitness. Notably, the phenotypes of culture-selected and patient-derived mutations were highly concordant, suggesting that these NA changes arose through natural selection rather than antiviral pressure. Re-analysis of patient sequences further revealed that many adaptive NA mutations co-occur with HA mutations that confer only weak, partial α2,6 Sia binding. Using reverse genetics, we found that such naturally occurring HA/NA mutation pairs acted cooperatively in a receptor-context-dependent manner to support α2,6-associated replication relative to HA-only mutants, placing these variants within a constrained "early-adaptation space" characterized by limited α2,6 engagement and moderately reduced NA activity. Together, these findings indicate that early human adaptation of clade 2.2 H5N1 involves not only HA and PB2, but also incremental, cooperative tuning of NA function. Monitoring coordinated HA-NA evolution may therefore improve risk assessment frameworks for zoonotic influenza viruses poised at early stages of human host adaptation.
## Introduction
The highly pathogenic avian influenza virus subtype H5N1 is currently prevalent worldwide, posing severe burdens on human public health systems. The H5N1 virus appeared in southern China in 1996, causing outbreaks in bird species and sporadic human infections, mainly in Egypt and Southeast Asia after expanding across the Eurasian continent via bird migration since 2005 [1,2]. In particular, H5N1 clade 2.2 was characteristic of single-clade epidemics in Egypt from 2006 to 2017 [3,4] causing a total of 359 human infections, the most for any country according to WHO (https://www.who.int). In contrast, in Southeast Asian countries such as Vietnam and Indonesia, multiple clades (e.g., 1.1, 2.3, and 7) caused a total of 412 cases of human infection [5,6]. Epidemics of these H5N1 clades have become sporadic since 2017 [7,8], coinciding with the emergence of other avian influenza virus subtypes such as H9N2 [9] and H5N8 [10]. However, the descendant H5N1 clade 2.3.4.4b emerged in Asia beginning in 2020, and has spread globally [11,12]. This clade is currently the most devastating zoonotic influenza virus and has caused numerous outbreaks in birds and several human infections worldwide.
Influenza A viruses express two surface proteins, hemagglutinin (HA) and neuraminidase (NA) [13]. HA functions to attach virions to host cells by binding to sialylglycans (Sia) on the target cell surface, whereas NA acts to dissociate virions from this receptor via sialidase activity. NA is thought to be necessary for the release of progeny viruses from infected cells, but recent studies noted that NA also influences initial binding and rolling of virus particles over the cell surface [14,15].
HA distinguishes between sialic acid-galactose linkages at the end of Sia [13]. Avian influenza viruses recognize the α2,3 linkage (α2,3 Sia), which is abundantly expressed in the avian intestine, whereas seasonal influenza viruses strongly bind to the α2,6 linkage (α2,6 Sia), strongly expressed in the human upper airway. For influenza viruses to cause a pandemic, receptor tropism has to change from avian-type to human-type [16]. To date, numerous H5N1 HA mutations that increase α2,6 Sia binding affinity have been reported [17][18][19][20][21][22], but this binding was still much weaker than that of seasonal influenza viruses [22][23][24][25]. Thus, avian influenza viruses would need to accumulate multiple HA mutations over time to undergo gradual fine tuning to fully switch receptor tropisms, as illustrated by the 1968 H3N2 pandemic [26].
A balance between the opposing functions of HA and NA is vital for successful viral infectivity [14,[27][28][29][30][31][32]. When avian influenza viruses with a typical bird-like tropism reach the upper airway in humans, they are exposed to an environment in which they would need to infect host cells through the inherently weak α2,6 Sia binding affinity. However, successful infection would need a change of the viral HA-NA balance for Sia. Thus, we hypothesized that adaptive mutations both in HA and NA that shift their functional balance would occur during avian influenza virus infection of humans. The human-adaptive mutations of influenza viruses have so far been analyzed mainly in HA and polymerase complexes [18,19,21,33,34]. Conversely, NA mutations have been analyzed exclusively for escape from NA inhibitors (NAIs) [35,36], with one recent study reporting adaptive NA mutations naturally selected in avian influenza virus-infected patients [37].
In Egypt, single clade 2.2 epidemics have caused human infections accumulating over time. This allowed us to search for adaptive mutations that may occur rarely in patients but would be detectable with high accuracy due to the prevalence of homogeneous single-clade viral gene sequences. This was not possible with cases from another hot-spot of H5N1 human infection, Southeast Asia, where multiple clades have circulated [5,6].
We have previously characterized adaptive HA mutations in clade 2.2 virus-infected patients by database searches [18]. Here, we aimed to characterize the NA mutations that were selected for in human infections. We identified a total of 20 adaptive NA mutations in clade 2.2-infected patients. In parallel, we identified 20 intra-cellular NA mutations occurring as a result of clade 2.2 diversification in primary human cells cultures in vitro. We then characterized the viruses with adaptive NA mutations by carefully comparing the phenotypic effects of the two sets of NA mutations. Our data reveal a mechanism involving complementary NA mutations in H5N1 influenza virus variants in patients which compensate for the weaker HA Sia binding in the human airway. These might also be applicable to other influenza virus adaptations to humans.
## Results
## Identification of human-adaptive NA mutations acquired by the H5N1 virus in vivo and in vitro
We first investigated whether the H5N1 virus NA gene can potentially acquire adaptive mutations during virus replication in humans. To this end, we analyzed clade 2.2 genetic diversification during single infection in primary human airway epithelial (HAE) cells and chicken embryo fibroblasts (CEFs) as a control (Fig 1A). The two cell strains were infected with recombinant clade 2.2 virus with a high degree of genetic homogeneity. The progeny viruses were analyzed 96 hours post-infection (hpi), as reported previously [19]. No amino acid mutations were observed in the internal viral protein NP in these cells (Fig 1B). On the other hand, approximately the same number of mutations were detected in both the HA and NA in HAE cells, but very few mutations were observed in CEFs, indicating selective pressure specific to human cells. These results suggested that human-adaptive mutations occur in the NA as frequently as in the HA during clade 2.2 replication in human airway cells.
We next sought to identify the NA mutations that were selected for in H5N1 virus-infected patients. To this end, we conducted a database search for the clade 2.2 virus gene sequence registered in the GISAID database (https://gisaid.org). This yielded 94 NA gene sequences derived from Egyptian patients (2006-2012) and 343 NA gene sequences derived from birds from the same region over the same period. We then searched for NA amino acid mutations characteristic of human-derived viruses by comparing each of the gene sequences with the consensus sequence determined by aligning all NA sequences. As a result, we identified 11 individual NA mutations in the infected patients (see Fig 2A for a flow-chart of identification and characterization of intra-patient NA mutations). These single NA mutations were classified into two categories. The first comprised NA mutations that were detected more frequently in human viruses than avian viruses (6 mutations), which had presumably emerged in the field and were subsequently transmitted from birds to humans with high efficacy. The second comprised NA mutations that were detected only in the human-derived viruses, but not in the avian viruses (5 mutations), which presumably emerged within the infected individuals. We additionally searched for combinations of these single NA mutations in all infected patients, and found 9 instances where multiple mutations co-occurred. A/ duck/Egypt/D1Br/2007 (EG/D1), one of the parental clade 2.2 strains [20], was used as the reverse-genetics (RG) backbone to generate recombinant viruses carrying intra-patient NA mutations. In total, 20 recombinant EG/D1 viruses were constructed: 11 bearing single NA mutations detected in patients, and 9 bearing naturally occurring combinations of these mutations. All viruses were successfully rescued by RG.
Clade 2.2 virus-infected patients had often been treated with NAIs in Egypt [4]. This raised the possibility that the identified intra-patient NA mutations potentially included NAI escape mutations, although their overall occurrence in the H5N1 virus is lower than in seasonal influenza viruses [35]. We therefore characterized the intra-cellular NA mutations Thirty-four single NA mutations were detected from 4 independent infections, with the mutations all existing as minor single variants at a frequency of 3%. We then generated recombinant EG/D1 viruses having each one of the intra-cellular NA mutations and succeeded in creating 20 replicative mutants. Of note was the N295S mutation, which was detected in both intra-patient and intra-cellular NA mutations. In total, 39 adaptive NA mutant viruses were analyzed in further studies. . Importantly, no serial passaging or iterative selection procedures were performed, ensuring that mutations reflected naturally arising diversity within a single infection cycle rather than laboratory-driven adaptation. Across four independent experiments, 34 NA mutations were detected as ≥3% minor variants. Numbers in parentheses indicate detection frequency among the four trials. Of these, 20 NA mutations (underlined) were successfully incorporated into recombinant EG/D1 viruses for evaluation of sialidase activity, Sia-dependent replication, NA protein expression, and susceptibility to NA inhibitors. Based on phenotypic characteristics, 8 mutants (5 intra-patient, 3 intra-cellular; N295S found in both groups) were selected for α2,6 Sia binding analysis by biolayer interferometry. The image was generated using Open AI (ChatGPT) and edited manually (licensed under CC BY 4.0). https://doi.org/10.1371/journal.ppat.1013863.g002
## Adaptive NA mutant viruses have slightly reduced sialidase activity
Because influenza viruses attach, enter, and disseminate as intact particles, we first quantified virion-level ("net") sialidase activity after normalizing purified virions by infectious particle number (focus-forming units; FFU). This approach measures the overall NA function per infectious virion-an essential determinant of HA-NA functional balance during entry and release-rather than the catalytic efficiency per NA molecule. We therefore considered FFU-normalized NA activity as the most biologically relevant readout to capture how each NA mutation alters the net receptor-dissociating capacity of virus particles. EG/D1 HA has an inherently strong binding affinity for α2,3 Sia, but only a weak affinity for α2,6 Sia [20], indicative of a typical avian tropism. We also included H275Y as a reference NA mutation, which significantly reduced H5N1 virus sialidase activity and serves as a representative example of resistance against oseltamivir and peramivir [28].
In addition to this virion-level measurement, we also assessed "NA amount-normalized (intrinsic) NA activity," defined as sialidase activity normalized by NA protein abundance determined by Western blotting. This complementary metric reflects the intrinsic catalytic efficiency of NA independently of its incorporation level into virions. Throughout this study, we distinguish between "FFU-normalized (virion-level, net) NA activity" and "NA amount-normalized (intrinsic) NA activity," using each parameter for the appropriate biological question (results shown in S4 and S5 Figs).
Most intra-patient-derived NA mutant viruses exerted reduced FFU-normalized (virion-level, net) NA activity relative to the NA wild-type (wt) (Fig 3A). Among them, the N295S and the N295/L224M/S339F viruses exerted the weakest FFU-normalized NA activity, at about one-fifth of the activity of the NA-wt. Nevertheless, the sialidase activity of all the NA mutants was still >10-fold higher than that of the H275Y virus. Similarly, most intra-cellular-derived NA mutants had reduced FFU-normalized NA activity relative to NA-wt (Fig 3B). A138V and D199G viruses had reduced FFU-normalized NA activity to an extent similar to the N295S virus, but again, all viruses still had activity >10-fold that of the H275Y virus. These results showed that both the intra-patient and intra-cellular NA mutants had modestly reduced FFU-normalized NA activity.
## Adaptive NA mutations enhance virus yields in an α2,6 Sia-dependent manner
We next evaluated Sia-dependent propagation of the NA mutant viruses using four different cell strains with distinct Sia expression patterns. Because the interpretation of Sia-dependent replication critically depends on the relative availability of α2,3-and α2,6-Sias on each cell line, we first confirmed their Sia expression profiles by fluorescent lectin staining (S1 Fig) . DF-1 cells showed strong Maackia amurensis lectin I (MAL-I) staining with negligible Sambucus nigra lectin (SNA) signal, indicating predominant α2,3 Sia expression. MDCK cells expressed both α2,3 and α2,6 Sias but remained α2,3-dominant. In contrast, 1A5 cells displayed robust α2,6 Sia expression with detectable, but weaker α2,3 Sia signals. These expression patterns were consistent with previous reports [38][39][40][41]. Importantly, α2,3-sialidase-treated 1A5 cells exhibited nearly complete loss of α2,3 Sia immediately after treatment and remained undetectable for at least 13 h. The α2,3 sialidase-treated 1A5 cells were thus biased to specifically express α2,6 Sia, which is not usually the case because α2,3 and α2,6 Sias are normally both expressed in tissues and cells at various different ratios [42]. We infected these cells with NA mutant viruses at low multiplicities of infection (MOI), and monitored their propagation kinetics by measuring FFU titers (S2 and S3 Figs).
Yields of NA mutant viruses at 12 hpi, which together with the 1-h adsorption period fall within the 13-h interval over which the α2,3-depleted state in sialidase-treated 1A5 cells was verified to persist, are shown in Fig 4 . In DF-1 cells, many of the intra-patient-derived NA mutants had lower virus yield than the wt virus, and no mutants with increased yields were observed (Fig 4A). The same phenomenon was observed in MDCK cells, although the decreases in viral yields were more subtle. In contrast, in 1A5 cells, many of the intra-patient NA mutants exhibited up to 4-fold increased viral yields relative to wt virus. Notably, these effects were more prominent in α2,3-sialidase-treated 1A5 cells, with virus yields reaching more than 10-fold above those of wt virus. The H275Y reference virus exhibited markedly reduced virus yields in all cell types. The effects of the intra-cellular NA mutations on virus yields across all four cell types mirrored those of the intra-patient NA mutations (Fig 4B). These results demonstrate that adaptive NA mutations enhanced virus yields in human airway cells in an α2,6 Sia-dependent manner.
## Slightly but not strongly reduced sialidase activity is correlated with improved NA mutant virus yields
Because the N295S, N295S/L224M/S339F, A138V and D199G viruses had noticeably lower sialidase activities and also relatively higher virus yields in human cells, we evaluated the relationship between these two factors in four types of cells. In DF-1 and MDCK cells, yields of the intra-patient and the intra-cellular NA mutants were correlated positively with their sialidase activity (Fig 5A and5B), and almost all NA mutants had reduced virus yields relative to the wt virus, depending quantitatively on the degree of reduction of sialidase activity. Data from the reference H275Y virus fitted the curve of the adaptive NA mutants. Conversely, in 1A5 cells, virus yields were negatively correlated with sialidase activity (Fig 5C ); almost all NA mutants had enhanced virus yields depending on the degree of reduction of sialidase activity. This relationship was more pronounced in α2,3 sialidase-treated 1A5 cells (Fig 5D). Data on the H275Y virus plotted outside the fitted curve for the adaptive NA mutants. However, all plots in general, including H275Y virus, revealed that NA mutant viruses with sialidase activity reduced to within a specific range (1/10 of wt virus) had maximized virus yields. Taken together, these results suggest that the adaptive NA mutant viruses slightly reduced sialidase activity to within an optimal range, thereby enhancing viral replication in cells predominantly expressing α2,6 Sia.
## Distinct classes of adaptive NA mutations differentially impact protein expression and intrinsic catalytic function
To further explore the mechanisms by which mutant NAs reduced "net" NA activity of virus particles, we evaluated the effects of adaptive NA mutations on NA protein expression. We quantified mutant NAs in plasmid-transfected cells, because previous studies had shown that the amounts of mutant NAs in the clade 2.2 virus virions reflected intracellular expression levels [37]. Consistent with this, our analysis demonstrated strong correlations between intracellular NA expression and NA levels incorporated into purified virions (R 2 = 0.7256 for intra-patient NA mutants and R 2 = 0.7863 for inter-cellular NA mutants) (S4 Fig) . As virion-based quantification exhibited substantially higher experimental variability, intracellular NA expression provided a more stable and reproducible surrogate for virion NA content. We therefore used NA expression levels in transfected cells as a surrogate measure of virion-associated NA levels to assess the effects of NA mutations on NA incorporation into virus particles. Western blotting showed that intra-patient mutant NAs were expressed in similar or decreased amounts relative to wt NA, and no increased expression was observed (Fig 6A and6B). Scatter plots showed a strong correlation between FFU-normalized (virion-level, net) NA activity and protein levels in the group of mutations presumably transmitted from birds to humans (Fig 6C, upper panel). In contrast, for the mutations presumably selected during replication in infected patients, FFU-normalized NA activity did not correlate with the level of expression (Fig 6C, lower panel).
Similar to these intra-patient mutant NAs, the intra-cellular mutant NAs were expressed at unchanged or decreased levels relative to wt NA (Fig 6D and6E). The amounts expressed also did not correlate with FFU-normalized NA activity (Fig 6F ), mirroring the pattern observed for NA mutations selected in patients. These results suggest that decreased sialidase activity in NA mutant viruses can be attributed to reductions in NA protein abundance and/or intrinsic enzymatic activity.
To further distinguish these possibilities, we additionally quantified NA amount-normalized (intrinsic) sialidase activity by normalizing virion sialidase activity to the NA protein abundance measured from purified virions (S5 Fig) . This analysis clarified that mutations with higher prevalence in human-derived viruses generally maintained intrinsic catalytic activity close to wt levels, indicating that their reduced virion-level activity primarily reflects decreased NA incorporation rather than impaired catalytic efficacy. In contrast, many mutations found exclusively in human-virus NAs or selected in human airway cells exhibited marked reductions in intrinsic catalytic activity, accompanied by modest decreases in NA protein abundance. Although both factors contribute to the overall reduction in virion-level sialidase activity, the magnitude of intrinsic activity loss was substantially greater than the change in protein amount, indicating that impaired enzymatic efficiency is the primary driver for these categories, with decreased NA abundance playing a secondary, reinforcing role. Collectively, these findings indicate that both reduced NA incorporation and reduced intrinsic activity shape NA adaptation, but that intrinsic catalytic impairment is disproportionately enriched among NA mutations selected in patients or human cells.
## Adaptive NA mutant viruses retain susceptibility to NAIs
Virus susceptibility to NAIs and sialidase activity are often altered concomitantly by introduction of NA mutations [28]. We therefore evaluated the susceptibility of NA mutant viruses to peramivir and laninamivir by a chemiluminescence NA inhibition assay using NA-XTD as the substrate. In accordance with the NA-XTD protocol and international NAI testing guidelines, virus input for each mutant was standardized by NA activity rather than by FFU; specifically, the dilution yielding a signal-to-noise ratio of 40 in the NA-XTD assay was used for IC 50 determination (see Methods). We also included the peramivir-resistant virus with mutation H275Y as a reference [43,44]. The IC 50 values of wt virus for peramivir and laninamivir were 0.35 nM and 0.062 nM, respectively (Fig 7 ), i.e., within the IC 50 range of many H5N1 viruses in the literature [44]. This confirmed that the wt virus was a typical H5N1 virus with high susceptibility to both NAIs.
Among the intra-patient NA mutations, N295S and N295S/L224M/S339F moderately increased the IC 50 for peramivir by 7.7-and 8.1-fold, respectively, and for laninamivir by 3.9-fold, but all the mutant viruses remained highly susceptible to both drugs (Fig 7A and7B). Among intra-cellular NA mutations, A138V and D199G increased IC 50 for peramivir by 5-and 4.2-fold, respectively, and for laninamivir by 3.5-and 3.4-fold, respectively, but here also all the mutant viruses remained highly susceptible to both drugs. These results suggest that the adaptive NA mutations did not lead to marked phenotypic changes regarding NAIs and the viruses retained inherent drug-susceptibility.
## Adaptive NA mutant viruses exhibit increased binding to α2,6 Sia
The balance between the opposing functions of HA-binding and NA-dissociation is important for influenza virus attachment to host cells [14,28]. We thus evaluated the effect of decreased sialidase activity of the adaptive NA mutants on virus binding to α2,6 Sia. To this end, we performed biolayer interferometry (BLI) using α2,6 Sia and purified mutant viruses as the ligand and analyte, respectively (Fig 8A). Because the assays were conducted in PBS (+) containing cations necessary for NA activity, we analyzed the binding profiles of virus particles as the sum of HA-binding and NA-dissociation. Based on the phenotypic effects of the adaptive NA mutations (Figs 3-5), 5 intra-patient NA mutants (N295S, L224M/ These results indicate that the NA mutant virus had reduced sialidase activity to compensate for low α2,6 Sia binding activity, thereby conferring an increase in Sia binding of virus particles.
## Co-occurring HA and NA adaptive mutations cooperatively promote α2,6 Sia-mediated viral fitness
Because HA and NA act in a coordinated manner during viral entry, we next examined whether the adaptive NA mutations identified in this study cooperate with HA mutations previously shown to enhance α2,6 Sia binding. In our earlier work on Egyptian clade 2.2.1 H5N1 viruses isolated from infected patients [18], multiple HA mutations conferring increased α2,6 Sia preference were identified. We therefore re-analyzed all available Egyptian patient-derived H5N1 sequences to determine whether these HA-adaptive mutations co-occur with the NA-adaptive mutations characterized in the present study. This survey revealed that 56 of 94 clinical virus isolates encoded at least one HA-adaptive and one NA-adaptive mutation simultaneously (Fig 9A and S1 Table), and 35 distinct HA/NA mutation combinations were detected (S2 Table ).
To experimentally assess the functional consequences of such simultaneous mutations, we selected 18 representative HA/NA combinations in which each HA mutation increased α2,6 Sia binding and the corresponding NA mutation reduced sialidase activity. Using reverse genetics, we generated recombinant viruses carrying each HA/NA double mutation, as well as the corresponding 18 HA-only mutant viruses for direct comparison. We first analyzed the relationship between altered HA α2,6 Sia binding preference and NA sialidase activity among these 18 HA/NA mutants (Fig 9B). Although all HA mutations modestly increased α2,6 Sia specificity relative to the wt H5N1, their α2,6 binding remained substantially lower than that of a representative seasonal H3N2 strain (A/Japan/434/2003). Conversely, the NA-adaptive mutations conferred a mild but significant reduction in sialidase activity, contrasting sharply with the high NA activity of the H3N2 reference virus. A centered second-order polynomial model explained 89.7% of the variance (R 2 = 0.8972), with a significant positive quadratic term, indicating that NA activity generally increases in an upward-curving manner as HA α2,6-binding specificity becomes stronger. Notably, within the limited α2,6-binding range characteristic of the simultaneous HA/NA mutants detected in patients, several variants occupied a locally concave region in which NA activity was moderately decreased. These findings suggest that the HA/NA mutation pairs detected in patients represent very early-stage human-adaptation intermediates, positioned in a transitional zone before acquisition of the high α2,6-binding and high NA-activity profile typical of seasonal human influenza viruses.
Finally, we compared the viral yields of the 18 HA-only and 18 HA/NA simultaneous mutants in α2,3-sialidase-treated 1A5 cells under conditions where α2,3 Sia depletion persisted for the entire 13-h assay window, as in [18] together with one or more NA-adaptive mutations identified in this study. From these, 35 unique HA/NA mutation combinations were identified, and 18 representative HA/NA pairs were selected for reconstruction by reverse genetics. (B) Relationship between HA α2,6 variants actually selected in infected patients, adaptive NA mutations augment the replication advantage conferred by α2,6 Sia-directed HA mutations, supporting a coordinated HA-NA adaptation process during the early stages of human infection.
## Adaptive mutations that affect sialidase activity and their marginal energetic effects are located near the tetramer interface or active site
To better understand the context of these NA mutations within the protein structure, we mapped the identified adaptive mutations onto an NA structure (Fig 10). Among the seven sets of mutations that we characterized experimentally (Fig 8 ), six sets were located at the interfaces of the tetramer, and one mutation (N295S) was located near the active site. Notably, A138V, D199G, and N295S exhibited the most pronounced loss of sialidase activity (Fig 3). This is likely due to the specific locations of these mutations: A138 is well-buried (relative solvent accessible surface area = 0) and situated at the center of the tetramer. Although the energetic effect of the A138V mutation is nearly marginal (computed ΔΔG = 0.6), mutating the Ala to Val might affect the equilibrium between the open and closed conformations of the tetramer. D199 forms a salt bridge with K150; therefore, mutating D199G would result in the loss of this interaction, leading to destabilization of the tetramer (computed ΔΔG = 40.3). N295 is located near the active site. Although the energetic effect of N295S on the stability of the NA tetramer itself was also marginal (computed ΔΔG = 3.8), a mutation in this region would directly affect sialidase activity. Interestingly, the mutation that caused the greatest deterioration, H275Y, a reference mutation in our study, also had a marginal energetic effect on the stability of the NA tetramer (computed ΔΔG = 1.0). Together with the experimental characterization mentioned above, these results imply a complex interplay between the catalytic activity, stability, and binding capability of the NA tetramer. These properties appear to counterbalance each other upon the introduction of adaptive mutations.
## Discussion
To date, the H5N1 virus has acquired multiple human-adaptive mutations in several genes during its infection in the field or in infected patients [18,20,34,45,46]. Many studies have searched for adaptive mutations in HA and polymerase [19,21,33,34], whereas studies on NA have intensively focused on NAI escape mutants [35,36], with a few exceptions investigating adaptive NA mutations [37,47]. In this context, to the best of our knowledge, the present report is the first to systematically characterize adaptive NA mutants selected for in H5N1 virus-infected patients, by comparing their phenotypic traits with adaptive NA mutants selected for in human cells in vitro.
In this study, no NAIs were added to the culture medium during cell growth and clade 2.2 diversification (Fig 1). Nevertheless, the observed phenotypes of intra-cellular-derived NA mutants were all consistent with NA mutants isolated from patients (Figs 345678). This implies that the identified intra-patient NA mutations included no, or at least negligible, Sia binding preference and FFU-normalized (virion-level, net) NA activity among the 18 reconstructed HA/NA simultaneous mutants. HA α2,6-binding preference was determined from previously reported solid-phase direct binding data [18], using the ratio K A (α2,6SLN2)/K A (α2,3SLN1) as an index of α2,6 specificity. As an external reference for a fully human-adapted phenotype, the seasonal H3N2 virus A/Japan/434/2003 was included for comparison. NA activity was measured by chemiluminescent NA assay and normalized to NA-wt. A centered second-order polynomial model accounted for 89.7% of the variance (R² = 0.8972). Although NA activity generally increased with greater HA α2,6-binding preference, HA/NA combinations within the limited α2,6-binding range characteristic of early human-adaptation variants exhibited locally reduced NA activity. (C) Comparison of viral yields of HA-only and HA/NA simultaneous mutants in α2,3-sialidase-treated 1A5 cells (α2,6-only). Cells were infected at an MOI of 0.05, and virus yields were measured at 12 hpi-the early replication phase corresponding to 13 h after sialidase treatment, during which α2,3 Sia depletion was experimentally validated to persist. Viral yields are shown relative to wt (= 1). Across all 18 mutation pairs, HA/NA simultaneous mutants exhibited significantly higher viral yields than their matched HA-only mutants, demonstrating cooperative enhancement of α2,6 Sia-dependent replication. (D) Summary comparison of HA-only versus HA/NA simultaneous mutants. Aggregate analysis confirmed significantly increased replication of HA/NA simultaneous mutants under α2,6-dominant conditions. *P < 0.01. https://doi.org/10.1371/journal.ppat.1013863.g009 escape mutations with the effects indistinguishable from adaptive mutations. In fact, among the single NA mutations identified in patients (Q136H, N222H, I223T, I223V, N295S), only N295S minimally altered IC 50 values to NAIs. N295S was already known to moderately reduce the susceptibility of H5N1 and seasonal H1N1 and H3N2 viruses to NAIs [36]. Intriguingly, however, in clade 2.2-infected patients, N295S was detected in NA of viruses from clinical swabs before NAI administration [48]. This implies that the N295S mutation occurred in the infected individuals without escape-selective pressure from NAIs. These findings further suggest that the intra-patient NA mutations we identified here arose as a result of selective pressure specific for adaptation to the human host. This is consistent with our null hypothesis, and also provides an answer to the question why the N295S mutation occurred in clade 2.2-infected patients before NAI administration. Nevertheless, because most medical records of the therapies received by the clade 2.2-infected patients, such as detailed information on NAI administration, were with a few exceptions not available to us [49], we cannot completely exclude the possibility that the intra-patient NA mutations included potential NAI escape mutations. To completely exclude this possibility, further analyses will be required that integrate clinical data of each infected patient and relevant viral gene sequences.
The intra-patient and intra-cellular NA mutants were shown to moderately decrease FFU-normalized (virion-level, net) NA activity to within a narrow range (about 1/10 of the wt virus) (Fig 3), which was advantageous for virus yields in human cells (Fig 5 ) and also virus binding profiles to α2,6 Sia (Fig 8). This is in line with an in vitro study [47] showing that H5N1 viruses acquired adaptive NA-E119V mutations to moderately reduce sialidase activity during serial passage in primary human cells. Conversely, when sialidase activity was further reduced by >10-fold by the H275Y mutation, virus yield was greatly reduced relative to NA-wt, consistent with other reports showing H5N1 virus attenuation by this mutation [50]. Also, in the present study, there appeared to be an optimal range of reduced sialidase activity that resulted in maximal virus titers in human cells ( Fig 5). Taken together, these findings suggest that the adaptive NA mutants optimally reduced NA sialidase activity in accordance with the weaker HA binding to α2,6 Sia during viral infection of human cells -a kind of "fine-tuning". Our Western blotting analyses, together with FFU-normalized (virion-level, net) NA activity assays and NA amount-normalized (intrinsic) NA activity measurements, revealed two mechanisms responsible for the reduced sialidase activity of adaptive NA mutants. The first is decreased protein expression due to NA mutations (Fig 6). Given the strong correlation between sialidase activity and protein expression, it is likely that group of mutations that facilitated transmission from birds to humans predominantly though this mechanism. This aligned with a previous report indicating that the human-like clade 2.2 NAs carried L204M mutation, which reduced protein expression on the virion, thereby lowering NA activity [37]. The second mechanism is a functional decline in sialidase activity itself due to NA mutations (S5 Fig) . This was observed sporadically in several NA mutations selected for in patients and human cells, which may be attributed to more direct selective pressure for human-adaptation in these environments. In summary, these findings suggest that mutant NA viruses in patients have decreased virion sialidase activity both quantitatively and qualitatively in H5N1 virus-infected patients.
The functional HA-NA balance is crucial for influenza virus attachment to host cells [14,[27][28][29][30][31][32]. Indeed, biolayer interferometry demonstrated that EG/D1, which possesses an avian-like HA with high α2,3 Sia affinity but only low α2,6 Sia affinity [20], increased the overall α2,6 Sia binding affinity of virus particles by reducing sialidase activity (Fig 8). In the same manner, when typical avian influenza viruses infect the human upper airway, they are inefficiently taken up, likely because their original NA activity is stronger than the weakened Sia binding activity in this environment. These findings suggest that H5N1 viruses acquire adaptive NA mutations that reduce virus-dissociating activity to counterbalance the weaker Sia binding ability in humans, thereby enhancing α2.6 Sia-tropism of virus particles. This notion aligns with a previous hypothesis regarding H5N1 virus adaptation to humans [47]. Also, because all replication measurements in this study were obtained from infectious virions released into the culture supernatant, these data inherently reflect successful progeny virion dissociation from sialylated cell surfaces. Thus, the comparable or enhanced titers observed for most adaptive NA mutants indicate that partial reductions in NA activity do not impair virion release under the cellular conditions tested.
In addition to these NA-driven adjustments in HA-NA balance, our analysis of clinical H5N1 isolates revealed that adaptive NA mutations frequently co-occur with HA mutations that confer only modest increases in α2,6 Sia binding. Among 94 Egyptian patient-derived viruses, 56 encoded at least one HA-and one NA-adaptive mutation simultaneously, forming 35 distinct HA/NA combinations. Functional evaluation of 18 naturally occurring HA/NA mutation pairs demonstrated that these variants occupied a narrow early-adaptation window, defined by weak but detectable α2,6 binding and locally reduced NA activity-levels far below those characteristic of human seasonal viruses. Under experimentally α2,6-dominant conditions in human airway-derived cells, viruses bearing both HA and NA mutations exhibited improved replication compared with their corresponding HA-only counterparts, whereas the parental HA background required NA adaptation to display measurable effects. Importantly, this cooperative phenotype was restricted to the weak-binding range representative of early adaptation intermediates and did not reflect the affinity profiles associated with fully humanadapted viral phenotypes.
Because α2,3-and α2,6-Sias are co-expressed in most human airway tissues at varying ratios [42], it is noteworthy that the most pronounced replication increases (>10-fold) were observed only under experimentally α2,3-depleted conditions (Figs 4 and5). In contrast, replication advantages under physiologically mixed receptor environments were modest. Thus, in natural infection settings, the phenotypic contribution of NA adaptation is likely incremental and less substantial than that of previously characterized HA or polymerase adaptations reported in ferret transmission studies [17,[51][52][53] or naturally occurring human isolates [18,33]. Nevertheless, the presence of naturally co-occurring HA-NA mutation pairs supports a model in which NA contributes to early-stage tuning of HA-receptor interactions during the initial steps of hostrange expansion.
For avian influenza viruses to achieve pandemic potential, a shift in receptor specificity from α2,3-to α2,6-Sia is required [16]. Numerous HA mutations affecting this property have been identified in ferret adaptation experiments [51,52] and patient-derived H5N1 strains [18]. But, importantly, most of these mutant HA viruses had only very low α2,6 Sia binding avidity relative to seasonal viruses [22][23][24][25]. This implies that avian influenza viruses undergo a transitional period, in which they infect humans with suboptimal avidity before gradually switching their receptor tropism, as seen in the 1968 pandemic caused by the H3N2 virus [26]. In fact, unlike bird-derived H5N1 viruses, human-derived H5N1 viruses often have lower sialidase activity (due to NA stalk deletion) [54,55], whereas seasonal influenza viruses have full-length NAs with strong sialidase activity. Our findings indicate that the complementary NA mutation mechanism may contribute only transiently during this early transition, supporting receptor-context-dependent replication before high-affinity HA adaptation is fully established.
Because this study investigates amino acid substitutions that modestly enhance H5N1 replication in human-like cellular environments, we carefully assessed its relevance to Dual-Use Research of Concern (DURC). All experiments were performed using recombinant EG/D1 viruses containing the avian-tropic clade 2.2 HA, which shows minimal α2,6 Sia engagement and does not support efficient infection or transmission in mammals. Notably, the phenotypic effects of NA mutations were detectable only under experimental condition with near-complete depletion of α2,3 Sia-a scenario not found in natural human tissues. In vivo, where α2,3-and α2,6-Sia coexist, their impact remained limited and substantially weaker than that of well-characterized human-adaptive HA substitutions (e.g., Q226L or G228S). While some NA mutations exhibited effects comparable to early weak HA adaptations identified in clade 2.2 patient isolates, multiple functional observations suggest that these effects remain auxiliary and do not provide evidence of enhanced transmissibility or pathogenicity. All work was performed under biosecurity level 3 (BSL3) containment and reviewed by our institutional biosafety and DURC oversight committees, which concluded that the study does not meet DURC criteria. We include this disclosure to ensure transparency and contextual framing of the public-health relevance of these findings.
Collectively, our results indicate that NA mutations identified in clade 2.2 human H5N1 infections do not independently confer human adaptation but instead act as incremental modifiers whose functional impact becomes apparent primarily when paired with HA variants exhibiting weak α2,6 specificity. By integrating evolutionary analysis, quantitative phenotyping, and receptor-context-dependent infection models, this study provides mechanistic insight into how early-stage HA-NA co-adaptation may proceed during the initial bottlenecks encountered when avian influenza viruses infect humans. These findings highlight the value of monitoring coordinated HA-NA evolutionary trajectories rather than evaluating each gene in isolation when assessing zoonotic influenza virus risk. Continued genomic surveillance and functional risk assessment incorporating HA-NA pairing will be essential to improve pandemic preparedness frameworks.
## Materials and methods
## Ethics statement
All experiments using live H5N1 viruses were conducted at BSL3 at Kyoto Prefectural University of Medicine. All studies using recombinant H5N1 viruses were conducted in accordance with relevant laws in Japan and approved by the Biological Safety Committee of Kyoto Prefectural University of Medicine (approval number 30-104), after a risk assessment by the Living Modified Organisms Committee of Kyoto Prefectural University of Medicine, and, if necessary, the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
## Cells
MDCK cells and chicken embryo fibroblast (DF-1) cells were obtained from the American Type Culture Collection. 293T cells were obtained from RIKEN BioResource Center Cell Bank. We previously established 1A5 cells from primary human airway cells, in which the predominant expression of α2,6 Sia over α2,3 Sia was confirmed [39]. CEFs were prepared from embryonated eggs obtained from Shimizu Laboratory Supplies, Japan, as previously reported [19]. They were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal calf serum. Primary human small airway epithelial cells (Lonza Corporation) were cultured under a non-air-liquid interface, according to the manufacturer's recommendation.
## Biosecurity and biosafety
All personnel conducting this study were vaccinated against H5N1 influenza virus. In this study, the infection investigations were conducted as a single infection as is routinely done in many other studies [20,33,56]. No passage or transmission experiments were performed in this study.
## Viruses and reverse genetics
Recombinant H5N1 viruses were generated by reverse genetics based on the genetic background of EG/D1, a representative clade 2.2 virus whose HA displays typical avian-type receptor binding (high α2,3 Sia affinity and only weak α2,6 Sia affinity) [20]. Mutant viruses were generated by site-directed mutagenesis, and rescued in 293T/MDCK-SIAT1 co-cultures. To avoid passaging-associated selection, rescued viruses were propagated only once in MDCK-SIAT1 cells, purified by ultracentrifugation, and titrated by focus-forming assay [33] on MDCK-SIAT1 cells.
For analyses of NA-specific adaptation (Figs 345678), NA mutations were selected as described below. For experiments assessing cooperative HA-NA adaptation (Fig 9), previously characterized HA mutations that modestly increase α2,6 Sia binding in Egyptian clade 2.2.1 H5N1 viruses [18] were incorporated into the EG/D1 HA backbone. Database screening of Egyptian clade 2.2.1 human isolates identified 35 HA/NA mutation combinations (S1 and S2 Tables), from which 18 representative pairs were selected based on (i) reported HA receptor-binding effects [18], and (ii) NA sialidase phenotypes in this study. For each selected HA/NA combination, two recombinant viruses were generated: (1) an HA-only mutant virus, and (2) an HA/NA simultaneous mutant virus.
The complete coding regions of both HA and NA in all recombinant viruses-including NA-only, HA-only, and HA/NA simultaneous mutants-were verified by Sanger sequencing prior to use. Because all recombinant viruses were derived from identical plasmid backbones and were not passaged beyond initial rescue, the six internal gene segments (PB2, PB1, PA, NP, M, and NS) remained genetically identical across all viruses. Therefore, all observed phenotypic differences can be attributed solely to the intentionally introduced HA and/or NA mutations.
## Identification of adaptive NA mutations in clade 2.2-infected patients
We sought NA mutations presumably selected in H5N1 virus-infected patients by database searching of NA sequences in human and bird clade 2.2 viruses in Egypt during 2006-2012. This yielded 94 human-derived NA sequences and 343 bird-derived NA sequences from the GISAID database (https://gisaid.org). NA mutations in the 94 human and 343 bird clade 2.2 strain were identified by comparing each NA sequence with the consensus sequence determined by aligning all the NA sequences. The criteria for selecting adaptive NA mutations were as follows. First, we selected NA mutations that were not detected in the bird-derived sequences, but were detected in human-derived sequences. This set of mutations was likely to have emerged within infected individuals. Second, we selected NA mutations that were detected more frequently in bird sequences than human sequences. Viruses with any of this set of mutations presumably emerged in birds and were subsequently transmitted to humans with high efficiency.
## Identification of adaptive NA mutations in primary human cells
HAE cells and CEFs as a control were infected with the recombinant EG/D1 virus at MOIs of 0.1 and 0.01, respectively. We employed low-MOI infection (0.01-0.1)-a widely used condition for studying influenza diversification and host adaptation in human airway epithelial cells-to enable reliable detection of de novo mutations arising during a near-single-round infection, while minimizing multiplicity re-infection that could obscure selective pressures acting on newly generated variants [57][58][59][60]. Before the experiment, the gene of the recombinant EG/D1 sample was sequenced in-depth by Sanger sequencing of 300 randomly selected clones obtained by RT-PCR. There were no clones with NA mutations, confirming sufficiently high gene homogeneity of the virus sample (the potential mutation rate was < 0.3%). 96 hpi, genetic heterogeneity of progeny virus quasispecies in the supernatant was analyzed by randomly selecting 30 clones obtained by RT-PCR and performing Sanger sequencing (mutation detection threshold was 3%).
## Sialidase assay
We performed a chemiluminescent sialidase assay using NA-XTD as the substrate (NA-XTD Influenza Neuraminidase assay Kit, Applied Biosystems) and NA mutant viruses that were normalized by FFU titers, as previously described [61]. We adopted this assay because its high sensitivity was advantageous for detecting potentially low sialidase activity of EG/ D1 viruses with an NA stalk deletion [20]. To additionally quantify "NA amount-normalized (intrinsic) NA activity," equivalent volumes of FFU-normalized virion samples were further normalized by NA protein abundance determined by Western blotting, and sialidase activity was expressed relative to NA quantity (i.e., activity per unit NA protein). This complementary analysis enables assessment of intrinsic catalytic efficiency of mutant NAs independent of differences in NA incorporation levels into virions.
## Fluorescent lectin staining and α2,3-sialidase treatment
To characterize the distribution and temporal stability of α2,3-and α2,6-Sias, DF-1, MDCK, untreated 1A5, and α2,3-sialidase-treated 1A5 cells were subjected to fluorescent lectin staining. For α2,3-sialidase treatment, 1A5 cells were incubated with α2,3-specific sialidase (Vector Laboratories) under previously established conditions [62] (1 h, 37°C), washed three times with PBS, and returned to culture medium. For the time-course analysis, 1A5 cells were fixed immediately (0 h) or 13 h after sialidase treatment to assess the persistence of α2,3 Sia depletion.
Cells were seeded into 96-well tissue culture plates, washed with PBS, and fixed with 4% paraformaldehyde for 15 min at room temperature. After three washes, cells were incubated with fluorescein-conjugated MAL-I (Vector Laboratories), which recognizes α2,3 Sia, and Cy3-conjugated SNA (Vector Laboratories), which recognizes α2,6 Sia. Lectins were diluted in PBS containing bovine serum albumin according to the manufacturers' instructions. Following lectin incubation, cells were washed with PBS and counterstained with Hoechst 33342 (ThermoFisher).
Fluorescence images for MAL-I, SNA, and Hoechst channels, along with merged images, were acquired using an FV3000 confocal microscope (Olympus) under identical exposure settings.
## Virus growth kinetics in cultured cells
MDCK cells and DF-1 cells were infected with wt or mutant NA virus at an MOI = 0.005. Untreated or α2,3-sialidase-treated 1A5 cells were infected at an MOI of 0.05. After 1-h virus adsorption, inocula were replaced with EMDM/F12 medium (ThermoFisher). Supernatants were collected over time for up to 72 h, and virus titers were determined by the FFU assay described above.
For comparative analyses of HA-only and HA/NA simultaneous mutants (Fig 10C), viral yields were quantified at 12 h post-infection (corresponding to 13 h after sialidase treatment), a time window during which α2,3 Sia depletion in 1A5 cells was confirmed to persist.
## Quantification of NA levels in transfected cells
Flag-tagged NA-wt and mutant plasmids were constructed in the pcXN2 backbone and transfected into 293T cells seeded in 24-well plates. Cells were harvested 24 h post-transfection and lysed in lysis buffer. Lysates were subjected to SDS-PAGE and Western blotting using anti-Flag antibody (Sigma-Aldrich) and Alexa Fluor 488-conjugated secondary antibody (ThermoFisher). β-actin was detected as a loading control. NA band intensities were normalized to β-actin and expressed relative to wt NA.
## Quantification of NA levels in virions
To quantify NA protein incorporated into virions, clarified culture supernatants were layered onto a 20% sucrose cushion and centrifuged at 28,000 × g for 2 h at 4°C. Pelleted virions were resuspended in PBS, and the preparations were normalized by FFU to ensure equivalent infectious units across samples. Equal FFU amounts of virions were mixed with SDS sample buffer, heated at 95°C for 5 min, and subjected to SDS-PAGE and Western blotting. NA was detected using an anti-NA antibody (Sino Biological), and NP was detected as a loading control. NA intensities were normalized to NP in the same lane and expressed relative to wt virus. Virion-associated NA levels were compared with intracellular NA expression levels obtained as described above.
## Western blotting and band quantification
For both transfected-cell and virion samples, bands were visualized using Amersham ECL Select reagent and imaged with an Amersham Imager 680 (Cytiva). Band intensities were quantified using Amersham Imager 680 software. All normalization procedures are described in the respective sections above.
## NAI susceptibility assay
Virus susceptibility to the NAIs peramivir (Peramivir Trihydrate, Selleck) and laminamivir (MedChemExpress), was determined using Chemiluminescent NA inhibition assays (NA-XTD Influenza Neuraminidase Assay Kit, Applied Biosystem), as described elsewhere [63]. In accordance with internationally accepted principles recommended by World Health Organization (WHO)-specifically, that inhibitor titrations should be performed using an activity-normalized virus input rather than a particle-normalized input-we followed the quantitative protocol of the National Institute of Infectious Diseases (NIID, Japan), which operationalizes this requirement. Serial dilutions of each virus preparation were first assayed to determine NA activity. The dilution yielding a signal-to-noise ratio of 40, ensuring sufficient enzyme activity and linear reaction kinetics, was identified for each sample. This activity-normalized dilution was then used as the standardized virus input for all subsequent inhibitor titrations. This procedure ensures that IC₅₀ values reflect the intrinsic inhibitor susceptibility of neuraminidase, independent of differences in FFU titers, particle infectivity, or NA incorporation levels among mutants. To evaluate virus susceptibility to NAIs, we used WHO criteria based on fold-change; for Influenza A viruses, an increase in IC 50 value of <10-fold is defined as normal susceptibility, 10-100-fold as decreased susceptibility, and >100-fold as highly reduced susceptibility.
## BLI analysis
Profiles of the binding between NA mutant viruses and α2,6 Sia were assessed by biolayer interferometry using the BLITz System (Sartorius/ForteBio). α2,6 Sia (α2,6 sialylglycopeptide, FUSHIMI Pharmaceutical Co Ltd) was biotin-labeled using EZ-Link Sulfo-NHS-LC-LC-Biotin (ThermoFisher), purified with Zeba Spin Desalting Columns 7K MWCO (ThermoFisher), and then dialyzed against PBS (+). Biotin labeling efficiency was calculated using the Fluorescence Biotin Quantitation Kit (ThermoFisher) to confirm that the labeling ratio was an ideal 1:1. The biotinylated α2,6 Sia was immobilized on Octet Streptavidin (SA) Biosensor (Sartorius/ForteBio) in a saturated state and used as a ligand. Apparent mole numbers of NA mutant viruses were calculated from FFU titers. The highest concentration shown in Fig 8 (e.g., "1.06 pM" for NA-wt) represents the calculated apparent virion concentration used as the starting point for a two-fold serial dilution series. The viruses and α2,6Sia were reacted in PBS (+) buffer, where HA binding and NA dissociating were assessed. Thus, virus binding to α2,6 Sia was measured as the sum of the HA binding activity and NA dissociating activity of virus particles. Apparent association constants (K A ) and on-rates were measured by global-fitting of the data.
## Assessment of HA α2,6 Sia binding preference
The α2,6 Sia binding properties of the HA mutants used in this study were obtained from our previous report [18] under CC-BY licensing. As described therein, receptor binding specificity was measured using a solid-phase direct binding assay with defined sialylglycopolymers. Binding curves were fitted to determine apparent K , and the ratio K A (α2,6SLN2)/ K A (α2,3SLN1) was used as the index of α2,6 Sia binding preference.
## Modeling of the EG/D1 N1 neuraminidase structure and energetic assessment of adaptive mutations
The amino acid sequence of N1 NA (EG/D1) was retrieved from GenBank (ID: AB497032), and the structure was modeled based on a crystal structure of N1 NA (A/Vietnam/1203/04) [64] as a template (PDB: 2HTY). Based on the template, different amino acids were modeled using the rotamer library in Rosetta 3.13 [65]. The initial N1 model structure was then refined further using the FastRelax protocol with all-atom constraints [66]. Finally, the energetics (ΔΔG = ΔG Mut -ΔG WT ) of the mutations were assessed using the cartesian_ddG protocol in Rosetta [67]. The structure was visualized with ChimeraX [68].
## Statistical analysis
Data analyses were conducted using GraphPad Prism 6 (GraphPad Software). Statistically significant differences between virus pairs were determined by ANOVA with Tukey's multiple comparison test. For pairwise comparisons (e.g., HA-only vs. HA/NA simultaneous mutants; NA amount-normalized activity analyses), two-tailed unpaired t-tests were performed.
## References
1. Chen, Smith, Zhang et al. (2005) "Avian flu: H5N1 virus outbreak in migratory waterfowl" *Nature*
2. Wang, Zhan, Li et al. (2008) "H5N1 avian influenza re-emergence of Lake Qinghai: phylogenetic and antigenic analyses of the newly isolated viruses and roles of migratory birds in virus circulation" *J Gen Virol*
3. Arafa, Naguib, Luttermann et al. (2014) "Emergence of a novel cluster of influenza A(H5N1) virus clade 2.2.1.2 with putative human health impact in Egypt" *Euro Surveill*
4. El-Shesheny, Kandeil, Mostafa et al. (2021) "H5 Influenza Viruses in Egypt. Cold Spring Harb Perspect Med"
5. Le, Nguyen (2014) "Evolutionary dynamics of highly pathogenic avian influenza A/H5N1 HA clades and vaccine implementation in Vietnam" *Clin Exp Vaccine Res*
6. Nguyen, Jang, Nguyen et al. (2017) "Shifting Clade Distribution, Reassortment, and Emergence of New Subtypes of Highly Pathogenic Avian Influenza A(H5) Viruses Collected from Vietnamese Poultry from 2012 to 2015" *J Virol*
7. Kandeil, Kayed, Moatasim et al. (2017) "Genetic characterization of highly pathogenic avian influenza A H5N8 viruses isolated from wild birds in Egypt" *J Gen Virol*
8. Salaheldin, El-Hamid, Elbestawy et al. (2017) "Multiple Introductions of Influenza A(H5N8) Virus into Poultry" *Emerging infectious diseases*
9. Naguib, Arafa, Parvin et al. (2017) "Insights into genetic diversity and biological propensities of potentially zoonotic avian influenza H9N2 viruses circulating in Egypt" *Virology*
10. Lee, Torchetti, Winker et al. (2015) "Intercontinental Spread of Asian-Origin H5N8 to North America through Beringia by Migratory Birds" *J Virol*
11. Tian, Bai, Li et al. (2021) "Highly Pathogenic Avian Influenza Virus (H5N1) Clade 2.3.4.4b Introduced by Wild Birds" *Emerg Infect Dis*
12. Youk, Torchetti, Lantz et al. (2021) "H5N1 highly pathogenic avian influenza clade 2.3.4.4b in wild and domestic birds: Introductions into the United States and reassortments" *Virology*
13. Watanabe, Ibrahim, Suzuki et al. (2012) "The changing nature of avian influenza A virus" *Trends Microbiol*
14. De Vries, Du, Guo et al. (2020) "Influenza A Virus Hemagglutinin-Neuraminidase-Receptor Balance: Preserving Virus Motility" *Trends Microbiol*
15. Guo, Rabouw, Slomp et al. (2018) "Kinetic analysis of the influenza A virus HA/NA balance reveals contribution of NA to virus-receptor binding and NA-dependent rolling on receptor-containing surfaces" *PLoS Pathog*
16. Tharakaraman, Raman, Viswanathan et al. (2013) "Structural determinants for naturally evolving H5N1 hemagglutinin to switch its receptor specificity" *Cell*
17. Linster, Van Boheemen, De Graaf et al. (2014) "Identification, characterization, and natural selection of mutations driving airborne transmission of A/H5N1 virus" *Cell*
18. Watanabe, Arai, Daidoji et al. (2015) "Characterization of H5N1 influenza virus variants with hemagglutinin mutations isolated from patients" *mBio*
19. Watanabe, Arai, Kawashita et al. (2018) "Characterization of H5N1 Influenza Virus Quasispecies with Adaptive Hemagglutinin Mutations from Single-Virus Infections of Human Airway Cells" *J Virol*
20. Watanabe, Ibrahim, Ellakany et al. (2011) "Acquisition of human-type receptor binding specificity by new H5N1 influenza virus sublineages during their emergence in birds in Egypt" *PLoS Pathog*
21. Yamada, Suzuki, Suzuki et al. (2006) "Haemagglutinin mutations responsible for the binding of H5N1 influenza A viruses to human-type receptors" *Nature*
22. Zhu, Viswanathan, Raman et al. (2015) "Structural Basis for a Switch in Receptor Binding Specificity of Two H5N1 Hemagglutinin Mutants" *Cell Rep*
23. De Vries, Zhu, Mcbride et al. (2014) "Hemagglutinin receptor specificity and structural analyses of respiratory droplet-transmissible H5N1 viruses" *J Virol*
24. Lu, Shi, Zhang et al. (2013) "Structure and receptor-binding properties of an airborne transmissible avian influenza A virus hemagglutinin H5 (VN1203mut)" *Protein Cell*
25. Xiong, Coombs, Martin et al. (2013) "Receptor binding by a ferret-transmissible H5 avian influenza virus" *Nature*
26. Liu, Bakker, Narimatsu et al. (2023) "H3N2 influenza A virus gradually adapts to human-type receptor binding and entry specificity after the start of the 1968 pandemic" *Proc Natl Acad Sci U S A*
27. Gambaryan, Matrosovich (2015) "What adaptive changes in hemagglutinin and neuraminidase are necessary for emergence of pandemic influenza virus from its avian precursor?" *Biochemistry (Mosc)*
28. Gaymard, Briand, Frobert et al. (2016) "Functional balance between neuraminidase and haemagglutinin in influenza viruses" *Clin Microbiol Infect*
29. Mitnaul, Matrosovich, Castrucci et al. (2000) "Balanced hemagglutinin and neuraminidase activities are critical for efficient replication of influenza A virus" *J Virol*
30. Wagner, Matrosovich, Klenk (2002) "Functional balance between haemagglutinin and neuraminidase in influenza virus infections" *Rev Med Virol*
31. Xu, Zhu, Mcbride et al. (2012) "Functional balance of the hemagglutinin and neuraminidase activities accompanies the emergence of the 2009 H1N1 influenza pandemic" *J Virol*
32. Yen, Liang, Wu et al. (2011) "Hemagglutinin-neuraminidase balance confers respiratory-droplet transmissibility of the pandemic H1N1 influenza virus in ferrets" *Proc Natl Acad Sci U S A*
33. Arai, Kawashita, Daidoji et al. (2016) "Novel Polymerase Gene Mutations for Human Adaptation in Clinical Isolates of Avian H5N1 Influenza Viruses" *PLoS Pathog*
34. Eggink, Spronken, Van Der Woude et al. (2020) "Phenotypic Effects of Substitutions within the Receptor Binding Site of Highly Pathogenic Avian Influenza H5N1 Virus Observed during Human Infection" *J Virol*
35. Ilyushina, Seiler, Rehg et al. (2010) "Effect of neuraminidase inhibitor-resistant mutations on pathogenicity of clade 2.2 A/Turkey/15/06 (H5N1) influenza virus in ferrets" *PLoS Pathog*
36. Kiso, Ozawa, Le et al. (2011) "Effect of an asparagine-to-serine mutation at position 294 in neuraminidase on the pathogenicity of highly pathogenic H5N1 influenza A virus" *J Virol*
37. Scheibner, Salaheldin, Bagato et al. (2023) "Phenotypic effects of mutations observed in the neuraminidase of human origin H5N1 influenza A viruses" *PLoS Pathog*
38. Adams, Lee, Lugovtsev et al. (2019) "Effect of influenza H1N1 neuraminidase V116A and I117V mutations on NA activity and sensitivity to NA inhibitors" *Antiviral Res*
39. Daidoji, Watanabe, Ibrahim et al. (2015) "Avian Influenza Virus Infection of Immortalized Human Respiratory Epithelial Cells Depends upon a Delicate Balance between Hemagglutinin Acid Stability and Endosomal pH" *J Biol Chem*
40. Lee, Jung, Jadhao et al. (2008) "Evaluation of chicken-origin (DF-1) and quail-origin (QT-6) fibroblast cell lines for replication of avian influenza viruses" *J Virol Methods*
41. Lin, Kappes, Chen et al. (2017) "Distinct susceptibility and applicability of MDCK derivatives for influenza virus research" *PLoS One*
42. Ulloa, Real (2001) "Differential distribution of sialic acid in alpha2,3 and alpha2,6 linkages in the apical membrane of cultured epithelial cells and tissues" *J Histochem Cytochem*
43. Memoli, Hrabal, Hassantoufighi et al. (2010) "Rapid selection of oseltamivir-and peramivir-resistant pandemic H1N1 virus during therapy in 2 immunocompromised hosts" *Clin Infect Dis*
44. Nguyen, Nguyen, Mishin et al. (2009) "Antiviral susceptibility of highly pathogenic avian influenza A(H5N1) viruses isolated from poultry" *Emerg Infect Dis*
45. Arai, Watanabe (2021) "Genetic tuning" of avian influenza virus host adaptation from birds to humans" *Biosaf Health*
46. Le, Sakai-Tagawa, Ozawa et al. (2009) "Selection of H5N1 influenza virus PB2 during replication in humans" *J Virol*
47. Ilyushina, Bovin, Webster (2012) "Decreased neuraminidase activity is important for the adaptation of H5N1 influenza virus to human airway epithelium" *J Virol*
48. Earhart, Elsayed, Saad et al. (2009) "Oseltamivir resistance mutation N294S in human influenza A(H5N1) virus in Egypt" *J Infect Public Health*
49. Younan, Poh, Elassal et al. (2013) "Microevolution of highly pathogenic avian influenza A(H5N1) viruses isolated from humans, Egypt, 2007-2011" *Emerg Infect Dis*
50. Hurt, Holien, Barr (2009) "In vitro generation of neuraminidase inhibitor resistance in A(H5N1) influenza viruses" *Antimicrob Agents Chemother*
51. Herfst, Schrauwen, Linster et al. (2012) "Airborne transmission of influenza A/H5N1 virus between ferrets" *Science*
52. Imai, Watanabe, Hatta et al. (2012) "Experimental adaptation of an influenza H5 HA confers respiratory droplet transmission to a reassortant H5 HA/H1N1 virus in ferrets" *Nature*
53. Liu, Van Kuppeveld, De Haan et al. (2023) "Gradual adaptation of animal influenza A viruses to human-type sialic acid receptors" *Curr Opin Virol*
54. Bender, Hall, Huang et al. (1999) "Characterization of the surface proteins of influenza A (H5N1) viruses isolated from humans in 1997-1998" *Virology*
55. Matrosovich, Zhou, Kawaoka et al. (1999) "The surface glycoproteins of H5 influenza viruses isolated from humans, chickens, and wild aquatic birds have distinguishable properties" *J Virol*
56. Hatta, Hatta, Kim et al. (2007) "Growth of H5N1 influenza A viruses in the upper respiratory tracts of mice" *PLoS Pathog*
57. Davis, Chertow, Kindrachuk et al. (1918) "Influenza receptor binding domain variants bind and replicate in primary human airway cells regardless of receptor specificity" *Virology*
58. Meganck, Edwards, Mallory et al. (2024) "SARS-CoV-2 variant of concern fitness and adaptation in primary human airway epithelia" *Cell Rep*
59. Pohl, Busnadiego, Kufner et al. (2021) "SARS-CoV-2 variants reveal features critical for replication in primary human cells" *PLoS Biol*
60. Smet, Catani, Ysenbaert et al. (2022) "Antibodies directed towards neuraminidase restrict influenza virus replication in primary human bronchial epithelial cells" *PLoS One*
61. Arai, Elgendy, Daidoji et al. (2020) "H9N2 Influenza Virus Infections in Human Cells Require a Balance between Neuraminidase Sialidase Activity and Hemagglutinin Receptor Affinity" *J Virol*
62. Saso, Yamasaki, Nakakita et al. (2022) "Significant role of host sialylated glycans in the infection and spread of severe acute respiratory syndrome coronavirus 2" *PLoS Pathog*
63. Takashita, Ejima, Ogawa et al. (2016) "Antiviral susceptibility of influenza viruses isolated from patients pre-and post-administration of favipiravir" *Antiviral Res*
64. Russell, Haire, Stevens et al. (2006) "The structure of H5N1 avian influenza neuraminidase suggests new opportunities for drug design" *Nature*
65. Leman, Weitzner, Lewis et al. (2020) "Macromolecular modeling and design in Rosetta: recent methods and frameworks" *Nat Methods*
66. Nivón, Moretti, Baker (2013) "A Pareto-optimal refinement method for protein design scaffolds" *PLoS One*
67. Park, Bradley, Greisen et al. (2016) "Simultaneous Optimization of Biomolecular Energy Functions on Features from Small Molecules and Macromolecules" *J Chem Theory Comput*
68. Goddard, Huang, Meng et al. (2018) "Meeting modern challenges in visualization and analysis" *Protein Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548425&blobtype=pdf | # Discovery of a novel double-stranded DNA virus associated with ant labial gland disease reveals its long-term interaction with ants
Shengqiang Jiang, Liangliang Zhang, Xingyu Guo, Jianchao Li, Jing Hu, Hong He, Hongying Chen
## Abstract
Large double-stranded DNA (dsDNA) viruses have been shown to have a wide host range in insects. However, their infection in ants has not yet been descri bed. In this study, we report the identification of a novel filamentous virus, Campono tus japonicus labial gland disease virus (CjLGDV), from the enlarged labial glands of Camponotus japonicus. Transmission electron microscopy revealed numerous non-envel oped nucleocapsids in the nuclei and enveloped envelopment intermediates in the cytoplasm of infected gland cells. Genome sequencing analysis showed that CjLGDV possesses a circular dsDNA genome of 142 kb. Comparative analysis identified a closely related virus from Anoplolepis gracilipes, which was named Anoplolepis gracilipes labial gland disease virus (AgLGDV). CjLGDV and AgLGDV have common genomic character istics and key conserved genes with the viral members in the order Lefavirales, class Naldaviricetes, as well as additional genes uniquely shared with Apis mellifera filamen tous virus (AmFV) and AmFV-like virus (AmFLV). Phylogenetic analysis places CjLGDV and AgLGDV in a distinct lineage within Naldaviricetes, probably representing two members of a novel virus family distantly related to AmFV and AmFLV. By mining public databases,LGDV-related endogenous viral elements were identified in the genomes of multiple ant species, suggesting widespread viral gene integration and a deep evolutionary association between these viruses and their ant hosts. Collectively, our findings reveal a previously unrecognized group of ant-infecting viruses that represents a new viral family within the order Lefavirales, class Naldaviricetes. IMPORTANCE Ants, as highly eusocial insects, play vital roles in ecosystems worldwide. While numerous RNA viruses have been documented in ants, no double-stranded DNA (dsDNA) virus has previously been confirmed to infect them. Labial gland dis ease, reported for decades, lacks a clearly defined cause until now. Here, we identify and characterize a large filamentous dsDNA virus, Camponotus japonicus labial gland disease virus (CjLGDV), from the swollen labial gland of C. japonicus, and a closely related Anoplolepis gracilipes labial gland disease virus in A. gracilipes. Phylogenetic and genomic analyses of the two viruses support the establishment of a new viral family within the order Lefavirales, class Naldaviricetes. The discovery of endogenous viral elements related to CjLGDV in multiple ant genomes suggests the historical infection of CjLGDV-like viruses in ants. These findings broaden the known host range of naldaviri cetes and shed new light on the diversity, evolution, and host interaction of large dsDNA viruses in arthropods.
R ecent advancements in molecular biology and high-throughput sequencing technologies have substantially enhanced our understanding of virus-host interactions and expanded our knowledge of the global virome (1)(2)(3)(4). To date, gene sequences of thousands of species of viruses have been identified in dozens of ant species (5)(6)(7). However, the infectivity and transmissibility of these viruses in ants remain largely speculative and unconfirmed. Among those confirmed, the majority are singlestranded RNA viruses belonging to the order Picornavirales (7). Some of these viruses, such as Solenopsis invicta virus 3, which alters foraging behavior and increases brood mortality in its host, have been explored as potential biological control agents targeting the invasive red imported fire ant (8)(9)(10). Despite extensive research on RNA viruses, documentation of DNA viruses that are infectious for ants is rare. To our best knowledge, the only DNA virus characterized in ants is Solenopsis invicta densovirus (SIDNV), a single-stranded DNA virus found in South American populations of Solenopsis invicta (11). Even so, the impact of SIDNV on its host remains unclear. Recent studies on ant viral diversity have uncovered numerous DNA viral genome fragments in several ant species, indicating a broader and more complex landscape of DNA viral infections in ants than previously thought (6).
Endogenous viral elements (EVEs), or viral fossils, are whole or fragmented viral sequences integrated into host genomes after the virus infection and preserved in the host through germline transmission. Recent findings on the distribution of EVEs in ant genomes have significantly advanced our understanding of the viral commun ity composition in ants (12)(13)(14). Notably, EVEs related to large double-stranded DNA (dsDNA) viruses have been identified in several ant species, suggesting a historical prevalence of DNA virus infections.
According to the International Committee on Taxonomy of Viruses report, classi fied members of Naldaviricetes are categorized into five families: Baculoviridae, Nudiviri dae, Hytrosaviridae, Filamentoviridae, and Nimaviridae (15,16). Unclassified filamentous viruses infecting bees, such as Apis mellifera filamentous virus (AmFV) and Apis mellifera filamentouslike virus (AmFLV), share some genetic characteristics with these five virus families. However, phylogenetic analyses suggest that these unclassified filamentous viruses belong to a new viral family (17)(18)(19).
A condition referred to as "labial gland disease, " characterized by swollen labial glands and malformed mesosomas, has been reported in more than 10 ant species from Europe, the USA, Japan, and China (20)(21)(22). Although the causative agent remains unknown, it has been postulated that the disease may be transmitted by trophallaxis behavior, cannibalism, and probably also vertically from queen to offspring (20,21). Assumption of a viral pathogen has been proposed to cause the distinct morphological changes, but it has not been confirmed (20,21).
Recently, we reported the observation of labial gland enlargement symptoms in Camponotus japonicus, an ant species widely distributed across East Asia (22). In this study, we identified a novel virus with a circular dsDNA genome from the enlarged labial gland and named it Camponotus japonicus labial gland disease virus (CjLGDV). Morphological characteristics of the virus particles were recorded by transmission electron microscopy. By searching the public database, we also identified a virus closely related to CjLGDV in Anoplolepis gracilipes, which we designated as Anoplolepis gracilipes labial gland disease virus (AgLGDV). Phylogenetic analysis was performed to elucidate the evolutionary relationships between LGDVs and other members in Naldaviricetes. Furthermore, CjLGDV-related EVEs were characterized in the genomes of various ant species, providing evidence for the long and frequent interaction history between CjLGDV and its ant hosts. These findings expand the known lineage of Naldaviricetes and broaden our understanding of their host range.
## RESULTS
## Observation of virus particles in the enlarged labial gland of Camponotus japonicus
In our recent report, we found that about onefifth of the minor workers in a mature nest of C. japonicus had swollen labial glands (Fig. 1A) (22). To investigate the cell structure in the enlarged labial gland, the tissue was sliced and examined using transmission electron microscopy. In some enlarged labial gland cells, we observed nuclei stuffed with long filaments, likely viral nucleocapsids in processing (Fig. 1B). In the cytoplasm, numerous virus particles, probably in intermediate stages of viral envelopment, were discovered. Some nucleocapsids appeared coiled within spherical envelopment vesicles around 200 nm, while others were enveloped by a membrane and had filamentous morphology measuring up to 1,000 nm in length and about 65 nm in width. Transitional stages between these two morphologies were also observed (Fig. 1C andD). These observations suggest that the causative agent of the swollen labial gland is probably an enveloped filamentous virus replicated in the nucleus, which is named as Camponotus japonicus labial gland disease virus.
## General features of the viral genome
Based on the observation of numerous nucleocapsid-like filaments in the cell nucleus and the fact that most DNA viruses replicate in the nucleus, we speculated that the causative agent of the labial gland disease could be a DNA virus. To confirm this speculation, DNA was, respectively, extracted from the normal labial gland (LG) and enlarged labial gland (ENLG) of Camponotus japonicus and subjected to Illumina sequencing. For the LG group, 91.7% of the reads were mapped to the Camponotus genome, while only 48.9% of the reads were mapped to the ant genome in the ENLG group (Fig. S1).
The unmapped ENLG reads were used for viral genome assembly. After removing low-abundance contigs and assembly of overlapping ones, ambiguous and gap regions were resolved by PCR amplification (Fig. S2) and Sanger sequencing. The primers for the PCR reactions, along with the length and position of the fragments, are listed in Table S1. Finally, we obtained a 142,484 bp circular dsDNA genome (Fig. 2A), which falls within the size range of naldaviricetes. The general features of the GjLGDV, in comparison with representative viruses of the class Naldaviricetes, are summarized in Table S2. The GC content in the GjLGDV genome is moderate at 52%. In contrast, most baculoviruses, nudiviruses, hytrosaviruses, and filamentoviruses with similar genome sizes have much lower GC contents.
Given the possibility that CjLGDV may infect multiple ant species, or that related viruses may exist, genetic sequences similar to CjLGDV could be present in ant sequenc ing data as a result of viral infections. Genome assemblies of 60 ant species from six subfamilies were retrieved from public databases and screened for similar viral sequen ces by BLAST. Candidate scaffolds representing exogenous viral sequences were expected to fall within the genome size range of naldaviricetes and to display distinct sequence characteristics compared to host genome scaffolds. Finally, a 104,242 bp scaffold (JAPWJP010001941.1) from the long-read sequencing genome assembly of Anoplolepis gracilipes (GCA_031304115.1) was identified as a CjLGDV-like sequence. Due to the unavailability of the original sequencing data for this assembly, the presence of the scaffold was then assessed in other publicly available A. gracilipes data sets (SRR21231523, SRR21232721, and SRR21232722), and no such sequence was detected. The absence of the scaffold in other sequencing samples suggests that it was derived from virus-infected individuals. Moreover, the scaffold lacks eukaryotic genes and exhibits a GC content distinct from that of host BUSCO scaffolds, further supporting its exogenous viral origin (Fig. S3). Although the sequence may be incomplete, its similar genome size and the presence of conserved genes shared with CjLGDV and members of Naldaviricetes suggest that the scaffold contains sufficient genomic information for the virus, which we designate as Anoplolepis gracilipes labial gland disease virus (Fig. 2B).
A TBLASTX similarity assessment revealed that CjLGDV shares more similarity with AgLGDV (22% coverage) and two honey bee filamentous viruses (AmFV, 7% coverage; AmFLV, 4% coverage) than any other viruses (Table S2). A total of 113 and 114 open reading frames (ORFs) with ATG start codons were predicted in the genomes of CjLGDV and AgLGDV, respectively. The average lengths of the predicted ORFs were 351 amino acids (aa) for CjLGDV and 232 aa for AgLGDV, with coding densities of 83.5% and 76.4%, respectively. These genome features are comparable to those of viruses with similar genome sizes in Naldaviricetes (Table S2). Among the predicted ORFs, 39 exhibit BLASTP sequence similarities between CjLGDV and AgLGDV, with amino acid identities ranging from 23.9% to 70.7% (Table S3A). Notably, the level of gene synteny between CjLGDV and AgLGDV was higher than that between CjLGDV and other members in Naldaviricetes (Fig. 2C andD). These findings suggest that AgLGDV is closely related to CjLGDV.
## Repeat regions as a common feature
Homologous regions (hrs), containing repeated sequences composed of imperfect palindrome sequences, are a common feature found in invertebrate dsDNA viral genomes. In baculoviruses, hrs function as origins of virus DNA replication and transcrip tion enhancers (23)(24)(25). In the CjLGDV and AgLGDV genomes, there are 17 and 14 tandem direct repeat (dr) sequences (Table S4), accounting for 4.4% and 3.6% of their genomes, respectively. Repeats in AgLGDV are scattered across the genome, whereas in CjLGDV, six repeats are notably clustered in the region 126,181-131,264 (Fig. 2A; Table S4). All repeats harbor clusters of imperfect palindrome motifs (Fig. S4). These repeats are highly conserved within each viral genome, but no similarities were detected between CjLGDV and AgLGDV or with any other viruses.
Inverted repeat (ir) sequences were detected in both viral genomes (Fig. 2A andB; Table S4). The paired repeats exhibit over 98% sequence similarity, with CG content ranging from 46% to 60%. Interestingly, the length of inverted repeats in AgLGDV ranges from 131 to 312 bp, whereas in CjLGDV, repeats are significantly longer, approximately 2,100 bp. The role of these repeats in virus replication remains to be elucidated. In summary, the presence of repeat sequences appears to be a common characteristic of the two ant dsDNA viruses, and the sequences of the repeats are largely virusspecific.
## LGDVs encode conserved core genes shared by viruses of Naldaviricetes
Functional annotations of ORFs in the CjLGDV and AgLGDV genomes were conducted by similarity searches based on amino acid sequences and protein structures. In the results, 34 ORFs in CjLGDV (Table S3B) and 29 ORFs in AgLGDV (Table S3C) were found to have homologs in other DNA viruses. The remaining ORFs exhibit either low or no similarity to sequences in available databases. Based on the assumption that viral homologs share similar functions, it was predicted that these genes are involved in DNA replication and processing (DNApol, helicase, helicase2, and integrase), transcription and processing (lef4, lef5, lef8, lef9, and methyltransferase), viral packaging and morphogenesis (ac81, p33, ATPase, trypsin-like serine protease, and odv-e18), viral infectivity (pif0/p74, pif1, pif2, pif3, pif4, pif5/odv-e56, chitin-binding protein-like, and odv-e66), and apoptosis inhibition (iap) (Table S3B andC).
Both CjLGDV and AgLGDV genomes contain seven core genes shared by naldaviri cetes, including genes for per os infectivity factors (pif0, pif1, pif2, pif3, and pif5), DNA polymerase gene (DNApol), and sulfhydryl oxidase gene (p33) (Fig. 3). The pif gene family is essential for oral infectivity and is recognized as conserved core genes of naldaviricetes (16). Except for pif4, which is absent in hytrosaviruses and filamentoviruses, five pif genes (pif0/p74, pif1, pif2, pif3, and pif5/odv-e56) are present in the genomes of LGDVs as well as in all sequenced viruses within the class Naldaviricetes.
Among the annotated genes, four genes (DNApol, helicase, helicase2, and integrase) are identified to be involved in viral DNA replication and processing. Both CjLGDV and AgLGDV are predicted to encode a type B DNA polymerase, a common feature of large dsDNA viruses. helicase genes, which are commonly found in members of Lefavirales, are also present. Similar to nudiviruses, AmFV, AmFLV, and some baculoviruses, two types of helicase genes are identified in CjLGDV and AgLGDV genomes. integrase, which has been reported to be involved in the excision and circularization of bracovirus DNA, is a conserved core gene in nudiviruses and bracoviruses (26,27). Here, it is also found in other members of Lefavirales except baculoviruses (Fig. 3).
Viruses of Lefavirales, comprising the virus families Baculoviridae, Nudiviridae, Hytrosaviridae, and Filamentoviridae, are characterized by the possession of conserved baculovirus transcription gene homologs (lef4, lef8, and lef9) and can be phylogenetically distinguished from Nimaviridae (16). Five genes (lef4, lef5, lef8, lef9, and ac81) conserved in lefavirales are detected in both CjLGDV and AgLGDV genomes (Fig. 3). In baculovirus, lef4, lef8, lef9, and p47 encode the four subunits of viral RNA polymerase (28,29). The homolog of p47 is absent in CjLGDV and AgLGDV, a pattern also observed in filamentovi ruses, hytrosaviruses, and AmFV/AmFLV (18,(30)(31)(32)(33). Additionally, we annotated several previously unidentified lef gene homologs in AmFV and AmFLV, including lef4 in AmFV, and lef4, lef5, lef8, and lef9 in AmFLV, based on their sequence similarities to CjLGDV (Fig. 3; Table S5).
## Other homologs shared by members of Naldaviricetes
CjLGDV and AgLGDV are predicted to encode a FtsJ-like methyltransferase, which is also found in some nudiviruses and baculoviruses (Fig. 3). The methyltransferase is reported to be expressed during the late phase of AcMNPV infection and is involved in the RNA capping process (34). Its removal, however, does not impact the production of budded or occluded viruses in AcMNPV (35). Phylogenetic analysis suggests that viruses may acquire the methyltransferase gene from eukaryotic hosts by horizontal gene transfer (Fig. S5A).
A putative ATPase from the AAA+ superfamily is detected in CjLGDV and AgLGDV, which is also conserved in AmFV, AmFLV, filamentoviruses, and hytrosaviruses (Fig. 3). The AAA+ superfamily of ATPases is widely distributed, where its members participate in diverse cellular processes, including membrane fusion, proteolysis, and DNA replication (36,37). In baculovirus infection, host ATPases play critical roles in the construction of the viral replication factory and virion morphogenesis (38). Phylogenetic analysis suggests that the viral ATPase was more likely acquired somehow from bacteria other than their eukaryotic host by horizontal transfer (Fig. S5B).
Insect DNA viruses commonly interfere with the host immune system by preventing apoptosis. Both CjLGDV and AgLGDV encode proteins containing the baculoviral IAP repeat domain, predicted to be homologs of inhibitors of apoptosis (IAPs). IAPs are anti-apoptotic regulators that prevent apoptosis by inhibiting the caspase family of proteases, and they are widely distributed in large dsDNA viruses, yeast, nematodes, insects, and mammals (39). Phylogenetic analysis indicates that the LGDVs iap genes may have eukaryotic origins (Fig. S5C). Both CjLGDV (orf81) and AgLGDV (orf4) are predicted to encode a protein containing a chitin-binding domain. Similar domains are also found in proteins of entomopoxviruses and baculoviruses, where they are known to play critical roles in oral infectivity (40,41). However, the LGDV homologs, especially CjLGDV ORF81, have larger molecular weights and more complex overall structures than AcMNPV chitin-binding proteins (Ac145 and Ac150) (Fig. S6), suggesting that they may perform additional or distinct functions.
Several multigene families are identified in CjLGDV and AgLGDV genomes. Three Baculovirus Repeated ORF (BRO) homologs with an N-terminal DNA-binding domain are detected in CjLGDV (Table S3B). The bro genes are prevalent in various viruses in the Naldaviricetes class, with baculoviruses containing 0-16 bro genes. However, the specific functions of these genes still remain unknown.
Two ORFs encoding putative trypsin-like serine proteases are conserved in CjLGDV and AgLGDV. Trypsin-like serine proteases are enzymatic proteins commonly found in various RNA and DNA viruses as well as in cellular organisms (2). Typically classified as nonstructural proteins in many viruses, these proteases play an indispensable role in viral maturation by facilitating proteolysis through serine-type endopeptidase activity (42,43).
Two baculovirus envelope structural protein homologs (ODV-E18 and ODV-E66) are detected in CjLGDV and AgLGDV. odv-e18 is one of the conserved core genes in baculoviruses and is essential for the budded virus production (44). odv-e66, identified in alphabaculovirus and betabaculovirus that infect Lepidoptera hosts, is also present in hytrosaviruses, AmFV, certain filamentoviruses, and nudiviruses. The copy number of odv-e66 varies among viruses, with most containing 0-5 copies. However, in bracovi ruses, it has expanded to 36 genes distributed across 10 genomic regions, likely playing a critical role in wasp adaptation (45). In baculoviruses, ODV-E66 is a major envelope protein with chondroitinase activity that degrades the larval peritrophic membrane, facilitating oral infection (46). In the CjLGDV and AgLGDV genomes, two adjacent odv-e66 homologs are annotated (Fig. 2A andB). Phylogenetic analysis indicates that ant virus odv-e66 homologs cluster with those of naldaviricetes, forming a monophyletic subclade from bacteria, suggesting that these viral odv-e66 genes might be acquired by an ancestral virus from bacteria (Fig. S7A). Additionally, we found that the viral ODV-E66 homologs share a conserved domain with bacterial chondroitin lyases, including three conserved amino acids previously identified crucial for the enzyme activity in bacteria and baculoviruses (47,48) (Fig. S7A). Of the two ODV-E66 homologs, ODV-E66a retains all three conserved amino acids, while ODV-E66b is mutated at two of the three sites (N to D and H to R). The mutations at the conserved sites are predicted to cause some structural changes in the enzyme activity center, which may affect the function of the protein (Fig. S7B).
## Conserved genes shared exclusively by LGDVs and AmFV/AmFLV
Eleven genes (CjLGDV_orf17, 18, 33, 41, 46, 47, 56, 67, 69, 80, and 86) are uniquely shared by CjLGDV and AgLGDV, and they show no significant similarity to any genes in other known viruses (Fig. 3). In addition, five conserved genes in LGDVs (CjLGDV_orf7, 23, 48, 77, and 94) are also found in all AmFV and AmFLV, but they have no homologs in other viruses. For all the conserved genes exclusively shared by LGDVs and AmFV/AmFLV, no significant sequence or structural similarity to known proteins was detected, preventing reliable functional annotation. Nevertheless, their conservation in LGDVs, and some also in AmFV and AmFLV, suggests that they may play important roles in the unique biology of these viruses.
## Phylogenetic position of LGDVs
CjLGDV shares several key characteristics with the viruses in Naldaviricetes, including their infection of arthropod hosts, the presence of long filamentous nucleocapsids in an envelope, a large circular dsDNA genome, and conserved core genes. These shared traits strongly support the classification of LGDVs as a member of Naldaviricetes. To elucidate the phylogenetic relationships of CjLGDV and AgLGDV, a highly supported phylogenetic tree was generated using the maximum likelihood method, based on the concatenated alignment of the 12 genes conserved in lefavirales (Fig. 4). In this tree, the interrelation ships between Naldaviricetes members are consistent with previously reported results (27,33,49). CjLGDV and AgLGDV cluster together, occupying a unique position within the phylogenetic tree. They are grouped with AmFV and AmFLV, forming a distinct clade adjacent to filamentoviruses and hytrosaviruses, but far from baculoviruses and nudiviruses.
Evolutionary distances within and between virus families in the Lefavirales were calculated based on the 12 genes conserved in lefavirales (Fig. S8). As expected, patristic distances within families are smaller (0.49-2.85) than those between families (3.08-4.38). The distance between CjLGDV and AgLGDV (1.14) falls within the range observed for intrafamily distances, being significantly lower than interfamily distances. Interestingly, the distances between honeybee filamentous viruses (AmFV and AmFLV) and LGDVs (2.49-2.71) are close to the upper limit but still fall within the range of intrafamily distances, suggesting a close evolutionary relationship.
## FIG 4 Phylogeny of
LGDVs with viruses in Naldaviricetes. The phylogenetic tree was constructed using maximum likelihood inference based on concatenated amino acid sequences of 12 conserved genes (p74, pif1, pif2, pif3, pif5, lef4, lef5, lef8, lef9, DNApol, p33, and ac81). Gene accession numbers are provided in Table S5. Node support values are indicated as SH-aLRT support (%)/Ultrafast bootstrap (%). The scale bar represents the average number of amino acid substitutions per site across the tree. Each viral family is denoted by a unique color, and icons next to each virus name indicate the arthropod order of the respective host. Overall, the phylogenetic analysis reveals a close relationship between CjLGDV and AgLGDV, and they may represent members of a novel virus family in the class Naldaviri cetes, order Lefavirales.
## Endogenous viral elements in ant genomes
Previous studies have demonstrated a rich diversity of EVEs in ant genomes, highlighting their evolutionary and functional significance (12,14). To expand our understanding of ant EVEs, we performed similarity searches against the genomes of 60 ant species available in NCBI databases, using the CjLGDV genome as a query.
Although the detection and characterization of EVEs in some species are constrained by the quality and completeness of available genome assemblies, a total of 1,844 loci were identified as candidate EVEs across the genomes of 36 ant species from five subfamilies, including Myrmicinae, Formicinae, Ponerinae, Dolichoderinae, and Pseudomyr micinae (Table S6). Of these, 1,407 loci were considered high-confidence endogenous sequences (Fig. S9; Table S6). In a previous comprehensive study of insect EVEs, 278 high-scoring AmFV-related EVEs were identified in ants (14). Among these, 248 were reidentified in our results (Table S6), highlighting the sensitivity and accuracy of the detection pipeline.
The EVEs identified in this study exhibit homology to 54 CjLGDV genes, including 21 genes with homologs in viruses of the class Naldaviricetes, 5 genes with homologs in AmFV and AmFLV, 10 genes with homologs in AgLGDV, and 18 genes unique to CjLGDV (Fig. 5). BLASTP analysis revealed amino acid sequence identities between the EVEs and viral homologs, ranging from 22% to 76.9% (Table S6). Phylogenetic analyses demonstrated a closer relationship between these EVEs and LGDVs than with AmFV, AmFLV, or other known viruses (Fig. 4; Fig. S101-28). These EVEs are distributed across various scaffolds in ant genomes, with scaffold sizes ranging from hundreds of base pairs to hundreds of millions of base pairs (Fig. S11). Each scaffold contains between 1 and 37 viral homologs. Among the ant species analyzed, the highest abundance of EVEs was detected in Harpegnathos saltator, where 1,378 loci spread across 174 scaffolds showed homol ogy to 29 CjLGDV genes. These homologs are represented by numerous paralogs in the H. saltator genome (Fig. 5), consistent with a previous observation (14). Phyloge netic analyses grouped these paralogs into three distinct clusters (Fig. S12A). Within each cluster, paralogs have low patristic distances (0.01-0.4) (Fig. S12B), share high sequence identity (81%-100%) (Fig. S12C), and have relatively conserved flanking regions (Fig. S12D). Comparative analysis of host gene content between EVE-containing scaffolds revealed distinct sets of eukaryotic genes (Fig. S13). These findings suggest that H. saltator experienced at least three independent viral endogenization events, each followed by lineagespecific duplications that expanded the copy number of the integrated viral sequences.
In Camponotus floridanus, a distinct EVE pattern was observed. Across three scaffolds, 37 loci showed homology to 31 CjLGDV genes, with 35 concentrated on one scaf fold with a length of about 100 kb (NW_020229367.1). These loci were confirmed as endogenous fragments based on several lines of evidence: the presence of identi cal sequences in another C. floridanus genome assembly (GCA_000147175.1) and the occurrence of multiple premature stop codons and transposable elements within these regions. RNA-seq analyses across multiple data sets did not detect any transcriptional activity in the EVE regions in C. floridanus.
Gene synteny analysis of five representative EVEs containing multiple viral gene homologs revealed that all the EVEs had some degree of gene collinearity with CjLGDV (Fig. S14A), supporting the evolutionary linkage between CjLGDV and the ant hosts. The 100 kb long region within the scaffold (NW_020229367.1) in the C. floridanus genome showed high synteny with the CjLGDV genome (Fig. S14B), suggesting that this scaffold was very likely derived from a virus closely related to CjLGDV.
To investigate the selective pressures acting on ant EVEs, we analyzed the ratio of nonsynonymous to synonymous substitution rates (d N /d S ) in the genes that appeared at least twice in the ant EVEs identified in this study. The results revealed that most CjLGDV homologs in ant EVEs have undergone strong purifying selection (d N /d S < 1; Fig. S15; Table S6). By searching available RNA-seq data, it was shown that at least 135 EVEs in 13 ant species were transcriptionally active (Table S6), of which 67 exhibited d N /d S < 1, suggesting evolutionary constraints on these EVEs and their potential functional importance.
## DISCUSSION
Substantial progress has been made in elucidating the viral landscape in invertebrates (2,50), with various insects identified as hosts to large DNA viruses (18,19,30,32,49,51). However, few DNA viruses infecting ants have been characterized so far (6,7,11). This study presents the first fully sequenced dsDNA virus from ants, which is tentatively named Camponotus japonicus labial gland disease virus. Although high copy num bers of viral genomes, especially RNA viruses, have been reported in some arthro pods without obvious pathology (52)(53)(54)(55), our findings suggest a potential association between CjLGDV and labial gland disease. Specifically, in individuals with visibly enlarged labial glands, we observed the presence of abundant viral particles and a high ratio of DNA sequencing reads (41.20% of CjLGDV vs 48.90% of ant host). While only a subset of ants harboring the virus showed gland enlargement, all ants with enlarged labial glands were detected to be infected with the virus. These observations indicate that high viral load may be correlated with labial gland enlargement.
The observation of elongated nucleocapsids stacked in the nucleus, along with coiled and uncoiled nucleocapsids covered with membrane in the cytoplasm, suggests a lifecycle of CjLGDV similar to that of most DNA viruses that replicate in the nucleus and mature in the cytoplasm. By sequence assembly, it is revealed that CjLGDV contains a 142 kb circular dsDNA genome. In addition, mining of publicly available genomic resources in this study uncovered a 104 kb CjLGDV-like scaffold in Anoplolepis gracilipes, which displays high gene synteny with the CjLGDV genome and is named AgLGDV. These findings point to the existence of a previously unrecognized family of large DNA viruses infecting ants.
CjLGDV and AgLGDV exhibit notable genomic similarities to viruses in the class Naldaviricetes, including comparable GC content, genome size, coding density, and the presence of repetitive sequences. The presence of conserved Naldaviricetes gene homologs in the LGDVs genomes supports their classification within this group. Although viruses in the class Naldaviricetes generally share similar genome sizes, LGDVs, filamentoviruses, hytrosaviruses, and AmFV have significantly longer nucleocapsids than baculoviruses and nudiviruses. Moreover, occlusion bodies, which are critical for oral infectivity of baculoviruses, are absent in LGDVs, filamentoviruses, hytrosaviruses, and AmFV.
The identification of pif gene homologs and three DNA-directed RNA polymerase subunits (lef4, lef8, and lef9) suggests that LGDVs are affiliated with the order Lefavirales. Phylogenetic analysis of 12 conserved genes in naldaviricetes positions CjLGDV and AgLGDV within a novel lineage of the order Lefavirales. These two ant viruses form a monophyletic clade with the unclassified bee-infecting viruses AmFV and AmFLV, whereas filamentoviruses and hytrosaviruses constitute a closely related sister group, and baculoviruses and nudiviruses are in distant branches on the phylogenetic tree. Notably, phylogenetic analysis of the ATPase gene, which is conserved across LGDVs, filamentoviruses, hytrosaviruses, and AmFV/AmFLV, suggests that this gene was very likely acquired from bacteria through a single ancestral horizontal gene transfer event, further reinforcing the shared evolutionary ancestry among these viral groups. Among these viruses, LGDVs also share a greater number of homologous genes and exhibit higher sequence identity with AmFV/AmFLV than with other members of Lefavirales, indicating that they have a closer evolutionary relationship.
However, there also exist some important distinctions between LGDVs and AmFV/ AmFLV. First, the genome sizes of LGDVs (142 kb for CjLGDV and 104 kb for AgLGDV) are much smaller than those of AmFVs (close to 500 kb). Second, while both CjLGDV and AmFV possess long filamentous nucleocapsids, the characteristic three figureeight looped nucleocapsids observed in AmFV virions are absent in CjLGDV (18,(56)(57)(58). Third, only about onefifth of the ORFs encoded by LGDVs have homologs in AmFV/AmFLV, and the genome sequence similarities between LGDVs and AmFV/AmFLV were clearly below 10%. The discovery of more viruses closely related to LGDVs and AmFV/AmFLV would be helpful to elucidate their evolutionary relationship and establish the taxonomy of these unclassified viruses.
The close relationship of LGDVs to AmFV/AmFLV raises intriguing questions about their evolutionary trajectory. Considering the intersecting ecological niches and interspecies interactions between ants and bees, ants have been proposed as poten tial vectors or reservoirs for pathogens that affect bee populations (59)(60)(61)(62)(63)(64). It is thus plausible that host ecology has played a role in shaping the co-evolution and crossspecies transmission of these large DNA viruses. The enlarged labial gland observed in C. japonicus is reminiscent of the salivary gland hypertrophy seen in flies infected with hytrosaviruses (22,65), probably indicating a shared pathological mechanism with a common evolutionary origin. Hytrosavirus-like viral genome fragments have been detected in Gigantiops destructor (6), indicating the possibility of hytrosavirus infection within ants and a potential connection between LGDVs and hytrosaviruses. Hytrosavirus has been considered an attractive candidate for fly biocontrol as the virus infection can reduce vitellogenesis and disrupt mating behavior in the infected host. The identifica tion of a novel DNA virus AgLGDV in Anoplolepis gracilipes, one of the most damaging invasive tramp ants globally, provides a potential candidate for its biocontrol. Future studies on the virus infection and its histopathological effects on the invasive host will be essential to evaluate its feasibility, efficacy, and safety as a biocontrol agent. LbFV is associated with the superparasitism behavior of Leptopilina boulardi (66,67), and the virus infection may contribute adaptive genes to parasitic wasps (68,69). Viral manip ulation of host behavior represents a sophisticated evolutionary strategy that optimi zes transmission conditions and enhances survival, illustrating the complex interplay between virus and host (70,71). Further research is required to determine whether
LGDVs induce behavioral modifications in ants, which could deepen our understanding of host-pathogen dynamics.
EVEs are the result of chromosomal integration of viral genes in the host germline cells. Non-retroviral EVEs represent rare remnants of ancient viral infections. Studying these EVEs can provide valuable insights into viral host range, ancestral viral diversity, and the timing of viral evolutionary events. In this study, an unprecedented number of EVEs, which were more closely related to CjLGDV and AgLGDV rather than AmFV or any other known viruses, were detected in the genomes of various ant species spanning five subfamilies (Myrmicinae, Formicinae, Ponerinae, Dolichoderinae, and Pseudomyrmicinae). Given the broad detection of EVEs in ants and the low probability of non-retroviral endogenization events occurring in the germline, this finding indicates prolonged and frequent interactions between the ancestral LGDVs and their hosts. The discovery of CjLGDV in C. japonicus and AgLGDV in A. gracilipes, along with reports of "labial gland disease" in various ant species, implies that such interactions have persisted in their descendants. Some EVEs identified in this study exhibit high gene collinearity and close phylogenetic relationships with CjLGDV, suggesting that some recent integration events or integrations from more closely related viruses have occurred. Furthermore, abundant EVEs and multiple independent endogenization events detected in H. saltator provide strong evidence of long and frequent interactions between ants and viruses. All these findings support the hypothesis of long-term and frequent interactions between LGDVs and ants. Further efforts on the direct detection and characterization of viruses from ants are necessary to conclusively establish the natural host range of LGDVs and to elucidate the relationship between LGDVs and LGDVs-related EVEs.
Studies have shown that EVEs can serve as reservoirs of immune memory in hosts and may function in antiviral defense (72)(73)(74)(75). In some parasitic wasps, endogenous viral elements have been discovered to produce "virus-like structures" in females' reproduc tive organs, which can deliver immunosuppressive DNA or proteins to modulate the immune system of their hosts (69,(76)(77)(78)(79)(80)(81). In our study, viral homologs in CjLGDV-related EVEs are found to be under strong purifying selection, and the transcription of some EVEs in some ant species can be spotted from the publicly available databases. These findings suggest the potential domestication of CjLGDV-related EVEs for functional roles. One such gene, odv-e66, which encodes a chondroitinase associated with viral infectivity, has been hypothesized to be a key factor in virus-host adaptation (45). Here, we reveal that homologs of odv-e66 are widely distributed in diverse ant species, and some of them are transcriptionally active. Whether ODV-E66 plays a functional role and contributes to ant evolution remains a question that requires further investigation.
Taken together, we identified two novel dsDNA viruses, CjLGDV and AgLGDV, in ants. Genomic features and phylogenetic analysis of the two viruses suggest that they belong to a new viral family in the order Lefavirales, class Naldaviricetes. The detection of LGDV-related endogenous viral elements in various ant genomes provides evidence that ants have been the host to LGDVs or their ancestors for a long time. These findings expand the known diversity of naldaviricetes, broaden our understanding of their host range, and reveal a long and frequent interaction between LGDVs and ant hosts.
## MATERIALS AND METHODS
## Ant collection
Colonies of Camponotus japonicus, in which minor workers were observed to have enlarged labial glands (22), were collected from Yangling, Shaanxi Province. The ants were refrigerated at -20°C for 10 minutes to reduce their activity, after which their labial glands were dissected in Ringer's physiological solution under an Olympus SZ51 microscope.
## Electron microscopy
Labial glands were initially fixed in 2.5% cold glutaraldehyde for at least 12 hours, then washed five times with PBS buffer (0.1 M, pH 7.2). Postfixation treatment was performed in 1% osmium tetroxide for 2 hours, followed by five washes with PBS. Samples were dehydrated through a graded ethanol series (30%, 50%, 70%, 80%, 90%, and 100%) and 100% acetone. The tissue was infiltrated with LR-White resin and acetone mixtures (1:3 for 2 hours, 1:1 for 5 hours, and 3:1 for 12 hours), then embedded in pure LR-White resin and polymerized over a period of 72 hours. The thin sections were obtained using a Leica EM UC7 ultramicrotome (Hitachi, Tokyo, Japan), and double-stained using uranyl acetate for 20 minutes followed by lead citrate for 10 minutes to enhance contrast. Observation of the stained sections was conducted using a Tecnai G2 Spirit Bio Twin electron microscope (FEI, Czech Republic and USA).
## Genomic DNA preparation and sequencing
Normal labial glands and enlarged labial glands were obtained from ants in the same colony. The genomic DNA was extracted from the entire labial glands with a genomic DNA extraction kit (BioTeke, Beijing, China) following the manufacturer's protocol. The quantity and quality of extracted DNAs were measured using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis, respectively. The extracted DNA was processed to construct metage nome shotgun sequencing libraries with an insert size of 400 bp by using Illumina TruSeq Nano DNA LT Library Preparation Kit. The library was sequenced on the Illumina HiSeq X-ten platform (Illumina, USA) with a PE150 strategy by Personal Biotechnology Co., Ltd. (Shanghai, China).
## Viral genome assembly
Paired-end sequencing reads were qualityfiltered and trimmed using Cutadapt version 4.4 (82). Genomes of three species closely related to Camponotus japonicus-Campono tus floridanus, Camponotus pennsylvanicus, and Camponotus vicinus-were merged into a composite reference genome. Sequencing data from the LG and ENLG groups were aligned to the reference using Bowtie2 version 2.5 (83). Reads that aligned to the reference genome were regarded as host sequences, and the remaining reads were utilized for viral genome assembly using Haploflow (84). The quality of the assembly was evaluated using QUAST version 5.2 (85). All reads used for genome assembly were mapped to the assembled contigs using Bowtie2, with coverage determined by bedtools version 2.31 (86). Low-abundance contigs were filtered out, and overlapping ones were merged. To resolve gaps and ambiguous regions between contigs, PCR was performed using primers listed in Table S1, and the sequences of the amplified fragments were determined by Sanger sequencing.
## Sequence analyses and genome annotation
Whole-genome similarities were evaluated by employing TBLASTX (87,88) against a set of representative invertebrate DNA virus genomes. Putative open reading frames (ORFs) were identified using ORF finder and Prodigal version 2.6 (89). ORFs were named based on their homologs or genomic location. BLASTP was used to identify ORF similarities against the NCBI nonredundant protein database. Domain identification within ORFs was performed using the NCBI Conserved Domain Search (90) and HMMER search against both public databases (CDD and PFAM) and local databases. The local databases were built using homologous sequences from nuclear arthropod large DNA viruses, which were aligned using MAFFT version 7.5 (91) and converted into HMMs using hmmbuild. Protein structure alignment and homologous structure searches were conducted using Foldseek (92). The sequence coding density was calculated as the ratio of the base number of all ORFs to the base number of the genome. Tandem direct repeats and imperfect palindromic motifs were identified using the etandem (93) and MEME suite (94), respectively, with a 100 score cutoff. The virus genome was graphically represented in a circular diagram using CGView (95).
## Phylogenetic analysis
The phylogenetic position of CjLGDV within Naldaviricetes was inferred by a maximum likelihood tree based on 12 conserved genes (p74, pif1, pif2, pif3, pif5, lef4, lef5, lef8, lef9, DNApol, p33, and ac81). Sequence accession numbers for the conserved genes used in the analysis are provided in Table S5. Specifically, amino acid sequences were aligned using MAFFT with the E-INS-I mode. These alignments were subsequently trimmed by trimAl version 1.4 (96) and concatenated into a single protein alignment by Sequence Matrix (97). Phylogenetic trees were then constructed using the maximum-likelihood method implemented in IQ-TREE version 2.2 (98). ModelFinder (99) was employed within IQ-TREE to identify the best models for each partition. White spot syndrome virus was selected as the outgroup. Node supports in the ML trees were determined using Ultra-fast bootstrap (100) and SH-aLRT (options -bb 1,000 and -alrt 1,000). The patristic distances within and between viral families were performed by the ape R package (101).
## Identification of endogenous viral elements
A BLAST-based approach complemented by systematic phylogenetic clade validations was utilized to identify endogenous viral sequences. Putative protein sequences of CjLGDV were subjected to a TBLASTN search against a database comprising 60 ant genomes. Information on the ant genomes is provided in Table S7. Only hits with an e-value smaller than 1e-5 and more than 20% sequence alignment coverage were maintained. Adjacent hits within a distance of 10 bp were merged into a single entity. Hits were then subjected to a BLASTX search against the NCBI nonredundant protein database. Sequences that clearly align with viral sequences were identified as potential EVEs. Several criteria were utilized to evaluate the endogenous characteris tics of candidate EVEs, including the presence of transposable elements, insect genes, premature stop codons, and sequencing depth.
To assess the authenticity of putative EVEs identified in ant genomes, we implemen ted a quantitative scoring scheme based on multiple genomic features indicative of viral integration. Each EVE was assigned points according to the following criteria: 2 points for the presence of premature stop codons in EVEs, suggesting non-functional or degraded viral sequences; 1 point if the sequencing depth was comparable to that of host scaffolds; 1 point for the presence of annotated eukaryotic genes on the scaffold; 1 point for the presence of transposable elements; 1 point for a GC content similar to that of host scaffolds; 1 point if the scaffold length substantially exceeded that of known exogenous Naldaviricetes viral genomes (>500 kb). Based on the total score, EVEs were classified as high confidence (>4 points), medium confidence (3-4 points), or low confidence (≤2 points).
Transposable elements on the scaffolds were predicted using EDTA version 2.0 (102), while genes were predicted using AUGUSTUS version 3.5 (103). Taxonomic assignments were performed on sequence similarity with the Uniprot/Swissprot database by BLASTP. Only genes assigned to insects were retained. Details on these EVE sequences are available in Table S6. To perform phylogenetic analyses, EVEs were aligned with related homologs from CjLGDV and NALDVs using MAFFT version 7.5 and subsequently refined with trimAl version 1.4. Phylogenetic trees were constructed using the maximum-like lihood method implemented in IQ-TREE version 2.2. To assess potential functional constraints on the EVEs, the ratio of nonsynonymous substitution rate (d N ) to synony mous substitution rate (d S ) was estimated using codeml on the PAML package (104).
## References
1. Rascovan, Duraisamy, Desnues (2016) "Metagenomics and the human virome in asymptomatic individuals" *Annu Rev Microbiol*
2. Shi, Lin, Tian et al. (2016) "Redefining the invertebrate RNA virosphere" *Nature*
3. Zhang, Shi, Holmes (2018) "Using metagenomics to character ize an expanding virosphere" *Cell*
4. Briese, Kapoor, Mishra et al. (2015) "Virome capture sequencing enables sensitive viral diagnosis and comprehensive virome analysis" *mBio*
5. Li, Lian, Zhang et al. (2024) "Virome of red imported fire ants by metage nomic analysis in Guangdong, southern China" *Front Microbiol*
6. Flynn, Moreau (2024) "Viral diversity and co-evolutionary dynamics across the ant phylogeny" *Mol Ecol*
7. Baty, Bulgarella, Dobelmann et al. (2020) "Viruses and their effects in ants (Hymenoptera: Formicidae)" *Myrmecol News*
8. Valles, Oi, Weeks et al. (2022) "Field evaluation of Solenopsis invicta virus 3 against its host Solenopsis invicta" *J Invertebr Pathol*
9. Valles (2023) "Solenopsis invicta virus 3 infection alters foraging behavior in its host Solenopsis invicta" *Virology (Auckl)*
10. Valles (2024) "Effect of Solenopsis invicta virus 3 on brood mortality and egg hatch in Solenopsis invicta" *J Invertebr Pathol*
11. Valles, Shoemaker, Wurm et al. (2013) "Discovery and molecular characterization of an ambisense densovirus from South American populations of Solenopsis invicta" *Biol Control*
12. Flynn, Moreau (2019) "Assessing the diversity of endogenous viruses throughout ant genomes" *Front Microbiol*
13. Gilbert, Belliardo (2022) "The diversity of endogenous viral elements in insects" *Curr Opin Insect Sci*
14. Guinet, Lepetit, Charlat et al. (2023) "Endoparasi toid lifestyle promotes endogenization and domestication of dsDNA viruses" *Elife*
15. Simmonds, Adriaenssens, Lefkowitz et al. (2024) "Changes to virus taxonomy and the ICTV statutes ratified by the international committee on taxonomy of viruses"
16. Van Oers, Herniou, Jehle et al. (2023) "Developments in the classification and nomenclature of arthropod-infecting large DNA viruses that contain pif genes" *Arch Virol*
17. Kadlečková, Saláková, Erban et al. (2024) "Discovery and characterization of novel DNA viruses in Apis mellifera: expanding the honey bee virome through metagenomic analysis"
18. Gauthier, Cornman, Hartmann et al. (2015) "The Apis mellifera filamentous virus genome" *Viruses*
19. Lepetit, Gillet, Hughes et al. (2016) "Genome sequencing of the behavior manipulating virus LbFV reveals a possible new virus family" *Genome Biol Evol*
20. Elton (1991) "Labial gland disease in the genus Formica (Formicidae, Hymenoptera)" *Ins Soc*
21. Laciny (2021) "Among the shapeshifters: parasite-induced morpholo gies in ants (Hymenoptera, Formicidae) and their relevance within the EcoEvoDevo framework" *Evodevo*
22. Zhang, Ma, Xu et al. (2023) "Comparative morphology and ultrastructure of the labial gland among castes of Camponotus japonicus (Hymenoptera: Formicidae)" *Arthropod Struct Dev*
23. Kool, Ahrens, Vlak et al. (1995) "Replication of baculovirus DNA" *Journal of General Virology*
24. Leisy, Rohrmann (1993) "Characterization of the replication of plasmids containing hr sequences in baculovirus-infected spodoptera frugiperda cells" *Virology (Auckl)*
25. Theilmann, Stewart (1992) "Tandemly repeated sequence at the 3′ end of the IE-2 gene of the baculovirus Orgyia pseudotsugata multicapsid nuclear polyhedrosis virus is an enhancer element" *Virology (Auckl)*
26. Burke, Thomas, Eum et al. (2013) "Mutualistic polydnavi ruses share essential replication gene functions with pathogenic ancestors" *PLoS Pathog*
27. Petersen, Burgess, Van Oers et al. (2024) "Nudiviruses in free-living and parasitic arthropods: evolutionary taxonomy" *Trends Parasitol*
28. Guarino, Xu (1998) "A virus-encoded RNA polymerase purified from baculovirus-infected cells" *J Virol*
29. Jin, Dong, Guarino (1998) "The LEF-4 subunit of baculovirus RNA polymerase has RNA 5'-triphosphatase and ATPase activities" *J Virol*
30. Abd-Alla, Cousserans, Parker et al. (2008) "Genome analysis of a Glossina pallidipes salivary gland hypertrophy virus reveals a novel, large, double-stranded circular DNA virus" *J Virol*
31. Wang, Bininda-Emonds, Van Oers et al. (2011) "The genome of Oryctes rhinoceros nudivirus provides novel insight into the evolution of nuclear arthropodspecific large circular doublestranded DNA viruses" *Virus Genes*
32. Guinet, Leobold, Herniou et al. (2025) *Full-Length Text Journal of Virology*
33. Varaldi, Bézier (2024) "A novel and diverse family of filamentous DNA viruses associated with parasitic wasps" *Virus Evol*
34. Kawato, Shitara, Wang et al. (2019) "Crustacean genome exploration reveals the evolutionary origin of white spot syndrome virus" *J Virol*
35. Jvi
36. Wu, Guarino (2003) "Autographa californica nucleopolyhedrovirus orf69 encodes an RNA cap (nucleoside-2'-O)-methyltransferase" *J Virol*
37. Ke, Wang, Deng et al. (2011) "Characterization of AcMNPV with a deletion of ac69 gene" *Microbiol Res (Pavia)*
38. Ogura, Wilkinson (2001) "AAA + superfamily ATPases: common structure-diverse function" *Genes Cells*
39. Snider, Thibault, Houry (2008) "The AAA+ superfamily of functionally diverse proteins" *Genome Biol*
40. Li, Hu, Chen et al. (2020) "Host AAA+ ATPase TER94 plays critical roles in building the baculovirus viral replication factory and virion morphogenesis" *J Virol*
41. Dubrez-Daloz, Dupoux, Cartier (2008) "IAPs: more than just inhibitors of apoptosis proteins" *Cell Cycle*
42. Takemoto, Mitsuhashi, Murakami et al. (2008) "The N-terminal region of an entomopoxvirus fusolin is essential for the enhancement of peroral infection, whereas the C-terminal region is eliminated in digestive juice" *J Virol*
43. Lapointe, Popham, Straschil et al. (2004) "Characterization of two Autographa californica nucleopolyhe drovirus proteins, Ac145 and Ac150, which affect oral infectivity in a host-dependent manner" *J Virol*
44. Bazan, Fletterick (1988) "Viral cysteine proteases are homologous to the trypsin-like family of serine proteases: structural and functional implications" *Proc Natl Acad Sci*
45. Babé, Craik (1997) "Viral proteases: evolution of diverse structural motifs to optimize function" *Cell*
46. Mccarthy, Theilmann (2008) "AcMNPV ac143 (odv-e18) is essential for mediating budded virus production and is the 30th baculovirus core gene" *Virology (Auckl)*
47. Gauthier, Boulain, Van Vugt et al. (2021) "Chromosomal scale assembly of parasitic wasp genome reveals symbiotic virus coloniza tion" *Commun Biol*
48. Hou, Kuang, Luo et al. (2019) "Baculovirus ODV-E66 degrades larval peritrophic membrane to facilitate baculovirus oral infection" *Virology (Auckl)*
49. Lin, Averett, Pritchard (1997) "Identification of a histidine residue essential for enzymatic activity of group B streptococcal hyaluronate lyase" *Biochem Biophys Res Commun*
50. Rodrigues, Peterson, De Oliveira et al. (2020) "Characterization of a novel alphabaculovi rus isolated from the Southern armyworm, Spodoptera eridania (Cramer, 1782) (Lepidoptera: Noctuidae) and the evolution of odv-e66, a bacterium-acquired baculoviral chondroitinase gene" *Genomics*
51. Bézier, Thézé, Gavory et al. (2015) "The genome of the nucleopolyhedrosis-causing virus from Tipula oleracea sheds new light on the Nudiviridae family" *J Virol*
52. Williams, Bergoin, Van Oers (2017) "Diversity of large DNA viruses of invertebrates" *J Invertebr Pathol*
53. Wang, Kleespies, Huger et al. (2007) "The genome of Gryllus bimaculatus nudivirus indicates an ancient diversification of baculovirus-related nonoccluded nudiviruses of insects" *J Virol*
54. Valles, Oi, Becnel et al. (2016) "Isolation and characterization of Nylanderia fulva virus 1, a positivesense, single-stranded RNA virus infecting the tawny crazy ant, Nylanderia fulva" *Virology (Auckl)*
55. Yue, Schröder, Gisder et al. (2007) "Vertical-transmission routes for deformed wing virus of honeybees (Apis mellifera)" *J Gen Virol*
56. Gisder, Aumeier, Genersch (2009) "Deformed wing virus: replication and viral load in mites (Varroa destructor)" *J Gen Virol*
57. Hernández-Pelegrín, Llopis-Giménez, Crava et al. (2022) "Expanding the medfly virome: viral diversity, prevalence, and sRNA profiling in mass-reared and fieldderived medflies" *Viruses*
58. Yang, Wang, Wang et al. (2022) "Genomics and proteomics of Apis mellifera filamentous virus isolated from honeybees in China" *Virol Sin*
59. Clark (1978) "A filamentous virus of the honey bee" *J Invertebr Pathol*
60. Sitaropoulou, Neophytou, Thomopoulos (1989) "Structure of the nucleocapsid of a filamentous virus of the honey bee (Apis mellifera)" *J Invertebr Pathol*
61. Payne, Shepherd, Rangel (2020) "The detection of honey bee (Apis mellifera)-associated viruses in ants" *Sci Rep*
62. Schläppi, Chejanovsky, Yañez et al. (2020) "Foodborne transmission and clinical symptoms of honey bee viruses in ants Lasius spp" *Viruses*
63. Celle, Blanchard, Olivier et al. (2008) "Detection of chronic bee paralysis virus (CBPV) genome and its replicative RNA form in various hosts and possible ways of spread" *Virus Res*
64. Levitt, Singh, Cox-Foster et al. (2013) "Cross-species transmission of honey bee viruses in associated arthropods" *Virus Res*
65. Lin, Lee, Nai et al. (2020) "Deformed wing virus in two widespread invasive ants: geographical distribution, prevalence, and phylogeny" *Viruses*
66. Dobelmann, Felden, Lester (2023) "An invasive ant increases deformed wing virus loads in honey bees" *Biol Lett*
67. Kariithi, Meki, Boucias et al. (2017) "Hytrosaviruses: current status and perspective" *Curr Opin Insect Sci*
68. Varaldi, Ravallec, Labrosse et al. (2006) "Artifical transfer and morphological description of virus particles associated with superparasitism behaviour in a parasitoid wasp" *J Insect Physiol*
69. Patot, Lepetit, Charif et al. (2009) "Molecular detection, penetrance, and transmission of an inherited virus responsible for behavioral manipulation of an insect parasitoid" *Appl Environ Microbiol*
70. Burke, Hines, Sharanowski (2021) "The presence of ancient core genes reveals endogenization from diverse viral ancestors in parasitoid wasps" *Genome Biol Evol*
71. Giovanni, Lepetit, Guinet et al. (2020) "A behavior-manipulating virus relative as a source of adaptive genes for Drosophila parasitoids" *Mol Biol Evol*
72. Varaldi, Fouillet, Ravallec et al. (2003) "Infectious behavior in a parasitoid" *Science*
73. Dheilly, Ravallec, Galinier et al. (2025) "Full-Length Text Journal of Virology October"
74. Thomas, Mitta (2015) "Who is the puppet master? Replication of a parasitic wasp-associated virus correlates with host behaviour manipulation" *Proc Biol Sci*
75. Whitfield, Dolan, Kunitomi et al. (2017) "The diversity, structure, and function of heritable adaptive immunity sequences in the Aedes aegypti genome" *Curr Biol*
76. Ligoxygakis (2017) "Immunity: insect immune memory goes viral" *Curr Biol*
77. Kloc, Halasa, Kubiak et al. (2024) "Invertebrate immunity, natural transplantation immunity, somatic and germ cell parasitism, and transposon defense" *Int J Mol Sci*
78. Lanz-Mendoza, Contreras-Garduño (2022) "Innate immune memory in invertebrates: concept and potential mechanisms" *Dev Comp Immunol*
79. Bézier, Herbinière, Lanzrein et al. (2009) "Polydnavirus hidden face: the genes producing virus particles of parasitic wasps" *J Invertebr Pathol*
80. Pichon, Bézier, Urbach et al. (2015) "Recurrent DNA virus domestication leading to different parasite virulence strategies"
81. Burke, Sharanowski (2024) "Parasitoid wasps" *Curr Biol*
82. Burke, Simmonds, Sharanowski et al. (2018) "Rapid viral symbiogenesis via changes in parasitoid wasp genome architecture" *Mol Biol Evol*
83. Wertheim (2022) "Adaptations and counter-adaptations in Drosophila host-parasitoid interactions: advances in the molecular mechanisms" *Curr Opin Insect Sci*
84. Salvia, Cozzolino, Scieuzo et al. (2022) "Identification and functional characterization of toxoneuron nigriceps ovarian proteins involved in the early suppres sion of host immune response" *Insects*
85. Martin (2011) "Cutadapt removes adapter sequences from highthroughput sequencing reads" *EMBnet J*
86. Langmead, Salzberg (2012) "Fast gapped-read alignment with Bowtie 2" *Nat Methods*
87. Fritz, Bremges, Deng et al. (2021) "Haploflow: strainresolved de novo assembly of viral genomes" *Genome Biol*
88. Gurevich, Saveliev, Vyahhi et al. (2013) "QUAST: quality assessment tool for genome assemblies" *Bioinformatics*
89. Quinlan, Hall (2010) "BEDTools: a flexible suite of utilities for comparing genomic features" *Bioinformatics*
90. Altschul, Madden, Schäffer et al. (1997) "Gapped BLAST and PSI-BLAST: a new generation of protein database search programs" *Nucleic Acids Res*
91. Altschul, Wootton, Gertz et al. (2005) "Protein database searches using compositionally adjusted substitution matrices" *FEBS J*
92. Hyatt, Chen, Locascio et al. (2010) "Prodigal: prokaryotic gene recognition and translation initiation site identification" *BMC Bioinformatics*
93. Wang, Chitsaz, Derbyshire et al. (2023) "The conserved domain database in 2023" *Nucleic Acids Res*
94. Nakamura, Yamada, Tomii et al. (2018) "Parallelization of MAFFT for large-scale multiple sequence alignments" *Bioinformatics*
95. Van Kempen, Kim, Tumescheit et al. (2024) "Fast and accurate protein structure search with Foldseek" *Nat Biotechnol*
96. Rice, Longden, Bleasby (2000) "EMBOSS: the European molecular biology open software suite" *Trends Genet*
97. Bailey, Williams, Misleh et al. (2006) "MEME: discovering and analyzing DNA and protein sequence motifs" *Nucleic Acids Res*
98. Stothard, Wishart (2005) "Circular genome visualization and exploration using CGView" *Bioinformatics*
100. Capella-Gutiérrez, Silla-Martínez, Gabaldón (2009) "trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses" *Bioinformatics*
101. Vaidya, Lohman, Meier (2011) "SequenceMatrix: concatenation software for the fast assembly of multi-gene datasets with character set and codon information" *Cladistics*
102. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol*
103. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: fast model selection for accurate phylogenetic estimates" *Nat Methods*
104. Hoang, Chernomor, Haeseler et al. (2018) "UFBoot2: improving the ultrafast bootstrap approximation" *Mol Biol Evol*
105. Paradis, Schliep (2019) "Ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R" *Bioinformatics*
106. Ou, Su, Liao et al. (2019) "Benchmarking transposable element annotation methods for creation of a streamlined, comprehensive pipeline" *Genome Biol*
107. Stanke, Diekhans, Baertsch et al. (2008) "Using native and syntenically mapped cDNA alignments to improve de novo gene finding" *Bioinformatics*
108. Yang (2007) "PAML 4: phylogenetic analysis by maximum likelihood" *Mol Biol Evol*
109. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793670&blobtype=pdf | # Abstract citation ID: ofaf695.2336 P-2173. Respiratory Virus Detections among Asymptomatic Students and Staff Members in a Large Public School District in Kansas City, Missouri, 2023-2025
Brian Lee, Jennifer Schuster, Brittney Fritschmann, Olivia Almendares, Hannah Kirking, Nibha Sagar, Dithi Banerjee, Anjana Sasidharan, Rangaraj Selvarangan, Jennifer Goldman
Background. While the epidemiology of acute respiratory illness (ARI) among those seeking medical care is well studied, less is known about ARI in non-medical settings (e.g., schools), especially in individuals not exhibiting ARI symptoms. We examined respiratory virus detections among asymptomatic students and staff in a public school district. Methods. Knowledge of Infectious Diseases in Schools (School KIDS) is a prospective respiratory virus surveillance program in a preK-12 th grade public school district in Kansas City, MO. From 2023-2025, student/staff participants self-collected monthly anterior nares swabs. Specimens were tested by multiplex PCR for adenovirus, human metapneumovirus, influenza, parainfluenza, respiratory syncytial virus, rhinovirus/enterovirus (RV/EV), seasonal coronaviruses (sCoV), and SARS-CoV-2. Prior to specimen collection, participants were asked about ARI symptoms (cough, fever, congestion, runny nose, shortness of breath, sore throat, and wheezing) in the past 7 days. Specimens from participants with no ARI symptoms (i.e., asymptomatic) were included in the analysis. Differences between positive (detecting ≥1 virus) and negative specimens were assessed. We used multilevel logistic models to compare odds of viral detection adjusting for school-level and season. 1) but overall showed little seasonal variation (Figure 1). Most commonly detected viruses were RV/EV (56%), sCoV (23%) and SARS-CoV-2 (7%) (Figure 2). Compared with staff from middle/ high-schools, increased odds of viral positivity were observed for preK (OR: 18.9 [4.9,72.4]), elementary (OR: 5.7 [3.0,10.7]), and middle-school students (OR: 2.4 [1.2, 4.7]) (Table 2).
Conclusion. Viral detections in asymptomatic students and staff were frequent (15%), with relatively consistent positivity (for any virus) throughout the school year. Public health strategies to mitigate respiratory viral transmission including cough/hand hygiene and staying home while sick, among others, may reduce both symptomatic and asymptomatic transmission Disclosures. Brian R. Lee, PhD, MPH, Merck: Grant/Research Support Rangaraj Selvarangan, PhD, Altona: Grant/Research Support|Biomerieux: Advisor/Consultant| Biomerieux: Grant/Research Support|Biomerieux: Honoraria|Cepheid: Grant/ Research Support|Hologic: Grant/Research Support|Hologic: Honoraria|Meridian: Grant/Research Support|Qiagen: Grant/Research Support |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12646013&blobtype=pdf | # Wǔhàn sharpbelly bornavirus infects and persists in cypriniform cells
Mirette Eshak, Angele Breithaupt, Birke Tews, Christine Luttermann, Kati Franzke, Marion Scheibe, Sören Woelke, Martin Beer, Dennis Rubbenstroth, Florian Pfaff
## Abstract
Our recent study using in silico data mining identified novel culterviruses (family: Bornaviridae) in fish, including a variant of Wuhan sharpbelly bornavirus (WhSBV) in grass carp kidney and liver cell lines. Here, metagenomic sequencing of different fish cell lines revealed WhSBV in two cell lines from grass carp (Ctenopharyngodon idella; order: Cypriniformes). Using these cell lines, we investigated the ability of WhSBV to infect and establish persistent infection in other cell lines from bony fish (Cypriniformes, Chichliformes, Salmoniformes, Centrarchiformes, and Spariformes), reptiles (Testudines and Squamata), birds (Galliformes), and mammals (Primates and Rodentia). WhSBV showed efficient replication and a time-dependent increase in viral RNA levels in cypriniform cells, whereas replication was limited, confined to single cells, and lacked a clear time-dependent increase in cells from other bony fish and reptiles. No replication was detected in avian and mammalian cells. In situ hybridization and electron micro scopy confirmed the presence of viral RNA and particles in infected cypriniform cells. Transcriptomic sequencing revealed minimal innate immune activation during early stages of infection and antiviral response only at later stages, suggesting that WhSBV establishes persistence by evading early immune recognition. In addition, we identified polycistronic viral mRNAs regulated by specific transcriptional start and termination sites and RNA splicing. Viral proteins were detected, confirming previous in silico predictions. These findings provide insights into the potential infectivity, persistence mechanisms, and transcriptional strategies of WhSBV. This study validates previous findings from in silico data mining, further reinforcing its effectiveness as a powerful tool for discovering hidden viruses. IMPORTANCE Understanding the diversity and host range of viruses is crucial for assessing their ecological role, associated diseases, and zoonotic potential. However, many newly discovered viruses are characterized using sequence data alone because isolates are often difficult to obtain. Using cell culture models, this study characterizes Wuhan sharpbelly bornavirus (WhSBV), a member of the genus Cultervirus. Here, we demonstrate its ability to establish persistent infection in cypriniform fish cell lines, while exhibiting restricted replication in certain non-cypriniform fish. The identifica tion of polycistronic transcription, splicing events, and immune evasion mechanisms advances our understanding of the molecular biology of WhSBV and culterviruses in general. By validating in silico predictions, this study highlights the power of computa tional approaches in uncovering viral diversity. As cypriniform fish include economically important species such as carp, understanding the dynamics of WhSBV host range and infection biology may be crucial for future aquaculture health management.
T he family Bornaviridae, within the order Mononegavirales, contains four genera: Orthobornavirus, Carbovirus, Cultervirus, and Cartilovirus (1)(2)(3). The genus Orthobor navirus has a remarkably broad host range, infecting avian, reptilian, and mammalian species (3). Furthermore, orthobornaviruses include zoonotic viruses, such as Borna disease virus 1 (BoDV1) and variegated squirrel bornavirus 1 (VSBV-1), both of which can cause fatal encephalitis in humans (4,5). Carboviruses and culterviruses, on the other hand, have only been found in reptiles and fish, respectively, and no zoonotic spillover has been reported (2,3,6,7).
Bornavirids possess enveloped virions that enclose a linear, negative-sense, monopartite RNA molecule of approximately 9,000 nucleotides (nt) (3,8). Their genomes encode for at least six open reading frames (ORFs) arranged in the order 3′-N-X/P-M-G-L-5′ for orthobornaviruses or 3′-N-X/P-G-M-L-5′ for carboviruses and culterviruses. These ORFs encode six viral proteins: nucleoprotein (N), accessory protein (X), phosphoprotein (P), matrix protein (M), glycoprotein (G), and the large protein (L) which contains an RNA-directed RNA polymerase (3,8). Recently, the largest bornavirid genome, compris ing 11,090 nt, was identified for the cartilovirus little skate bornavirus (LSBV). The LSBV genome encodes two additional ORFs, potentially encoding viral proteins 1 and 2 (vp1 and -2), and represents a unique genomic architecture 3′-N-vp1-vp2-X/P-G-M-L-5′ (2). Studies of BoDV-1 have demonstrated that its RNA is transcribed and its genome replicated in the host cell nucleus (9). In bornavirids, the generation of a diverse array of mRNAs is regulated by differential use of transcription start and termination sites (10), as well as alternative splicing of polycistronic primary transcripts to increase and control their transcriptional complexity (11,12).
Wuhan sharpbelly bornavirus (WhSBV, NC_055169; species Cultervirus hemicultri), a member of the genus Cultervirus, was initially identified by RNA sequencing of the gut, liver, and gill tissues from a freshwater sharpbelly [Hemiculter leucisculus (Basilewsky, 1855), family Xenocyprididae, order Cypriniformes] from China (3,6). Our recent study using in silico data mining of publicly available sequencing data sets led to the identifi cation of novel fish culterviruses, including the first complete genome of the Murray-Dar ling carp bornavirus (MDCBV, BK063521) from a goldfish [Carassius auratus (Linnaeus, 1758), family Cyprinidae, order Cypriniformes] tissue pool data set and a WhSBV variant (BK063520) detected in grass carp [Ctenopharyngodon idella (Valenciennes, 1844), family Xenocyprididae, order Cypriniformes] CIK (kidney) and L8824 (liver) cell lines (2). The two viruses belong to the same bornavirid species (Cultervirus hemicultri) and share 78% nt overall identity. However, current knowledge of these culterviruses remains limited to in silico genetic discoveries and bioinformatic predictions, with no virus isolates available for experimental research.
To deepen our understanding of the biology and abundance of culterviruses, we screened 48 fishderived cell lines for the presence of known or novel culterviruses using metagenomic RNA sequencing. This revealed the presence of WhSBV in two grass carp swim bladder cell lines, with genomes nearly identical to the aforementioned WhSBV variant BK063520 from the liver (L8824) and kidney (CIK) cell lines. This study represents the first investigation into the biology of WhSBV and lays the groundwork for future research of culterviruses.
## RESULTS
## Metagenomic screening of different fish cell lines reveals the presence of WhSBV
During metagenomic sequencing of 48 fish cell lines, the cypriniform cell lines GCSB1441 and GCSB1542 were found to contain sequences matching WhSBV (Fig. S1). Viral genomes were assembled from both cell lines, sharing 99.9% nt identity with each other and 99.9% and 87.8% nt identity with WhSBV variants CIK (BK063520) and DSYS4497 (NC_055169), respectively. Both currently published WhSBV reference genomes (NC_055169 and BK063520) were identified through metagenomic sequenc ing and de novo assembly. The NC_055169 genome was 8,989 nt in length, while the BK063520 variant was 8,985 nt long. Here, specific sequencing extended the WhSBV genome to 9,008 nt (Fig. S2). These results did not alter the internal genome sequence but revealed additional nucleotides beyond those captured in the metage nomic assembly. Compared to the current reference genome sequence in GenBank (RefSeq NC_055169), the complete WhSBV genome sequence presented here is 7 and 11 nt longer at the 3' and 5' ends, respectively.
Based on the complete WhSBV genome sequences, genespecific reverse transcrip tion quantitative polymerase chain reaction (RT-qPCR) assays targeting the N, P, G, M, and L gene regions were established. All five assays were evaluated using total RNA from the persistently WhSBV-infected GCSB1441 and GCSB1542 and uninfected GK cells as the control. The Cq value of the individual WhSBV gene-specific RT-qPCRs was comparable in GCSB1441 and GCSB1542, while no amplification was observed in GK cells (see Fig. S3). Among the targets, the G genespecific RT-qPCR assay demonstrated the highest sensitivity and consistently produced the lowest Cq values. It was therefore selected for further screening and for subsequent analyses. Using serial dilutions, a Cq cutoff value of 37 was defined for the G genespecific RT-qPCR, corresponding to the highest dilution that consistently produced detectable amplification.
Subsequent testing of the 48 cell lines used for metagenomic analysis with the G gene assay identified WhSBV RNA in GCSB1441 and GCSB1542 cell lines, while all other cell lines tested negative (see Table S1). Furthermore, qPCR assays targeting the WhSBV N, P, G, M, and L gene regions without prior reverse transcription were negative for GCSB1441 and GCSB1542 cells, indicating that the detected WhSBV sequences likely represent genuine viral RNA rather than endogenous viral elements integrated into the host DNA genome (see Fig. S3).
## Cell lysis is necessary for efficient release of WhSBV viral particles
Five inoculum preparation methods were compared to assess their ability to induce WhSBV replication in CCB cells (Table S2). In the inoculated cultures, WhSBV and cellular ACTB RNA levels were monitored using RT-qPCR over a time period of 240 hours post-infection (hpi) (Fig. 1A). Three cycles of freeze/thawing without subsequent centrifugation appeared to yield the highest amounts of infectious virus as higher viral RNA loads were detectable in the inoculated cells already at early time points, as compared to methods employing sonication with or without freeze/thaw cycles or hypertonic treatment (Fig. 1B). Cells infected with the untreated supernatant barely showed any increase in viral RNA over time, suggesting that the measured viral RNA might have been the residual inoculum rather than a result of an active viral replication (Fig. 1A). Based on its performance, consistency, and practical ease, freeze/thaw lysis was selected for use in subsequent infection experiments.
## WhSBV establishes persistency only in cypriniform fish cells
In order to test the infectivity of WhSBV and its potential in vitro host range, we used cell lysates from persistently WhSBV-infected GCSB1441 cells to inoculate various cell lines from bony fish, reptiles, birds, and mammals (see Table S2). Following inoculation, the cells were propagated for several passages, and both cell pellets and supernatants were tested for presence of viral RNA using specific RT-qPCR.
WhSBV RNA was detected in the supernatant and cell pellet of all fish cell lines tested during the first one to two passages. In cells from the order Cypriniformes, including cyprinid (CaPi, CCB, and GTS-9), leuciscid (EPC), and danionid (ZF4) species, viral RNA was retained over all six consecutive passages. The amount of viral RNA detected started to increase compared to freshly inoculated cells (T1) and eventually reached a plateau at around Cq 22 in both supernatants and cell pellets. This stabilization occurred between passages 2 and 4 for the CaPi, EPC, CCB, and GTS-9 cell lines. For zebrafish cell line ZF4, the amount of viral RNA initially decreased after inoculation and passaging until passages 2-3, but then recovered, reaching Cq values of about 27 in the supernatant and cell pellets (Fig. 2A)
In contrast, the non-cypriniform fish, centrarchid (RT/F), and sparid (SAF-1) showed no evidence of WhSBV replication after inoculation (Fig. 2B). Immediately after inoculation (T1), viral RNA levels in the supernatant were similar to those of the cypriniform cell lines (Fig. 2A), but dropped below the detection limit in cell pellets and supernatant after passage 2 (Fig. 2B). Although viral RNA was detected in SAF-1 cell pellets at passage 4, the levels remained below the defined Cq cutoff and were not indicative of active replication. The cichlid (TiB) and salmonid (RTG-2/f and CHSE-214) cell lines exhibited reduced cell growth over the course of the experiment, which was reflected by a decline in the measured ACTB RNA levels, a trend also observed in corresponding uninfected controls. Viral RNA remained detectable in cell pellets until the end of the experiment up to passage 6 for (TiB) and passage 4 for (RTG-2/f and CHSE-214), but without a detectable increase in levels. In several instances, Cq values fluctuated near or fell below the established detection cutoff, indicating the absence of sustained viral replication (Fig. 2B).
In the reptilian cell lines SKL-R, VH-2, and CDSK, WhSBV RNA levels in the culture supernatant decreased after inoculation but remained detectable up to passage 6, with Cq values stabilizing around 28 and 35; however, viral RNA levels fell below the defined Cq cutoff at passages 5 and 6 in SKL-R (Fig. 3). In the SKL-R cell pellets, viral RNA was still detectable up to passage 6, although consistently at low levels, yet above the cutoff. In contrast, CDSK and VH-2 pellets showed an increase in viral RNA levels during passages 5 and 6, reaching Cq values of 28 and 30, respectively. A decline in ACTB RNA levels was also observed in the reptilian cell lines, mirroring the trend in cichlid and salmonid cell lines. Similar decreases in ACTB expression were also noted in the corresponding uninfected controls.
In the mammalian Vero and C6 cell lines and the avian QM7 and DF-1 cell lines, WhSBV RNA levels in the cell pellets or supernatants dropped below or near the detection limit already after passages 1-3 (Fig. 3). All negative controls used in this experiment showed no Cq values for the viral glycoprotein.
No apparent cytopathic effect was observed in any of the cell cultures used through out the experiment. Using transmission electron microscopy, putative virus particle-like structures were observed in WhSBV-inoculated ZF4 cell lines, whereas similar structures were absent in non-inoculated controls. These structures appeared spherical to pleomorphic without visible peplomers and a diameter of approximately 150-180 nm (Fig. 4A). Potential membrane-associated budding was rarely observed at the plasma membrane of the persistently WhSBV-infected GCSB1441 cells (Fig. 4B)
## WhSBV replicates efficiently only in cypriniform cell lines
In order to gain deeper insights into viral propagation in different fish cell lines, a kinetic infection experiment was performed over 10 days using both cypriniform (EPC and CCB) and non-cypriniform fish cells (RTG-2/f and TiB). The level of viral RNA load was assessed by RT-qPCR, and the presence of viral RNA at the cellular level was assessed by RNA in situ hybridization (RNA ISH) at different time points.
In the cypriniform EPC and CCB cell lines, viral RNA levels increased exponentially until 12 hpi, reaching Cq values of about 18.6 in EPC and 19.5 in CCB. In contrast, the viral RNA levels in salmonid RTG-2/f and cichlid TiB cell lines showed stagnation or a linear increase (Fig. 5A). The presence of viral (-)RNA was detected by specific RNA ISH in both cypriniform (EPC and CCB) and non-cypriniform fish cell lines (RTG-2/f and TiB) from 4 hpi onward. In detail, RNA ISH confirmed the RT-qPCR results from the kinetic experiment as the number of infected cells, indicated by the presence of viral (-)RNA, increased in a time-dependent manner for the cypriniform cell lines EPC and CCB. At early time points (4-72 hpi), infected single cells appeared, followed by foci of positive cells, later confluent and finally diffuse labeling of the cell pellet (Fig. 5B). Percentage quantification of individual cell pellets for EPC showed widespread infection, with 95% of the fields analyzed showing positive detection as early as 4 hpi. Infection remained widespread (63-100% of infected test fields) throughout the course of the experiment until 240 hpi (see Fig. S4). In CCB cells, only 29% of the test fields were infected at 4 hpi and longer cultivation correlated with a higher proportion of positive test fields, eventually reaching 100% at 96, 120, and 240 hpi (see Fig. S4). Similar observations were made using a custom-designed probe against WhSBV (+)RNA on CCB cell lines (Fig. S5). In terms of the labeling signal, infected cells harvested at early time points were more likely to show finegranular cytoplasmic labeling, and larger, globular signals were consistently found at 48 hpi and later (Fig. 5B). In clear contrast, the infection pattern in the non-cypriniform fish cell lines was essentially confined to single cells and showed no evidence of increasing infection abundance over time (Fig. 5B). Only the intensity of the chromogen signal within the cell increased in TiB from 72 hpi, but in RTG-2/f; this phenomenon was limited to 72 and 96 hpi. RTG-2/f and TiB cells ultimately yielded 28% and 16% infected test fields at 240 hpi, respectively (see Fig. S4).
For mammalian Vero and avian DF-1 cell lines, only a single time point at 240 hpi was collected, alongside CCB which was used as a positive control. RNA ISH analysis of mammalian and avian DF-1 cell lines showed no detectable WhSBV RNA at either the control 0 hpi or at 240 hpi (data not shown).
Using routine hematoxylin-eosin staining, neither viral inclusion bodies nor other morphological evidence of a cytopathogenic effect could be identified in EPC, CCB, RTG-2/f, TiB, Vero, and DF-1 cell lines (see Fig. S6).
## Cellular response to early and late WhSBV infection is limited in cypriniform cells
To investigate cellular transcriptional changes associated with WhSBV infection, we performed RNA sequencing (RNAseq) of RNA transcripts at early and late time points following viral inoculation in CCB cells (Fig. 6A). Quantitative RT-qPCR and sequencing confirmed the presence of WhSBV RNA and the onset of infection at 4 hpi, with viral RNA levels progressively increasing at all subsequent time points (Fig. 6B). This trend was consistent with the RNA ISH data, which showed a gradual increase in the percentage of infected CCB cells over time, reaching 100% infection at 96, 120, and 240 hpi (Fig. 6B) The number of differentially expressed genes (DEGs) remained minimal at early and late time points but became more pronounced at 240 hpi, indicating a delayed transcriptional response to infection (Fig. 6C). Principal component analysis (PCA) of early time points revealed clear transcriptional differences between WhSBV-infected and control cells at 4 and 8 hpi, which diminished at 24 and 48 hpi, with both groups becoming more similar (Fig. 6C). In contrast, the PCA of late time points experiment showed no significant transcriptional changes at 72 and 96 hpi. However, minimal transcriptional alterations were observed at 120 hpi, followed by distinct clustering of infected and control samples at 240 hpi, suggesting a change in the host transcriptional response at later stages of infection (Fig. 6D).
Differential gene expression analysis of early time points identified three DEGs at 4 or 8 hpi. Specifically, RASAL2 (Ras GTPase-activating activator like 2 [LOC109087896]) and IRF1B (interferon regulatory factor 1b) were upregulated at 4 hpi and 8 hpi (Fig. 6E), whereas EDEM3 (ER degradation enhancing alpha-mannosidase-like 3 [LOC109095466]) was upregulated exclusively at 8 hpi. Conversely, ABCC1 (multidrug resistance-associated 1-like [LOC109090229]) was downregulated at 4 hpi. The expression of these genes returned to baseline levels by 24 and 48 hpi. At later stages of infection (72 and 96 hpi), no DEGs were detected between infected and control cells. At 120 hpi, eight DEGs were upregulated, including IRF7 (interferon regulatory factor 7) and MxB (MX dynamin like GTPase 2) (Fig. 6E). At 240 hpi, 113 DEGs were upregulated, including IRF7, MxB, and several other key antiviral genes involved in early innate immune responses in fish, such as IRF3, ISG15 (interferon-stimulated gene 15), and RSDA2 (viperin, radical S-adeno syl methionine domai-containing 2), along with other genes involved in interferon pathways, RIG-I (antiviral innate immune response receptor RIG-I [LOC109081892]) and GVINP1 (interferon-induced very large GTPase 1-like [LOC109058642]) (Fig. 6E; Table S3 for complete list of DEGs).
## Detection of spliced WhSBV gene transcripts
Mapping of sequence reads from the persistently WhSBV-infected GCSB1441 and GCSB1542 cells to the complete WhSBV genome (Fig. 7A) revealed the presence of three potential introns (In1-3) located at the beginning of the L ORF. Conventional RT-PCR and Sanger sequencing confirmed the presence of In1-3 within the L ORF and also revealed an additional, unpredicted splice site, In4, within the L ORF (Fig. 7B). Based on the above findings, we predicted several potential transcripts for the M and L genes, which share the same start site S3 and terminate at either T3 or T4. The M gene could produce a short transcript encoding the M ORF, which can either remain unspliced (M t1; size: 741 nt) or undergo splicing at In1 to produce a shorter transcript (M t2; 565 nt), both theoretically terminating at T3. Although the unspliced transcript M t1 was detected by RT-PCR, it appears to be in the minority of transcripts based on RNAseq results. In addition, four transcripts were predicted, termed M-L t1-4, all sharing the same start site S3 and termination site T4, but differing in their splicing patterns. The longest unspliced transcript, M-L t1, contains the full-length M and L ORFs (5,680 nt), whereas the other transcripts (M-L t2-4) are progressively truncated by splicing at different introns, resulting in transcripts of size 5,218 nt, 5,150 nt, and 4,945 nt, respectively (Fig. 7C). These transcripts may encode either truncated versions of the L ORF or only the M ORF.
## Detection of multicistronic WhSBV gene transcripts
The observed sequence coverage across the genome was uneven, with abrupt increases and decreases observed in certain potential intergenic regions (Fig. 7A). These coverage variations corresponded in part to transcription start and termination sites identified by motif prediction (see Table S4). In detail, three transcriptional start sites (S1-3) were located upstream of the N, X, and M ORFs. Start sites S1 and S3 were marked by an increase in read coverage, indicating transcription initiation. Four transcription termination sites were detected by motif prediction: T1, T2, and T4 were predicted downstream of the N, G, and L ORFs, respectively, while T3 was found within the L ORF.
The termination sites T2 and T3 were characterized by a marked decrease in sequence coverage, and termination sites T1-3 coincided with reads transitioning to poly(A)-tails. S2 and S3 were located immediately adjacent to T1 and T2, respectively. Sequencing of 5′ ends of cDNA identified the likely transcription start site S1 at nucleotide positions 33 to 35 in the viral genome, marked by the sequence "GAA, " matching the predicted sequence of S2 and S3. Regions of the genome with comparable levels of sequence coverage between adjacent start and termination sites were interpreted as belonging to the same viral RNA transcript or mRNA. We observed no coverage drop at termination signal T1, and the majority of transcripts seem to skip T1, indicating that N, X/P and G genes can be expressed from polycistronic mRNA (Fig. 7A). However, a small number of reads transitioning to poly(A)-tails at T1 indicate the presence of monocistronic N mRNA. The mRNAs containing the M and L genes shared S3 as the common start site, but their transcript levels differed, with the M gene being comparably highly expressed to the L gene (Fig. 7A). We did not detect a possible short upstream ORF (uORF) in the X-P-G transcript as observed for BoDV-1 (13).
A complementary Northern blot analysis of RNA from the persistently WhSBV-infec ted GCSB1441 and the experimentally WhSBV-infected EPC cells supported some findings from the sequence data analysis (Fig. 8A through C). All probes (N, X, P, G, M, or L ORF specific) detected the full genome of WhSBV from the GCSB1441 and EPC cells at the expected size of 9,008 nt. Probes X, P, and G probably labeled the same RNA species at approximately 2,100 nt, matching the predicted length for the polycistronic X-P-G mRNA. The M and L probes both labeled an RNA between 6 and 8 kb, likely representing the M-L t1 transcript. A faint band at 500 nt was detected using the M probe, but not with the L probe, which may represent the transcripts encoding the M ORF alone. The N probe identified an RNA species that was consistently higher than expected and was also recognized by the X, P, and G probes. This observation suggests the presence of an RNA spanning N, X, P, and G, while the monocistronic N mRNA was not detected at the expected size.
High-resolution mass spectrometry was performed on the persistently WhSBV-infec ted GCSB1441 cell suspension to provide an overview of the viral proteins. The analysis revealed peptides that matched all the predicted proteins of WhSBV, except for the L ORF was performed on the RNAseq gene expression data using R (v4.3.1). Orange triangles and blue dots represent replicates of WhSBV-infected and mock-infected CCB cells, respectively. (E) Trends in transcript levels of selected genes known to be key antiviral genes in the early innate immune response to viral infection in fish. These genes include RAS GTPase-activating protein 2-like (LOC109087896) (RSAL2), interferon regulatory factor 1 (IRFIB), interferon regulatory factor 7 (IRF7), interferon regulatory factor 3 (IRF3), interferon-induced GTP-binding protein MxB-like (MxB), interferon-induced very large GTPase 1-like (LOC109058642) (GVINP1), interferon-stimulated gene 15 (ISG15), antiviral innate immune response receptor RIG-I (LOC109081892) (RIG 1), and viperin, radical S-adenosyl methionine domain-containing 2 (RSDA2). Asterisks indicate time points with significantly differentially gene expression as compared to uninfected controls. The complete list of DEGs can be found in Table S3. Results are presented as the arithmetic/geometric mean (± standard deviation) of three replicate wells. where no matching peptides were detected. No fused peptides were detected either, suggesting no additional splicing (Fig. 9).
## DISCUSSION
This study represents the first in vitro molecular characterization of WhSBV, a mem ber of the family Bornaviridae, genus Cultervirus (3,6), and thus the first biological ). Our RT-qPCR assay targeting the G gene was the most reliable and sensitive method of detecting WhSBV RNA. This is consistent with the findings of Northern blot analysis and genome read coverage, which showed that the X-P-G coding RNA was expressed at higher levels than other viral RNAs.
Our comparative analysis across cypriniform cell lines revealed that WhSBV can establish a persistent infection without inducing cytopathic effects and maintains stable viral RNA levels over multiple passages (Fig. 2A). Similar behavior has been observed for BoDV-1, VSBV-1, and several avian orthobornaviruses, where some inoculated cell lines became persistently infected without cytopathic effect and could be sub-cultivated without loss of infectivity (19)(20)(21)(22)(23)(24)(25)(26). As some orthobornaviruses show a relatively broad host spectrum in vitro, we tested WhSBV inoculation of non-cypriniform, reptile, avian, and mammalian cell lines.
Avian and mammalian cells were clearly not susceptible to WhSBV infection. This may be due to the absence of essential host factors in avian and mammalian cells required for WhSBV replication, the activation of an effective antiviral response, or the virus's potential intolerance to the elevated incubation temperature of 37°C. In some non-cypri niform fish (RTG-2/f, CHSE-214, and TiB) and reptilian cell lines, susceptibility to WhSBV infection was difficult to assess because ACTB levels declined over several passages, with eventual loss of cultures. This phenomenon was also observed in uninfected control cultures, suggesting the deterioration was likely due to suboptimal culture conditions, rather than WhSBV infection. Nevertheless, in these cell lines, viral RNA levels over the successive passages fluctuated near or below the detection Cq cutoff value. While these low-level signals may reflect residual inoculum, RNA ISH confirmed the presence of viral RNA within some cells for RTG-2/f and TiB, indicating that these cells were indeed infected by WhSBV. However, infection remained confined to individual cells, and there was no progressive increase in the percentage of infected cells, unlike in cypriniform cells where infection spread over time. As RNA ISH was not possible for reptilian cells, there is currently no conclusive evidence whether these are susceptible to WhSBV infection, and further investigations will be necessary.
The difference in WhSBV infectivity between cypriniform and non-cypriniform fish cell lines may be attributed, in part, to inherent cytological and physiological distinc tions between these cells. The fish cell lines used in this study originate from various tissues and primarily represent epithelial or fibroblast lineages. These cells may differ substantially in the membrane composition, receptor expression profiles, and basal innate immune activity, all of which can influence their permissiveness to viral entry, replication, and persistence (27). Furthermore, variation in metabolic activity, tempera ture tolerance, and stress responses can further modulate viral propagation in vitro (28)(29)(30)(31). The restricted replication of WhSBV in non-cypriniform fish cells may also indicate an evolutionary history of adaptation to cypriniform hosts, which is supported by the absence of confirmed detections in other taxonomic groups to date (2,6,15). For the production of virus preparations, active disruption of the infected cells by, e.g., freeze/thawing was necessary, while comparably little infectious virus appeared to be present in the supernatant of infected culture. This is in agreement with ortho bornaviruses, such as BoDV-1 (32,33). We observed virus particle-like structures and budding in WhSBV-inoculated ZF4 cells, which were absent in the respective controls. Although these structures cannot be definitively attributed to WhSBV, their morphology is consistent with that of viral particles. With 150-180 nm in diameter, their detected size exceeds the typical size range reported for BoDV-1 (70-130 nm) (34,35) and avian bornaviruses (60-104 nm) (36)(37)(38). While the observed structures may represent viral particles and budding, it is currently not possible to confirm their identity as WhSBV without specific staining tools, such as immunogold labeling.
In order to investigate the cellular response to WhSBV infection and to gain insights into the possible mechanisms underlying its ability to establish a persistent infection without inducing cytopathic effects, we analyzed the host gene expression at different time points of infection. The innate antiviral immune system, including the interferon (IFN) pathway, is well conserved between mammals and teleost fish (39)(40)(41). In teleosts, as in higher vertebrates, viral nucleic acids are sensed by pattern recognition receptors (PRRs), notably RIG-I-like receptors such as RIG-I and MDA5 (42)(43)(44). Upon sensing, these receptors initiate signaling cascades that culminate in the production of type I IFNs and proinflammatory cytokines, which then upregulate ISGs to restrict viral replication and enhance viral clearance (44)(45)(46)(47)(48). Transcriptomic analysis of the experimentally WhSBVinfected CCB cells indicated a limited initial activation of the innate immune response shortly after inoculation at 4 and 8 hpi by upregulation of the genes IRF1B and RASAL2 (Fig. 6E). Interferon regulatory factors, such as IRF1B, are key regulators of the innate antiviral response in vertebrates, controlling the expression of type I IFNs and ISGs (49,50). In zebrafish, IRF1B alone was shown to induce the expression of IFNs and ISGs, resulting in effective protection against viral infection (51). While the role of RASAL2 in fish remains unclear, in mammals, it has been characterized as a tumor and metastasis suppressor (52). Despite this initial immune activation, no significant changes in host cell gene expression were observed between 24 and 96 hpi (Fig. 6C), even as viral RNA levels and the proportion of infected cells increased steadily during this period (Fig. 6B). These findings suggest that WhSBV may evade or suppress early innate immune signaling through yet-unknown mechanisms, thereby enabling efficient viral replication and the establishment of persistent infection. In comparison, orthobornaviruses have been shown to evade IFN induction by possessing monophosphorylated genome ends (53,54). BoDV-1 further suppresses the interferon response via P-protein-mediated inhibition of TANK-binding kinase 1 (55) and N-protein interference with IRF7 activation (56).
Only at 120 and 240 hpi did we observe the upregulation of RIG-I and IRFs other than IRF1B, which may have contributed to a delayed innate immune response at later stages. This response was marked by significant upregulation of many ISGs, such as MxB, GVINP1, ISG15, and RSDA2, all of which play antiviral roles in teleost fish (57). This suggests that the host's innate immune system did not mount a response until the infection had spread to nearly all cells. However, this late response appeared insufficient to clear the infection as the cells remained persistently infected, suggesting either the activation of a secondary immune response aimed to control viral load or that the increasing intracellular viral burden eventually triggered the innate immune system of the host. Despite the upregulation of ISGs, no significant induction of apoptosisor necrosis-related genes was observed throughout infection, suggesting that WhSBV establishes infection without inducing programmed cell death. This may contribute to continued viral replication and long-term or even persistent infection. In contrast, common carp hepatocytes exposed to ammonia stress undergo apoptosis via the P53-BAC/BCL-2 pathway, demonstrating the species' ability to induce cell death under certain conditions (58). The absence of such responses in WhSBV-infected cells raised the possibility that the virus actively suppresses apoptosis. Our study provides a temporal overview of host gene expression during a 10-day infection course in the cypriniform CCB cells. To fully elucidate the molecular strategies employed by WhSBV to evade and/or modulate the host immune response, future investigations should focus on identifying the specific viral genes, proteins, or genomic structures involved in immune suppression and persistence. The establishment of persistence without cytopathic effects is a common feature of orthobornaviruses (3) and has also been observed in other mononegaviruses, such as arenavirids (family Arenaviridae), which suppress host antiviral responses, particularly by inhibiting type I interferon production, to maintain chronic infections (59,60).
One strategy to evade the host's immune response might be the strict regulation of WhSBVs transcription that involves expression gradients, polycistronic mRNAs, and splicing, as seen in some orthobornaviruses (11,12,61,62). Some regulatory features, such as the splicing of M-L transcripts, and the position and sequence of transcription start and termination sites (11,33), seem to be highly conserved among culterviruses and therefore suggest conserved mechanisms. This transcription strategy is in contrast to the predominantly monocistronic transcription seen in other mononegaviruses (63), such as paramyxovirids (family Paramyxoviridae), where the viral genome is transcribed into monocistronic unspliced mRNAs, each encoding a single protein (64). However, there are some exceptions, such as Ebola virus (family Filoviridae), that uses RNA editing of its glycoprotein gene in order to generate multiple variants, thereby optimizing replication and pathogenicity without increasing the genome size (65).
Northern blot and sequence coverage analysis indicated that the N proteins are expressed not only from a monocistronic mRNA but also co-expressed with X/P and G from a polycistronic mRNA. However, it remains to be clarified whether the detected RNA of approximately 3,000-4,000 nt in length is actually mRNA or a sub-genomic RNA. For BoDV-1, the non-coding region between the N and X genes shows structural similarities to that of WhSBV, where T1 is located downstream of the transcription start site S2 in BoDV-1 (66). The consensus motif of the BoDV-1 termination sites, U(A) 6-7 (67), serves as a core signal sequence but is insufficient for efficient transcription termination on its own, as the surrounding upstream and downstream nucleotides significantly influence T1 utilization. Consequently, when transcription is initiated at S1 and terminated at T1, these regulatory elements contribute to the production of a monocistronic (1,200 nt) mRNA encoding the N protein (66). However, studies have demonstrated that the viral RNA-directed RNA polymerase frequently bypasses the T1 transcription termination site, resulting in the production of a longer (1,900 nt) polycistronic mRNA containing the N, X, and P ORFs (66). The presence of such transcripts in BoDV-1-infected cell cultures and rat brains suggests that they are likely authentic mRNAs, rather than by-products of aberrant replication, and their production may be related to viral fitness (66). Conversely, other studies have suggested that these long mRNAs may represent aborted replication products or an extended form of "leader RNA, " similar to those described in certain other negative-strand RNA viruses (10,68).
In BoDV-1, the 5′ untranslated region of the X/P mRNA contains a short uORF that modulates ribosomal scanning and translation initiation, contributing to controlled protein expression (13). A comparable feature was not detected for WhSBV, suggesting that the translation of the multicistronic mRNAs is regulated by leaky scanning rather than ribosomal termination/reinitiation, as seen in other mononegaviruses (69). Future studies are needed to investigate the function of these multicistronic mRNAs, particularly those co-expressing N with X/P and G, and to determine whether they encode proteins or contribute to viral fitness.
The relatively low expression level of the L gene observed in WhSBV may be explained by a combination of transcriptional and posttranscriptional mechanisms. One potential factor is the presence of a transcription termination signal (T3) located within the L ORF. In BoDV-1, the T3 signal mediates premature transcription termination of the L ORF, resulting in lower transcript abundance and consequently lower protein expression (10). A similar regulatory structure in WhSBV could result in early termination of L gene transcripts, as supported by our transcriptional profiling, which shows high coverage starting at S3 and ending at T3. Additionally, PCR and Sanger sequencing revealed the presence of introns within the L ORF, suggesting a splicing event that may disrupt the integrity of the L ORF. This could serve to downregulate L expression while allowing the expression of the overlapping M ORF. This phenomenon has also been reported in BoDV-1, where alternative splicing within the L gene results in expression of the overlapping G ORF (8,11,61). Although L transcripts were detected by Northern blot, proteomic analysis did not identify the L protein, further supporting its comparable low expression levels. Collectively, these findings suggest that the L gene in WhSBV may be tightly controlled to finetune L protein levels. Future studies using more sensitive proteomic techniques and functional assays are warranted to clarify whether L protein expression is truly low or simply undetectable under current conditions. Additionally, further investigation is needed to elucidate the functional relevance of the observed introns. Based on the current data, possible hypotheses include the following: facilitat ing increased expression of the M gene by truncating the L ORF, regulating L expres sion at low levels, or generating an alternative transcript encoding a yetunidentified protein. A similar mechanism has been observed for vesicular stomatitis virus (VSV; family Rhabdoviridae), which controls L gene expression by polycistronic transcription, co-expressing it with other viral genes (70). Ultimately, the transcriptional regulation of WhSBV, including both polycistronic and spliced transcripts, appears to play a role in balancing viral protein levels and minimizing host cell stress, while optimizing viral replication and assembly.
## Conclusion
This study provides a molecular characterization of WhSBV, highlighting its evolution ary adaptation to cypriniform hosts. The virus replicates efficiently in cypriniform cell cultures, establishing persistent infections without inducing any visible cell damage and suppressing the host's early innate immune responses. Its limited or absent replication in non-cypriniform cultures highlights the need for in vivo validation to determine its full host range. Additionally, this study sheds light on the transcriptional mechanisms of WhSBV, revealing its regulatory strategies for modulating gene expression. Further research is essential to unravel the molecular mechanisms underlying WhSBV persis tence and to assess its ecological impact, particularly in aquaculture, where persistent viral infections could significantly impact fish health and disease management. Our findings also emphasize the importance of transitioning from in silico screening to in vitro validation-a critical step in bioinformatics-driven virology research. Although metagenomic approaches facilitate virus discovery, a key limitation is often the lack of direct validation from real virus samples.
## MATERIALS AND METHODS
## Nucleic acid extraction and metagenomic sequencing
A total of 48 fish cell lines were provided by the Collection of Cell Lines in Veterinary Medicine (CCLV) at the FriedrichLoefflerInstitut, Riems, Germany (see Table S1 for a complete list of cell lines). Fresh-frozen cell lines were lysed in 1 mL TRIzol LS reagent (Life Technologies) and shaken vigorously for 10 min at room temperature. After the addition of 200 µL chloroform (Carl Roth) and centrifuging at 13,000 × g, 10 min, 4°C, 200 µL of the upper aqueous phase was used for RNA extraction. Total RNA was extracted using the Agencourt RNAdvance Tissue Kit (Beckman Coulter) with a KingFisher Flex Purification System (Thermo Fisher Scientific) according to the manufacturer's instruc tions. DNase I digestion (Qiagen) was carried out prior to the final elution to remove genomic DNA residues. The quantity of total RNA was determined using the NanoDrop Lite Spectrophotometer (Thermo Fisher Scientific). Poly(A) RNA was isolated using the Dynabeads mRNA DIRECT Purification Kit (Invitrogen) following the manufacturer's instructions. Fragmentation and strandspecific construction of whole-transcriptome libraries was performed using the Collibri Stranded RNA Library Prep Kit for Illumina Systems (Invitrogen). The quality and integrity of the RNA and the final library were verified using the Agilent TapeStation 4150 (Agilent Technologies) with appropriate chips and reagents. The final libraries were quantified using the Qubit dsDNA HS Assay Kit (Invitrogen) in combination with the Qubit 2.0 Fluorometer (Invitrogen). Equimolar amounts of the library were then sequenced using Novaseq 6000 with SP flow cell in the single-end 101 bp mode (CeGaT) according to the manufacturer's instructions.
## Metagenomic analysis
The raw reads obtained from the 48 fish cell lines were trimmed using TrimGalore (v0.6.10 [71]) and cutadapt (v4.0 [72]). The metagenomic pipeline SqueezeMeta (v1.6.4 [73]) was then used to individually assemble the trimmed reads from each data set and then sequentially merge them into a single transcriptome ("seqmerge" mode). Poly(A/T) sequences at the sequence termini of the assembled and merged transcripts were then removed using cutadapt (v4.0 [72]). The polished transcriptome was then used as input for taxonomic assignment and abundance estimation using SqueezeMeta (v1.6.4 [73], database version Sep-3-2023). The results were further analyzed using the R (v4.3.1 [74]) package SQMtools (v1.6.3 [73]).
## Validation of WhSBV genome termini
Persistently WhSBV-infected grass carp swim bladder (GCSB) cell lines GCSB1441 (CCLV no. 1441) and GCSB1542 (1542; see Table S1) were grown in non-vented T25 tissue culture flasks (Corning) in a suitable growth medium supplemented with 10% fetal bovine serum (FBS) to 95%-100% confluence. Total RNA was extracted using the TRIzol-chloroform method, followed by purification using the RNeasy Mini Kit (Qiagen) and DNase I digestion prior to elution. The quality and quantity of the RNA were assessed using the Agilent TapeStation 4150 in combination with appropriate chips and reagents, and the NanoDrop Lite spectrophotometer.
The 5′-Rapid Amplification of cDNA Ends (RACE) kit, version 2.0 (Invitrogen), was used to determine the 3' end of the WhSBV genome. Reverse transcription (RT) was performed on the extracted RNA using a genespecific primer WhSBV-945-R (0.4 µM) (see Table S5 for a complete list of primers). The resulting cDNA was then digested with RNase cocktail (Invitrogen) at 37°C for 20 min and subsequently purified using AMPure XP beads (Beckman Coulter). A tailing reaction was then performed using dCTP and dATP (New England Biolabs). A second round of amplification was performed using gene-spe cific primers WhSBV-777-R (0.2 µM) and the AAP primer (0.2 µM) for purified C-tailed cDNA or the adapter primer (AP) (0.2 µM) for purified A-tailed cDNA (see Table S5). The hemi-nested PCR was performed with genespecific primers WhSBV-545-R (0.2 µM) and the AUAP primer (0.2 µM) (see Table S5). To identify the 5' end, total RNA samples were treated with the E. coli Poly(A) Polymerase Kit (New England Biolabs) according to the manufacturer's instructions in order to add poly(A) tails. The poly(A) tailed RNA was then purified using AMPure XP beads. This purified poly(A) tailed RNA was used as input to the 3′-RACE system for rapid amplification of cDNA ends (Invitrogen), using the AP primer (see Table S5) for RT. The cDNA was then digested with an RNase cocktail and purified using AMPure XP beads. For the second round of amplification, six different genespecific primers (WhSBV_7388F, WhSBV_7528F, WhSBV_7911F, WhSBV_8215F, WhSBV_8737F, and WhSBV_8901F; 0.2 µm) were combined with the AUAP primer (see Table S5), for individual reactions. The resulting PCR amplicons were separated by 2% agarose gel electrophoresis. The remaining product was purified with AMPure XP beads and sequenced using the BigDye Terminator v1.1 Cycle Sequencing Kit on a 3500 Genetic Analyzer (Applied Biosystems). The final WhSBV genome sequences were uploaded to GenBank under accession number PV171101.
## RT-qPCR and qPCR assays for specific detection of WhSBV
For the following assays, persistently WhSBV-infected GCSB1441 and GCSB1542 cells and the WhSBV-uninfected GK (0119) cell line (see Table S1) were grown to 95%-100% confluence in non-vented T25 tissue culture flasks using appropriate growth medium supplemented with 10% FBS. Cell pellets were harvested using Alsever's trypsin-versusversus-solution (ATV; FLI, Riems) after discarding the supernatant, and RNA/DNA was extracted using the NucleoMag VET kit (Macherey-Nagel) according to the manufactur er's instructions.
WhSBVspecific RT-qPCR assays were established by designing primers and probe sets targeting the viral N, P, G, M, and L genes (see Table S5 for a complete list of primers and probes). WhSBVspecific RNA (and potentially DNA) was detected using the AgPath-ID one-step RT-PCR reagents (Thermo Fisher Scientific). Briefly, 5 µL of extracted RNA was reverse-transcribed and amplified in a reaction mix of 25 µL total volume containing the specific primer/probe mix (final primer and probe concentration: 0.4 and 0.2 µM). For the internal control, primers ACT-1030-F (final concentration: 0.2 µM) and ACT-1135-R (0.2 µM) along with probe ACT-1081-HEX (0.2 µM) were included in each reaction to target the cellular actin beta (ACTB) gene (75). The reaction was performed on a Bio-Rad CFX96 qPCR cycler (Bio-Rad) using the following cycling conditions: 45°C for 10 min, 95°C for 10 min, 42 cycles of 95°C for 15 s, 57°C for 20 s and 72°C for 30 s. The efficiency and limit of detection of the WhSBV G gene RT-qPCR assay were tested using serial dilutions. A Cq cutoff value of 37 was set as the limit of detection for all performed assays.
To exclude the presence of WhSBV sequences as endogenous viral elements (EVE) integrated into the cellular genome of GCSB1441, GCSB1542, and GK, we specifically tested extracted DNA from these cells using a non-reverse transcription qPCR system. WhSBV cDNA from the 3′-RACE experiment, as previously described, was used as a positive control. The Maxima Probe/ROX qPCR Master Mix (Thermo Fisher Scientific) was used according to the manufacturer's protocol, targeting the viral genes N, P, G, M, and L (final primer and probe concentrations: 0.4 and 0.2 µM) and the cellular ACTB gene (final primer and probe concentrations: 0.2 and 0.2 µM). The reaction was performed on a Bio-Rad CFX96 qPCR cycler using the following cycling conditions: 50°C for 2 min, 95°C for 10 min, 40 cycles of 95°C for 15 s, 60°C for 30 s, and 72°C for 30 s.
## Optimization of WhSBV inoculum preparation
Persistently WhSBV-infected GCSB1441 cells were maintained in non-vented T75 culture flasks (75 cm²; Corning) at 26°C until reaching 95%-100% confluency in growth medium supplemented with 10% FBS. Five different methods were evaluated to prepare the WhSBV inoculum in order to determine the most efficient and reliable protocol for subsequent experiments: (i) 5 mL of the supernatant (equivalent to 1 mL per 15 cm² of infected cell layer) was retained, and the cells were subjected to three freeze/thaw cycles freezing at -80°C and thawing at room temperature for 30 minutes each. (ii) After three freeze/thaw cycles as described above, the resulting cell suspension was collected in a Falcon tube (Sarstedt), sonicated (Branson Sonifier 450), and centrifuged at 1,200 × g for 20 minutes at 4°C, and only the supernatant was collected. (iii) Cell lines were suspended in 5 mL supernatant, which was subsequently sonicated and centrifuged as described above, and only the supernatant was collected. (iv) The culture medium was replaced with 5 mL hypertonic medium containing 150 mM NaCl. After 2 h of incubation at 26°C, the cells were scraped using a cell scraper, and the cell suspension was centrifuged as previously described, and only the supernatant was collected. (v) 5 mL of the untreated supernatant was collected. All inocula were stored at -80°C until further use.
In order to test the infectivity of the different inocula, CCB cells (see Table S2) were seeded in 12-well plates (3.8 cm 2 per well; Corning) and grown to 95%-100% confluency in the appropriate medium supplemented with 10% FBS at 26°C. Inoculation was performed at an infection ratio of 1:0.8, corresponding to the ratio between the area of the cell layer from which the inoculum originated and the area of the cell layer inoculated. Specifically, 300 µL of the medium was replaced by 300 µL of WhSBV inoculum. Infections were conducted in three independent replicate wells for each time point. Cell pellets were harvested at 0,4,8,24,48,72,96,120,144,168,192,216, and 240 hpi. At each time point, supernatants were discarded, and cell pellets were collected by trypsinization using ATV. RNA was extracted using the NucleoMag VET kit as described previously. The extracted nucleic acids were used for RT-qPCR analysis, as described above, using specific primer/probe mixes for WhSBV G and ACTB. A Delta-Delta Cq (ΔΔCq) analysis was conducted using R (version 4.3.1; [74) to compare different inoculation methods, with ACTB serving as the reference gene and method (i) as the reference method. To ensure consistency, a single batch of inoculum prepared by method (i) was used throughout most assays.
## Experimental infection of different cell cultures with WhSBV
Bony fish cells RT/F (CCLV no. 0088), CaPi (0112), EPC (0173), RTG-2/f (0686), CCB (0816), SAF-1 (0826), CHSE-214 (1104), GTS-9 (1388), ZF4 (1492), and TiB (1550), the reptilian cells SKL-R (0484), VH-2 (1092), and CDSK (1536), the avian cells QM7 (0466) and DF-1 (1529), and the mammalian cells Vero (0015) and C6 (1452) were obtained from CCLV (see Table S2 for a complete list of cell lines). All fish and reptile cell lines were grown in non-vented T25 flasks (25 cm²; Corning) to 95%-100% confluence in an appropriate 10 mL growth medium supplemented with 10% FBS. Incubation temperatures were set at 20°C for the fish cells CHSE-214 and RTG-2/f, 26°C for all other fish cells and reptilian cell SKL-R, and 28°C for the reptile cells VH-2 and CDSK. Mammalian and avian cells were grown in vented T25 flasks at 37°C in a humidified atmosphere containing 5% CO 2 .
To initiate the infection experiment, cells were grown to confluence in T25 flasks, as described above. Prior to inoculation (T0), 1 mL of the supernatant was collected from each cell culture. The cells were then inoculated with 1 mL of the WhSBV inoculum, prepared using the optimized protocol involving three freeze/thaw cycles, as previously described. The inoculum was applied at a ratio of 1:1.7, as described above. Cultures were then incubated for 1 h under the above conditions. After this incubation, a further 1 mL of the supernatant was collected (T1). Cells were then incubated and split at regular intervals of 2-3 days at a 2-fold splitting factor for a total of 4-6 passages (P1-6). The splitting process involved collecting the supernatant before each passage, washing and detachment of the cells with ATV, and collecting half of the detached cells for analysis. The remaining cells were used for further culture by adding fresh growth medium supplemented with 10% FBS. The collected supernatants and cell pellets were stored at -80°C for subsequent RNA extraction and RT-qPCR analysis. Negative controls were treated in the same way as infected samples, except that the WhSBV inoculum was replaced by 1 mL RNase-free water (Qiagen). The infection experiment was performed in two independent runs for each cell line. As the infection of reptile, avian, and mammalian cell lines was performed separately from the comparison of different fish cell lines, the cypriniform cell line GTS-9 was used as a positive control to validate the infectivity of the inoculum.
Nucleic acid extraction from supernatant and cell pellet samples was performed using the NucleoMag VET kit according to the manufacturer's instructions. RT-qPCR was then performed using WhSBV G gene and ACTB specific RT-qPCR primer/probe mixes, as described earlier.
## Electron microscopy
WhSBV-infected and uninfected ZF4 cells (see Table S2) from the previous infec tion experiment and persistently WhSBV-infected GCSB1441 cells were transferred to non-vented T75 flasks for electron microscopy analysis. Cells were fixed in 2.5% glutaraldehyde buffered in 0.1 M sodium cacodylate pH 7.2 (SERVA Electrophoresis), and 1% aqueous OsO 4 was used for postfixation and 2.5% uranyl acetate in ethanol for en bloc staining (SERVA Electrophoresis). After a graded dehydration in ethanol, the samples were cleared in propylene oxide and infiltrated with Glycid Ether 100 (SERVA Electrophoresis). For polymerization, samples were incubated at 60°C for 3 days. Ultrathin sections were transferred to formvar-coated nickel grids (slot grids; Plano). All grids were counterstained with uranyl acetate and lead citrate before examination on a Talos F200i transmission electron microscope (FEI) at an accelerating voltage of 80 kV.
## Specific detection of viral RNA using RNA in situ hybridization
The cypriniform cell lines (EPC and CCB), the non-cypriniform fish cells (RTG-2/f and TiB), the mammalian Vero line, and the avian DF-1 cells (see Table S2) were maintained in non-vented T75 flasks in a suitable medium supplemented with 10% FBS until they reached 95%-100% confluence. Cultures were maintained under the conditions described in the previous infection experiment. For infection, 3 mL of the medium was replaced with an equal volume of the inoculum prepared via three freeze/thaw cycles, to achieve an infection ratio of 1:1.7, as previously described. For cypriniform and non-cypriniform fish, cells were harvested from two independent experiments at early and late time points: 0, 4, 8, 24, and 48 hpi for the early time points and 0, 72, 96, 120, and 240 hpi for the late time points. In a separate experiment, the Vero and DF-1 cell lines were harvested exclusively at 0 and 240 hpi, alongside the CCB cell line, which served as a positive cypriniform control. All cells were harvested using a cell scraper. The harvested cells were collected in 50 mL Falcon tubes, centrifuged at 3,500 rpm for 15 minutes, and the supernatants discarded. The cell pellets were then fixed in 10% neutral buffered formalin (Carl Roth) for at least 24 h. After paraffin embedding, 4-µm-thick sections were processed for subsequent RNA in situ hybridization (RNA ISH) using RNAScope 2-5 HD Reagent Kit Red and custom-designed probes targeting WhSBV (-)RNA (genomic RNA; WhSBV-L-O2, catalog number 1567161-C1, Advanced Cell Diagnostics, Hayward, USA) on all samples according to the manufacturer's instructions. In addition, custom-designed probes targeting WhSBV (+)RNA (mRNA and antigenomic RNA; WhSBV-G-O1, catalog number 1567171-C1) were used on CCB samples. As technical controls, probes against the genes for peptidylprolyl isomerase B (PPIB) and dihydrodipicolinate reductase (DapB) were included in each run. In addition, routine staining using hematoxylin-eosin was performed on all samples taken at 0 and 240 hpi for histopathologic evaluation. All slides were scanned using a Hamamatsu S60 scanner and evaluated using NDPview.2 plus software (version 2.8.24, Hamamatsu Photonics, K.K. Japan) by a boardcertified pathologist (AB, DiplECVP). The abundance of infected cells was recorded as a percent age in a 10 × 10 grid (size 100 × 100 µm per grid cell). Thus, a positive signal within a grid cell resulted in one percentage point.
In a separate kinetics experiment to correlate with the RNA ISH results, the same fish cell lines as above were seeded in 12-well plates (Corning) and maintained to 95%-100% confluence in appropriate medium supplemented with 10% FBS. Fish cultures were maintained at 26 or 20°C (for RTG-2/f ). Inoculation was performed at a 1:0.8 infection ratio as described above, with 300 µL of medium replaced by 300 µL of WhSBV inoculum. Infection of each cell line was performed in three independent replicate wells for each time point. Cell pellets were harvested at 0,4,8,24,48,72,96,120,144,168,192,216, and 240 hpi. At each time point, supernatants were discarded, and cell pellets were collected by trypsinization using ATV. RNA was extracted using the NucleoMag VET kit as described previously. RT-qPCR was then performed using G gene and ACTBspecific RT-qPCR primer/probe mixes.
## Gene expression analysis of early and late WhSBV infection
The CCB cells (see Table S2) were used to study the effect of WhSBV on cellular gene expression. Cells were seeded in 6-well plates (9.6 cm 2 per well; Corning) and cultured for 24 hours to 95%-100% confluence in 10% FBS supplemented medium. They were maintained under the conditions described in the previous kinetics experiment. Infection with WhSBV was performed at a 1:1.3 ratio as described above by replacing 500 µL of the culture medium in each well with 500 µL of WhSBV inoculum prepared via three freeze/ thaw cycles. Negative controls were inoculated with 500 µL of the same medium used for persistently WhSBV-infected GCSB1441 cultivation, replacing 500 µL of the cell culture medium. Cell pellets were collected in two independent experiments to assess early and late infection stages. In the early infection experiment, mock-infected and WhSBV-infec ted cell pellets were collected at 4, 8, 24, and 48 hpi. For the late-stage experiment, samples were collected at 72, 96, 120, and 240 hpi. In addition, pre-infection cell pellets were collected at 7 and 24 hours before inoculation, the latter representing the 0 hpi time point. Infection was performed in three independent replicate wells for each time point. Cell pellets were collected by trypsinization after supernatants were discarded. Total RNA was extracted using Trizol, chloroform, and the Agencourt RNAdvance Tissue Kit, as previously described. RT-qPCR was performed on all samples using G-gene and ACTBspecific RT-qPCR primer/probe mixes to validate samples prior to sequencing. Transcriptomic libraries were prepared as previously described, but including ERCC spike-in mix (Invitrogen) as an internal control. Subsequently, the libraries were pooled and sent for sequencing on an Illumina NovaSeq 6000 system (Novogene GmbH) running in the paired-end 150 bp mode. Raw reads were subsequently trimmed using TrimGalore (v0.6.10; [71]) and then analyzed using the nf-core "RNAseq" pipeline (version 3.14.0; [76,77]). In detail, the common carp reference genome GCF_018340385.1 was used for splice-aware mapping with STAR and subsequent quantification with Salmon. The quantification data were then further analyzed in R (version 4.3.1; [74]) and differentially expressed genes (DEGs) were deduced using the DESeq2 package (v1.40; [78]). Log2-fold changes were shrunk using the function "lfcShrink" with the "apeglm" shrinkage estimator (79). The cutoff values for DEGs were set to absolute shrinked log2-fold change > 1 and adjusted P-value < 0.05. To explore global gene expression patterns, PCA was performed using R (version 4.3.1; [74]). Transcriptomic data have been deposited in ArrayExpress under accession E-MTAB-14894.
## Prediction of regulatory and splice sites in the WhSBV genome
The trimmed raw reads from the metagenomic RNA data sets from persistently WhSBVinfected GCSB1441 and GCSB1542 cells (see Table S1) were mapped back to the RACE corrected WhSBV genome (PV171101) using the "bbmap" function from BBTools (version 39.33 [80]). Duplicates were removed from mapped reads using the "MarkDuplicates" function from the Picard toolkit (version 2.20.4 [81]), and reads were separated by mapping orientation using Samtools (version 1.22.1 [82]). The genome coverage for forward and reverse deduplicated reads was obtained using Bedtools (version 2.31.1 [83]) and was plotted using R (version 4.3.1 [74]). Poly(A)-addition sites were deduced using the same data set and reference with ContextMap (version 2.7.9 [84]) and bowtie (version 1.3.1 [85]). Potential introns were detected using STAR (version 2.7.11b [86]) running in the "2-pass" mode, enabling sensitive novel junction discovery. The function "FIMO" from the MEME Suite (v5.5.2 [87]) was used to identify the conserved motif "AKUUAAYAAAAACAUGAA" (2) of potential transcription termination and start sites.
## Validation of predicted splice sites in the WhSBV genome
To experimentally validate each predicted splice site, primers were designed to amplify the specific genomic regions by RT-PCR. Purified RNA was extracted by using the RNeasy Mini kit as described above. cDNA from poly(A) RNA was synthesized using the Protoscript II First Strand cDNA Synthesis Kit and oligo-dT primers (New England Biolabs). The cDNA products were then digested with an RNase cocktail and purified using AMPure XP beads. Conventional Taq-PCR was performed using the Accuprime Taq DNA Polymerase System (Invitrogen) with primers specific for the predicted splice sites within the viral M and L genes (see Table S5). RNase-free water was used instead of cDNA as a negative control. The amplified DNA products were run on a 2% agarose gel, and bands were excised. The DNA was extracted using a QIAquick Gel Extraction Kit (Qiagen) and then sequenced using the BigDye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems) on a 3500 Genetic Analyzer (Applied Biosystems).
## Specific detection of viral RNA using Northern blot
RNA was extracted from the persistently WhSBV-infected GCSB1441, WhSBV-infected, and uninfected EPC cells (see Tables S1 andS2) from the previous infection experiment using the TRIzol-chloroform method and the RNeasy Mini Kit with DNase I (Qiagen), as previously described. RNA quality and quantity were assessed using the TapeStation System 4150 (Agilent Technologies) and NanoDrop Lite spectrophotometer, as described previously. Subsequently, 2-4 µg of total RNA was glyoxylated using Glyoxal (Sigma-Aldrich) at 56°C for 45 min before separation on a 0.9% denaturing agarose gel (4.7% formaldehyde, Carl Roth) in a phosphate buffer system. RNA was transferred overnight to Hybond-N membranes (Cytiva Amersham) using SSC transfer buffer. RNA was crosslinked to the membrane, and viral RNA (mRNA and genomic RNA) was detected using DNA probes. Probes were generated by RT-PCR using specific forward and reverse primers against each gene of WhSBV (see Table S5) using the Platinum Script III One-step RT-PCR Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. The amplified cDNA products were then run on a 2% agarose gel. After amplification, the products were purified using AMPure XP beads according to the manufacturer's protocol. The quantity of purified products was measured using a NanoDrop Lite spectropho tometer and labeled with 32 P-CTP (Hartmann-Analytic) using a Nick translation kit (GE Healthcare). Hybridization was performed overnight at 60°C-65°C, and the signal was analyzed after washing using the CR35 phosphoimager system (Dürr Medical).
## Specific detection of WhSBV proteins
Approximately 3 × 10 5 persistently WhSBV-infected GCSB1441 cells were harvested, resuspended in 1 × LDS Buffer (Thermo Fisher Scientific) containing 100 mM DTT (Sigma-Aldrich), and heated for 10 min at 70°C shaking at 1,400 rpm in a Thermomixer (Eppendorf ). Reconstitution buffer (50 mM HEPES/KOH, pH 7.5, 100 mM NaCl, 1 mM EDTA, pH 8.0, 0.5% [wt/vol] Sodium deoxycholate, 1% [vol/vol] Triton X-100) was added to a final volume of 100 µL. Samples were reduced in 10 mM DTT for 30 min at 60°C, followed by alkylation in 25 mM iodoacetamide (Sigma-Aldrich) for 25 min at RT in the dark with orbital shaking at 800 rpm in a Thermomixer (Eppendorf ). The reaction was quenched by incubation with 10 mM DTT f.c. for 5 min at RT. For protein binding, 125 µg hydrophilic and 125 µg hydrophobic Sera-Mag CarboxylateModified Magnetic Particles (GE Healthcare) together with 150 µL absolute ethanol (Carl Roth) were added to each sample and incubated for 15 min at RT at 1000 rpm in a Thermomixer (Eppendorf ). Beads were washed three times with 80% LC-/MS-grade ethanol (Supelco). Residual ethanol was removed prior to adding 1 µg LC-grade trypsin (Serva) in 100 µL of 50 mM ammonium bicarbonate buffer pH 8. After incubation for 4 h at 37°C and 1,000 rpm in a Thermomixer (Eppendorf ), peptides were collected and desalted on StageTips (88). The amount of 400 ng digested peptides was analyzed with a nanoElute2 HPLC system (Bruker) coupled to a TimsTOF HT mass spectrometer (Bruker). Peptides were separated on an Ultimate CSI 25 × 75 C18 UHPLC column (Ionopticks) by running an optimized 100 min gradient of 2 to 95% MS-grade acetonitrile with 0.1% formic acid at 250 nl/min at 50°C. The mass spectrometer was operated with the manufacturer-provi ded DDA_PASEF_1.1 sec_cycletime method. Mass spectrometry raw files were processed with MaxQuant (version 2.4.13.0) (14) using a custom virus database (six entries) and the C. idella database (C_idella_female_genemodels.v1; 32,811 entries) with standard settings.
## References
1. Kuhn, Dürrwald, Bào et al. (2015) "Taxonomic reorganization of the family Bornaviridae" *Arch Virol*
2. Eshak, Rubbenstroth, Beer et al. (2023) "Diving deep into fish bornaviruses: uncovering hidden diversity and transcriptional strategies through comprehensive data mining" *Virus Evol*
3. Rubbenstroth, Briese, Dürrwald et al. (2021) "ICTV Virus Taxonomy Profile: Bornaviri dae" *J Gen Virol*
4. Niller, Angstwurm, Rubbenstroth et al. (2020) "Zoonotic spillover infections with Borna disease virus 1 leading to fatal human encephalitis, 1999-2019: an epidemiological investigation" *Lancet Infect Dis*
5. Hoffmann, Tappe, Höper et al. (2015) "A variegated squirrel bornavirus associated with fatal human encephalitis" *N Engl J Med*
6. Shi, Lin, Chen et al. (2018) "The evolutionary history of vertebrate RNA viruses" *Nature*
7. Hyndman, Shilton, Stenglein et al. (2018) "Divergent bornaviruses from Australian carpet pythons with neurological disease date the origin of extant Bornaviridae prior to the end-cretaceous extinction" *PLoS Pathog*
8. Briese, Schneemann, Lewis et al. (1994) "Genomic organization of Borna disease virus" *Proc Natl Acad Sci*
9. Briese, De La Torre, Ludwig et al. (1992) "Borna disease virus, a negative-strand RNA virus, transcribes in the nucleus of infected cells" *Proc Natl Acad Sci*
10. Schneemann, Schneider, Kim et al. (1994) "Identification of signal sequences that control transcription of borna disease virus, a nonsegmented, negative-strand RNA virus" *J Virol*
11. Schneider, Schneemann, Lipkin (1994) "RNA splicing in Borna disease virus, a nonsegmented, negative-strand RNA virus" *J Virol*
12. Tomonaga, Kobayashi, Lee et al. (2000) "Identification of alternative splicing and negative splicing activity of a nonsegmented negative-strand RNA virus, Borna disease virus" *Proc Natl Acad Sci*
13. Watanabe, Ohtaki, Hayashi et al. (2009) "Autoge nous translational regulation of the Borna disease virus negative control factor X from polycistronic mRNA using host RNA helicases" *PLoS Pathog*
14. Cox, Mann (2008) "MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification" *Nat Biotechnol*
15. Costa, Mifsud, Gilligan et al. (2021) "Metagenomic sequencing reveals a lack of virus exchange between native and invasive freshwater fish across the Murray-Darling Basin" *Australia. Virus Evol*
16. Brandolini, Taddei, Marino et al. (2021) "Correlating qRT-PCR, dPCR and viral titration for the identification and quantification of SARS-CoV-2: a new approach for infection management" *Viruses*
17. Hitchman, Siaterli, Nixon et al. (2007) "Quantitative real-time PCR for rapid and accurate titration of recombinant baculovirus particles" *Biotechnol Bioeng*
18. Falsey, Formica, Treanor et al. (2003) "Comparison of quantitative reverse transcription-PCR to viral culture for assessment of respiratory syncytial virus shedding" *J Clin Microbiol*
19. Herzog, Rott (1980) "Replication of Borna disease virus in cell cultures" *Med Microbiol Immunol*
20. Rinder, Ackermann, Kempf et al. (2009) "Broad tissue and cell tropism of avian bornavirus in parrots with proventricular dilatation disease" *J Virol*
22. Pham, Leacy, Deng et al. (2020) "Isolation of Ontario aquatic bird bornavirus 1 and characterization of its replication in immortalized avian cell lines" *Virol J*
23. Rubbenstroth, Rinder, Kaspers et al. (2012) "Efficient isolation of avian bornaviruses (ABV) from naturally infected psittacine birds and identification of a new ABV genotype from a salmon-crested cockatoo (Cacatua moluccensis)" *Vet Microbiol*
24. Rubbenstroth, Schmidt, Rinder et al. (2014) "Discovery of a new avian bornavirus genotype in estrildid finches (Estrildidae) in Germany" *Vet Microbiol*
25. Rubbenstroth, Rinder, Stein et al. (2013) "Avian bornaviruses are widely distributed in canary birds (Serinus canaria f. domestica)" *Vet Microbiol*
26. Schlottau, Nobach, Herden et al. (2020) "First isolation, in-vivo and genomic characterization of zoonotic variegated squirrel Bornavirus 1 (VSBV-1) isolates" *Emerg Microbes Infect*
27. Nowotny, Kolodziejek, Jehle et al. (2000) "Isolation and characterization of a new subtype of Borna disease virus" *J Virol*
28. Davidovich, Dishon, Ilouze et al. (2007) "Susceptibility of cyprinid cultured cells to cyprinid herpesvirus 3" *Arch Virol*
29. Marsella, Buratin, Pascoli et al. (2025) "Temperature impact on replication and virulence of Full-Length Text Journal of Virology November"
30. "European infectious hematopoietic necrosis viruses" *Aquaculture*
31. Li, Yang, Ye et al. (2021) "Integrated metabolomics and transcriptomics analyses reveal metabolic landscape in neuronal cells during JEV infection" *Virol Sin*
32. Barton, Iwama (1991) "Physiological changes in fish from stress in aquaculture with emphasis on the response and effects of corticoste roids" *Annu Rev Fish Dis*
33. Ramnani, Powell, Shetty et al. (2023) "Viral hemorrhagic septicemia virus activates integrated stress response pathway and induces stress granules to regulate virus replication" *Viruses*
34. Matsumoto, Hayashi, Omori et al. (2012) "Bornavirus closely associates and segregates with host chromosomes to ensure persistent intranuclear infection" *Cell Host Microbe*
35. Tomonaga, Kobayashi, Ikuta (2002) "Molecular and cellular biology of Borna disease virus infection" *Microbes Infect*
36. Zimmermann, Breter, Rudolph et al. (1994) "Borna disease virus: immunoelectron microscopic characterization of cell-free virus and further information about the genome" *J Virol*
37. Kohno, Goto, Takasaki et al. (1999) "Fine structure and morphogenesis of Borna disease virus" *J Virol*
38. Wünschmann, Honkavuori, Briese et al. (2011) "Antigen tissue distribution of Avian bornavirus (ABV) in psittacine birds with natural spontaneous proventricular dilatation disease and ABV genotype 1 infection" *J Vet Diagn Invest*
39. Hoppes, Gray, Payne et al. (2010) "The isolation, pathogenesis, diagnosis, transmission, and control of avian bornavirus and proventricular dilatation disease" *Vet Clin North Am Exot Anim Pract*
40. Gancz, Kistler, Greninger et al. (2009) "Experimental induction of proventricular dilatation disease in cockatiels (Nymphicus hollandicus) inoculated with brain homogenates containing avian bornavirus 4" *Virol J*
41. Altmann, Mellon, Distel et al. (2003) "Molecular and functional analysis of an interferon gene from the zebrafish, Danio rerio" *J Virol*
42. Schultz, Kaspers, Staeheli (2004) "The interferon system of nonmammalian vertebrates" *Dev Comp Immunol*
43. Zou, Tafalla, Truckle et al. (2007) "Identification of a second group of type I IFNs in fish sheds light on IFN evolution in vertebrates" *J Immunol*
44. Takeuchi, Akira (2010) "Pattern recognition receptors and inflamma tion" *Cell*
45. Rao, Su (2015) "Insights into the antiviral immunity against grass carp (Ctenopharyngodon idella) reovirus (GCRV) in grass carp" *J Immunol Res*
46. Chang, Collet, Nie et al. (2011) "Expression and functional characterization of the RIG-I-like receptors MDA5 and LGP2 in Rainbow trout (Oncorhynchus mykiss)" *J Virol*
47. Zou, Bird, Secombes (2010) "Antiviral sensing in teleost fish" *Curr Pharm Des*
48. Biacchesi, Leberre, Lamoureux et al. (2009) "Mitochondrial antiviral signaling protein plays a major role in induction of the fish innate immune response against RNA and DNA viruses" *J Virol*
49. Verrier, Langevin, Benmansour et al. (2011) "Early antiviral response and virus-induced genes in fish" *Dev Comp Immunol*
50. Chang, Zou, Nie et al. (2013) "Intracellular interferons in fish: a unique means to combat viral infection" *PLoS Pathog*
51. Nehyba, Hrdlicková, Bose (2009) "Dynamic evolution of immune system regulators: the history of the interferon regulatory factor family" *Mol Biol Evol*
52. Negishi, Taniguchi, Yanai (2018) "The interferon (IFN) class of cytokines and the IFN regulatory factor (IRF) transcription factor family" *Cold Spring Harb Perspect Biol*
53. Feng, Zhang, Zhang et al. (2015) "Zebrafish IRF1 regulates IFN antiviral response through binding to IFNϕ1 and IFNϕ3 promoters downstream of MyD88 signaling" *J Immunol*
54. Mclaughlin, Olsen, Dake et al. (2013) "The RasGAP gene, RASAL2, is a tumor and metastasis suppressor" *Cancer Cell*
55. Habjan, Andersson, Klingström et al. (2008) "Processing of genome 5' termini as a strategy of negative-strand RNA viruses to avoid RIG-I-dependent interferon induction" *PLoS One*
56. Reuter, Ackermann, Kothlow et al. (2010) "Avian bornaviruses escape recognition by the innate immune system" *Viruses*
57. Unterstab, Ludwig, Planz et al. (2005) "Viral targeting of the interferon-β-inducing Traf family member-associated NF-κB activator (TANK)-binding kinase-1" *Proc Natl Acad Sci*
58. Song, Kao, Zhai et al. (2013) "Borna disease virus nucleoprotein inhibits type I interferon induction through the interferon regulatory factor 7 pathway" *Biochem Biophys Res Commun*
59. Ortega-Villaizan, Del, Chico et al. (2022) "Fish innate immune response to viral infection-an overview of five major antiviral genes" *Viruses*
60. Xue, Lin, Han et al. (2021) "Ammonia stress-induced apoptosis by p53-BAX/BCL-2 signal pathway in hepatopancreas of common carp (Cyprinus carpio)" *Aquacult Int*
61. Meyer, Ly (2016) "Inhibition of innate immune responses is key to pathogenesis by arenaviruses" *J Virol*
62. Kane, Zang, Rihn et al. (2016) "Identification of interferon-stimulated genes with antiretroviral activity" *Cell Host Microbe*
63. Cubitt, Oldstone, Valcarcel et al. (1994) "RNA splicing contributes to the generation of mature mRNAs of Borna disease virus, a non-segmented negative strand RNA virus" *Virus Res*
64. Cubitt, Ly, De La Torre (2001) "Identification and characterization of a new intron in Borna disease virus" *J Gen Virol*
65. Amarasinghe, Ayllón, Bào et al. (2019) "Taxonomy of the order Mononegavirales: update" *Arch Virol*
66. Wignall-Fleming, Hughes, Vattipally et al. (2019) "Analysis of paramyxovirus transcription and replication by high-throughput sequencing" *J Virol*
67. Yu, Weng, Wu et al. (2017) "The lifecycle of the Ebola virus in host cells" *Oncotarget*
68. Poenisch, Wille, Staeheli et al. (2008) "Polymerase readthrough at the first transcription termination site contributes to regulation of borna disease virus gene expression" *J Virol*
69. (2025) *Full-Length Text Journal of Virology*
70. Schneemann, Schneider, Lamb et al. (1995) "The remarkable coding strategy of borna disease virus: a new member of the nonseg mented negative strand RNA viruses" *Virology (Auckl)*
71. Banerjee, Barik, De (1991) "Gene expression of nonsegmented negative strand RNA viruses" *Pharmacol Ther*
72. Firth, Brierley (2012) "Non-canonical translation in RNA viruses" *J Gen Virol*
73. Barr, Whelan, Wertz (2002) "Transcriptional control of the RNAdependent RNA polymerase of vesicular stomatitis virus" *Biochim Biophys Acta*
74. Krueger, James, Ewels et al. (2023) "FelixKrueger/TrimGalore: v0.6.10 -add default decompression path"
75. Martin (2011) "Cutadapt removes adapter sequences from highthroughput sequencing reads" *EMBnet J*
76. Tamames, Puente-Sánchez (2018) "SqueezeMeta, a highly portable, fully automatic metagenomic analysis pipeline" *Front Microbiol*
77. (2024) "R: a language and environment for statistical computing"
78. Wernike, Hoffmann, Kalthoff et al. (2011) "Development and validation of a triplex real-time PCR assay for the rapid detection and differentiation of wild-type and glycoprotein E-deleted vaccine strains of Bovine herpesvirus type 1" *J Virol Methods*
79. Ewels, Peltzer, Fillinger et al. (2020) "The nf-core framework for communitycurated bioinformatics pipelines" *Nat Biotechnol*
80. Patel, Ewels, Manning et al. (2024)
81. Love, Huber, Anders (2014) "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2" *Genome Biol*
82. Zhu, Ibrahim, Love (2019) "Heavy-tailed prior distributions for sequence count data: removing the noise and preserving large differences" *Bioinformatics*
83. Bushnell, Rood, Singer (2017) "BBMerge -accurate paired shotgun read merging via overlap" *PLoS One*
84. Picard Toolkit (2019)
85. Danecek, Bonfield, Liddle et al. (2021) "Twelve years of SAMtools and BCFtools. Gigascience 10:giab008"
86. Quinlan, Hall (2010) "BEDTools: a flexible suite of utilities for comparing genomic features" *Bioinformatics*
87. Bonfert, Kirner, Csaba et al. (2015) "ContextMap 2: fast and accurate context-based RNA-seq mapping" *BMC Bioinformatics*
88. Langmead, Trapnell, Pop et al. (2009) "Ultrafast and memoryefficient alignment of short DNA sequences to the human genome" *Genome Biol*
89. Dobin, Davis, Schlesinger et al. (2013) "STAR: ultrafast universal RNA-seq aligner" *Bioinformatics*
90. Bailey, Elkan (1994) "Fitting a mixture model by expectation maximization to discover motifs in biopolymers" *Proc Int Conf Intell Syst Mol Biol*
91. Rappsilber, Mann, Ishihama (2007) "Protocol for micropurification, enrichment, pre-fractionation and storage of peptides for proteomics using StageTips" *Nat Protoc*
92. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12697135&blobtype=pdf | # SERINC5 co-expressed with HIV-1 Env or present in a target membrane destabilizes small fusion pores leading to their collapse
Ruben Markosyan, Mariana Marin, Gregory Melikyan
## Abstract
Serine incorporator protein 5 (SERINC5) is a multipass membrane protein that reduces infectivity of retroviruses and other enveloped viruses through incorporat ing into budding virions and inhibiting their ability to fuse with target cells. Several mechanisms for anti-HIV activity of SERINC5 have been reported, including binding to Env glycoprotein, induction of conformational changes and destabilization of Env, as well as disruption of transmembrane asymmetry of viral envelope. All these reported anti-HIV mechanisms involve SERINC5 incorporation into progeny virions, while there is very little information regarding a potential antiviral activity of SERINC5 in target cells. Here, we show that SERINC5 expressed in target cells efficiently inhibits fusion with cells expressing Env glycoproteins of sensitive but not SERINC5-resistant HIV-1 strains. This activity of SERINC5 does not result from downregulation of CD4 or co-receptors on target cells or interference with the formation of prefusion Env/CD4/co-receptor complexes. We further demonstrate that SERINC5 destabilizes small fusion pores, causing their collapse, and that this block can be rescued by incorporation of phosphatidylserine into the plasma membrane of either effector (Env-expressing) or target cells. Interestingly, we did not detect a significant reduction in HIV-1 pseudovirus fusion with SERINC5-express ing target cells, while virus-mediated cell-cell fusion (fusion from without) was inhibi ted, suggesting a potential role of the virus's entry pathway in sensitivity to SERINC5 restriction. Collectively, our results reveal a novel mechanism of inhibition of HIV-1 fusion by SERINC5 through destabilization of small fusion pores in a manner that depends on lipid composition.
IMPORTANCE SERINC5 incorporates into virions produced by infected cells and inhibitsHIV-1 fusion with target cells through a poorly understood mechanism. Here, we show that SERINC5 blocks HIV-1 Env-mediated cell fusion when expressed in either effector (Env-expressing) or target cells. Inhibition of Env-mediated fusion by SERINC5 expressed in target cells is not through reduction in receptor or co-receptor expres sion or interference with Env's ability to engage a requisite number of receptors and co-receptors. We demonstrate that the block of fusion is at a post-hemifusion stage of small fusion pores that collapse when SERINC5 is present in effector or target membrane and that this block is rescued by incorporation of specific anionic lipids, such as phosphatidylserine. These findings reveal a previously unappreciated mode of HIV-1 restriction through destabilization of small fusion pores that occurs irrespective of SERINC5 localization in either fusing membranes.
S erine incorporator protein 3 and serine incorporator protein 5 (SERINC5) are 10-transmembrane domain proteins that reduce retrovirus infectivity by incorporat ing into virions and diminishing their ability to fuse with target cells (1)(2)(3)(4)(5). While SERINC expression is not increased upon interferon stimulation, endogenous levels of these proteins in CD4 T cells restrict HIV-1 infection (3,5). The cellular functions of the SERINC family proteins are not well understood (6,7), but SERINC5 has been recently shown to exhibit a lipid scrambling activity manifested in phosphatidylserine (PS) exposure on the virion surface (8,9). Interestingly, recent studies documented a broad antiviral activity of SERINC5 (hereafter abbreviated SER5) against diverse enveloped viruses, including the influenza A virus (IAV), SARS-CoV-2, classical swine fever virus, and others (10)(11)(12)(13)(14). Viruses counteract SER5 restriction by encoding proteins that bind to and degrade and/or remove SER5 from plasma membrane, thereby preventing its incorporation into virions. These include HIV-1 Nef, murine leukemia virus GlycoGag, equine infectious anemia virus S2, and SARS-CoV-2 ORF7a proteins (3,5,10,(15)(16)(17).
In the absence of Nef, HIV-1's resistance to SER5 restriction maps to Env glycoprotein, with lab-adapted Tier 1 strains being sensitive and primary isolates being relatively resistant to this protein (1, 3-5, 10, 18-20). The gp120 variable loop 3 (V3), which controls virus co-receptor tropism and Env stability (21), has been identified as the major determinant of SER5 resistance (1). The gp41 cytoplasmic tail has also been implicated in Env's sensitivity to SER5 (22), but this notion has not been confirmed by other groups, including ours (4,23). The mechanism of SER5's antiviral activity and virus resistance is not well understood and appears to be multifaceted. SER5 has been shown to alter the conformation of Env on virions, promote its spontaneous inactivation, and sensitize to neutralizing antibodies (1,4,18,19,24,25). These SER5 effects have been proposed to occur through direct SER5-Env interaction that is promoted upon Env-CD4 engagement (20). However, our two-color super-resolution microscopy results do not support direct interaction of Env and SER5 on single virions (26) but instead reveal SER5-mediated disruption of Env clusters on mature HIV-1 that may diminish the virus's fusion compe tence. A recent study has linked the lipid scrambling activity of SER5 to reduction of HIV-1 infectivity (8), although it remains unclear whether loss of the viral membrane's lipid asymmetry strictly correlates with reduction in infectivity (9). Regardless of the exact mechanism of HIV-1 restriction, the antiviral activity of SER5 against other enveloped viruses is indicative of a universal mechanism of inhibition of viral fusion, perhaps through modification of lipid membranes, as has been proposed in reference 27.
Studies of SER5-mediated restriction have focused on the effects of this protein incorporated into the viral membrane. Only two groups have examined the anti-IAV activity of SER5 expressed in target cells but arrived at discordant conclusions regarding its ability to inhibit IAV HA-mediated fusion/infection (13,14). It is thus unclear whether SER5 can inhibit viral fusion when present in target cells, which could prompt a revision of the current models of SER5-mediated virus restriction. Here, we conducted a series of experiments to assess the effects of SER5 expressed in effector (fusion protein-express ing) and target cells on HIV-1 Env-mediated infection and membrane fusion. These experiments revealed a marked inhibition of cell-cell fusion but not virus-cell fusion or infection by SER5 in target cells. Inhibition of cell-cell fusion was not associated with downregulation of CD4 or co-receptor levels or with inhibition of Env binding to CD4 or co-receptors. We also found that SER5 in target cells diminished cell-cell fusion mediated by Env expressed on effector cells or Env incorporated into virions. Notably, the sensitivity of HIV-1 strains to SER5 restriction from effector cells was mirrored by SER5 expressed in target cells. We found that SER5 arrests Env-mediated cell-cell fusion at a small pore stage and promotes pore closure. Addition of exogenous PS to either effector or target cells antagonized the SER5 effect on cell fusion. These results demonstrate the ability of SER5 to interfere with HIV-1 Env-mediated fusion when present in target or effector cells, suggesting a novel mechanism of antiviral activity, likely involving a local modification of membrane properties at the fusion site.
## RESULTS
## SER5 expressed both in effector and target cells inhibits cell-cell fusion mediated by sensitive HIV-1 env
Using a dual-split protein assay (28), we have previously reported inhibition of HIV-1 Env-mediated cell-cell fusion by SER5 expressed in effector cells (4). Here, we confirmed the restriction activity of SER5 transiently expressed in the effector (Env-expressing) HeLa cells, using a fluorescence microscopy-based assay that detects redistribution of small cytosolic dyes between fused cells (29, 30) (Fig. S1). As expected, co-expression of Nef rescued fusion with SER5-expressing effector cells. To test if SER5 can inhibit HIV-1 fusion when present in the target cells, we compared fusion of TF228.1.16 effector cell line stably expressing a CXCR4-tropic HIV-1 Env (31) to JurkatTag (JTag) CD4 T cells endogenously expressing SER3 and SER5 and double-knockout (S3/S5-/-) JTag cells (5). To improve HIV-1 fusion efficiency, these cells were additionally transduced with a CD4-expressing vector. Env-expressing TF228.1.16 cells fused with JTag depleted of SER3 and SER5 significantly better than with parental cells (Fig. 1A). The enhanced fusion with JTag/CD4/S3/S5-/-cells was not caused by increases in CD4 or CXCR4 expression (Fig. S2). Thus, endogenous levels of SER3/SER5 in JTag cells suppress HIV-1 Env-mediated cell fusion.
Consistent with the results obtained with JTag cells, overexpression of SER3 or SER5 in target TZM-bl cells inhibited Env-mediated fusion with the TF.228.1.16 effector cells (Fig. 1B). In contrast to SER3 and SER5, expression of the inactive member of the SERINC family proteins, SER2 (4,19), in target cells was without effect on cell fusion (Fig. 1B). We note that SER3 in target cells less potently inhibited cell-cell fusion compared to SER5, despite the much higher expression of the former protein in TZM-bl cells (Fig. S3). The inhibition of cell-cell fusion by ectopically expressed SER5, but not SER2, observed in a microscopy-based assay was confirmed using a dual-split-GFP/luciferase cell-cell fusion assay (Fig. S4). Note that this assay detected a slight reduction in cell-cell fusion upon SER2 expression in effector or target cells, which was, however, far less potent than expression of SER5.
We verified the HIV-1 restriction by SERINC proteins using pseudovirus fusion and single-cycle infectivity assays. HXB2 pseudoviruses lacking or containing SER2, SER3, or SER5 were used to infect reporter TZM-bl cells, and the resulting viral fusion (betalactamase activity [32,33]) and infectivity (luciferase signal) were measured. Neither SER2 nor SER3 inhibited HXB2 pseudovirus fusion or infectivity, while SER5 potently sup pressed both signals (Fig. 1C andD), despite incorporating into virions less efficiently than SER2 or SER3 (Fig. 1E). In fact, SER2 incorporation significantly enhanced HIV-1 pseudovirus infectivity (Fig. 1D). Our results thus reveal a differential effect on cell-cell fusion of SER3 expressed in target cells, as compared to fusion of pseudoviruses contain ing this protein.
The HIV-1 sensitivity to virion-incorporated SER5 maps to Env glycoprotein, specifi cally to the gp120 V3 loop (1). We, therefore, asked if the sensitivity of Env-mediated fusion to SER5 expressed in target cells exhibits the same HIV-1 strain dependence as when SER5 is expressed in effector cells. HEK293T cells transiently expressing Env (and SER5, when indicated) were co-cultured with TZM-bl cells transfected with SER5 or SER2 or mock-transfected. The resulting cell-cell fusion driven by HXB2 (lab adapted, CXCR4 tropic) or ADA (lab adapted, CCR5 tropic) Env was equally efficiently suppressed by SER5 in effector or target cells (Fig. 2A andB). By contrast, fusion mediated by JRFL, JR2, and AD8 Env glycoproteins that are relatively resistant to SER5 restriction (1, 3-5, 10, 18-20) was unaffected by SER5 expression in effector or target cells (Fig. 2A andB). In control experiments, SER2 failed to inhibit cell fusion mediated by any of these Env glycopro teins. Unless stated otherwise, all subsequent experiments were performed using HXB2 Env chosen for its high sensitivity to SER5-mediated restriction (see references 4 and 18; Fig. 1A, B, and3B). Virus-cell fusion was measured using a beta-lactamase assay (see Materials and Methods) and normalized to the HXB2.Vector control. Data are the mean and SD of two independent viral preparations. (D) TZM-bl target cells were infected with equal amounts (p24) of HXB2 pseudoviruses containing or lacking HA-tagged SERINCs. The resulting luciferase signal was measured (Continued on next page) Next, we asked if SER5 in target cells was active against non-HIV fusion proteins. Unlike the HIV-1 Env-mediated cell fusion (Fig. 3B), the influenza virus X:31 HA-mediated fusion triggered by exposure to low pH was not inhibited by SER5 expressed in effector or target cells (Fig. 3C). The resistance of HA-mediated cell fusion to SER5 was also observed for another IAV HA strain, PR8/34 (Fig. S5). We concluded that SER5 selectively inhibits cell-cell fusion mediated by sensitive HIV-1 Env strains and not strains known to be resistant to restriction or IAV HA-mediated fusion.
## SER5 expression in target cells does not downregulate CD4/co-receptors or interfere with Env/CD4/co-receptor interactions
To determine if SER5 in target cells inhibits HIV-1 Env-mediated fusion by downregulat ing CD4 or co-receptor expression, we compared the levels of these proteins on control and SER5-expressing target cells. Immunofluorescence labeling for CD4 and CXCR4 and flow cytometry analysis did not reveal significant changes in surface expression of these proteins in control and SER5-or SER2-transfected TZM-bl cells (Fig. S6). Further, overex pression of CD4 or co-receptors (CXCR4 or CCR5) did not rescue HXB2 or ADA Envmediated fusion with SER5-expressing target cells (Fig. S7). These results rule out downregulation of CD4 or co-receptor as the mechanism for inhibition of Env-mediated fusion by SER5 expressed in target cells.
Another potential mode of SER5-mediated inhibition of HIV-1 fusion is through interference with the binding of Env to CD4 and/or co-receptors on the cell surface, perhaps through restriction of their lateral mobility. To explore this possibility, we took advantage of the ability to capture/stabilize HIV-1 fusion at a step when a fraction of Env glycoproteins engage a requisite number of CD4 and co-receptors. We have demonstra ted that pre-incubation of effector and target cells at a sub-threshold temperature for HIV-1 fusion (typically, 23°C) allows Env to form ternary complexes with CD4 and cognate co-receptors but not to proceed to fusion (34,35). The formation of Env/CD4/co-receptor complexes at this temperature-arrested stage (TAS) is manifested by acquisition of resistance to inhibitors of CD4 and CXCR4 binding (34,35). As expected, prolonged preincubation of HXB2 Env-expressing effector cells and target cells at 23°C rendered the subsequent fusion partially resistant to a fully inhibitory concentration of AMD3100 (CXCR4 binding inhibitor) added after pre-incubation (Fig. 4A). In contrast, the gp41derived T20 peptide that prevents the formation of the final six-helix bundle structure driving the formation of a fusion pore (34)(35)(36) abrogated fusion equally potently when present throughout the experiment or added at TAS (Fig. 4A). Like Env-mediated fusion with control cells, fusion with SER5-expressing target cells arrested at TAS was partially resistant to AMD3100 (Fig. 4B). In fact, protection of fusion from AMD3100 added at TAS was independent of SER5 expression, with 61% and 67% of fusion of untreated cells occurring in the presence of AMD3100 for control and SER5-expressing cells, respectively (Fig. 4A andB). These results imply that SER5 does not interfere with the engagement of CD4 or CXCR4 by Env on the cell surface, at least on the time scale of establishing TAS.
## SER5 does not capture HIV-1 Env-mediated fusion at a hemifusion stage
Viral protein-mediated fusion proceeds through a hemifusion intermediate (e.g., see references [37][38][39][40][41][42][43][44][45][46][47][48]. Depending on the viral glycoprotein and experimental conditions, a significant fraction of fusion events may not progress beyond a hemifusion stage (38)(39)(40)(41). These dead-end hemifusion structures can be converted to full fusion by a brief treatment with chlorpromazine (CPZ), which we have shown to partition into and destabilize the hemifusion diaphragm (Fig. 3A), thereby promoting fusion pore forma tion (40). In agreement with our previous findings (49-51), a sizeable fraction of HA-mediated cell fusion remained at a hemifusion stage, as evidenced by a nearly 30% increase in cell fusion after CPZ treatment (Fig. 3C, hatched bars). By comparison, HIV-1 Env is less prone to form dead-end hemifusion structures, as judged by a marginal increase in cell-cell fusion after exposure to CPZ (Fig. 3B). Importantly, the lack of rescue of Env-mediated fusion with SER5-expressing cells by CPZ implies that SER5 does not arrest HIV-1 Env-mediated cell fusion at a hemifusion stage and is, thus, likely to block downstream steps of fusion.
## Phosphatidylserine and phosphatidylglycerol selectively rescue HIV-1 Env-mediated fusion of SER5-expressing cells
Since the lipid composition modulates viral protein-mediated fusion and HIV-1 Envmediated fusion, in particular (34,52,53), we tested whether the SER5-imposed block of Env-mediated fusion was lipid dependent. An additional motivation was the recently discovered lipid scrambling activity of SER5, which has been proposed to contribute to the antiviral activity of this protein (8). Co-cultures of effector and target cells were pre-treated with exogenously added phosphatidylcholine (PC), PS, phosphatidyl glycerol (PG), or bis(monoacylglycero)phosphate (BMP), and the resulting cell-cell fusion was measured after further incubation at 37°C. Env-mediated fusion with control and SER2-expressing target cells was modestly promoted by exogenous PS and, to a lesser extent, PG (Fig. 5A). In contrast, PS and PG potently restored fusion with SER5-express ing cells almost to the level of SER5-negative cells (Fig. 5A). This selective antagonism of PS and PG with SER5-mediated inhibition of fusion can be readily appreciated by plotting the fold inhibition of cell-cell fusion by SER5 in cells pre-treated with different lipids (Fig. 5B). Thus, PS and PG specifically rescue fusion with SER5-expressing cells, whereas PC and BMP are without effect. The same lipid dependence was observed upon SER5 co-expression with Env in effector cells. Here too, PS, but not PC or BMP, rescued fusion between SER5-expressing effector cells with target cells, while PC or BMP did not significantly modulate Env-mediated fusion (Fig. 5C andD).
We next asked if selective enhancement of fusion with SER5-expressing cells occurs through incorporation of PS in effector vs target cells. Target or effector cells were separately pre-treated with exogenous lipids followed by removal of excess lipids and co-culture to allow fusion. Addition of PS selectively promoted fusion with SER5-express ing cells, irrespective of whether this lipid was added to effector or target cells (Fig. S8). Such a "symmetric" enhancing effect of PS suggests that late stages of Env-mediated fusion with SER5-expressing cells, perhaps downstream of the membrane merger, are lipid dependent. We hypothesized that SER5 and PS can modulate the stability of small fusion pores.
## SER5 slows down the formation of fusion pores and prevents their dilation
To resolve the formation of fusion pores in real time, we synchronized Env-mediated fusion by first capturing this process at TAS (pre-incubation at 23°C) and quickly shifting to 37°C using a custom-made temperature jump setup (see references 29 and 54 and Fig. 6A). Owing to the formation of functional Env/CD4/co-receptor complexes, fusion from TAS is much faster than uninterrupted cell-cell fusion at 37°C. Effector and target cells were co-cultured at 23°C for 2.5 h. The formation of nascent fusion pores after shifting to 37°C resulted in redistribution of calcein loaded into effector cells to target cells (Fig. 6C). These experiments revealed that the kinetics of fusion pore formation with SER5-expressing target cells captured at TAS was markedly slower than with control cells (Fig. 6B). This result, together with the observation that SER5 does not interfere with the formation of Env/CD4/CXCR4 complexes at TAS (Fig. 4), supports the notion that SER5 interferes with the late steps of Env-mediated membrane fusion, namely, the formation of fusion pores.
We have previously deduced the effective size and dynamics of fusion pores medi ated by HIV-1 Env based upon the rate of calcein redistribution between the effector/ target cell pairs (29). Strikingly, whereas calcein redistribution from effector to control target cells was typically completed within 1-2 min, dye redistribution to SER5-express ing cells was much slower and never reached completion (Fig. 6C through F; Fig. S9). In other words, none of the fusion events with SER5-expressing target cells allowed calcein to equilibrate between the donor and recipient cells, suggesting a premature closure of small fusion pores. Measurements of the time course and extent of small dye redistribu tion between a cell pair allow for calculation of a relative pore permeability (29). Plotting multiple pore permeability profiles for either target or effector cells expressing SER5 confirms the virtual lack of pore dilation compared to control cell pairs (Fig. 7A through C). Ensemble-averaged profiles of the fusion pore permeability for control and SER5positive target cells (Fig. 7D) further illustrate this notion. These results demonstrate that SER5 strongly impairs the dilation of Env-mediated fusion pores and, ultimately, pro motes pore closure.
## HIV-1 pseudovirus fusion is not inhibited by SER5 expressed in target cells
We next asked if target cells expressing SER5 are less conducive to HIV-1 pseudovirus fusion. Parental TZM-bl cells and cells ectopically expressing SER2-GFP or SER5-GFP (to assess the efficiency of SERINC transfection) were inoculated with pseudoviruses bearing SER5-sensitive (HXB2) Env or SER5-resistant (JRFL) Env, and the resulting fusion was measured using a beta-lactamase (BlaM) assay (32,33,55). In parallel experiments, we verified that GFP-tagged SER5 retained anti-HIV activity in a cell-cell fusion assay (Fig. S10). In stark contrast to robust inhibition of cell-cell fusion by SER5-GFP, neither SER2-GFP nor SER5-GFP expression significantly affected fusion of particles bearing sensitive or resistant Env (Fig. 8A). This result prompted us to test if SER5 in target cells modulates single-cycle infection. Although infection of SER5-GFP-expressing cells was modestly reduced compared to control cells, SER2-GFP expression had the same modest inhibitory effect on both HXB2 and JRFL Env-mediated infection (Fig. 8B). This unexpected effect of SER2 on HIV-1 infection, but not fusion, points to a nonspecific effect of SERINC expression on post-fusion steps of HIV-1 infection. We further assessed the effect of endogenous SER5 and SER3 expression on HIV-1 entry using parental JTag/CD4 and JTag/CD4/S3/S5-/-cells (see Fig. 1A). Consistent with HIV-1 fusion with TZM-bl cells, we did not detect a significant increase in HIV-1 infection of SER3/SER5-negative cells compared to parental cells (Fig. 8C), although two out of four independent experiments did show a modest enhancement of infectivity (Fig. S11).
## SER5 can inhibit HIV-1 pseudovirus-mediated fusion of adjacent cells (fusion from without)
To address the discrepancy between the effects of SER5 on cell-cell fusion vs virus-cell fusion/infection, we asked whether SER5 can inhibit fusion from without (FFWO) -cellcell fusion mediated by viruses bound to juxtaposed cells (56)(57)(58). FFWO involves fusion of two plasma membranes mediated by viral particles fusing with both adjacent cells, without a contribution from internalized virions. We examined FFWO mediated by pseudoviruses displaying SER5-sensitive (HXB2) or SER5-resistant (JRFL) Env using TZMbl target cells stably expressing GFP-tagged SER2 or SER5 to readily identify SERINCexpressing cells. SER2-GFP or SER5-GFP expressing TZM-bl cells were non-enzymatically detached from plates, split in two samples that were separately labeled with different cytoplasmic fluorescent dyes. Labeled cells were co-cultured to create a confluent monolayer, followed by virus binding in the cold and incubation at 37°C to allow FFWO. Microscopic examination of fused (double-positive for both dyes) cells revealed a marked reduction of HXB2 pseudovirus-mediated FFWO of SER5-expressing target cells com pared to SER2-expressing target cells (Fig. 9A). By contrast, JRFL pseudovirus-mediated FFWO of SER5-expressing cells was not significantly reduced relative to SER2-expressing cells (Fig. 9A).
We next verified the effect of SER5 on FFWO using an independent split-luciferase assay. Toward this goal, we used TZM-bl cells stably expressing one of the complemen tary domains of dual-split GFP-luciferase (28) and either SER2-GFP or SER5-GFP. HXB2 or JRFL pseudovirus-mediated fusion between these cells results in reconstitution of luciferase activity. We found that SER5, but not SER2, expression markedly inhibited FFWO mediated by HXB2 particles (Fig. 9B). By comparison, JRFL pseudovirus-mediated FFWO was modestly and equally reduced by both SER2 and SER5 (Fig. 9B). Because SER2 and SER5 expression caused a similar modest reduction in JRFL-mediated FFWO, the effect of SER5 is likely nonspecific. Our results thus confirmed that, in contrast to regular HXB2 entry in TZM-bl cells, confining viral fusion to the plasma membrane (FFWO) renders it sensitive to SER5 inhibition.
## DISCUSSION
Here, we demonstrated the ability of SER5 expressed in target membranes to inhibit HIV-1 Env-mediated cell-cell fusion and virus-mediated fusion from without. By compari son, SER3 only modestly suppressed cell-cell fusion (Fig. 1B), despite its much higher expression compared to SER5 (Fig. S3B). SER5's inhibitory effect on cell-cell fusion phenocopied the impact of virus-incorporated SER5 on HIV-1 fusion and infectivity, with fusion mediated by the known SER5-resistant HIV-1 strains being unaffected by this restriction factor. As with SER5 in the viral membrane, the potency with which SER5 expressed in target cells suppresses fusion is HIV-1 Env strain dependent (Fig. 2A andB). Note, however, that a recent study did not observe inhibition of NL4-3 Env-mediated cell-cell fusion by SER5 expressed in effector cells (59). Another study also failed to detect the effect of endogenous SER5 levels in HIV-infected Jurkat cells on their fusion with primary macrophages (60). While the reasons for these discrepant results are unclear, we note that both studies examined late fusion products (6-24 h) that may mask possible kinetic effects of SER5 expression. Also, the second study (60) did not knock out both SER5 and SER3 in Jurkat cells, despite the fact that endogenous SER3 in these cells exhibits a modest antiviral activity (5). Neither of the two studies tested the effect of SER5 expressed in target cells.
We also found that SER5 in target or effector cells does not significantly reduce IAV HA-mediated cell-cell fusion (Fig. 3). The lack of SER5 effect on IAV HA-mediated fusion in our experiments agrees with the recent article (14). However, other groups have reported HA subtype-and glycosylation-dependent inhibition of HA-mediated fusion/infection by SER5 in the viral/effector membrane (12)(13)(14). SER5 expressed in target cells has been reported to inhibit IAV lipid mixing (hemifusion), apparently, by relocating from the plasma membrane to endosomes upon infection (13). The reasons for these discrepant results are currently unclear but may be related to the levels of SER5 expression.
An important finding of this study is the striking rescue of HIV-1 Env-mediated fusion with SER5-expressing target or effector cells by exogenous PS and PG, but not PC or BMP (Fig. 5). Failure of BMP to antagonize SER5-mediated restriction of Env-mediated fusion indicates that negative charge is not the main determinant of the effect of PS. Of course, we do not know the distribution of exogenously added lipids in cells and, thus, cannot rule out the unlikely possibility that a fraction of PS and PG remain at the plasma membrane, while other lipids are effectively removed from the cell surface. Our result in Fig. 5 also shows a modest increase in fusion efficiency of control cells by exogenous PS and PG, in excellent agreement with our previous finding that cell surface exposure of PS stimulates HIV-1 Env-mediated fusion (53). Considering the recently reported SER5's lipid scrambling activity (8), we surmise that excess PS may favor a lipid scrambling conformation of SER5, which may be less inhibitory for membrane fusion.
## Possible mechanism of inhibition of Env-mediated fusion by SER5
Several lines of evidence presented in this study support the notion that SER5 exerts its effect on the nascent fusion pores formed by HIV-1 Env without affecting the upstream steps of fusion. First, SER5 in the target membrane does not interfere with the formation of functional Env-CD4/CXCR4 complexes at a reduced temperature that is not permissive for fusion (Fig. 4). Note, however, that we cannot rule out the possibility that SER5 can slow down CD4 or co-receptor engagement by Env on a timescale that is shorter than that required to establish TAS (>2 h). Second, SER5 does not capture Env-mediated fusion at a hemifusion stage (see reference 4 and Fig. 3). Third, exogenous PS promotes Env-mediated fusion with SER5-expressing cells, regardless of whether it was added to effector or target cells (Fig. 5). This observation is consistent with the SER5's involvement at the point where two membranes merge and exchange lipids. Finally, we observed a potent destabilization of Env-mediated pores and their collapse upon SER5 expression in either target or effector cells (Fig. 6 and7). Since SER5 similarly impacts HIV-1 Env-medi ated fusion regardless of whether it is present in effector or target cells, we propose that SER5 increases the energetic penalty for sustaining small fusion pores and thereby promotes their collapse.
Inhibition of full HIV-1 fusion with plasma membrane-derived vesicles containing SER3 and SER5 has been observed by cryo-electron tomography (cryo-ET) (61). The authors have reported HIV-1 pseudovirus fusion arrest at hemifusion and early fusion intermediates for viruses containing SER3 or SER5. A major distinction with our results is that these early intermediates were featuring fusion pores comparable to the virus diameter (~100 nm), with fusion products having an hourglass-shaped membrane. In contrast, our definition of small/nascent fusion pores is pores that limit fluorescent dye diffusion and are thus likely in the range of a few nanometers. By monitoring the dynamics of fusion pore in real time, we demonstrate potent destabilization of nascent fusion pores and their collapse caused by SER5 expression in the context of Env-medi ated cell-cell fusion. It is possible that these transient events going back to two separate membranes could have been missed by cryo-ET analysis.
We note that our published results obtained by single virus tracking did not detect the formation of fusion pores permeable to mCherry (viral fluidphase marker) between HIV-1 pseudoviruses and target cells (4). By comparison, we detect partial transfer of cytoplasmic dyes between fusing cells expressing SER5. These discordant results can be reconciled by considering the differences in the sensitivity of small pore detection. Cell-cell fusion is monitored based upon transfer of calcein, which is considerably smaller than mCherry used to detect viral fusion (~0.6 vs ~27 kDa). Additionally, the large contact area between the effector and target cells may allow for formation of multiple fusion pores, some of which may not be fully blocked by SER5.
Note that the proposed model does not explain the reason for the lack of an effect of SER5 on virus-cell fusion with or infection of TZM-bl or JTag cells (Fig. 8). Surprisingly, SER5 in target cells exhibits distinct effects on HXB2 fusion and on cell-cell fusion mediated by this virus (FFWO, Fig. 9). To reconcile these differences, we propose that these reflect the preferred HIV-1 entry pathway in TZM-bl cells. Our studies have revealed a strong preference for HXB2 entry via endocytosis and fusion with pH-neutral intracellular compartments (62). It is thus possible that SER5 effectively restricts HIV-1 Env-mediated fusion at the plasma membrane (cell-cell fusion and FFWO) but fails to interfere with virus entry through an endocytic pathway.
In conclusion, our results suggest a universal mechanism of inhibition of HIV-1 Env-mediated fusion by SER5 present in the effector or target membrane. Destabilization of small fusion pores by SER5 in the target membrane is unlikely to occur through direct interactions with HIV-1 Env, which has been proposed as the mechanism for restriction by virus-incorporated or effector cell-expressed SER5 (20). Since the inhibitory effect of SER5 in target cells is not related to downregulation of CD4 or co-receptors (Fig. S2), this protein must promote collapse of nascent fusion pores by modulating the properties of fusing membranes. This mechanism of HIV-1 restriction is consistent with the report that SER5 incorporation raises membrane tension of HIV-1 pseudovirus membrane (27), although our results using lipid order probes did not reveal a significant effect of SER5 on lipid packing relative to SER2 (63). It is thus possible that SER5 exerts a local effect on membrane properties at the fusion site, perhaps through altering the distribution of lipid domains (27). We surmise that SER5-resistant HIV-1 Env glycoproteins, which are mostly from primary isolates, overcome the resistance of SER5-containing membranes through forming more robust prefusion complexes, as evidenced by the lower minimal number of Env trimers required for their fusion compared to Envs of laboratory-adapted strains (64,65).
## MATERIALS AND METHODS
## Plasmids, cell lines, and reagents
The expression vectors pCAGGS-HXB2, pCAGGS-JRFL, pcRev, pR9ΔEnvΔNef, pMDG-VSV-G, pBJ5-SER2-HA, pBJ5-SER3-HA, pBJ5-SER5-HA, pBJ-SER2-GFP, and pBJ-SER5-GFP have been described previously (3,4). JR2, AD8, and ADA HIV-1 Env expression vectors were previously described (18). Expression vectors for the influenza hemag glutinin strains PR8/34 and X:31 were kindly provided by Dr. A. Brass (University of Massachusetts) and Judith White (University of Virginia), respectively. NL4-3.Luc.E-R-(HRP-3418) (66) and pMM310-BlaM-Vpr (ARP-11444) (33) expression plasmids were from the BEI Resources, NIH HIV Reagent Program. The cloning of pQCXIP-SER2-GFP and pQCXIP-SER5-GFP was done as follows. The SER2-GFP fragment was amplified by PCR using TaqDNA Polymerase High Fidelity (Invitrogen, Waltham, MA, USA; 11304-011), pBJ5-SER2-GFP expression vector as template, and the forward SER2-SbfI 5′-GGCCTGCAGGGCCATGGACGGGAGGATGATGAG-3′ and the reverse GFP-NotI 5′-GCTG CGGCCGCTTACTTGTACAGCTCGTCCATGCCGA-3′ primers. The SER5-GFP fragment was amplified using the pBJ5-SER5-GFP plasmid as template and the forward SER5-SbfI 5′-GGCCTGCAGGGCCATGTCAGCTCAGTGCTGTGC-3′ and the same reverse primer as for SER2-GFP. The amplified fragments and the retroviral vector pQCXIP (Clontech, Mountain View, CA, USA; 631516) were digested with SbfI and NotI, purified, and ligated. The lentiviral vectors pLenti-DSP1-7 and pLenti6.2-DSP8-11 were a gift from Dr. Z. Matsuda (University of Tokyo) (67).
HEK293T/17 cells were purchased from ATCC (Manassas, VA, USA). HeLa-derived TZM-bl cells (donated by Drs. J.C. Kappes and X. Wu) (68) and TF228.1.16, a mamma lian Burkitt's lymphoma cell line that stably expresses the HIV-1 envelope glycoprotein (BH-10 clone of HIV-1 LAI, contributed by Drs. Zdenka Jonak and Steve Trulliy [31]), were obtained from BEI Resources. JTag and JurkatTag SER3 -/-SER5 -/-(abbreviated JTag S3/S5-/-) double-knockout cells were a gift from Dr. H.G. Göttlinger (University of Massachusetts). HeLa cells expressing the HIV-1 ADA Env (HeLa-ADA cells) were a gift from Dr. Marc Alizon (Pasteur Institute, France) (69).
HEK293T/17, HeLa, TZM-bl, and COS7 cells were cultured in high-glucose Dulbecco's Modified Eagle's Medium (Mediatech, Manassas, VA, USA), while TF228.1.16 cells and JTag-derived cells were grown in RPMI-1640 (Life Technologies, Grand Island, NY, USA). Media were supplemented with 10% FBS (GIBCO BRL) or 10% Cosmic Calf Serum (HyClone, Logan, UT, USA) and 100 units/mL penicillin/streptomycin (from Hyclone or Gibco). For HEK293T/17 cells, the growth medium was additionally supplemented with 0.5 mg/mL G418 (Life Technologies, 1013-027). The JTAg/CD4 versions were obtained by retroviral transduction with pCXbsrCD4 and selection with blasticidin (5). TZM-bl.SER-GFP stable cell lines were obtained by transducing with VSV-G/pQCXIP-SER-GFP pseudotyped viruses and selecting with 1 µg/mL puromycin (InvivoGen, San Diego, CA, USA; ant-pr-1). TZM-bl.SERs-GFP-DSPs stable cell lines were obtained by lentiviral transduction with DSP1-7 or DSP8-11 and selection with 5 µg/mL blasticidin (Research Products International, Mount Prospect, IL, USA; B12200-0.05).
EnduRen and Bright-Glo luciferase were from Promega (Madison, WI, USA; E6481), whereas BlaM substrate, CCF4-AM, was from Invitrogen (K1089). The Micro BCA Protein Assay Kit was from Thermo Scientific (Rockford, IL, USA; 23235). The HIV-1 gp41-derived C34 peptide was a gift from Dr. L. Wang (University of Maryland). The HIV-IG serum (ARP-3957) and HIV-1 gp41-derived T20 peptide (HRP-12732) were from BEI Resources, NIH HIV Reagent Program. RIPA buffer was obtained from Abcam (Cambridge, MA, USA; ab288006), and the cOmplete Mini protease inhibitor cocktail was from Roche Diagnostics (Mannheim, Germany; 11836153001). Rabbit anti-HA polyclonal antibody (H6908) was from Sigma (St. Louis, MO, USA). Mouse-antialpha tubulin antibody (T6074) and mouse-antirabbit-HRP (AP188P) were from Millipore/Sigma (Burlington, MA, USA). Rabbit-antimouse-HRP (6170-05) was purchased from SouthBiotech (Birmingham, AL, USA). All lipids, 1,2-dioleoyl-sn-glycero-3-phosphocholine (850375C), 1,2-dioleoyl-snglycero-3-phospho-L-serine (840035C), 1,2-dioleoyl-sn-glycero-3-[phospho-rac-(3-lysyl(1glycerol))] (840521P), and BMP (S,R Isomer, 857133C) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). AMD3100 was from Millipore/Sigma (A5602).
## Pseudovirus production and characterization
All the pseudoviruses were generated by transfecting HEK293T/17 cells grown to ~75% confluency in 10 cm dishes with JetPRIME transfection reagent (Polyplus Transfec tion, IllkirchGraffenstaden, France; 114-15). For generating the pseudoviruses used in virus-cell fusion BlaM assay and FFWO, HEK293T/17 cells were transfected with 4 µg pR9ΔEnvΔNef, 2 µg BlaM-Vpr, 0.5 µg pcRev, and 3 µg HIV-1 Env expression vector. For luciferase-encoding pseudoviruses used in infectivity assays, HEK293T/17 cells were transfected with 6 µg pNL4-3.Luc.R-E-, 0.5 µg pcRev, and 3 µg HIV-1 env expression vector. The viral supernatants were collected at 40-48 h post-transfection, filtered through 0.45 µm polyethersulfone filters (VWR), and concentrated 10× using Lenti-X concentrator (Takara, San Jose, CA, USA; 631232). The p24 content of pseudoviruses was determined by ELISA (70), and the infectious titer was determined by β-Gal assay in TZM-bl cells (71).
## Western blot analysis
For preparation of cell lysates, cells were seeded at ~2 × 10⁵ cells per well in six-well plates 24 h prior to transfection. Transfection was performed with 2 µg of DNA per well using JetPRIME transfection reagent, according to the manufacturer's protocol. Thirtysix hours post-transfection, cells were harvested and lysed in RIPA buffer containing cOmplete protease inhibitor cocktail, and the total protein content of lysates was quantified using BCA protein kit. For viral lysates, equal amounts of pseudoviruses were lysed in 2× Laemmli buffer (Bio-Rad, Hercules, CA, USA; 1610737). Samples were heated at 42°C for 10 min, loaded onto 4%-15% polyacrylamide gel (Bio-Rad, 4561083), and transferred onto a 0.45 µm nitrocellulose membrane (Amersham, Marlborough, MA, USA; 10600002). The membranes were blocked in 10% dry milk in 0.1% Tween 20 in phosphatebuffered saline solution for 1 h at room temperature. The membranes were next incubated overnight at 4°C with different primary antibodies diluted in a buffer containing 5% dry milk/0.1% Tween 20 as follows: HIV-IG (1:2,000 dilution), rabbit anti-HA (1:500 dilution), and mouse antitubulin (1:1,000). Secondary antibody incuba tion was performed using either goat antihuman HRP (1:1,000), mouse antirabbit HRP (1:1,000), or rabbit antimouse HRP (1:1,000) for 1 h at room temperature. The membranes were developed using a chemiluminescence reagent (Amersham, RNP2232), and the chemiluminescence signal was recorded on a ChemiDoc Imaging system (Bio-Rad) using Image Lab software version 6.1 (Bio-Rad).
## Virus-cell fusion, FFWO, and infection assays
The virus-cell fusion using a β-lactamase (BlaM) assay was done, as previously described (71). Briefly, HIV-1 pseudovirus (MOI ~1) bearing respective envelope glycoprotein and β-lactamase fused to Vpr (BlaM-Vpr) was bound to a confluent monolayer of target cells in a black, clear bottom 96-well plate by centrifugation at 4°C for 30 min at 1,550 × g. The unbound virus was washed out; growth medium was added; and samples were incubated at 37°C with 5% CO 2 for 90 min, after which the cells were loaded with CCF4-AM substrate and incubated at 12°C overnight. The blue to green fluores cence ratio was measured using the Synergy HT fluorescence microplate reader (Agilent Bio-Tek, Santa Clara, CA, USA).
For FFWO experiments, two different assays were used: (i) a microscopy-based assay using cytosolic dyes and (ii) a luciferase assay using DSP-expressing cells. For micro scopic detection of FFWO, TZM-bl cells stably expressing SER2-GFP or SER5-GFP, each pre-loaded with either a blue dye CMAC (10 µM, ThermoFisher; C2110) or 2 µM Calcein Red-Orange AM (C34851) for 30 min at 37°C. A 1:1 mixture of cells loaded with CMAC or Calcein Red-Orange was co-cultured overnight. The next day, HIV-1 pseudoviruses (MOI ~4) were added to confluent cultures and spinoculated at 10°C for 30 min at 1,500 × g. Cells were washed and incubated for 2.5 h at 37°C. Fused (double-positive) cells were quantified microscopically and normalized to the total number of cells. For luciferasebased detection, TZM-bl.SER-GFP.DSP1-7 and TZM-bl.SER-GFP.DSP8-11 were seeded at a 1:1 ratio in black, clear bottom 96-well plate on the day before the experiment. On the day of the experiment, cells were loaded with 60 µM EnduRen substrate for 1 h at 37°C with 5% CO 2 . Next, HIV-1 pseudoviruses (MOI ~4) were added to cells and spinoculated at 10°C for 30 min at 1,500 × g. After washing out the unbound virus, fresh growth medium was added, and cells were incubated for 3 h at 37°C. The luciferase signal was measured using a Wallac1420 multilabel counter (PerkinElmer, Turku, Finland).
For the infectivity assay, 1•10 5 JTag/CD4 or JTag/CD4 S3/S5-/-cells were inoculated with 2 IU/cell HXB2 pseudoviruses bearing luciferase by centrifugation at 4°C for 30 min at 1,550 × g and incubated at 37 °C with 5% CO 2 for 48 h. Cells were lysed using Bright-Glo luciferase, and the luciferase signal was measured on a TopCount NXT reader (PerkinElmer Life Sciences, Shelton, CT, USA)
## Cell-cell fusion
Cell-cell fusion was measured, as previously described (29,30). Briefly, effector cells were loaded with Calcein AM (Thermo Fisher, C3099), and target cells were loaded with CMAC Blue (Thermo Fisher, C2110). The cells were detached from plates using PBS supplemented with EDTA and EGTA, resuspended in PBS++, mixed at a 1:1 ratio, and allowed to attach to poly-L-lysine-coated eight-well chamber slides (Millipore/Sigma, P1274; Lab-Tek, 177402) for 30 min at 23°C. Fusion was triggered by applying an acidic (pH 5.0) buffer at 37°C. Where indicated, cells were treated for 1 min at room temperature with CPZ (Millipore/Sigma, 215921).
## Monitoring fusion pore enlargement
Effector TF228.1.16 cells were loaded with calcein-AM, mixed with target TZM-bl cells, and allowed to adhere to poly-L-lysine-coated coverslips for 30 min at room temperature. Fragments of coverslips containing cells were transferred to a custom-built imaging chamber and placed onto an IR-absorbing coverslip at the bottom. Local temperature was rapidly raised and maintained at 37°C by illuminating the absorbing coverslip with an IR diode, as described in reference 29. Dye transfer was monitored using a Fluoview300 laser-scanning confocal microscope (Olympus IX70, Melville, NY, USA) with a UPlanApo ×60/1.20 NA water-immersion objective, and pore permeability as a function of time was calculated, as described previously (29).
## References
1. Beitari, Ding, Pan et al. (2017) "Effect of HIV-1 Env on SERINC5 antagonism" *J Virol*
2. Firrito, Bertelli, Vanzo et al. (2018) "SERINC5 as a new restriction factor for human immunodeficiency virus and murine leukemia virus" *Annu Rev Virol*
3. Rosa, Chande, Ziglio et al. (2015) "HIV-1 Nef promotes infection by excluding SERINC5 from virion incorporation" *Nature*
4. Sood, Marin, Chande et al. (2017) "SERINC5 protein inhibits HIV-1 fusion pore formation by promoting functional inactivation of envelope glycoproteins" *J Biol Chem*
5. Usami, Wu, Göttlinger (2015) "SERINC3 and SERINC5 restrict HIV-1 infectivity and are counteracted by Nef"
6. Inuzuka, Hayakawa, Ingi (2005) "Serinc, an activity-regulated protein family, incorporates serine into membrane lipid synthesis" *J Biol Chem*
7. Trautz, Wiedemann, Lüchtenborg et al. (2017) "The host-cell restriction factor SERINC5 restricts HIV-1 infectivity without altering the lipid composition and organization of viral particles" *J Biol Chem*
8. Leonhardt, Purdy, Grover et al. (2023) "Antiviral HIV-1 SERINC restriction factors disrupt virus membrane asymmetry" *Nat Commun*
9. Raghunath, Abbott, Marin et al. (2024) "Disruption of transmembrane phosphatidylserine asymmetry by HIV-1 incorporated SERINC5 is not responsible for virus restriction" *Biomolecules*
10. Ahi, Zhang, Thappeta et al. (2016) "Functional interplay between murine leukemia virus glycogag, serinc5, and surface glycoprotein governs virus entry, with opposite effects on gammaretro viral and ebolavirus glycoproteins" *mBio*
11. Li, Ahmad, Shi et al. (2019) "Murine leukemia virus glycosylated gag reduces murine SERINC5 protein expression at steady-state levels via the endosome/lysosome pathway to counteract SERINC5 antiretroviral activity" *J Virol*
12. Diehl, Guney, Vanzo et al. (2021) "Influence of different glycoproteins and of the virion core on SERINC5 antiviral activity" *Viruses*
13. Zhao, Xu, Liu et al. (2022) "SERINC5 restricts influenza virus infectivity"
14. Lai, Munro, Shi et al. (2022) "Restriction of influenza A virus by SERINC5"
15. Timilsina, Umthong, Ivey et al. (2022) "SARS-CoV-2 ORF7a potently inhibits the antiviral effect of the host factor SERINC5" *Nat Commun*
16. Ahmad, Li, Li et al. (2019) "The retroviral accessory proteins S2, Nef, and glycoMA use similar mecha nisms for antagonizing the host restriction factor SERINC5" *J Biol Chem*
17. Chande, Cuccurullo, Rosa et al. (2016) "S2 from equine infectious anemia virus is an infectivity factor which counteracts the retroviral inhibitors SERINC5 and SERINC3" *Proc Natl Acad Sci*
18. Kirschman, Marin, Chen et al. (2022) "SERINC5 restricts HIV-1infectivity by promoting conformational changes and accelerating functional inactivation of Env" *Viruses*
19. Schulte, Selyutina, Opp et al. (2018) "Localization to detergent-resistant membranes and HIV-1 core entry inhibition correlate with HIV-1 restriction by SERINC5" *Virology (Auckl)*
20. Zhang, Shi, Qiu et al. (2019) "CD4 expression and Env conformation are critical for HIV-1 restriction by SERINC5" *J Virol*
21. Bowder, Hollingsead, Durst et al. (2018) "Contribution of the gp120 V3 loop to envelope glycoprotein trimer stability in primate immunodeficiency viruses" *Virology (Auckl)*
22. Haider, Snetkov, Jolly (2021) "HIV envelope tail truncation confers resistance to SERINC5 restriction" *Proc Natl Acad Sci*
23. Anang, Zhang, Ennis et al. (2025) "Characterization of full-length and cytoplasmic tailtruncated envelope glycoproteins incorporated into human immunode ficiency virus (HIV-1) virions and virus-like particles"
24. Staropoli, Dufloo, Ducher et al. (2020) "Flow cytometry analysis of HIV-1 Env conformations at the surface of infected cells and virions: role of Nef, CD4, and SERINC5" *J Virol*
25. Featherstone, Aiken (2020) "SERINC5 inhibits HIV-1 infectivity by altering the conformation of gp120 on HIV-1 particles" *J Virol*
26. Chen, Sood, Marin et al. (2020) "Super-resolution fluorescence imaging reveals that serine incorporator protein 5 inhibits human immunodeficiency virus fusion by disrupting envelope glycoprotein clusters" *ACS Nano*
27. Ward, Sokovikova, Waxham et al. (2023) "Serinc5 restricts HIV membrane fusion by altering lipid order and heterogeneity in the viral membrane" *ACS Infect Dis*
28. Wang, Kondo, Long et al. (2009) "Monitoring of HIV-1 envelope-mediated membrane fusion using modified split green fluorescent proteins" *J Virol Methods*
29. Markosyan, Cohen, Melikyan (2003) "HIV-1 envelope proteins complete their folding into six-helix bundles immediately after fusion pore formation" *Mol Biol Cell*
30. Melikyan, Markosyan, Brener et al. (2000) "Role of the cytoplasmic tail of ecotropic moloney murine leukemia virus Env protein in fusion pore formation" *J Virol*
31. Jonak, Clark, Matour et al. (1993) "A human lymphoid recombinant cell line with functional human immunodeficiency virus type 1 envelope" *AIDS Res Hum Retroviruses*
32. Marin, Du, Giroud et al. (2015) "Highthroughput HIV-cell fusion assay for discovery of virus entry inhibitors" *Assay Drug Dev Technol*
33. Cavrois, Noronha, Greene (2002) "A sensitive and specific enzyme-based assay detecting HIV-1 virion fusion in primary T lymphocytes" *Nat Biotechnol*
34. Melikyan, Markosyan, Hemmati et al. (2000) "Evidence that the transition of HIV-1 gp41 into a sixhelix bundle, not the bundle configuration, induces membrane fusion" *J Cell Biol*
35. Mkrtchyan, Markosyan, Eadon et al. (2005) "Ternary complex formation of human immunodeficiency virus type 1 Env, CD4, and chemokine receptor captured as an intermediate of membrane fusion" *J Virol*
36. Lawless, Barney, Guthrie et al. (1996) "HIV-1 membrane fusion mechanism: structural studies of the interactions between biologically-active peptides from gp41" *Biochemistry*
37. Kemble, Danieli, White (1994) "Lipid-anchored influenza hemagglutinin promotes hemifusion, not complete fusion" *Cell*
38. Chernomordik, Chanturiya, Green et al. (1995) "The hemifusion intermediate and its conversion to complete fusion: regulation by membrane composition" *Biophys J*
39. Melikyan, White, Cohen (1995) "GPI-anchored influenza hemagglutinin induces hemifusion to both red blood cell and planar bilayer membranes" *J Cell Biol*
40. Melikyan, Brener, Ok et al. (1997) "Inner but not outer membrane leaflets control the transition from glycosylphosphatidylino sitol-anchored influenza hemagglutinin-induced hemifusion to full fusion" *J Cell Biol*
41. Chernomordik, Frolov, Leikina et al. (1998) "The pathway of membrane fusion catalyzed by influenza hemagglutinin: restriction of lipids, hemifusion, and lipidic fusion pore formation" *J Cell Biol*
42. Gaudin (2000) "Rabies virus-induced membrane fusion pathway" *J Cell Biol*
43. Dutch, Lamb (2001) "Deletion of the cytoplasmic tail of the fusion protein of the paramyxovirus simian virus 5 affects fusion pore enlargement" *J Virol*
44. Melikyan, Barnard, Markosyan et al. (2004) "Low pH is required for avian sarcoma and leukosis virus Env-induced hemifusion and fusion pore formation but not for pore growth" *J Virol*
45. Stiasny, Heinz (2004) "Effect of membrane curvature-modifying lipids on membrane fusion by tick-borne encephalitis virus" *J Virol*
46. Zaitseva, Mittal, Griffin et al. (2005) "Class II fusion protein of alphaviruses drives membrane fusion through the same pathway as class I proteins" *J Cell Biol*
47. Floyd, Ragains, Skehel et al. (2008) "Singleparticle kinetics of influenza virus membrane fusion" *Proc Natl Acad Sci*
48. Desai, Marin, Chin et al. (2014) "IFITM3 restricts influenza A virus entry by blocking the formation of fusion pores following virus-endosome hemifusion" *PLoS Pathog*
49. Borrego-Diaz, Peeples, Markosyan et al. (2003) "Completion of trimeric hairpin formation of influenza virus hemagglutinin promotes fusion pore opening and enlargement" *Virology (Auckl)*
50. Li, Markosyan, Zheng et al. (2013) "IFITM proteins restrict viral membrane hemifusion" *PLoS Pathog*
51. Melikyan, Markosyan, Roth et al. (2000) "A point mutation in the transmembrane domain of the hemagglutinin of influenza virus stabilizes a hemifusion intermediate that can transit to fusion" *Mol Biol Cell*
52. Melikyan, Egelhofer, Laer (2006) "Membrane-anchored inhibitory peptides capture human immunodeficiency virus type 1 gp41 conformations that engage the target membrane prior to fusion" *J Virol*
53. Zaitseva, Zaitsev, Melikov et al. (2017) "Fusion stage of HIV-1 entry depends on virus-induced cell surface exposure of phosphatidyl serine" *Cell Host Microbe*
54. Markosyan, Cohen, Melikyan (2005) "Time-resolved imaging of HIV-1 Env-mediated lipid and content mixing between a single virion and cell membrane" *Mol Biol Cell*
55. Cavrois, Neidleman, Yonemoto et al. (2004) "HIV-1 virion fusion assay: uncoating not required and no effect of Nef on fusion" *Virology (Auckl)*
56. Clavel, Charneau (1994) "Fusion from without directed by human immunodeficiency virus particles" *J Virol*
57. Kondo, Marin, Kim et al. (2015) "Distinct requirements for HIV-cell fusion and HIV-mediated cell-cell fusion" *J Biol Chem*
58. Marin, Kushnareva, Mason et al. (2019) "HIV-1 fusion with CD4+ T cells is promoted by proteins involved in endocytosis and intracellular membrane trafficking" *Viruses*
59. Yu, Jiang, Li et al. (2025) "HIV-1 and BLV are insensitive to SERINC5 restriction under the cell-cell infection" *Microbiol Spectr*
60. Woottum, Yan, Durringer et al. (2025) "HIV-1 cell-to-cell infection of macrophages escapes type I interferon and host restriction factors, and is resistant to antiretroviral drugs" *PLoS Pathog*
61. Ward, Kiessling, Pornillos et al. (2020) "HIV-cell membrane fusion intermediates are restricted by Serincs as revealed by cryo-electron and TIRF microscopy" *J Biol Chem*
62. Sharma, Marin, Wu et al. (2023) "Human immunodeficiency virus 1 preferentially fuses with pH-neutral endocytic vesicles in cell lines and human primary CD4+ T-cells" *ACS Nano*
63. Raghunath, Chen, Marin et al. (2022) "SERINC5mediated restriction of HIV-1 infectivity correlates with resistance to cholesterol extraction but not with lipid order of viral membrane" *Viruses*
64. Brandenberg, Magnus, Rusert et al. (2015) "Different infectivity of HIV-1 strains is linked to number of envelope trimers required for entry" *PLoS Pathog*
65. Magnus, Rusert, Bonhoeffer et al. (2009) "Estimating the stoichiometry of HIV entry" *J Virol*
66. Marzio, Choe, Ebright et al. (1995) "Mutational analysis of cell cycle arrest, nuclear localization and virion packaging of human immunodeficiency virus type 1 Vpr" *J Virol*
67. Teeranaipong, Hosoya, Kawana-Tachikawa et al. (2013) "Development of a rapid cell-fusion-based phenotypic HIV-1 tropism assay" *J Int AIDS Soc*
68. Wei, Decker, Liu et al. (2002) "Emergence of resistant human immunodefi ciency virus type 1 in patients receiving fusion inhibitor (T-20) monotherapy" *Antimicrob Agents Chemother*
69. Pleskoff, Tréboute, Brelot et al. (1997) "Identification of a chemokine receptor encoded by human cytomegalo virus as a cofactor for HIV-1 entry" *Science*
70. Hammonds, Chen, Zhang et al. (2007) "Advances in methods for the production, purification, and characterization of HIV-1 Gag-Env pseudovirion vaccines" *Vaccine (Auckl)*
71. Miyauchi, Kim, Latinovic et al. (2009) "HIV enters cells via endocytosis and dynamin-dependent fusion with endosomes" *Cell* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12578771&blobtype=pdf | # Correlation Between Infectivity and qRT-PCR Values for Murine Norovirus Recovered from Frozen Berries
Daniel Plante, Julio Bran Barrera, Maude Lord, Jennifer Harlow, Irène Iugovaz, Neda Nasheri
## Abstract
in risk groups, such as children under five, elderly or immunocompromised people, norovirus infections can lead to severe outcomes (Bartsch et al., 2020). In the past two decades, numerous HuNoV outbreaks have been attributed to contaminated berries in particular frozen berries (Bozkurt et al., 2021; Nasheri et al., 2019). The most frequently implicated berry types are strawberries and raspberries as there is evidence that norovirus is very adherent to the surface of these berries (Tian et al., 2011; Trudel-Ferland et al., 2021).In order to reduce the risk of foodborne HuNoV infections, food safety authorities implement regulations and surveillance programs to monitor compliance to the regulations that utilize standardized molecular tests, such as quantitative reverse transcription polymerase chain reaction (qRT-PCR), to assess the levels of HuNoV contamination in high-risk foods (ISO, 2017;Lowther et al., 2019). However, Neda Nasheri
## Introduction
Human norovirus (HuNoV) is the leading agent of acute gastroenteritis, causing millions of infections annually (Havelaar et al., 2015). Although characterized as mild infections, detecting HuNoV in food samples is complicated, mainly due to the absence of routine in vitro cultivation methods to propagate the virus from naturally-contaminated samples. The general approach for detection of enteric viruses in food samples consists of two major steps: (i) sample preparation (virus concentration and nucleic acid extraction); and (ii) detection and quantification of viral genome (usually by qRT-PCR). In this approach, viral detection is based on the amplification of a fragment of the viral genome, and is unable to distinguish between infectious and non-infectious viral particles (Jaykus et al., 2025).
This approach, combined with epidemiological evidence has been implemented for outbreak control and post outbreak analysis (Papafragkou et al., 2025;Raymond et al., 2022;Summa et al., 2024). However, qRT-PCR results in the absence of epidemiological data, pose a significant challenge for comprehensive risk assessments, as multiple studies have shown that food samples that are positive by qRT-PCR might not contain infectious virus. For example, in a human challenge study, none of the 20 participants that consumed berries contaminated with 120-252 genome copies per gram reported any symptoms (Eshaghi Gorji et al., 2021), suggesting the common presence of non-infectious virus or viral RNA fragments. Data regarding the correlation between viral genome copy detected in food samples and human illness is scarce, but it appears that food samples associated with outbreaks and illnesses, have higher viral load compared to the surveillance studies that did not result in human illnesses, for example, one study on naturally contaminated oysters demonstrated that the samples that were associated with human illnesses had approximately one log higher viral load compared to the ones that were not (Lowther et al., 2012). Therefore it is important to know the minimum level of viral genome copy number in food samples that could lead to infection. The aim of this study is to determine the minimum viral titer from frozen berries detected by qRT-PCR that would lead to successful norovirus infection. The current human intestinal enteroid system (HIE) for HuNoV cultivation requires high inoculum levels and does not produce consistent results for viral loads that are often found in naturally contaminated berries (2-3 log genome copies per g (Wales et al., 2023). For this reason, we used a wide range of concentrations of murine norovirus (MNV), as a surrogate for HuNoV, to inoculate frozen raspberries and strawberries in order to determine the lowest inoculum levels that would lead to a successful MNV replication and quantification in cell culture.
## Materials and Methods
## Cells and Viruses
Murine BV-2 cells were maintained in Dulbecco's Modified Eagle Medium (DMEM; GIBCO, Cat. No. 12800-058) supplemented with 2 mM L-glutamine (GIBCO, Cat. No. 25030-081), 100 U/mL penicillin,100 µg/mL streptomycin (GIBCO, Cat. No. 15140-122), 3 g/L sodium bicarbonate and 10% (v/v) fetal bovine serum (FBS; GIBCO, Cat. No. 16140-071).
MNV-1 was kindly provided by Dr. Virgin, Washington University School of Medicine, St. Louis, MO. MNV-1 stock were prepared as previously described (Nasheri et al., 2021). Briefly, MNV-1 was propagated in murine BV-2 cells. When cytopathic effects reached approximately 80% as determined microscopically, the cells were lysed with two cycles of freezing at -80 °C and thawing at room temperature. The virus suspension was clarified by centrifugation at 400 x g, 20 min then passed through a 0.22 µM filter to remove cell debris and stored at -80 °C until use. Virus titer was determined by plaque assay conducted as described previously (Fallahi & Mattison, 2011).
## Inoculation of Frozen Berries
Frozen berries (raspberries and strawberries) were purchased from local grocery stores (Longueuil, QC). Portions of 25 g were weighed in petri dishes and allowed to thaw at room temperature. The berries were inoculated with 100 µL of MNV suspension dispersed in multiple droplets. The inoculum concentrations ranged from 7.1 to 1.0 log PFU, with additional concentrations between 4.1 and 3.1 log PFU as the limit of detection was expected to be in this range (Wales et al., 2023). The samples were placed in a biological safety cabinet until the inoculum was visibly absorbed (approximately 30 min) then transferred into mesh filter bags and processed as described below.
## Virus Recovery with ISO 15216 Methodology
Viruses were recovered as described in ISO 15,216 :2017(ISO, 2017). Briefly, 40 mL of Tris-Glycine-Beef extract buffer (TBGE) with 30 units of pectinase from A. niger was poured into the bags and incubated at room temperature for 20 min under constant agitation at 60 rpm. pH was adjusted to 9.5 with NaOH at the beginning of the incubation and every 10 min thereafter. The liquid was recovered in 50 mL screw-cap tubes, clarified by centrifugation at 10 000xg, 30 min, 4 °C and the pH was lowered to 7.0 with HCl. The volume of rinsate was measured and 0.25 volume of 5X PEG/NaCl [500 g/L PEG 8000, 87 g/L NaCl] was added, followed by incubation at room temperature for 60 min at 60 rpm constant agitation. The tubes were then centrifuged as mentioned above and the supernatant was gently discarded. As the pellets were very loose, residual supernatant was recovered and discarded with a second round of centrifugation. The pellets were resuspended in 1 mL Phosphate Buffered Saline (PBS) and used for plaque assay. For RNA extraction, an equal volume of chloroform: butanol was added to the virus suspension. The aqueous phase was recovered by centrifugation at 10 000xg, 30 min at 4 °C and extracted with the NucleoSpin RNA virus funnel kit from Macherey-Nagel as per manufacturer's instructions. Final elution volume was 100 µL.
## qRT-PCR Detection
The extracted RNA was tested by qRT-PCR as published previously (Bae & Schwab, 2008;Plante et al., 2021), using undiluted and 1/10 diluted RNA.
## Plaque Assay
The viral suspension was serially diluted in Dulbecco's Modified Eagle Medium (DMEM) without amino acids and without Fetal Bovine Serum (FBS) and used for plaque assays. BV-2 murine microglial cells were seeded in 12-well plates at a density of 8 × 10⁵ cells per well in 2 mL of DMEM and incubated for 24 h at 37 °C with 5% CO₂. On the day of infection, three wells per dilution were inoculated with 150 µL of viral suspension and incubated for 1 h at 37 °C, with gentle rocking every 10 min to facilitate viral adsorption. The inoculum was then removed, and the cells were washed 2 times with PBS (8 g/L NaCl, 0.2 g/L KCl, 1.44 g/L Na₂HPO₄, 0.24 g/L KH₂PO₄, pH 7.4). An overlay medium consisting of equal parts 1.4% agarose and 2× DMEM was applied (2 mL per well), and plates were left at room temperature for 15-20 min to allow the overlay to solidify. Plates were then incubated for 48 h at 37 °C with 5% CO₂.
After incubation, the cells were fixed with 2 mL of 3.7% paraformaldehyde for a minimum of 4 h at room temperature. The agar overlay was then removed and the wells were stained with 0.1% crystal violet for 20 min. Plates were rinsed with tap water and air-dried. Plaques were counted manually, and viral titers were calculated as plaque-forming units per milliliter of inoculum (PFU/mL), then converted to PFU per gram of the starting sample.
## Recovery Rate Calculations
The recovery rate for each method was determined as the ratio between the recovered viral titre (PFU) from each sample to the inoculated viral titre (PFU) (recovered P F U/inoculated P F U ) × 100.
## Determination of the Limit of Detection
Each sample was artificially inoculated with decreasing viral titers, in triplicate. At T 0 , which is the time of sample processing immediately after inoculation, the virus was extracted and assayed by plaque assay as described above. The plaques were counted for each inoculation titer and the results were analyzed to determine the lowest titre for which plaques were still obtained.
## Statistical Analysis
Statistical analysis was performed using GraphPad Prism v9.0 (GraphPad Software) and the one-way ANOVA method for extraction efficiency comparisons at each inoculation level. Analysis of comparisons of viral extraction from frozen berries at two inoculation levels was performed by t-test.
## Results
## Determining the Limit of Detection (LOD) on Raspberries
In order to determine the LOD for both plaque assay and qRT-PCR, frozen raspberries were inoculated with varying concentrations of MNV from 7.1 to 1.0 log PFU/25 g and subjected to viral extraction procedure, using the ISO 15,216 protocol. As shown in Fig. 1, the number of plaques and calculated virus recovery followed a linear relationship (R2 = 0.974) with the inoculated MNV concentration down to inoculation level of 3.1 log PFU, where 2 out of 3 replicates yielded plaques. At inoculation level of 2.1 log PFU, no plaque was detected, therefore the LOD for the detection of infectious MNV on raspberries, defined as the inoculated concentration that yields fractional recovery, was determined to be 3.1 log PFU/25 g. The recovery rates for infectious MNV across the inoculation levels above the LOD varied between 10% and 45% as demonstrated in Fig. 2. Overall the recovery rates for extraction of infectious virus across the inoculation levels were significantly higher than the acceptable threshold, which was 1%. The highest recovery rate was achieved at 6.1 log PFU inoculation level and the difference is statistically significant compared to other inoculation levels.
Viral nucleic acid was detected by qRT-PCR from samples inoculated in parallel to those used for cell culture. In all qRT-PCR experiments, undiluted and 1/10 diluted RNA ultimately detrimental to the estimation of LOD in a qualitative detection context. Therefore, the results from undiluted RNA were retained for analysis.
Detection of MNV by qRT-PCR was found to be more sensitive than cell culture. There is a linear relationship between inoculation level and Ct value (R 2 = 0.941) until 1.0 log PFU/25 g (Fig. 3; Table 1), since at this inoculation were tested, and undiluted RNA produced better results (Table 1). Since the Ct value difference between the undiluted and 1:10 diluted samples across all inoculation levels is much greater than 2, the inhibition rate is considered negligible (Table 1) (ISO, 2017). Dilution of RNA might alleviate some potential PCR inhibition if present but, as Ct values were never improved by dilution, this practice was Viral Extraction from Strawberries level, only 2 out of 3 replicates produced results. This means that at Ct values higher than 36.7, the viral genome could not be consistently detected from all the replicates. Detection by qRT-PCR therefore extends about 2 log lower than cell culture.
An important consideration when comparing qRT-PCR and cell culture is that both protocols use different volumes of starting material. Considering all volumes and dilutions, each PCR reaction represented 2% of the starting material (5µL out of 100µL), while cell culture used a maximum of 1.5% per well (150µL out of 1000µL). This slight advantage in testing volume for qRT-PCR is not sufficient to explain the 2-log difference in sensitivity when compared with cell culture. significant proportion of genetic material detected by qRT-PCR does not correspond to infectious particles. Many of the RNA targets that are detected by qRT-PCR might not belong to an infectious virus, furthermore, one PFU might be comprised of an aggregate of many particles, which contain amplifiable genomes.
It is important to note the ratio between infectious and non-infectious viral particles could be different between MNV-1 and HuNoV and depend on the viral strain and the culture system that is employed for propagation. For example, the RNA: PFU ratio for MNV-1 is reported to range from 100 to 10,000 genome copies (Baert et al., 2008;Budicini et al., 2024). However, determining the infectious and non-infectious ratio for HuNoV is much more complicated as no plaque assay exists and infectivity is determined as TCID 50 . It is estimated that the minimum infectious dose for HuNoV in enteroids is around 3 log genome copies (Wales et al., 2023) but even that estimate can vary between strains (Ettayebi et al., 2024). Nevertheless the ratio between infectious and non-infectious viral particles is not drastically different between MNV and HuNoV (Atmar et al., 2014). Furthermore, ratio of non-infectious to infectious particles may change overtime on a matrix as the virus could lose infectivity but still detected by qRT-PCR (Nasheri et al.,2021). Therefore the LOD for qRT-PCR is always lower compared to infectivity assays.
In a Canadian surveillance study, involving berries and pomegranate arils, the HuNoV positivity rate ranged between and 6.1% with the Ct values ranged between 33.9 and 42.2 (Steele et al., 2022), while in a European surveillance study on berries the HuNoV positivity rate ranged between 0.1 and 0.3% with Ct values ranged from 34.2 to 39.3 (Jaykus et al., 2025). In an FDA surveillance study involving frozen berries, the Ct values of the positive samples ranged between 40. 75 and 49.98 (US-FDA, 2025). These Ct values are too high for determination of infectious viruses (data from this study and (Wales et al., 2023), and for effective sequencing, as it has been shown that at Ct values above 35, HuNoV sequencing cannot be consistently achieved (Yang et al., 2024). In our study, the cell culture LOD for MNV in BV-2 cells happened 6.3 Ct earlier than the qRT-PCR LOD. Even though the conditions could be different for HuNoV, it's quite possible that qRT-PCR may generate positive results when the presence of infectious particles is questionable. Therefore interpretation of surveillance data with high Ct values, in the absence of epidemiological data, is challenging. Nevertheless, the authors acknowledge that contamination of berries with HuNoV often occurs at very low concentrations and the viral distribution is non-homogeneous, thus determination of a cut-off Ct value has not been recommended, although at elevated We next examined the extraction efficiency for both infectious MNV and its genetic material from frozen strawberries at inoculation levels higher than the LOD. As shown in Fig. 4A andB, the viral recovery from frozen strawberries is very similar to viral recovery from frozen raspberries with no significant differences between them at medium inoculation levels and by both plaque assay and qRT-PCR. The difference is only significant by qRT-PCR at high inoculation level. This observation demonstrates that there is no marked differences between these two matrices regarding recovery of infectious MNV.
## Discussion
HuNoV is responsible for 54% of the foodborne outbreaks associated with fresh produce, and frozen berries are the major culprits (Chatziprodromidou et al., 2018). For this reason, various non-culture-based testing methods have been implemented to examine the prevalence of foodborne viruses in these high-risk commodities. Positive samples are defined by the presence of a Ct value indicating a successful amplification of the target genomic RNA. Historically these data have been employed in outbreak investigations, supply chain management, and public health protection (Jaykus et al., 2025).
In this study, we employed the ISO 15,216 method for viral isolation, which is widely accepted in the field and is often used for qualitative (presence/absence) determination for berries (Jaykus et al., 2025). We also used MNV as a surrogate for HuNoV to allow detection of infectious virus with cell culture. Overall the extraction efficiencies obtained in this study were significantly higher than the minimum requirement by ISO 15,216 (1%) and inhibition rate was negligible as for most of the tested concentrations, the delays in Ct values between 1:1 and 1:10 diluted RNA are greater than 3.3, which is an indication of 100% efficiency in qPCR (Karlen et al., 2007), also the ΔCt is greater than 2 for all the tested concentrations, which is the acceptable threshold for the ISO 15,216 method (ISO, 2017). These results were consistent between the two types of berries examined in this study.
Herein at the Ct value of 29.8 ± 0.6 or higher (i.e. 2.1 log PFU/25 g), we could not replicate the virus in cell culture. This observation is consistent with the data obtained from culturing HuNoVs in the enteroid system (Wales et al., 2023). However, we still obtained positive qRT-PCR results at Ct value of 36.7 ± 0.6 which translates into 1.0 log PFU of inoculum per 25 g. The proportion of starting material tested by each method were different but similar (1.5% for cell culture vs. 2% for qRT-PCR), yet the LOD of both methods differed by approximately 2 log. This is an indication that a
## References
1. "using plaque assay, transfection assay, and real-time reverse transcription-PCR before and after heat exposure" *Applied and Environment Microbiology*
2. Bartsch, O'shea, Lee (2020) "The clinical and economic burden of norovirus gastroenteritis in the United States" *The Journal of Infectious Diseases*
3. Bozkurt, Phan-Thien, Van Ogtrop et al. (2021) "Outbreaks, occurrence, and control of norovirus and hepatitis a virus contamination in berries: A review" *Critical Reviews in Food Science and Nutrition*
4. Budicini, Rodriguez-Irizarry, Maples et al. (2024) "Murine Norovirus mutants adapted to replicate in human cells reveal a post-entry restriction" *Journal of Virology*
5. Chatziprodromidou, Bellou, Vantarakis et al. (2018) "Viral outbreaks linked to fresh produce consumption: A systematic review" *Journal of Applied Microbiology*
6. Eshaghi Gorji, Tan, Zhao et al. (2021) "No clinical symptom experienced after consumption of berry fruits with positive RT-qPCR signals of human norovirus" *Pathogens (Basel Switzerland)*
7. Ettayebi, Kaur, Patil et al. (2024)
8. Fallahi, Mattison (2011) "Evaluation of murine norovirus persistence in environments relevant to food production and processing" *Journal of Food Protection*
9. F P - (1981)
10. Havelaar, Kirk, Torgerson et al. (2015) "World health organization global estimates and regional comparisons of the burden of foodborne disease in 2010" *PLoS Medicine*
11. Iso (2017) "Microbiology of the food chain-Horizontal method for determination of hepatitis A virus and norovirus using real-time RT-PCR-Part 1: Method for quantification" *ISO*
12. Jaykus, Bidawid, Bosch et al. (2025) "Detection of foodborne viruses. Berries -State of Science and Future Considerations Food Control"
13. Karlen, Mcnair, Perseguers et al. (2007) "Statistical significance of quantitative PCR" *BMC Bioinformatics*
14. Lowther, Bosch, Butot et al. (2019) "Validation of EN ISO method 15216 -Part 1 -Quantification of hepatitis A virus and Norovirus in food matrices" *International Journal of Food Microbiology*
15. Lowther, Gustar, Hartnell et al. (2012) "In conclusion, this study further confirms that positive qRT-PCR test results do not assure detection of infectious viruses, particularly at higher Ct values. From a risk management perspective, it could be argued that positive qRT-PCR results suggest a risk of exposure and not necessarily a risk of infection. Therefore caution should be taken in interpretation of surveillance data"
16. Atmar, Opekun, Gilger et al. (2014) "Determination of the 50% human infectious dose for Norwalk virus" *Journal of Infectious Diseases*
17. Bae, Schwab (2008) "Evaluation of murine Norovirus, feline calicivirus, poliovirus, and MS2 as surrogates for human Norovirus in a model of viral persistence in surface water and groundwater" *Applied and Environmental Microbiology*
18. Baert, Wobus, Van Coillie et al. (2008) "Detection of murine norovirus 1" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12835948&blobtype=pdf | # Epidemiological indicators of accidental laboratory-origin outbreaks
Sandhya Dhawan, Wirichada Pan-Ngum, Raina Chandini, Macintyre, Stuart Blacksell
## Abstract
Accidental escapes of pathogens from laboratories continue to cause outbreaks in the community today, posing significant risks to the general public, animal communities and the environment. These incidents, as well as the uncertainties surrounding the origins of the COVID-19 pandemic, highlight the need to consider unnatural origins as part of emerging outbreak surveillance and detection. Identifying recurring patterns and distinctive factors of laboratory-associated disease outbreaks can aid in successfully preventing and mitigating these occurrences. Seventy incidents of laboratory-associated leaks that led to outbreaks in the wider public have been reported (Supplementary Appendix S1). Seven renowned cases that have been comprehensively studied were selected for review: (i) 1955 Polio vaccine incident in western USA, (ii) 1977 H1N1 influenza virus re-emergence in China and the Soviet Union, (iii) 1979 Anthrax release in Sverdlovsk, Soviet Union, (iv) 1995 Venezuelan equine encephalitis epidemics in Venezuela and Colombia, (v) 2003-4 SARS-CoV-1 escapes from Singapore, Taiwan and China, (vi) 2007 Foot-and-Mouth disease virus outbreak in Pirbright, England and (vii) 2019 Brucella leak in Lanzhou, China. These outbreaks were selected because data on their geographical spread, genetics, phylogeny, epidemiological factors (including attack rates, infectious dose, time, location and season of spread) and governmental and institutional responses to the incidents had been previously analysed and published. Thematic analysis of these lines of evidence revealed seven recurring insights described in historically confirmed laboratoryassociated outbreaks: unusual strain characteristics, peculiar clinical manifestations or affected demographics, unusual geographical features, atypical epidemiological patterns, delayed government action and communication to the public, misinformation and disinformation spread to the public and biosafety concerns/incidents predating the event. The outbreaks exhibited between 13 and 19 retrospectively identified indicators. These indicators were used to develop preliminary risk criteria intended to support structured, hypothesis-generating assessment of outbreaks, rather than to establish origin.
## Introduction
Despite continual advances in biosafety and biosecurity policies, accidental pathogen escapes from laboratories continue to cause disease outbreaks in the community. The question is not if a pathogen will escape, but rather which pathogen will and what measures are in place to contain an escape with serious consequences [1]. Past laboratory-origin epidemics [2][3][4] and outbreaks of unknown origin [5], namely the Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) pandemic (2019-2023) [6], underscore the need to consider unnatural origins when identifying outbreaks.
Different lines of evidence, including phylogenetics, epidemiology, seroepidemiology and criminal or geopolitical intelligence, are required to determine whether an outbreak is of unnatural origin [7]. Phylogenetics alone may not identify pathogens of laboratory origin because serial passaging a pathogen through an animal host will produce genetic markers that appear to be of natural origin [7]. To investigate distinctive factors of laboratory-origin outbreaks, historically confirmed incidents should be studied to identify emerging themes and indicators.
A total of 70 incidents of accidental laboratory leaks have been reported, with the earliest recorded in 1901 and the most recent in 2024 (Appendix, Supplementary Table 1). Of these incidents, 56 (80.0%) resulted in community cases and 29 (41.4%) resulted in fatalities. Seven well-known cases that have bee comprehensively studied were selected for review: (i) 1955 Polio vaccine incident in western USA, (ii) 1977 H1N1 influenza virus re-emegence in China and the Soviet Union, (iii) 1979 Anthrax release in Sverdlovsk, Soviet Union, (iv) 1995 Venezuelan Equine Encephalitis epidemics in Venezuela and Colombia, (v) 2003-4 SARS-CoV-1 escapes from Singapore, Taiwan, and China, (vi) 2007 Foot-and-Mouth disease virus outbreak in Pirbright, England, and (vii) 2019 Brucella leak in Lanzhou, China, were comprehensively evaluated for insights into their geographical, epidemiological, phylogenetic and other characteristics. These indicators were used to explore whether similar epidemiological features have been discussed in relation to the SARS-CoV-2 pandemic, without implying causality.
## Methods
A literature review was conducted across the PubMed, ProMED-Mail, Scopus and Web of Science databases using keywords to identify published literature on the seven laboratory-confirmed outbreaks. Public information accessible on the World Health Organization (WHO) and the Centres for Disease Control (CDC) platforms was gathered, as well as relevant news articles, government reports, correspondence and grey literature published during the respective outbreaks. The information collected included historical facts, witness accounts, outbreak investigations, characteristics of the outbreak and strain, epidemiological parameters and descriptive statistics. An article or report was excluded if it contained no information on any risk variables analysed (geographical spread, genetics and phylogeny, epidemiological factors, timely and accurate reporting/infodemic).
## Summary of case studies
The Cutter Laboratories polio vaccine trials across the western United States
Poliomyelitis epidemics plagued the world in the 1950s, leading to intensive research into the development of inactivated or live-attenuated vaccines for poliovirus [8]. In April 1955, Cutter Laboratories in California was licenced to produce the Salk formaldehyde-inactivated polio vaccine (IPV), following successful trials [9]. However, some batches produced by Cutter Laboratories were insufficiently inactivated and contained live poliovirus [10,11]. Multiple children received these contaminated doses, leading to tens of thousands of abortive infections, dozens of paralytic cases and several deaths, including secondary transmission within families and communities [11,12]. The incident shook public trust in vaccines, reshaped vaccine policy and became a defining moment in the history of vaccine safety.
## Geographical spread
Approximately 120,000 contaminated doses were administered to primarily grade-school children, and roughly 400,000 people received the Cutter vaccine during a 10-day period in mid-April. A majority of them developed abortive polio [8]. Most cases occurred between late April and May 1955, then declined sharply by June, aligning with the vaccination window [11]. Ultimately, at least 220,000 people were exposed, including 100,000 household contacts of immunized children, resulting in 164 cases of severe paralysis and 10 deaths [11].
Infections clustered in states where the Cutter vaccine was widely used. California and Idaho experienced the highest numbers, while nearby states saw smaller outbreaks. Idaho typically reported very low polio incidence (11 cases) during April-June in 1950 to 1954. In 1955, however, an eightfold increase (88 cases) was observed, of which 84 were attributable to vaccine-associated cases [13].
## Genetic and clinical evidence
Paralytic poliomyelitis developed 4 to 10 days after vaccination in patients [11], with paralysis typically beginning in the inoculated arm, a pattern less common in natural poliomyelitis cases [12,13]. All early cases were linked to Cutter vaccine recipients, with secondary cases in their families and community contacts [13,14]. Severe neurological complications involving the central nervous system were noted in those who were administered the vaccine, as compared to family contacts as well [14]. The incidence of paralytic disease peaked in children aged 7 across the western US, reflecting an unusually high concentration of poliomyelitis in children, likely because they were heavily vaccinated under school programs in this region [15].
Laboratory investigations revealed that live type 1 poliovirus (the causative agent of the Cutter vaccine-associated cases) was isolated from 7 of 8 vaccine lots, demonstrating failed inactivation. Additional findings of types 2 and 3 poliovirus in certain lots underscored multiple strains of poliovirus circulating at the time, even though types 2 and 3 did not result in clinical cases [14,16].
## Epidemiological factors
The number of cases among Cutter vaccine recipients and their contacts far exceeded what would be expected for natural poliomyelitis at the time, and vaccine recipients from other manufacturers showed no such pattern [13]. Moreover, Cutter vaccineassociated cases declined as cases of seasonal poliomyelitis began to increase [13].
Incubation periods among vaccinated children ranged from 4 to 15 days, shorter than the typical interval for naturally occurring polio, whereas contact cases showed incubation periods consistent with secondary transmission [11]. Vaccinated cases peaked approximately 1 week after vaccination, and secondary cases peaked 3 weeks after the midpoint of vaccination. The appearance of cases in waves is suggestive of a common-source outbreak [13].
Attack rate analysis of the vaccine lots revealed that two of three production pools were inadequately inactivated and accounted for a more than 10-fold increase in paralytic cases [11]. Secondary attack rates in household contacts were similar to those seen in natural poliomyelitis, with limited spread to wider community contacts [13].
## Timely and accurate reporting/Infodemic
The Salk vaccine was licenced within a day under political pressure and distributed widely within two weeks [10]. Although the vaccine had passed required safety testing [16], biosafety concerns at Cutter Laboratories were identified, including the use of the highly virulent Mahoney strain, insufficient viral inactivation and inadequate safety testing [10]. The vaccine was swiftly withdrawn on April 27 after cases rose sharply [11].
Overall, communication with other scientists and the government was poor. Swedish researchers, such as Sven Gard, presented research showing that Salk's vaccine inactivation procedure was ineffective [17][18][19]. Despite these concerns, Salk did not make the proposed changes, and the vaccine trial was launched in 1954, even though regulators lacked the capacity to validate each dose during production, relying on manufacturers for quality assessment [10,17].
In fact, a similar incident was documented in which another company using the Salk IPV, Wyeth Laboratories, was responsible for 37 vaccine-associated poliomyelitis cases. Yet, the report was kept confidential from public health authorities and the public [10,17].
Media further contributed to misinformation and infodemics, which eroded public trust in vaccines. Vaccine rates significantly dropped across the world [17,20]. The Cutter polio vaccine incident contributed to the shift toward Sabin's oral polio vaccine (OPV) in the 1960s [8]. Both the Salk and Sabin vaccines were also found to have been contaminated with Simian Vacuolating Virus (SV40) due to inadequate formaldehyde inactivation of the monkey kidney cells used to cultivate poliovirus [21]. This led to SV40-contaminated polio vaccines being administered to millions of people between 1955 and 1963 [8].
The re-emergence of the H1N1 influenza virus in China and the Soviet Union Every pandemic influenza strain has replaced its predecessor strain [22,23]. However, in 1977, two serotype A viruses were recorded to co-circulate for the first time in history: the dominant H3N2 subtype and the previously extinct human influenza A H1N1 virus [24]. This situation was due to an accidental release during laboratory activities.
## Geographical spread
The H1N1 influenza re-emerged in northeast China in May 1977 and soon after in the eastern Soviet Union [25]. The Soviet Union reported the outbreak to the WHO in September 1977, followed by Chinese reports in May 1978 [26]. The c.1957 H1N1 virus strain initially spread in the Soviet Union, Hong Kong and China, then rapidly worldwide, causing mild infections in individuals under 21 while excess mortality was largely confined to older populations, with global death estimates varying widely. [26,27].
## Genetic and clinical evidence
Genetic analysis showed that the 1977 H1N1 virus outbreak strain was closely related to strains from 1949-1950 but distinct from the 1947 or 1957 strain [27], suggesting it had likely been preserved since 1950 [27] and accidentally released when population immunity to H1 and N1 antigens declined [8].
Many isolates from the outbreak were temperature-sensitive, a marker of laboratory manipulation distinctive to live attenuated influenza vaccine (LAIV) studies. However, not all strains were temperature-sensitive [28]; a mixed population of strains suggests a possible escape event during the temperature-sensitivity selection experiment [26,28]. The outbreak strain had low virulence, varied attack rates within the same region and a low mortality rate, likely owing to attenuation and pre-existing immunity in the older population [1,26].
## Epidemiological factors
The H1N1 virus spread more slowly nationwide in Liaoning (May-October) than the Asian H2N2 pandemic in 1957 (February to March) [26]. This unusual disease progression may have been due to an unfavourable season, although off-season outbreaks suggest an unnatural origin.
Two factors suggest that an incompletely attenuated vaccine strain caused the outbreak: ongoing research on LAIV at the time [1] and the renewed interest in prophylaxis following the 1976 H1N1 outbreak at Fort Dix, New Jersey [29]. It is plausible that a Chinese or Russian vaccine facility thawed and cultivated a c.1950 H1N1 influenza virus in response to the US 'swine flu' program launched in the aftermath of the Fort Dix outbreak [1].
## Timely and accurate reporting/Infodemic
The source was debated, with suggestions of an accidental escape being refuted by Chinese and Soviet virologists [26]. Western scientists refrained from discussing the laboratory-origin theory to foster collaboration amid Cold War tension [1]. Natural-origin hypotheses included the possibility of viral latency in an unspecified animal reservoir. In 2006, a paper suggested that the virus emerged from migratory birds at Siberian lakes, after isolating strains mistaken for avian H1N1 influenza virus from meltwater. The paper was criticised in 2008, where direct evidence demonstrated that the meltwater strain, ironically, was also leaked from a laboratory [30]. In 2009-2010, the laboratory release theory became widely accepted [1].
## The release of inhalational anthrax from an exhaust vent in Sverdlovsk, Soviet Union
After WWII, the Soviet Union established an anthrax production plant [31] in their Military Research Facility: Compound 19 in Sverdlovsk. The causative agent, Bacillus anthracis, mainly affects domestic animals and, occasionally, humans via cutaneous transfer or, rarely, through ingestion or inhalation [32]. Natural anthrax is almost always cutaneous, and inhalational anthrax should raise suspicion towards a deliberate event [32].
A clogged filter in the exhaust vent was removed but not replaced; machines ran for several hours as usual [31]. Anthrax aerosols escaped to a ceramic plant and a town nearby, where many workers were discovered ill. Within a week, most exposed workers had died, and hospitals received an influx of patients from different towns [31].
## Geographical spread
According to Soviet reports, the epidemic began in late March, took place from 4 th April to 18 th May 1979, and caused a total of 96 cases with 66 fatalities [2]. Witnesses claimed a death toll of ~105 [31], and an article quoted as many as a thousand deaths [32]. The actual number of human fatalities or cases remains unknown, as it was reported that the KGB destroyed most hospital records [31,33]. In an attempt to conceal the truth, the incident was falsely attributed to gastrointestinal anthrax (a rare manifestation) resulting from consumption of anthrax-contaminated meat [31].
## Genetic and clinical evidence
Genetic studies dated the accident to April 3 rd or 4 th, 1979, consistent with the observed anthrax incubation period. The Soviet officials falsely reported the start date as March 30 th 1979, manipulated medical records of early cases and issued fabricated death reports to the victims' families as part of the cover-up [31]. They also denied inhalational anthrax, although this was later confirmed from autopsy data [33].
## Epidemiological factors
It was estimated that victims were exposed to a far lower infectious dose (~1-10 or 100-2,000 spores) than observed for naturally occurring inhalational anthrax (8,000-10,000 spores) [34], signifying a potentially weaponised strain. This is consistent with early clinical studies, where more than four virulent strains of B. anthracis circulating during the accident were all traced to the biological weapons facility [35].
Most infected patients worked or lived close to the military facility (within 4 km) 35 , with animal cases detected up to 60 km away [2]. The aerosol size was estimated to be <5-10 μm to have allowed for extended dispersal and prolonged infection, more extensive than that observed in the 2001 'Amerithrax' attacks (<5 μm) [34].
Autopsies confirmed that fatal cases resulted from inhalation exposure [33], a rare clinical form of natural anthrax [34]. The mean incubation period of the Sverdlovsk accident (~10 days, with some cases appearing after 43 days) 2 was longer than that of naturally occurring anthrax outbreaks and even the 2001 Amerithrax intentional anthrax release (4-6 days) 35 .
## Timely and accurate reporting/infodemic
In November 1979, a Russian magazine reported that in April, 'an explosion in the military facility of Sverdlovsk had released a cloud of deadly bacteria' [31]. Western agencies later picked up coverage of the outbreak [2], alleging it was a violation of the 1972 Biological Weapons Convention, although all claims of a laboratory accident were denied.
Workers in Compound 19 raised biosafety concerns about airborne spores in the laboratory, clogged filters and neglected maintenance checks [31]. After the accident, senior officers were alerted, but city officials and the Ministry of Defence in Moscow were not informed.
On June 12, 1980, residents of Sverdlovsk were informed that the outbreak was caused by contaminated meat from illegal wet market stalls, leading to the culling of more than 100 stray dogs and animals in the vicinity [31].
Soviet authorities denied requests to permit independent scientists to investigate the incident [2]. An 'information war' arose between those doubting a natural outbreak and those convinced of its natural origin. It took nine years for Soviet medical experts to disclose information about the Sverdlovsk incident to the US and thirteen years for then-Soviet President Boris Yeltsin to admit to the accident [1].
## The VEE epidemics in Venezuela and Colombia
In 1995, one of the largest epidemics of VEE was documented in Venezuela and Colombia [8]. VEE is an arboviral disease transmitted by mosquitoes that causes intermittent epizootics and sometimes human epidemics across the Americas. Equine disease is severe, while human infections range from asymptomatic to acute febrile illness with neurological complications, and fatality rates of 4 to 14% [1,8]. Naturally, VEE circulates at low levels in enzootic cycles. The enzootic strains (ID, IE, IF, II-VI) rarely cause major outbreaks, which occur only when an enzootic strain mutates into an epizootic subtype (IAB or IC) that efficiently amplifies in equines and drives widespread human spillover [36,37].
Epizootic strains have mutated from enzootic only three times (ID!IAB in the 1930s; ID!IC in 1963 and again in 1992). However, many VEE outbreaks reported from the late 1930s through early 1970s were traced to inadequately inactivated veterinary vaccines derived from the 1938 IAB strain [38,39]. Residual live virus in the vaccines repeatedly sparked outbreaks until the seed strain was replaced with an attenuated variant in 1973, after which epizootics ceased for nearly 20 years [38,40] (Figure 1). Unlike the IAB strains, there is no record that subtype IC strains were ever used in vaccine production, hence an unlikely source of the 1995 outbreak.
The 1995 outbreak in Venezuela and Colombia was unusual because this strain matched an IC virus used in diagnostic reagents in a local virology laboratory, one previously shown to contain live virus. Many investigators concluded that the 1995 epidemic most likely resulted from an inadvertent laboratory escape rather than natural evolutionary emergence [41,42].
## Geographical spread
In April 1995, veterinarians in Venezuela first detected equine deaths suggestive of VEE, followed by human febrile cases [43]. The outbreak began in eastern Falcón State, Venezuela and spread westward across states by mid-July. Transmission intensified in rural areas by late August, and by September-October, large numbers of human and equine cases were reported in the Colombian state of La Guajira [42,44]. Overall, the epidemic caused ≥ 100,000 human cases and ~300 deaths [41][42][43].
VEE outbreaks typically emerge in regions with known enzootic subtype ID circulation, and within localized equine-mosquito amplification cycles [36,37]. The only historical exception was the 1969-1971 outbreak originating in the Guajira peninsula, although the area had prior enzootic ID activity [45]. In contrast, the 1995 outbreak began abruptly in Falcón State, a region with no record of circulation of closely related enzootic ID progenitor strains [46], or of laboratories or vaccine production facilities in close proximity [47].
Heavy rainfall in the normally arid Guajira region increased vector densities and expanded the spread [42]. However, unlike natural transmission patterns, the outbreak spread rapidly through rural areas with limited equine populations, suggesting that human-mosquito-human transmission was also occurring.
Importantly, the 1995 virus was identical to a subtype IC antigen strain that was in regular use for antigen preparation in laboratories near the outbreak area at the time [41].
## Epidemiology and Infection
Genomic analyses showed that the 1995 outbreak virus was subtype IC, which had previously caused two other major epidemics (1962 to 1963 and 1992 to 1993). It was a genetic match to a strain isolated in 1963, which had disappeared from nature 30 years ago [41]. The 1995 viral sequence showed almost no evolutionary change during the interepidemic period, inconsistent with estimates of epidemic and enzootic VEE virus evolution rates, which indicate a relatively steady rate of nucleotide substitutions, on the order of 2-4 × 10 À4 substitutions/nucleotide/year [48][49][50][51].
Phylogenetic analysis demonstrated sequence identity to the P676-ag virus, isolated from a 1982 antigen preparation used for diagnostic testing in Venezuela [41], explaining the genetic conservation between the epidemic events.
Clinically, infected humans exhibited high viremias comparable to those in equines, sufficient to infect the epidemic mosquito vector [42,52,53]. Higher disease incidence was observed in unimmunized equines, particularly donkeys, due to low equine immunization rates in the region [43].
## Epidemiological factors
The 1995 VEE epidemic displayed attack rates of ~36%, with some communities reporting rates as high as ~93%, far exceeding typical VEE epizootic patterns [43]. However, secondary attack rates were low in Colombian communities, and no secondary infections occurred among Venezuelan healthcare workers, indicating low person-to-person transmission despite extensive community spread [54].
Field studies, before and after the epidemic, found no evidence of ongoing circulation of epizootic IAB or IC strains [45,46,55], or enzootic ID viruses genetically related to the 1995 IC strains in northern Venezuela, demonstrating an absence of local natural reservoirs for the disease [46].
Natural transmission often generates genetic shift or drift due to low-dose mosquito transmission, which was not observed with the implicated strain [56,57]. Furthermore, the much faster geographic spread of the 1995 outbreak compared to the natural 1962-1964 epizootic suggests an atypical introduction rather than gradual local emergence [42].
## Timely and accurate reporting/Infodemic
The Colombian Ministry of Health and Animal Health Service deployed timely surveillance across La Guajira, generating realtime intelligence to guide vector control, equine vaccination and movement restrictions. Implementation of early interventions, in line with animal health regulations, helped prevent wider domestic or international spread at a time when heavy rainfall had increased vector density and heightened epidemic risk [44].
While the spread of infodemic during the outbreak was limited, scientific discourse on the origins of the virus emerged years after the epidemic [41]. Between 2000 and 2003, outbreaks of a subtype IC strain genetically identical to the 1995 virus were reported in Venezuela, despite the strain no longer being widely used in laboratories (Figure 1) [47]. From 1995 to 2000, this IC lineage showed a ≈ 10-fold slower evolutionary rate, implying limited replication compared with typical mammal, mosquito, or equine transmission cycles [47]. Field investigations failed to identify reservoir hosts or vectors, and the outbreaks occurred at the end of the rainy season, which is not typical of the VEE epidemic pattern [47]. The 2000 strains also did not cluster phylogenetically with the P676-ag strain, and the 2000s outbreak locations were not near any diagnostic or vaccine production laboratories that work with VEE virus either [47]. These findings suggest that the 2000 outbreaks involved naturally circulating strains that have remained genetically stable since the 1995 laboratory release. While these anomalies led the VEE working group to reconsider their suggestion of a laboratory origin, direct evidence of a natural mechanism causing prolonged genomic stasis in the subtype IC lineage was not identified.
The SARS-CoV-1 escapes in Singapore, Taiwan and China
The initial risk of contracting SARS-CoV-1 through laboratory exposure is very higheven a single mishap could lead to a potential pandemic [58]. This was evidenced by six documented escapes from high-containment virology laboratories: one from a Biological Safety Level (BSL) 3 in Singapore, one from a BSL-4 in Taipei and 4 from the same BSL-3 in Beijing [1]. Despite raising public health alarms, these escapes are not referenced in historical and official reviews of SARS-CoV infections.
A laboratory exposure incident in Singapore
In August 2003, a graduate student at the National University of Singapore (NUS) contracted SARS in a BSL-3 laboratory at the Institute of Environmental Health (EHI) Singapore despite handling West Nile Virus (WNV). Examination of the vials revealed that the WNV samples had been cross-contaminated with a SARS-CoV-1 isolate [59]. The student's sample-preparation techniques were speculated to be the cause of infection. The infected student exposed 8 household contacts, 2 community contacts, 32 hospital contacts and 42 work contacts, of whom 25 were placed under home quarantine [59]. No secondary cases occurred.
Investigations in the laboratory revealed poor record-keeping, missing or defective equipment, a lack of freezers and HEPA/air filter problems, all of which were exacerbated by the student receiving insufficient training in BSL-3 procedures [60].
## A laboratory spill in Taiwan
In December 2003, research scientist Lieutenant-Colonel (LTC) Chan Jiacong at the Taiwan Military Institute of Preventive Medical Research (IPMR) acquired a SARS-CoV-1 infection, before travelling to Singapore for a conference, where he exposed fellow passengers and airline staff [61].
While working with SARS-CoV-1, LTC Chan found a leaking waste bag; in a hurry to travel, he inadequately disinfected the spill and incorrectly disposed of the waste without appropriate personal protective equipment [62]. WHO investigations revealed numerous safety violations in the laboratory, including poor record-keeping, long work shifts (12-14 h) and the absence of incident-reporting protocol [61]. Original reports cited 95 contacts placed in quarantine, while WHO investigations reported only 74 contacts [1].
## A laboratory-origin SARS-CoV-1 outbreak in China
In April 2004, reports of a nurse with a hospital-acquired SARS-CoV-1 infection came from Beijing, China. She had contracted the illness from a graduate student who was admitted for pneumonia in March. Eventually, the disease spread among their family contacts and healthcare workers over three generations, causing one death [63]. Official reports initially accounted for 9 total cases; however, investigations revealed 2 additional cases from February 2004 [64] (Figure 2).
## Geographical spread
The graduate student was interning at the viral diarrhoea department of the Chinese National Institute of Virology (NIV) in Beijing, a part of the China Centre for Disease Control (CCDC), and did not work with SARS-CoV nor in a BSL-3 laboratory; the exact mechanism of infection is unknown [1]. She travelled home by train while ill, where her mother developed a severe infection and died as a consequence of attending to her. The nurse who had contracted the illness from the student also transferred it to an additional five individuals (Figure 2) [63]. Investigations found another postdoctoral researcher at NIV who had been infected with SARS-CoV on April 17, 2004 [64]. By the end of April, officially, 747 people were quarantined at NIV [1] and unofficially, over a thousand people [63]. Further investigation found two more graduate students from the same department at NIV who contracted SARS-CoV independently in February [64] (Figure 2). Official reports suggest that one of the doctoral students improperly inactivated a SARS-CoV sample, contaminating the electron microscopy room, from which the second student also acquired the infection [63]. Neither student caused secondary cases, and both recovered. The deactivation solution prepared to inactivate SARS had not been verified or recommended by the Ministry of Health [63].
## Epidemiological factors
Healthcare workers account for almost 16% of probable SARS-CoV cases with attack rates of >56% [65]. The attack rate (4.23%) and case fatality rate (9.1%) observed in the Beijing escapes were lower than the standard [65], perhaps owing to timely interventions, such as quarantine and isolation, that prevented the outbreak from spreading further. The disease outbreak also occurred in the summer, a low season for virus spread.
## Timely and accurate reporting/Infodemic
A joint WHO-China report reviewed the cases, although it did not mention two primary cases from February, which were officially discovered in May via IgG testing [64]. As both students were hospitalized in February, the LAIs were known prior to the antigen testing. Perhaps these cases were not disclosed by the institution in the April report [63], as they were later recognised in a WHO report in October 2004 [66]. The WHO highlighted biosafety shortcomings with handling live SARS-CoV and surveillance of LAIs at NIV [1].
## The Foot-and-Mouth virus leak from drainage pipes in Pirbright, England
The United Kingdom was free of Foot-and-Mouth Disease (FMD) for six years until its re-emergence in 2007 [67]. FMD virus (FMDV) predominantly infects cattle, sheep and pigs, with rare cases of mild illness in humans [68]. It is highly transmissible and can spread via contaminated surfaces, aerosols (up to 250 km) and fomites [1,68]. The FMD outbreak in 2001 incurred a $16 billion loss to the British economy [1]. On 3rd August 2007, the United Kingdom reported an FMD outbreak detected on a cattle farm in Surrey [69]. The pathogen escaped from the Pirbright campus, the only authorized facility in the UK for storing FMDV, specifically the Institute for Animal Health (IAH) and Merial, a vaccine manufacturer.
Initial investigations ruled out aerosol or surface water transmission from Pirbright [70]. Eventually, they revealed a damaged wastewater pipe connecting the Merial vaccine plant to the waste treatment plant in IAH, leaking partially treated waste into the ground and surface water. FMDV-contaminated mud was carried from the campus to the farms via flooding, roads and the tyres of construction vehicles parked at the site [69]. FMDV likely spread further via windborne and fomite transmission and was exacerbated by visitor car parks located near livestock areas [69].
## Geographical spread
The 2007 FMD outbreak infected 8 premises and 278 animals, necessitating the culling of 1,578 animals and resulting in estimated losses of £200 million [71]. The epidemic comprised two distinct clusters: two farms in the first cluster and six in the second [72]. The intermediary farm between the two infection clusters was missed during initial surveillance [69].
## Genetic and clinical evidence
The outbreak-causing virus was FMDV serotype O subtype BFS 1860 isolated in 1967, a strain no longer circulating globally [69,70] but used in large quantities (10,000 l) at the Merial facility and in microquantities at IAH [70]. Both facilities fault the other, leaving the exact location of the outbreak uncertain [1]. Genomic analysis indicated a single escape of FMDV from Pirbright, dating between 13 and 26 July 2007 [71], which caused the August outbreak and reemerged in mid-September 2007.
The first cases of FMD were discovered in cattle, although pigs are more sensitive to FMDV infection in natural or introduced outbreaks and are often the first animals infected [73]. Additionally, ruminants are more susceptible to airborne infections [74], and the FMDV vaccine production guidelines also focus on efficacy testing and immunization in cattle [75]. The outbreak strain, which shows a greater affinity for ruminants, may indicate a released FMD vaccine strain.
## Epidemiological factors
The second outbreak cluster began after a significant temporal lag [72]; the epidemic was prematurely deemed over, restrictions on livestock movement were lifted and farm surveillance was eased [69]. Molecular analysis revealed that one of the two originally classified primary farms had been infected by the other, initially overlooked because it lay beyond the 10 km radius for animal epidemic monitoring [76].
Furthermore, risk mapping of the 2007 outbreak indicated an extremely low likelihood of local spread compared to the 2001 FMD outbreak [72], with the sparse livestock density in Surrey raising suspicion of unnatural origins. Moreover, the R 0 for the 2007 outbreak was ~15, much higher than that for the 2001 outbreak (~4 [71]. While the 2007 strain was more transmissible, there is evidence that a very low level of virus was circulating in infected animals [71].
## Timely and accurate reporting/Infodemic
The outbreak had a limited infodemic, as animal disease reporting is more strictly enforced and standardized, with delays in reporting risking immediate trade bans and fines. Under the World Organisation for Animal Health (WOAH, formerly known as the OIE) Animal Health Code, member states must notify within 24 h of confirmation [77], while the WHO International Health Regulations (IHR) allow 48 hours [78] with no tangible penalties. Animal health surveillance covers a wider radius with multiple detection points, e.g., mandatory farm reporting, abattoir checks, etc. [79], whereas human health surveillance relies on public health data sharing, often hindered by resource constraints or patient privacy laws [80].
Systemic gaps in communication and reporting were mirrored in site conditions, where long-term damage to drainage pipework was known but unaddressed until FMD emerged in nearby farms [81].
The Brucella-contaminated waste-gas leak in Lanzhou, China
## Incident description
The Lanzhou Brucella leak is the largest and longest recorded laboratory-origin outbreak [82], surpassing the 1977 H1N1 and 1979 Sverdlovsk anthrax events [2]. Brucellosis, a globally prevalent zoonotic disease, poses significant public and animal health concerns despite its low human mortality rate [83]. It is transmitted to humans via contact with infected livestock, ingestion of unpasteurized dairy or undercooked meat, and, less commonly, aerosol inhalation, often from laboratory accidents or releases during microbiologic technique [84].
The first cases of Brucella spp. infections were detected in November 2019 at the Lanzhou Veterinary Research Institute (LVRI) in Gansu, China, with 181 positive individuals by December [85]. The outbreak originated from the Zhongmu Lanzhou Biopharmaceutical Plant, a state-owned facility producing Brucella vaccines for animals [86]. Expired disinfectants led to contaminated waste gas leaking from fermentation tanks downwind to LVRI and neighbouring communities [82].
## Geographical spread
Although the factory's manufacturing licence was revoked immediately, transmission continued for at least 12 months [82]. By 30 November 2020, the Health Commission of Lanzhou reported 10,528 cases among 68,571 tested [86]. No deaths were reported, but the full extent of cases is inconclusive due to limited published official data [82,83].
## Genetic and clinical evidence
The leaked strain was a Brucella abortus A19 vaccine strain, [83]. Owing to its low infectious dose Brucella spp. accounts for almost 2% of all LAIs [87], with many documented in China since 1936 [83].
## Prior incidents
In 2017, a similar outbreak caused by Brucella suis S2 vaccines was reported in Gansu, where 51 animal epidemic prevention controllers were positive (attack rate: 24.8%) [85]. In December 2020, another incident occurred at a biological products company in Chongqing, where 61 workers were positive (attack rate: 43.6%). The infection spread to nearby departments [85]. Neither facility complied with biosafety regulations: improper handling techniques, ineffective PPE and ineffective emergency measures were identified [83].
## Epidemiological factors
Brucella was transmitted by aerosols from late July to August 2019 [82,83], but the attack rate observed with the Lanzhou leak was much lower than typical laboratory exposures (~30%-100%) [88], ranging from ~12.9% to 15.4%. Moreover, no deaths were reported despite a case-fatality rate of 1-2% for brucellosis.
Brucellosis outbreaks in Lanzhou are unusual, with low seroprevalence from 2013 to 2018, and has never been at risk of a Brucella epidemic [89]. The sudden increase in cases cannot be attributed to improved surveillance [82], as low levels of brucellosis were detected in high-risk individuals [90].
## Timely and accurate reporting/Infodemic
The absence of official clinical data from the incident raises concerns about the effectiveness of the response by Chinese authorities [82]. One study revealed that 96 initially exposed individuals were asymptomatic but seroconverted, without mention of the total patients tested or follow-ups [91]. The incident has often been referenced in Chinese publications unrelated to the topic [92,93]. There were considerable delays in state action, and substandard biosafety and biosecurity regulations remain unresolved [82].
## Summary of risk factors across laboratory-associated outbreaks
Thematic analysis of the outbreaks identified 19 biological and epidemiological indicators, and 14 institutional, state and social indicators (Table 2). Across the seven outbreaks, the lowest number of indicators (n = 13) was noted in the 2003 SARS escapes, while the highest were observed in the 1955 Cutter Laboratories Polio incident (n = 19) (Supplementary Appendix S2). The number of biological and epidemiological indicators across these outbreaks ranged from 9 to 14, while the institutional state, and social indicators ranged from 3 to 9. The indicators observed across the seven laboratory-associated outbreaks were evaluated to develop risk criteria for flagging possible unnatural origins. The framework consists of five primary criteria, including: unusual strain characteristics, geographical features, epidemiological factors, peculiarities in clinical manifestation and/or affected population(s) and communication to the public and predating biosafety incidents of concern (Table 3).
## Discussion
The study examined recurring epidemiological, operational and governance features that distinguish laboratory-origin outbreaks from natural ones, drawing on historical cases and applying this analytical framework to the origins of SARS-CoV-2 without implying causality. These events are rarely attributable to a single technical failure, but rather an interplay of immediate laboratory-level breaches and systemic deficiencies in governance, oversight and risk communication.
Consistent indicators emerged across the examined outbreaks. Technical failures, such as inadequate handling and transfer of inactivated pathogens and poor maintenance, underpinned the majority of outbreaks. Sudden, unexplained deaths in animal populations within an area have historically served as sentinels for the release of infectious agents and have preceded human case recognition in outbreaks of Anthrax, VEE, Brucella and FMD (Table 2). This reinforces the value of animal surveillance data for outbreak identification, as most pathogens released intentionally or unintentionally are zoonotic.
Epidemiological, spatial and geographical anomalies were recurrent across outbreaks [1], as were atypical strain features [94] (Table 2). Importantly, aside from the Anthrax escape, pathogens were often circulating prior to detection, demonstrating fragmented reporting and disease surveillance systems. Certain outbreaks lacked many biological and epidemiological indicators, but institutional/state/social factors pointed to outbreaks of laboratory origin, i.e., the 2003 SARS-CoV-1 escapes and the 2019 Brucella leak. The opposite was also observed in the 2007 Pirbright FMDV leak, the 1955 Cutter Laboratories Polio incident and the 1995 VEE epidemic, where biological and epidemiological indicators were more abundant than institutional and state factors (-Supplementary Appendix S2), suggesting that both classes of indicators should be considered when assessing laboratory origins.
While natural or unnatural origin cannot always be conclusively distinguished, such anomalies provide strong signals of possible unnatural origin [7]. The identified indicators were used to develop a framework of risk criteria for identifying such incidents (Table 3).
Using this framework, the emergence of SARS-CoV-2 exhibited several indicators warranting structured assessment. We emphasise that the presence of these indicators does not establish origin and should not be interpreted independently of virological, epidemiological, and institutional investigations. Initial WHO missions in 2020 and 2021 concluded that a laboratory origin was highly improbable. Subsequent evaluations by the WHO Scientific Advisory Group for the Origins of Novel Pathogens (SAGO) reported Note: The listed themes include: strain characteristics (yellow), geographical features (green), clinical manifestation (dark blue), epidemiological factors (blue), mis/disinformation (pink), communication to the public (red) and biosafety concerns/incidents predating the event (purple). † SARS-CoV-2 is included for comparative application of the risk-assessment framework only. Inclusion does not imply confirmation of a laboratory-associated origin.
that zoonotic spillover remains the most supported hypothesis, however a laboratory-associated incident cannot be excluded due to incomplete access to requested information [95]. The earliest recognized cluster occurred in Wuhan, which hosts two major coronavirus research facilities, including the Wuhan Institute of Virology (WIV), where research on SARS-related coronaviruses had been conducted [96,97]. Several early cases lacked exposure to the Huanan seafood market [98]. Although environmental sampling detected SARS-CoV-2 contamination at the market, including in wildlife-associated stalls [99], these findings could not distinguish between contamination arising from infected animals and introduction by infected humans. [99]. No intermediate host has been definitively identified to support zoonotic spillover despite extensive sampling. Reports suggesting pre-December 2019 circulation in several countries remain unconfirmed owing to the absence of virus-neutralisation or sequencing data [95]. Clinical anomalies were also present, in particular, unique neuropathological and cardiovascular symptoms were observed in young adults [6,100]. Moreover, high rates of presymptomatic and asymptomatic transmission were seen in SARS-CoV-2 patients in comparison to the 2002-3 SARS-CoV-1 outbreak, where asymptomatic infection was rare. SARS-CoV-2 achieved effective dissemination due to its widespread asymptomatic carriage in the population leading to undetected community spread. Genomic features attracting scientific interest include the presence of a furin cleavage site not observed in the closest known sarbecovirus [101], early markers of high virulence [102] and rapid adaptation to human transmission [103]. Although unusual, current analyses demonstrate that these features can arise through natural evolutionary mechanisms. The modified Grunow-Finke tool produced scores • The R 0 (may be higher or lower than their natural counterparts, as an attenuated vaccine strain may be released) consistent with an unintentional laboratory-related event [6]. However, these tools rely on incomplete datasets and assumptions and cannot substitute for direct virological or epidemiological evidence. Consistent with our risk criteria, recurrent themes included unconfirmed biosafety concerns, uncertainties in early transmission dynamics, marked clinical impact, genomic features of interest and multiple early clusters. Although an accidental laboratory origin of SARS CoV-2 has been suggested, no direct evidence supports this scenario, and many experts continue to view natural zoonotic spillover as the more likely pathway. A definitive resolution will require access to the missing data and continued investigation. All events detailed here were shaped by systemic weaknesses. Fragmented legislation, inadequate oversight and poor governance allowed breaches to go unreported. Unlike animal health systems, international frameworks such as the IHR lack standardized implementation of guidelines for human outbreak reporting. Risk communication failures, whether through delay, omission or disinformation, were a persistent feature undermining public trust and hindering containment efforts.
These findings demonstrate that laboratory-origin outbreaks share a recognizable epidemiological and operational fingerprint, which, paired with systemic governance insights, can strengthen outbreak surveillance. Prevention demands a shift from reliance on technical safeguards to a systems-based approach, with cooperative governance, integrated One Health surveillance, transparent data sharing and proactive risk communication. By embedding biosafety into these broader systems, accidental releases can become rare exceptions rather than recurrent events. Supplementary material. The supplementary material for this article can be found at http://doi.org/10.1017/S0950268825100915.
$$•$$
## References
1. Furmanski (2014) "Self-Fulfilling Prophecy" *Epidemics. Center for Arms Control and Nonproliferation*
2. Meselson (1994) "The Sverdlovsk anthrax outbreak of 1979" *Science*
3. Qi (2021) "Scientist in Taiwan contracts COVID-19 after P3 lab exposure, repeats SARS mistake" *Global Times*
4. Duizer (2022) "Wild poliovirus type 3 (WPV3)-shedding event following detection in environmental surveillance of poliovirus essential facilities, the Netherlands" *Eurosurveillance*
5. Chen, Chughtai, Macintyre (2020) "Application of a risk analysis tool to Middle East respiratory syndrome coronavirus (MERS-CoV) outbreak in Saudi Arabia" *Risk Analysis*
6. Chen (2024) "Use of a risk assessment tool to determine the origin of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)" *Risk Analysis*
7. Macintyre (2018) "Converging and emerging threats to health security" *Environmental Systems Decisions*
8. Berche (2025) "Laboratory-associated infections and biosafety" *La Presse Médicale*
9. Meldrum (1998) "A calculated risk": The Salk polio vaccine field trials of 1954" *British Medical Journal*
10. Offit (2007) "The Cutter Incident: How America's First Polio Vaccine Led to the Growing Vaccine Crisis"
11. Nathanson, Langmuir (1963) "The Cutter Incident. Poliomyelitis following formaldehyde-inactivated poliovirus vaccination in the United States during the spring of 1955: I. Background" *American Journal of Epidemiology*
12. Offit (2005) "The Cutter Incident, 50 years later" *New England Journal of Medicine*
13. Nathanson, Langmuir (1963) "The Cutter Incident. Poliomyelitis following formaldehyde-inactivated poliovirus vaccination in the United States during the spring of 1955: II. Relationship of poliomyelitis to cutter vaccine" *American Journal of Epidemiology*
14. Nathanson, Langmuir (1963) "The Cutter Incident. Poliomyelitis following formaldehyde-inactivated poliovirus vaccination in the United States during the spring of 1955: III. Comparison of the clinical character of vaccinated and contact cases occuring after use of high rate lots of cutter vaccine" *American Journal of Epidemiology*
15. Hall, Nathanson, Langmuir (1957) "The age distribution of poliomyelitis in the United States in 1955" *American Journal of Hygiene*
16. Thrupp (1958) "Surveillance of poliomyelitis in the United States in 1957"
17. Axelsson (2011) "The cutter incident and the development of a Swedish polio vaccine, 1952-1957" *Dynamis*
18. Norrby, Prusiner (2007) "Polio and Nobel prizes: Looking back 50 years" *Annals of Neurology*
19. Clark (1955) "Poliomyelitis: Papers and discussions presented at the third international poliomyelitis conference"
20. Day (2009) "An American tragedy'. The cutter incident and its implications for the Salk polio vaccine in New Zealand 1955-1960" *Health and History*
21. Mccormick, Da, Stratton et al. (2002) "Immunization safety review: SV40 contamination of polio vaccine and cancer"
22. Webster (1992) "Evolution and ecology of influenza a viruses" *Microbiological Reviews*
23. Zimmer, Burke (2009) "Historical perspective -Emergence of influenza a (H1N1) viruses" *New England Journal of Medicine*
24. Finkelman (2007) "Global patterns in seasonal activity of influenza a/H3N2, a/H1N1, and B from 1997 to 2005: Viral coexistence and latitudinal gradients" *PLoS One*
25. (1978) "H1N1 virus Influenza"
26. Kung (1978) "Influenza in China in 1977: Recurrence of influenza virus a subtype H1N1" *Bulletin of the World Health Organization*
27. Scholtissek (1978) "Genetic relatedness between the new 1977 epidemic strains (H1N1) of influenza and human influenza strains isolated between 1947 and 1957 (H1N1)" *Virology*
28. Oxford (1980) "Naturally occurring temperature-sensitive influenza A viruses of the H1N1 and H3N2 subtypes" *Journal of General Virology*
29. Lessler (1976) "Transmissibility of swine flu at fort dix" *Journal of the Royal Society Interface*
30. Worobey (2008) "Phylogenetic evidence against evolutionary stasis and natural abiotic reservoirs of influenza a virus" *Journal of Virology*
31. Alibek, Handelman (1999) "Biohazard: The Chilling True Story of the Largest Covert Biological Weapons Program in the World-Told from the Inside by the Man Who Ran It"
32. Riedel, York (2004) "Biological warfare and bioterrorism: A historical review" *Baylor University Medical Center Proceedings*
33. Abramova (1993) "Pathology of inhalational anthrax in 42 cases from the Sverdlovsk outbreak of 1979" *Proceedings of the National Academy of Sciences USA*
34. Wilkening (2006) "Sverdlovsk revisited: Modeling human inhalation anthrax" *Proceedings of the National Academy of Sciences USA*
35. Jackson (1998) "PCR analysis of tissue samples from the 1979 Sverdlovsk anthrax victims: The presence of multiple bacillus anthracis strains in different victims"
36. Weaver (2004) "Venezuelan equine encephalitis" *Annual Review Entomology*
37. Powers (1997) "Repeated emergence of epidemic/epizootic Venezuelan equine encephalitis from a single genotype of enzootic subtype ID virus" *Journal of Virology*
38. Weaver (1999) "Genetic evidence for the origins of Venezuelan equine encephalitis virus subtype IAB outbreaks" *American Journal of Tropical Medicine and Hygiene*
39. Kinney (1992) "Molecular evidence for the origin of the widespread Venezuelan equine encephalitis epizootic of 1969 to 1972" *Journal of General Virology*
40. Pittman (1996) "Long-term duration of detectable neutralizing antibodies after administration of live-attenuated VEE vaccine and following booster vaccination with inactivated VEE vaccine" *Vaccine*
41. Brault (2001) "Potential sources of the 1995 Venezuelan equine encephalitis subtype IC epidemic" *Journal of Virology*
42. Weaver (1996) "Re-emergence of epidemic Venezuelan equine encephalomyelitis in South America" *The Lancet*
43. Rivas (1995) "Epidemic Venezuelan equine encephalitis in La Guajira, Colombia" *Journal of Infectious Disease*
44. (1995) "Pan American Health Organization-(1995) Outbreak of Venezuelan equine encephalitis" *PAHO Epidemiological Bulletin*
45. Walder, Suarez, Calisher (1984) "Arbovirus studies in the Guajira region of Venezuela: Activities of eastern equine encephalitis and Venezuelan equine encephalitis viruses during an interepizootic period" *American Journal of Tropical Medicine and Hygiene*
46. Salas (2001) "Ecological studies of enzootic Venezuelan equine encephalitis in north-Central Venezuela, 1997-1998" *American Journal of Tropical Medicine and Hygiene*
47. Navarro (2005) "Postepizootic persistence of Venezuelan equine encephalitis virus" *Venezuela. Emerging Infectious Diseases*
48. Weaver (1994) "Evolution of alphaviruses in the eastern equine encephalomyelitis complex" *Journal of Virology*
49. Cilnis, Kang, Weaver (1996) "Genetic conservation of highlands J viruses" *Virology*
50. Brault (1999) "Genetic and antigenic diversity among eastern equine encephalitis viruses from north, central, and South America" *American Journal of Tropical Medicine and Hygiene*
51. Estrada-Franco (2004) "Venezuelan equine encephalitis virus, southern Mexico" *Emerging Infectious Diseases*
52. Suárez, Bergold (1968) "Investigations of an outbreak of Venezuelan equine encephalitis in towns of eastern Venezuela" *American Journal of Tropical Medicine and Hygiene*
53. Sudia (1972) "Epidemic Venezuelan equine encephalitis in North America in 1971: vector studies" *American Journal of Epidemiology*
54. (1995) "Update: Venezuelan equine encephalitis-Colombia" *MMWR Morbidity and Mortality Weekly Report*
55. Dickerman (1983) "Venezuelan equine encephalitis virus activity in northern Colombia during April and May" *Bulletin of the Pan American Health Organization*
56. Weaver (1999) "Genetic and fitness changes accompanying adaptation of an arbovirus to vertebrate and invertebrate cells" *Journal of Virology*
57. Weaver, Scott, Lorenz (1990) "Patterns of eastern equine encephalomyelitis virus infection in Culiseta melanura (Diptera: Culicidae)" *Journal of Medical Entomology*
58. (2003) "WHO post-outbreak biosafety guidelines for handling of SARS-CoV specimens and cultures"
59. Lim (2004) "Laboratory-acquired severe acute respiratory syndrome" *The New England Journal of Medicine*
60. (2003) "Biosafety and SARS Incident in Singapore"
61. (2003) "Severe acute respiratory syndrome in Taiwan, China"
62. (2003) "Officials: garbage bag likely cause of infection" *Sun Journal*
63. Demaneuf (2020) "The Good, the Bad and the Ugly: a review of SARS Lab Escapes"
64. Liang (2006) "Severe acute respiratory syndrome-retrospect and lessons of 2004 outbreak in China" *Biomedical and Environmental Sciences*
65. Liang (2003) "Severe acute respiratory syndrome" *Emerging Infectious Diseases*
66. (2010) "List of FMD Free Members"
67. Blacksell (2023) "The Biosafety Research Road Map: The search for evidence to support practices in the laboratory-foot and mouth disease virus" *Applied Biosafety*
68. Anderson (2008) "Foot and mouth disease 2007: a review and lessons learned. House of Commons Paper HC 312. The Stationery Office"
69. (2007) "Initial report on potential breaches of biosecurity at the Pirbright site"
70. Ryan (2008) "Clinical and laboratory investigations of the outbreaks of foot-and-mouth disease in southern England in 2007" *Veterinary Record*
71. Bessell (2009) "Spatial epidemiology of Foot and Mouth Disease in Great Britain"
72. Alexandersen (2003) "The pathogenesis and diagnosis of footand-mouth disease" *Journal of Comparative Pathology*
73. Alexandersen, Donaldson (2002) "Further studies to quantify the dose of natural aerosols of foot-and-mouth disease virus for pigs" *Epidemiology and Infection*
74. Stenfeldt (2016) "The pathogenesis of foot-and-mouth disease in pigs" *Frontiers in Veterinary Science*
75. Ethelberg (2009) "Outbreak of non-O157 Shiga toxin-producing Escherichia coli infection from consumption of beef sausage" *Clinical Infectious Diseases*
76. (2024) "Terrestrial Animal Health Code. Chapter 1.1 notification of diseases, infections and infestations, and provision of epidemiological information"
77. (2005) "Third Edition. Articles 6 & 7 (Notification) and Article 56 (Settlement of disputes)" *International Health Regulations*
78. (2023) "Terrestrial animal health code" *Animal Health Surveillance*
79. Wilson, Halabi, Lo (2020) "The international health regulations (2005), the threat of populism and the COVID-19 pandemic" *Globalization and Health*
80. Walker (2007) "Foot and mouth reports blame lab site drains. The Guardian"
81. Pappas (2022) "The Lanzhou Brucella leak: The largest laboratory accident in the history of infectious diseases?" *Clinical Infectious Diseases*
82. Pappas (2023) "Correction to: The Lanzhou Brucella Leak: The Largest Laboratory Accident in the History of Infectious Diseases?" *Clinical Infectious Diseases*
83. Ackelsberg (2015) "Brucella exposure risk events in 10 clinical laboratories" *Journal of Clinical Microbiology*
84. Liu (2022) "A systematic analysis of and recommendations for public health events involving brucellosis from 2006 to 2019 in China" *Annals of Medicine*
85. (2020) "Bulletin of the Handling of the Brucella Antibody Positive Incident of Lanzhou Institute. Lanzhou Municipal People's Government Office"
86. Robichaud, Libman, Behr (2004) "Prevention of laboratory-acquired brucellosis" *Clinical Infectious Diseases*
87. Yagupsky (2000) "Exposure of hospital personnel to Brucella melitensis and occurrence of laboratory-acquired disease in an endemic area" *Scandinavian Journal of Infectious Diseases*
88. Zhu (2020) "continual spread and expansion from northern to southern China and its relationship to worldwide lineages" *Emerging Microbes and Infections*
89. Lin (2019) "Serological prevalence survey among the high-risk populations of brucellosis-endemic areas -China" *China CDC Weekly*
90. Li (2023) "Follow-up outcomes of asymptomatic brucellosis: a systematic review and meta-analysis" *Emerging Microbes and Infections*
91. Zhang (2022) "MicroRNA-155 expression with Brucella infection in vitro and in vivo and decreased serum levels of MicroRNA-155 in patients with brucellosis" *Scientific Reports*
92. Chen (2021) "Key issues affecting the current status of infectious diseases in Chinese cattle farms and their control through vaccination" *Vaccine*
93. Treadwell (2003) "Epidemiologic clues to bioterrorism" *Public Health Reports*
94. (2025) "Independent Assessment of the Origins of SARS-CoV-2 from the Scientific Advisory Group for the Origins of Novel Pathogens (SAGO)"
95. Ge (2013) "Isolation and characterization of a bat SARS-like coronavirus that uses the ACE2 receptor" *Nature*
96. Latinne (2020) "Origin and cross-species transmission of bat coronaviruses in China" *Nature Communications*
97. Li (2020) "Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia" *New England Journal of Medicine*
98. Worobey (2022) "The Huanan seafood wholesale market in Wuhan was the early epicenter of the COVID-19 pandemic" *Science*
99. Akhtar (2023) "The impact of COVID-19 and COVID vaccination on cardiovascular outcomes" *European Heart Journal Supplements*
100. Coutard (2020) "The spike glycoprotein of the new coronavirus 2019-nCoV contains a furin-like cleavage site absent in CoV of the same clade" *Antiviral Research*
101. Mclean (2022) "The impact of evolving SARS-CoV-2 mutations and variants on COVID-19 vaccines" *MBio*
102. Piplani (2021) "In silico comparison of SARS-CoV-2 spike protein-ACE2 binding affinities across species and implications for virus origin" *Scientific Reports* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548409&blobtype=pdf | # Adjuvant-dependent protection of SARS-CoV-2 spike vaccines: comparative immunogenicity of human-applicable formulations
Zhendong Pan, Liangliang Jiang, Yingying Chen, Haoran Peng, Yangang Liu, Xu Zheng, Yanhua He, Ying Wang, Xiaoyan Zhang, Zhongtian Qi, Cuiling Ding, Jianqing Xu, Ping Zhao
## Abstract
Although recombinant vaccines with various adjuvant systems are widely deployed in global coronavirus disease 2019 immunization, their distinct immune profiles have not been fully elucidated. In this study, we evaluated immune responses induced by four clinically validated adjuvants-aluminum hydroxide (Al), two water-inoil emulsions (Montanide ISA720 and ISA51), and an oil-in-water emulsion (Sepivac SWE) -formulated with the prefusion-stabilized ancestral severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike trimer (S-2P) in mouse models. Both S-2P:ISA720 and S-2P:ISA51 elicited potent humoral immunity, including cross-neutralizing antibod ies against Omicron variants, along with robust interferon gamma-producing T-cell responses in the spleen and lung. Immunization with S-2P:ISA720 conferred complete protection against lethal challenge with the ancestral virus as well as Omicron sub variants BA.5 and BF.7. In contrast, S-2P:Al and S-2P:SWE elicited substantially weaker antibody responses, undetectable T-cell immunity, and markedly reduced protective efficacy. Prime-boost immunization with S-2P:ISA720 induced sustained peak antibody titers against both homologous SARS-CoV-2 and the Omicron BA.2 subvariant, afford ing complete protection for up to 17 weeks post-vaccination. While a third dose of S-2P:Al following a prime-boost regimen triggered a robust increase in antibody levels, titers declined rapidly thereafter, recapitulating the decay kinetics observed after the second dose. Both ISA720 and Al formulations exhibited age-dependent declines in antibody responses and protective efficacy. Notably, aged mice displayed markedly attenuated neuroinflammatory responses following SARS-CoV-2 challenge and significantly compromised protection after adoptive transfer of immune serum. These results underscore adjuvantspecific determinants of broad and durable immunity and reveal novel age-related constraints in vaccine-induced protection against SARS-CoV-2.IMPORTANCE Persistent viral evolution, rapid waning of vaccine-induced immunity, and the heightened vulnerability of elderly populations remain major challenges for COVID-19 vaccination strategies. In this study, we systematically assessed immune responses elicited by the ancestral spike protein formulated with four distinct adju vants in mouse models. We demonstrate that an optimized adjuvant formulation markedly enhances the magnitude and breadth of antibody responses, potentiates T-cell immunity, and rapidly induces sustained peak antibody titers against both homologous virus and Omicron variants. Vaccine-induced antibody responses were significantly attenuated in aged mice, and furthermore, both the protective efficacy of antibodies and inflammatory cytokine responses upon viral challenge were impaired in aged animals. These results provide compelling evidence that rational adjuvant selection is critical for enabling recombinant vaccines to achieve rapid-onset, broad, and durable immune protection. Furthermore, our study offers new mechanistic insights into the reduced vaccine efficacy observed in the elderly.
KEYWORDS SARS-CoV-2, spike protein, adjuvant, cross-protection, booster immuniza tion, immunosenescence T he coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has led to great damage to human health and the world's global social and economic development and still remains a threat to public health worldwide. Since the beginning of 2025, a resurgence of SARS-CoV-2 infections has occurred in numerous locations worldwide. Vaccination is the most effective strategy for controlling infectious diseases. The SARS-CoV-2 spike protein enables viral cell entry by binding its S1 subunit's receptor-binding domain (RBD) to the human cellular receptor angiotensin-converting enzyme 2 (ACE2) (1,2). Thus, the spike protein and its RBD function as principal targets for neutralizing antibodies and serve as critical vaccine antigens (3)(4)(5). Several vaccines were rapidly developed and approved for mass vaccination worldwide. These vaccines have shown efficacy against SARS-CoV-2 infection, particularly in reducing severe disease, hospitalizations, and mortality (6)(7)(8)(9)(10). However, immune-evading variants that largely occurred in the viral spike protein continue to emerge. Notably, since late 2021, Omicron and its subvariants have shown pronounced antibody evasion, posing a significant challenge to worldwide vaccination campaigns (11)(12)(13). Furthermore, the neutralizing antibody elicited by natural infection and vaccination generally rapidly decreases over time (14)(15)(16). Older people are particularly susceptible to SARS-CoV-2 infection and develop severe disease due to the impaired adaptive immunity to vaccination, dysregulated inflammatory response, and age-related chronic diseases (17)(18)(19). Multiple strategies have been adopted to combat the challenges of rapidly evolving viral variants and waning immune protection, including booster vaccinations, antigen updates in vaccine formulations, and the development of multivalent antigen-based vaccines (13,(20)(21)(22).
Currently, three distinct vaccine platform technologies-inactivated virus vaccine (6), mRNA-based lipid nanoparticle vaccine (7,8), replication-incompetent adenovi rus vector vaccine (9), and adjuvanted recombinant protein vaccines (10)-are com mercially available for COVID-19 prevention. Among these, adjuvanted recombinant protein vaccines exhibit unique advantages, including superior safety profiles, enhanced stability, greater costeffectiveness, and improved manufacturing scalability. These benefits position recombinant protein vaccines as a particularly valuable option for global immunization efforts.
Adjuvants serve as crucial components of recombinant vaccines, orchestrating both antigen presentation to immune cells and delivery of immunostimulatory signals required for effective immune priming (23). Adjuvants profoundly influence the magnitude, breadth, and quality of immune responses to vaccine antigens. Although diverse adjuvant systems, from conventional aluminum salts to newer formulations like the oil-in-water (O/W) emulsion AS03 and saponin-based Matrix-M nanoparticles, have been employed in recombinant COVID-19 vaccines (10,(24)(25)(26), the respective unique immune characteristics induced by these distinct adjuvants remain to be fully elucidated.
Aluminum salt adjuvants (principally aluminum hydroxide and aluminum phosphate) are the most widely used adjuvants in human vaccines. They function through multiple mechanisms, including the formation of an antigen depot enabling sustained release, the induction of local inflammatory responses, and the activation of innate immune pathways (23,27). While highly effective at enhancing antibody-mediated immunity, these adjuvants predominantly elicit Th2-skewed immune responses with weak cellular immunity induction. This Th2 bias raises potential safety considerations for respira tory virus vaccines, as it may potentially contribute to vaccine-associated enhanced respiratory disease (VAERD) in certain cases (28).
Water-in-oil (W/O) emulsion adjuvants demonstrate superior immunogenicity compared to aluminum salt adjuvants, particularly in eliciting T-cell-mediated immunity, and their application in human vaccines has been steadily increasing (23,29). Among these, Seppic's Montanide ISA 720 and ISA 51 represent well-established W/O emulsion platforms that have undergone extensive clinical evaluation over several decades for various vaccine candidates against infectious diseases and cancers (29,30). These adjuvants have proven effective in enhancing both antigenspecific antibody produc tion and cytotoxic T-lymphocyte responses. The immunostimulatory properties of these adjuvants primarily involve antigen depot formation, localized inflammatory responses, and lymphocyte trapping at the injection site. While both are W/O emulsions, they differ substantially in composition. ISA 720 employs metabolizable squalene as its oil phase, offering excellent biocompatibility, whereas ISA 51 uses non-metabolizable mineral oil that typically provokes stronger local inflammation, a characteristic that may contribute to its enhanced capacity to stimulate T-cell-mediated immunity. Sepivac SWE represents a next-generation O/W emulsion platform that is similar to another O/W adjuvant MF59 (31). All three adjuvants (ISA 720, ISA 51, and SWE) have demonstrated favorable safety profiles and immunogenicity in I/II clinical trials of SARS-CoV-2 vaccine candidates (32)(33)(34)(35).
In this study, we evaluated aluminum hydroxide and Seppic's Montanide ISA 720, ISA 51, and Sepivac SWE-combined with the prefusion-stabilized ancestral SARS-CoV-2 spike protein trimer (S-2P) in wild-type C57BL/6 mice and congenic humanized ACE2 transgenic mice. Our assessment included multiple protective immunity parameters: (i) magnitude and breadth of immune responses, (ii) durability of immunity, (iii) protective efficacy in aged mice, and (iv) effects of booster immunization.
## MATERIALS AND METHODS
## Cells, viruses, and proteins
HEK293T and HEK293 cells were obtained from the American Type Culture Collection (ATCC). The HEK293 cells stably expressing human ACE2 (HEK293-ACE2) were generated by infecting HEK293 cells with a human ACE2 lentivirus, followed by puromycin selection. Vero E6 cells were kindly provided by Dr. Rong Zhang of Fudan University. HEK293T, HEK293-ACE2, and Vero E6 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA), 1% penicillin-streptomycin (Thermo Fisher Scientific), 1% L-glutamine (Thermo Fisher Scientific), and 1% non-essential amino acids (Thermo Fisher Scientific) at 37°C with 5% CO₂. FreeStyle CHO-S suspension cells (Invitrogen, USA) were cultured in EX-CELL 325 PF CHO Serum-Free Medium (Sigma-Aldrich) supplemented with 1% L-glutamine and 1% penicillin-streptomycin. Cells were maintained in nonpyrogenic, vented polycarbonate Erlenmeyer flasks (Thermo Fisher Scientific) at 37°C with 5% CO₂ and constant shaking at 130 rpm in an incubator shaker. All SARS-CoV-2 viruses were isolated from naso pharyngeal swab samples collected from nucleic acid-positive cases and subsequently propagated and titrated in Vero E6 cells as previously described (36). The viral strains included the following: SARS-CoV-2 ancestral strain (GenBank: MT622319.1), Omicron BA.2 (GenBank: MT627325.1), Omicron BA.5 (GenBank: PQ213095.1), and Omicron BF.7 (GenBank: PQ351184.1).
The SARS-CoV-2 S1 protein, S2 protein, and RBD protein corresponding to the ancestral, Delta, and Omicron variants were obtained from Sino Biological (Beijing, China).
## Spike protein expression
The mammalian cell codon-optimized ectodomain of ancestral SARS-CoV-2 spike trimer cDNA was synthesized by Generay Biotech Co., Ltd (Shanghai, China), in which a trimeric foldon from T4 phage fibritin (Tfd) was fused to the C-terminus of SARS-CoV-2 spike ectodomain (residues 1-1,211; GenBank: NC_045512.2), with the furin cleavage site RRAR replaced by GSAS and two proline mutations (K986P/V987P) introduced to stabilize the prefusion conformation, along with a C-terminal polyhistidine tag for purification (1). This cDNA was cloned into a modified pCI-GS CHO expression vector and electroporated into CHO cells, followed by selection of stable spike trimer-expressing clones using glutaminedeficient medium containing 25 µM methionine sulfoximine (Sigma). A selected clone was gradually scaled up to 200 mL culture volume, after which supernatants were harvested and purified sequentially through ultrafiltration, nickel affinity chromatography, and desalting. The purified recombinant protein (designated S-2P) was quantified by BCA assay (Thermo Fisher Scientific), with purity verified by Coomassie blue staining and S1domainspecific western blotting under both denatur ing and non-denaturing conditions.
## Assay of S-2P binding to human ACE2
The binding of S-2P to human ACE2 was assessed by enzyme-linked immunosorbent assay (ELISA). Briefly, S-2P protein (2 µg/mL) was coated onto high-binding 96-well plates (Thermo Fisher Scientific) and incubated overnight at 4°C. After washing with PBST (PBS containing 1% Tween-20), plates were blocked with blocking buffer (PBST containing 1% BSA) for 2 hours at room temperature. Following another PBST wash, serially diluted human ACE2 protein (mFc Tag; Sino Biological, Beijing, China) was added to the plates and incubated for 2 hours at room temperature. After washing, HRP-conjugated goat anti-mouse IgG1 cross-adsorbed secondary antibody (Invitrogen; 1:1,000 dilution) was added and incubated for 2 hours at room temperature. Following final PBST washes, the reaction was developed using 1-Step Slow TMB solution (Thermo Fisher Scientific) and stopped with ELISA Stop Solution (Solarbio, China). Absorbance was measured at 450 nm.
## Animals and ethics statement
C57BL/6 mice were obtained from Shanghai Jihui Laboratory Animal Care Co., Ltd. (Shanghai, China), and CAG-hACE2 transgenic mice on the C57BL/6 background were acquired from Shanghai Model Organisms Center, Inc. (Shanghai, China). Unless otherwise specified, the experimental mice were 6-8 weeks of age. For the aged group, 18-month-old mice were used, as the standard laboratory mouse has an average lifespan of about 2 years, with 18-24 months generally considered old age (37). All animals were housed in individually ventilated cages under specific pathogen-free conditions. All animal experiments were reviewed and approved by the Institutional Committee on Ethics of Medicine of Navy Medical University and conducted in accordance with China's Regulations for the Administration of Affairs Concerning Experimental Animals.
## Mouse immunization
The recombinant S-2P protein was formulated with alum (Croda, Denmark), Montanide ISA 720 VG, Montanide ISA 51 VG, or Sepivac SWE (all from Seppic, France), designated as S-2P:Al, S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively. Each mouse was intramusc ularly injected with 100 µL of vaccine formulation containing 10 µg of S-2P protein, with adjuvant volume ratios of 70%, 50%, and 50% for S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively, while S-2P:Al contained 100 µg of aluminum hydroxide. Unless otherwise specified, C57BL/6J or hACE2 mice were immunized twice at 3 week intervals with S-2P:Al, S-2P:ISA720, S-2P:SWE, or S-2P:ISA51, while control animals received PBS alone.
## Serum antibody assay
Mouse serum antibodies against SARS-CoV-2 were evaluated by ELISA as described (38). Briefly, recombinant SARS-CoV-2 S1, S2, or RBD protein (1 µg/mL) was coated onto high-binding 96-well plates overnight at 4°C, followed by blocking with blocking buffer. Serially diluted serum samples were then added to the plates and incubated at room temperature for 2 hours. After extensive washing, HRP-conjugated secondary antibodies (goat anti-mouse IgG, goat anti-mouse IgG1, or goat anti-mouse IgG2b; Thermo Fisher Scientific) were added. Endpoint titers were determined as the reciprocal of the highest dilution showing an absorbance at 450 nm ≥2.1fold higher than control mouse sera.
For the blocking ELISA assessing antibody inhibition of RBD-hACE2 binding, RBDcoated (1 µg/mL) 96-well plates were first incubated with serially diluted serum samples for 2 hours at room temperature before adding hACE2 protein (0.25 µg/mL). After washing, rabbit anti-hACE2 antibody (Sino Biological) was added for 2 hours at room temperature, followed by HRP-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific). The remaining steps were performed as described above.
## Serum neutralization of authentic SARS-CoV-2
Vero E6 cells were seeded in 96-well plates (1 × 10⁴ Vero E6 cells per well) and cultured overnight. Serially diluted serum samples were mixed with either 0.1 MOI (multiplicity of infection) of SARS-CoV-2 ancestral strain or 0.01 MOI of Omicron BA.2 strain at 37°C for 30 minutes, then the virus-serum mixtures were added to Vero E6 cells in the presence of 2 µg/mL TPCK-treated trypsin. At 24 hours post-infection, cells were fixed with methanol at -20°C for 30 minutes and blocked with 3% BSA. Rabbit polyclonal antibodies against SARS-CoV-2 nucleocapsid protein (NP; Sino Biological, Beijing, China) were added and incubated overnight at 4°C, followed by PBS washing and incubation with Alexa Fluor 488-conjugated goat anti-rabbit IgG (Thermo Fisher Scientific) at room temperature for 1.5 hours. After DAPI staining (Sigma-Aldrich), cells were imaged using a Cytation 5 system (BioTek, USA). Infected cell numbers were quantified using Gen5 3.10 software, with neutralization percentage calculated relative to virus-only controls. The 50% inhibitory concentration (IC₅₀) was determined as the reciprocal serum dilution showing 50% infection inhibition and calculated using GraphPad Prism 8.0 (GraphPad Software, Inc.).
## Serum neutralization of pseudotyped virus
HEK293T cells were co-transfected using Lipofectamine 2000 reagent (Thermo Fisher Scientific) with HIV capsid packaging plasmids, a transfer plasmid harboring the EGFP reporter gene, and a plasmid harboring the spike gene of SARS-CoV-2 or SARS-CoV. In the SARS-CoV-2 spike expression plasmid, the fragment encoding a 19-amino-acid at the C-terminal of spike was truncated. Following 48 hours of incubation, pseudovirus-containing supernatants were harvested, centrifuged at 1,000 × g for 10 minutes, and stored at -80°C until use. For neutralization assays, serially diluted samples of the test sera were pre-incubated with pseudoviruses at 37°C with 5% CO₂ for 30 minutes, then the mixtures were added to 293T-ACE2 target cells (seeded at 1 × 10⁴ cells/well). After 4 hours, the medium was replaced with fresh DMEM supplemented with 2% FBS, and EGFP-positive cells were quantified at 48 hours post-infection using a Cytation 5 cell imaging multimode reader (BioTek). Neutralization titers (ID50) were calculated using nonlinear regression analysis in GraphPad Prism 8.0 (GraphPad Software).
## Enzyme-linked immunosorbent spot assay
T-cell-mediated immune responses in mice were evaluated using an interferon gamma (IFN-γ) ELISPOT kit (Dakewe, China) as described (38). Spleen and lung cells were isolated 14 days post-secondary immunization. Briefly, IFN-γ monoclonal antibody-precoated ELISpot plates were blocked with 200 µL/well of serum-free ELISPOT medium (Dakewe, China) for 5-10 minutes at room temperature. Spleen or lung cells (2 × 10⁵/well) were stimulated with ancestral SARS-CoV-2 S1 or S2 protein at 400 µg/mL for 20 hours at 37°C with 5% CO₂. IFN-γ-producing cells were detected using biotinylated detection antibody and streptavidin-HRP, followed by spot visualization with an S6 Ultra Immu noSpot Reader (Cellular Technology Ltd.) and quantification using ImmunoSpot 5.1.36 software (Cellular Technology Ltd.). Spot-forming cells were enumerated as a measure of antigenspecific T-cell responses.
## Challenge with SARS-CoV-2
Under isoflurane anesthesia, hACE2 transgenic mice were intranasally inoculated with 50 µL of viral suspension containing SARS-CoV-2 ancestral strain (1 × 10² PFU), Omicron BA.5 (1 × 10 4 PFU), or Omicron BF.7 (1 × 10 4 PFU) as described (38). Mice were monitored daily for body weight changes and survival rates. At 4 days post-infection (dpi), a subset of mice was humanely euthanized for tissue collection (nasal turbinates, lungs, and brains) to assess viral load and inflammatory factors, while the remaining mice were maintained for longitudinal survival analysis.
## Measurement of viral load and inflammatory cytokines
Viral RNA levels and inflammatory cytokine gene expression were analyzed by quan titative reverse transcription PCR (qRT-PCR). Tissue samples (brain, lung, and nasal turbinates) from mice were homogenized in TRIzol Reagent (Invitrogen) for total RNA extraction according to the manufacturer's protocol. First-strand cDNA synthesis was performed using PrimeScript RT Master Mix (TaKaRa, Japan), followed by quantitative PCR amplification with TB Green Premix Ex Taq II (TaKaRa, Japan) and the following genespecific primers (5′→3′): SARS-CoV-2 nucleocapsid (N) gene (forward: AAGGCGTTC CAATTAACACCA, reverse: TGCCGTCTTTGTTAGCACCA); mouse β-actin (forward: GGCTGTA TTCCCCTCCATCG, reverse: GCACAGGGTGCTCCTCAG); IL-6 (forward: TCGGAGGCTTAATTA CACA, reverse: TCATACAATCAGAATTGCCAT); CCL2 (forward: TCGGAACCAAATGAGATCAG A, reverse: TAGCTTCAGATTTACGGGTCA); and CXCL10 (forward: AATTTAATGAAAGCGTTTA GCC, reverse: ATTAGGACTAGCCATCCAC). All data were normalized to β-actin expression levels.
## Serum adoptive transfer
Passive immunization was performed through adoptive transfer of immune sera to evaluate protective efficacy against SARS-CoV-2 challenge in recipient hACE2-transgenic mice, as previously described (39). Briefly, hACE2 transgenic mice were intravenously administered 0.1 mL of pooled sera collected 14 days post-booster immunization (from vaccinated mice) or PBS-treated control mice. Twenty-four hours post-serum transfer, mice were intranasally challenged with live SARS-CoV-2 ancestral strain. Body weights were monitored daily until day 4 post-infection, when mice were humanely euthanized for virological and immunological analyses. Viral RNA loads and inflammatory factor expression profiles in brain tissues were quantified by qRT-PCR.
## Statistical analysis
Statistical analyses were conducted using GraphPad Prism 8.0 (GraphPad Software). Between-group comparisons were performed using either unpaired or paired two-tailed Student's t-tests as appropriate, while multiple group comparisons were analyzed by one-way ANOVA followed by Tukey's post hoc test. Statistical significance was defined as P < 0.05, with the following notation: *P < 0.05, **P < 0.01, and ***P < 0.001; ns (not significant) for P ≥ 0.05.
## RESULTS
## Immune responses to ancestral SARS-CoV-2 spike trimers formulated with different human-compatible adjuvants in C57BL/6 mice
The prefusion-stabilized spike trimer of ancestral SARS-CoV-2 was expressed in CHO cells and purified via affinity chromatography. The purified protein, designated S-2P, was confirmed by native and denaturing SDS-PAGE and western blot using a mono clonal antibody (mAb) targeting the S1 subunit (Fig. S1A). ELISA analysis demonstra ted dose-dependent binding of soluble human ACE2 to immobilized S-2P (Fig. S1B). Furthermore, S-2P effectively inhibited SARS-CoV-2 infection in Vero-E6 cells (Fig. S1C).
Aluminum hydroxide (Croda, Denmark) and Seppic's Montanide ISA 720, ISA 51, and Sepivac SWE (Seppic, France) were evaluated for their ability to enhance immune responses to recombinant S-2P protein in C57BL/6 mice. Mice were immunized intramuscularly twice at a 3-week interval with 10 µg S-2P protein formulated with either aluminum hydroxide or Seppic's adjuvants (denoted as S-2P:Al, S-2P:ISA720, S-2P:ISA51, and S-2P:SWE, respectively). Serum samples were collected 2 weeks after each immu nization and analyzed for antigenspecific IgG and neutralizing antibodies (Fig. 1A). Comparative analysis showed significant differences in immunogenicity depending on the adjuvant used. Both S-2P:ISA720 and S-2P:ISA51 induced stronger antibody responses compared to S-2P:Al or S-2P:SWE. Sera from S-2P:ISA720-immunized mice showed the highest levels of S1-and S2specific IgG (Fig. 1B andC), the strongest inhibition of RBD binding to hACE2 (Fig. 1D) and the most potent neutralizing activity against the ancestral SARS-CoV-2 (4.3-, 5.1-and 2.2-fold higher than S-2P:Al, S-2P:SWE and S-2P:ISA51 groups at week 5, respectively) (Fig. 1E).
To assess antigenspecific T-cell responses, immunized mice were sacrificed 3 weeks after the booster dose. Single-cell suspensions were prepared from both spleens and lungs, then stimulated ex vivo with ancestral S1 or S2 proteins. IFN-γ-producing cells were quantified using enzyme-linked immunospot (ELISpot) assay. The results showed that mice immunized with S-2P:ISA720 or S-2P:ISA51 exhibited significantly stronger antigen specific T-cell responses in both splenic and pulmonary compartments, whereas S-2P:Al and S-2P:SWE failed to induce detectable antigenspecific IFN-γ-producing T cells (Fig. 1E and F; Fig. S1D andE).
## Adjuvant-dependent protection of S-2P vaccines against lethal SARS-CoV-2 challenge in humanized ACE2 mouse model
The CAG-hACE2 transgenic mouse represents a valuable small animal model for evaluat ing candidate vaccine efficacy. Although SARS-CoV-2 infection affects both upper and lower respiratory tracts and induces tissue damage, viral invasion into the central nervous system emerges as the primary cause of mortality in this model (40). We immunized CAG-hACE2 transgenic mice (C57BL/6 background) twice at 3-week intervals with 10 µg S-2P protein formulated with different adjuvants (14 mice/group), along with an additional S-2P:ISA720 single-dose group (10 mice) (Fig. 2A). Consistent with observa tions in wild-type C57BL/6 mice, all formulations efficiently induced anti-RBD IgG (Fig. 2B andC) and virus-neutralizing antibodies in hACE2 mice (Fig. 2D andE), with S-2P:ISA720 showing superior immunogenicity, exhibiting 10.8-, 7.8-, and 3.9-fold higher neutralizing titers versus S-2P:Al, S-2P:SWE, and S-2P:ISA51, respectively at week 5. Remarkably, single-dose S-2P:ISA720 outperformed two-dose S-2P:Al and S-2P:SWE in neutralizing antibody induction, while S-2P:Al showed consistently weaker responses than ISA51-or ISA720-formulated vaccines (Fig. 2C andE). Moreover, S-2P:ISA51 and S-2P:ISA720 induced Th1-skewed immunity (evidenced by elevated IgG2b:IgG1 ratios; Fig. S2A andB) that correlated with a robust IFN-γ-producing T-cell response observed in wild-type C57BL/6 mice.
Three weeks after booster immunization, mice were intranasally challenged with 100 PFU of ancestral SARS-CoV-2. Unvaccinated control mice exhibited rapid disease progression, with severe weight loss and 100% mortality within 5 days post-challenge (Fig. 2F andG). All vaccinated groups showed varying degrees of protection: the S-2P:ISA720 group demonstrated complete protection (100% survival, 14/14), while S-2P:ISA51 showed 85.7% survival (12/14) (Fig. 2F andG). Comparatively, S-2P:SWE and S-2P:Al groups exhibited relatively lower protection rates of 42.9% (6/14) and 35.7% (5/14), respectively. Notably, even a single dose of S-2P:ISA720 conferred 90% protection (9/10 survivors), highlighting its superior efficacy.
To elucidate the protective role of vaccine-induced neutralizing antibodies, we analyzed the correlation between humoral immune responses and survival rates in vaccinated mice. The S-2P:Al and S-2P:ISA720 groups demonstrated strong correlations between anti-RBD IgG and neutralizing antibody titers, while the S-2P:SWE and S-2P:ISA51 group showed a relatively weaker correlation (Fig. S2C through G), suggesting adjuvants critically influence both the magnitude and functional quality of antibody responses. Protection analysis revealed that neutralizing antibody titers ≧3 × 10 3 conferred complete protection (5/5) in S-2P:Al-immunized mice and 71% protection (5/7) in S-2P:SWE-immunized mice (Fig. S2C through G). Since neither S-2P:Al nor S-2P:SWE immunization induced detectable T-cell response, the observed protection can be primarily attributed to neutralizing antibodies. We further confirmed the protective efficacy of vaccine-induced antibodies through passive immunization in naïve hACE2 mice. The mice (n = 10 per group) were adminis tered 100 µL of pooled sera from immunized or control mice 1 day prior to SARS-CoV-2 challenge. Five mice per group were sacrificed at 4 days post-infection for analysis of brain viral loads and cytokine profiles, while the remaining five mice were monitored for clinical outcomes. All mice that received control sera or sera from S-2P:Al-immunized mice exhibited significant weight loss and mortality. In contrast, Seppic-adjuvanted vaccine immune sera conferred partial protection, with 1, 3, and 3 mice receiving sera from mice immunized with SWE-, ISA51-, and ISA720-adjuvanted vaccine surviv ing, respectively (Fig. 2H). The observed reduction in viral titers and proinflammatory cytokine levels in murine brain tissue demonstrates that vaccine-induced antibodies effectively inhibit SARS-CoV-2 neuroinvasion and mitigate associated neuropathological damage (Fig. 2I andJ).
## ISA720-adjuvanted ancestral S-2P conferred complete protection against challenge with Omicron variants
Pseudovirus-based neutralization assays were employed to quantify the cross-neutraliz ing antibody induced by various adjuvanted S-2P vaccines against SARS-CoV-2 variants of concern (VOCs) in murine models. While Delta variant neutralization showed modest (0.8-to 1.3-fold) titer reduction versus ancestral strain, Omicron BA.1 neutralization was substantially impaired across sera of all groups (Fig. S3A). The majority of sera from S-2P:Al-and S-2P:SWE-immunized mice showed no detectable neutralization against BA.1 at 100-fold dilution, while 84.6% (11/13) of S-2P:ISA51 and 100% (11/11) of S-2P:ISA720 immune sera maintained neutralizing activity, despite a 17-fold decrease in geometric mean neutralization titers compared to neutralization against ancestral strain for the S-2P:ISA720 group. Importantly, both ISA-adjuvanted vaccine groups demonstra ted cross-neutralization capacity against SARS-CoV-1 pseudovirus at low serum dilutions (Fig. S3B).
Next, hACE2-transgenic mice were immunized with either S-2P:Al or S-2P:ISA720 to evaluate the in vivo protective efficacy against Omicron variants BA.5 and BF.7. Consistent with the findings in Results 1 and 2, immunization with S-2P:ISA720 elicited significantly higher titers of anti-RBD antibody and neutralizing antibody against the ancestral strain compared to S-2P:Al (Fig. 3A andB). Notably, in authentic virus neutraliza tion assays, sera from the majority of S-2P:ISA720-immunized mice potently neutralized Omicron subvariants BA.2, BA.5, and BF.7. In contrast, only minimal neutralizing activity was detected in sera from S-2P:Al-immunized mice, with significantly lower titers (Fig. 3B). Following viral challenge, all of the S-2P:Al-immunized mice exhibited progressive weight loss and succumbed (Fig. 3C andD; Fig. S3C andD). Conversely, all of the S-2P:ISA720-immunized mice maintained normal body weight and survived (Fig. 3C andD; Fig. S3C andD). Quantitative analysis demonstrated that S-2P:ISA720 vaccination significantly reduced viral loads in lungs, brains, and nasal turbinates (Fig. 3E andF). Correspondingly, cytokine mRNA profiling in brain tissues revealed marked inflammatory responses in PBS-control and S-2P:Al groups, but minimal response in S-2P:ISA720-immu nized mice (Fig. 3G andH).
In summary, ISA720-adjuvanted S-2P not only broadened the neutralization capacity against diverse SARS-CoV-2 variants but, more significantly, conferred robust crossprotection against antigenically distinct VOCs. (G and H) Brain inflammatory cytokines mRNA levels assessed by RT-qPCR assay. Data are represented as the mean ± SEM.
Groups were compared using one-way ANOVA with Tukey's multiple comparisons test.
## Sustained ceiling-level antibody responses and durable protection following ISA720-adjuvanted S-2P prime-boost vaccination
Booster immunization, whether homologous or heterologous, is a widely adopted strategy to enhance vaccine efficacy against rapidly evolving SARS-CoV-2 variants (20).
To assess the impact of booster doses on antibody responses elicited by differently adjuvanted spike proteins, we immunized hACE2 mice three times (weeks 0, 3, and 11) with either S-2P:Al or S-2P:ISA720. Serum samples were collected at 2 and 8 weeks after the second and third immunizations (weeks 5, 11, 13, and 19), enabling a longitudinal assessment of antibody response kinetics (Fig. 4A). Consistent with Results 1 and 2, S-2P:ISA720 induced significantly higher RBDspecific IgG and neutralizing antibody titers against the ancestral strain than S-2P:Al at all indicated timepoints (Fig. 4B). After the second and third immunizations, anti-RBD IgG antibody titers in S-2P:Al-immunized mice showed faster decay kinetics compared to the S-2P:ISA720 group, indicating superior durability of the S-2P:ISA720-induced humoral response. Notably, while the S-2P:Al group showed a significant increase in IgG titers after the third dose, the S-2P:ISA720 group reached peak IgG levels after two doses, with no further increase after the third dose (Fig. 4B). Similarly, the third dose of S-2P:Al markedly enhanced neutralizing antibody titers against the ancestral strain, whereas the third dose S-2P:ISA720 showed no significant improvement in serum neutralization postbooster (Fig. 4C). These results suggest that antibody responses to S-2P:ISA720 vaccina tion reached a ceiling effect after prime-boost immunization. Moreover, the S-2P:Al group showed a modest (1.6-fold) increase in IgG antibodies against the Omicron BA.2 RBD after the third dose, while S-2P:ISA720 failed to enhance RBDspecific IgG titers postbooster (Fig. 4D). Neutralization of Omicron BA.2 improved slightly with the third S-2P:Al dose, with the number of sera exhibiting neutralizing titers > 100 increasing from 2 to 3. In contrast, sera from S-2P:ISA720-immunized mice displayed stronger neutralization against Omicron BA.2 after the second dose, but the third dose provided no additional benefit (Fig. 4E).
To evaluate long-term immunity, we compared antibody persistence and protective efficacy between the S-2P:Al and S-2P:ISA720 in hACE2 mice immunized at weeks 0 and 3. The S-2P:Al group exhibited a 6.3-fold decline in anti-RBD IgG titers between weeks 5 and 20, compared to only a 2.4-fold decline in the S-2P:ISA720 group (Fig. S4A). Even worse, the S-2P:Al group showed a more than 7.6-fold decline in neutralizing titers, whereas the S-2P:ISA720 group exhibited only a 2.7-fold decline (Fig. S4B). Following the challenge at week 21 with ancestral SARS-CoV-2, all S-2P:Al-immunized mice ultimately succumbed (albeit with delayed mortality versus PBS controls), whereas the S-2P:ISA720 group achieved complete protection (Fig. 4G andH).
## ISA720-adjuvanted S-2P provided superior protective efficacy compared to Alum-adjuvanted S-2P in aged mice
The elderly population exhibits heightened susceptibility to severe COVID-19 outcomes (17,18). To evaluate vaccine protection in aging populations, we compared immune responses in young (3-month-old) and aged (18-month-old) hACE2 mice immunized with either S-2P:Al or S-2P:ISA720. Serum antibody responses were measured at 2 weeks after prime and booster vaccination. Following prime immunization, aged mice exhibited significantly lower anti-RBD IgG levels compared to young adults (1.7-fold lower for S-2P:Al and 1.9-fold lower for S-2P:ISA720). This age-dependent reduction persisted after booster immunization, with aged mice showing 2.1-fold (S-2P:Al) and 1.8-fold (S-2P:ISA720) lower antibody levels than their younger counterparts (Fig. 5A). Neutraliz ing antibody levels after the booster immunization exhibited a similar pattern, being 3.6fold lower in aged S-2P:Al recipients and 2.1-fold lower in S-2P:ISA720 recipients compared to young adults (Fig. 5B). Consistent with observations in young mice, S-2P:ISA720 similarly elicited a Th1-skewed immune response in aged animals, as demonstrated by significantly elevated IgG2b:IgG1 ratios (Fig. 5A andB). Following SARS-CoV-2 challenge, control and S-2P:Al-immunized aged mice showed rapid weight loss and 100% mortality. In contrast, aged mice immunized with S-2P:ISA720 demonstrated prolonged survival and a 40% survival rate (Fig. 5C andD). These findings demonstrate that while both vaccines displayed substantially reduced efficacy in aged mice, S-2P:ISA720 provided significantly better protection compared with S-2P:Al. Full-Length Text Interestingly, although some aged mice immunized with S-2P:ISA720 developed higher anti-RBD IgG titers than young survivors receiving S-2P:Al immunization, they nevertheless ultimately succumbed to infection (Fig. 5E). To investigate this apparent disconnect between antibody levels and protection in aged animals, we performed passive immunization experiments. Naïve young and aged hACE2 mice received 100 µL of pooled sera from PBS-, S-2P:Al-, or S-2P:ISA720-immunized young donors prior to SARS-CoV-2 challenge. The mice were sacrificed at 4 days post-infection for analysis of brain viral loads and cytokine profiles. Compared to serum from S-2P:Al-vaccinated mice, serum from S-2P:ISA720 vaccinated mice demonstrated superior effect in reducing viral load in both aged and young mouse brains (Fig. 5F). Notably, the antiviral efficacy of serum from either S-2P:Al-or S-2P:ISA720-vaccinated mice was significantly attenuated in aged recipients compared to young recipients, indicating age-dependent impairment of serum-mediated protection. We also observed that young naïve mice exhibited significantly stronger neuroinflammatory responses post-challenge than aged mice, with 14.2-fold higher CXCL10 (P = 0.043), 10.3-fold higher CXCL2 (P = 0.047), 2.9-fold higher TNF-α (P = 0.049), and 9.9-fold elevated IL-6 (P = 0.019) levels in brains (Fig. 5G andH). IFN-γ levels showed comparable changes (P = 0.126) between age groups. This suggests that the attenuated innate immune response in aged animals may contribute to impaired antibody-mediated protection, ultimately leading to poorer outcomes despite comparable or higher antibody titers.
## DISCUSSION
Recombinant vaccines account for a substantial share of global COVID-19 vaccine supplies. Adjuvants act as indispensable components in mediating effective induction of adaptive immune responses to recombinant vaccines. In this study, we characterized the immune responses induced by recombinant prefusion-stabilized ancestral SARS-CoV-2 spike protein S-2P formulated with aluminum or one of three Seppic emulsion adjuvants (ISA 720, ISA 51, or SWE). Our findings demonstrate that adjuvants critically influence vaccine efficacy across magnitude, breadth, and pattern of immune responses and reveal key insights into the age-dependent decline in vaccine effectiveness as well as the establishment of a sustained vaccination ceiling effect.
Using the ancestral SARS-CoV-2 S-2P protein as antigen, all four adjuvanted vaccines effectively elicited S1-, S2-, and RBDspecific IgG antibodies with neutralizing activity in mice. The S-2P:ISA720 formulation demonstrated superior immunogenicity, closely followed by S-2P:ISA51, with both inducing robust Th1-biased immune responses and strong IFN-γ-producing T-cell responses in both spleen and lung tissues. In contrast, neither S-2P:Al nor S-2P:SWE induced detectable IFN-γ-producing T-cell responses. Two-dose immunization with S-2P:ISA720 conferred complete protection against lethal challenge with homologous SARS-CoV-2, and one dose achieved 90% protection rates, while S-2P:ISA51 also demonstrated excellent protective efficacy. In contrast, alum-and SWE-adjuvanted vaccines exhibited markedly reduced protective effectiveness.
Since alum-and SWE-adjuvanted formulations did not induce detectable T-cell responses, protection was likely mediated primarily by neutralizing antibodies. This is supported by the positive correlation between reduced brain viral loads, decreased brain inflammatory cytokine levels, and higher serum neutralizing antibody titers in individual mice. The protective efficacy of passive transfer of immune sera experi ments further underscores the important protective role of vaccine-induced neutraliz ing antibodies against homologous SARS-CoV-2 infection. It is noteworthy that both SWE-and ISA51-adjuvanted groups exhibited delayed neutralizing antibody activity, potentially due to the lower affinity of antibodies generated during the early immune response. These initial antibodies primarily originate from extrafollicular responses, which typically produce loweraffinity antibodies with diminished neutralizing capacity. Affinity maturation, a process requiring prolonged antigen exposure in germinal center (GC), evolves over time (41,42). In contrast, the ISA720 adjuvant appears to enhance GC reactions more effectively by activating key cellular components such as follicular helper T (Tfh) cells and follicular dendritic cells (29,30), thereby facilitating the rapid production of highaffinity antibodies.
VAERD is characterized by pulmonary eosinophil infiltration and enhanced type 2 cytokine responses (28). Recent studies in rodent models have demonstrated VAERD following immunization with SARS-CoV-2 vaccines (28), including inactivated formula tions adjuvanted with aluminum hydroxide, which triggered Th2-type inflammatory cytokines and exacerbated respiratory pathology upon heterologous challenge (43). In this study, although key Th2 cytokines such as IL-4 were not quantified, the absence of detectable IFN-γ-producing T-cell responses and a reduced IgG2a/IgG1 ratio in the aluminum adjuvant group collectively suggest a Th2-skewed immune response. By comparing IFN-γ-producing T-cell activity and IgG2a/IgG1 ratios across adjuvant groups, and in light of the essential role of Th1 immunity in vaccine-mediated protection, these findings underscore the translational promise of ISA 720 and ISA 51 adjuvants.
The remarkable ability of SARS-CoV-2 to undergo rapid and continuous evolution has led to recurrent global outbreaks, predominantly driven by Omicron variants (13). Adjuvants are known to enhance the breadth of the immune response to vaccina tion. For example, AS03 adjuvant induced robust RBDspecific memory B cells with increased cross-neutralization activity (44). In this study, we observed that the magnitude of antibody responses enhanced by various adjuvants correlated with their breadth of neutralization. Sera from hACE2 mice immunized with Alum-or SWE-adjuvanted formulations showed minimal neutralization against the Omicron BA.1 variant, while sera of mice receiving ISA720-or ISA51-adjuvanted formulations demonstrated moderate neutralizing activity. The majority of sera from S-2P:ISA720-immunized mice effectively neutralized multiple Omicron subvariants, whereas only a minority of S-2P:Al immune sera showed detectable activity against these variants. Importantly, vaccination with the ISA720 formulation conferred complete protection against lethal challenge with BA.5 and BF.7 variants, whereas none of the mice receiving the aluminum formulation survived challenge. The significantly reduced viral load and inflammatory cytokine mRNA levels in mouse brains further demonstrated the superior protective efficacy of the ISA720 formulation.
Αlthough S-2P:ISA720 elicited cross-neutralizing antibodies against Omicron variants, their titers were substantially lower than those against the ancestral strain and were also significantly reduced compared to the ancestralspecific antibodies induced by S-2P:Al. Given that both active immunization with S-2P:Al and passive transfer of immune serum conferred partial protection against homologous challenge, the modest cross-neutralizing antibody response induced by S-2P:ISA720 is unlikely to account for the complete protection observed against lethal challenge with Omicron BA.5 and BF.7 variants. By integrating neutralizing antibody titers, T-cell responses, and survival rates after challenge, these results strongly suggest that T-cell immunity serves as the primary mechanism underlying the superior protective efficacy provided by ISA720based vaccination against Omicron variants. As documented in numerous studies, T-cell epitopes-particularly CD8+ T-cell epitopes-are highly conserved across SARS-CoV-2 variants (16,39). Moreover, CD8+ T cells have been shown to mediate complete protection independently of neutralizing antibodies (45,46).
Viral loads were reduced more effectively in the brain than in the nasal turbinates or lungs in S-2P:ISA720 immunized mice upon challenge with Omicron BA.5 and BF.7 variants. This disparity in tissuespecific protection may be attributed to two mecha nisms. First, the vaccine does not provide sterilizing immunity but instead limits viral replication. After intranasal challenge, infection is initially established in the respiratory tract, which triggers clonal expansion of memory T cells and a rapid effector response in vaccinated mice. By the time the virus disseminates to the brain, effector T cells have been mobilized and are efficiently recruited to the brain tissue, where they mediate robust antiviral effector functions. Second, strong suppression of viral replication in the respiratory tract of S-2P:ISA720-immunized mice reduces the viral load available to disseminate to the brain.
Consistent with our observation that ISA720 exhibits potent immunostimulatory activity, the AKS-452 vaccine, which comprises an ancestral RBD-human Fc fusion protein adjuvanted with ISA 720, elicited robust neutralizing antibody responses and conferred strong protection in both murine and non-human primate models (47,48). Further more, in a randomized Phase I/II clinical trial, two 45 µg doses administered 28 days apart induced high titers of IgG and broad-spectrum neutralizing antibodies capable of targeting ancestral, Alpha, and Delta variants, alongside sustained IFN-γ-producing T-cell responses that remained detectable through Day 180 (32).
The age-related immunological alterations not only increase susceptibility to severe COVID-19 and mortality, particularly in those with comorbidities, but also impair vaccine responsiveness (49,50). In this study, we systematically compared the immunogenicity and protective efficacy of Alum-versus ISA720-adjuvanted S-2P vaccines in young adult and aged hACE2 mice. The S-2P:ISA720 formulation elicited significantly higher anti-RBD IgG and neutralizing antibody titers in both age groups relative to S-2P:Al. While both adjuvanted formulations showed age-related reductions in antibody responses, S-2P:Al demonstrated significantly greater decreases in both anti-RBD and neutralizing antibody titers in aged mice compared to S-2P:ISA720. Notably, protection against ancestral SARS-CoV-2 challenge dropped from 40% to 0% in Alum-adjuvanted groups when comparing young versus aged mice, whereas S-2P:ISA720 maintained 40% protection in aged mice. These findings highlight the significant impact of immunosenescence on vaccine efficacy but also pose new challenges for developing vaccination strategies for elderly populations.
Some aged mice received immunization with S-2P:ISA720, despite producing relatively high levels of neutralizing antibodies, still succumbed to SARS-CoV-2 infection, whereas young mice exhibited clear antibody titer-related protection. The experiment of passive transfer of immune sera revealed two additional striking differences between aged and young mice. First, aged mice mounted a substantially attenuated inflamma tory cytokine response, characterized by significantly lower production of key chemo kines, including CXCL10 and CCL2. During viral infection, CXCL10 and CCL2 often function synergistically to orchestrate immune cell recruitment: CCL2 mediates the initial recruitment of monocytes and macrophages, which, upon activation, produce IFN-γ, which subsequently induces CXCL10 expression. CXCL10 then promotes the recruitment of T cells and NK cells into infected tissues, thereby enhancing the cellu lar immune response and amplifying antiviral immunity (51). Second, aged mice that received passive antibody transfer also showed significantly reduced protection against viral challenge compared to young recipients. Taken together with the T-cell-mediated protection observed in this study and the well-established role of T cells in vaccineinduced immunity, these findings suggest that the age-related decline in vaccine efficacy is multifactorial in nature. Nonetheless, a substantially impaired innate immune response and the concomitant weakening of cellular immunity likely represent key mechanisms underlying reduced vaccine protection in aged individuals. Additionally, the findings provide new insights into the development of antibody-based immunotherapy strategies against viral infections in the aged population.
Although direct evidence linking neuroinflammation to viral clearance remains limited, these inflammatory mediators are known to enhance antiviral immunity by promoting the recruitment and activation of monocytes and T cells (51). The attenuated neuroinflammatory response observed in aged mice following SARS-CoV-2 infection may therefore compromise viral clearance. Furthermore, accumulating evidence indicates that persistent SARS-CoV-2 components (such as RNA and proteins) in the brain (52), combined with chronic inflammation and immune dysregulation due to sustained SARS-CoV-2 infection (53), can lead to prolonged cytokine production, a mechanism strongly implicated in long COVID. Thus, the diminished neuroinflammatory response in aged mice after viral challenge may contribute to the development of long COVID.
The attenuated inflammatory response observed in aged mice following viral challenge appears to contradict the well-established paradigm of elevated baseline inflammation in elderly populations. By employing young and aged mice as parallel reference groups to analyze post-challenge inflammatory dynamics, our findings reveal distinct age-dependent patterns in inflammatory magnitude, potentially reflecting differential engagement of protective antiviral immunity mechanisms. These results align with recent reports demonstrating similar age-related divergence in inflammatory responses to virus challenge (54,55).
Booster vaccinations are primarily administered to restore waning antibody levels and broaden immune responses against emerging variants (20). However, repeated immunization may lead to a ceiling effect-a maximum IgG titer, where additional doses fail to elicit significantly higher antibody responses. This immune plateau phenomenon has been well-documented in recipients of frequent influenza vaccines and has similarly been observed in populations receiving multiple SARS-CoV-2 vaccinations (56)(57)(58). In this study, two doses of S-2P:ISA720 vaccination induced a ceiling effect, characterized by peak anti-RBD and neutralizing antibody titers against both the homologous strain and heterologous Omicron BA.2 variant that were not enhanced by a third dose. These antibodies sustained at high levels with minimal waning. In contrast, mice immunized with S-2P:Al showed significantly increased antibody titers to ancestral and Omicron BA.2 after the third dose, though the levels remained substantially lower than those achieved with just two doses of S-2P:ISA720. Furthermore, the S-2P:Al group exhibited similar waning kinetics following the third dose as observed after the second dose. The capacity of S-2P:ISA720 to rapidly elicit high antibody titers that plateau satisfies the requirements for durable immune protection. This long-lasting protection was evidenced by 100% survival rates in S-2P:ISA720 immunized mice following viral challenge at 17 weeks post the second dose, while all S-2P:Al-immunized mice succumbed to infection. Our results suggest that adjuvants optimized for both rapid kinetics and broad-spectrum antibody responses could represent a strategic approach for combating continuously evolving SARS-CoV-2.
Collectively, our study demonstrates that optimal adjuvant formulation in recombi nant SARS-CoV-2 spike protein vaccine not only significantly enhances the magnitude and breadth of antibody responses but also robustly boosts T-cell immunity. Further more, it rapidly induces and sustains peak antibody levels against both the homolo gous strain and Omicron variants. Notably, aged mice exhibited not only markedly reduced antibody production post-vaccination but also significantly compromised antiviral functionality of the elicited antibodies. These findings provide critical insights for developing next-generation SARS-CoV-2 vaccine strategies tailored for broad and durable protection.
## References
1. Wrapp, Wang, Corbett et al. (2020) "Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation" *Science*
2. Lan, Ge, Yu et al. (2020) "Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor" *Nature*
3. Zost, Gilchuk, Case et al. (2020) "Potently neutralizing and protective human antibodies against SARS-CoV-2" *Nature*
4. Cao, Su, Guo et al. (2020) "Potent neutralizing antibodies against SARS-CoV-2 identified by high-throughput single-cell sequencing of convalescent patients' B cells"
5. Xia, Zhang, Wang et al. (2021) "Safety and immunogenicity of an inactivated SARS-CoV-2 vaccine, BBIBP-CorV: a randomised, double-blind, placebo-controlled, phase 1/2 trial" *Lancet Infect Dis*
6. Baden, Sahly, Essink et al. (2021) "Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine" *N Engl J Med*
7. Polack, Thomas, Kitchin et al. (2020) "Safety and efficacy of the BNT162b2 mRNA COVID-19 vaccine" *N Engl J Med*
8. Sadoff, Gray, Vandebosch et al. (2021) "Safety and efficacy of single-dose Ad26.COV2.S vaccine against COVID-19" *N Engl J Med*
9. Heath, Galiza, Baxter et al. (2021) "Safety and efficacy of NVX-CoV2373 COVID-19 vaccine" *N Engl J Med*
10. Planas, Saunders, Maes et al. (2022) "Considerable escape of SARS-CoV-2 Omicron to antibody neutralization" *Nature*
11. Kirsebom, Andrews, Stowe et al. (2022) "COVID-19 vaccine effectiveness against the omicron (BA.2) variant in England"
12. Pajon, Rose, Shen et al. (2022) "SARS-CoV-2 Omicron variant neutralization after mRNA-1273 booster vaccination" *N Engl J Med*
13. Ibarrondo, Fulcher, Goodman-Meza et al. (2020) "Rapid decay of anti-SARS-CoV-2 antibodies in persons with mild Covid-19" *N Engl J Med*
14. Levin, Lustig, Cohen et al. (2021) "Waning immune humoral response to BNT162b2 COVID-19 vaccine over 6 months" *N Engl J Med*
15. Guo, Zhang, Gu et al. (2024) "Durability and cross-reactive immune memory to SARS-CoV-2 in individuals 2 years after recovery from COVID-19: a longitudinal cohort study" *Lancet Microbe*
16. Li, Guan, Wu et al. (2020) "Early transmission dynamics in Wuhan, China, of novel coronavirus-infected pneumonia" *N Engl J Med*
17. Goyal, Choi, Pinheiro et al. (2020) "Clinical characteristics of Covid-19 in"
18. Bartleson, Radenkovic, Covarrubias et al. (2021) "SARS-CoV-2, COVID-19 and the ageing immune system" *Nat Aging*
19. Barouch (2022) "Covid-19 vaccines -immunity, variants, boosters" *N Engl J Med*
20. Ma, Surie, Lauring et al. (2023) "Effectiveness of updated 2023-2024 (monovalent XBB.1.5) COVID-19 vaccination against SARS-CoV-2 Omicron XBB and BA.2.86/JN.1 lineage hospitalization and a comparison of clinical severity-IVY Network, 26 hospitals" *Clin Infect Dis:ciae*
21. Pulendran, Arunachalam, Hagan (2021) "Emerging concepts in the science of vaccine adjuvants" *Nat Rev Drug Discov*
22. Dai, Gao, Tao et al. (2022) "Efficacy and safety of the RBD-dimerbased COVID-19 vaccine ZF2001 in adults" *N Engl J Med*
23. Richmond, Hatchuel, Dong et al. (2021) "Safety and immunogenicity of S-Trimer (SCB-2019), a protein subunit vaccine candidate for COVID-19 in healthy adults: a phase 1, randomised, double-blind, placebo-controlled trial" *Lancet*
24. Dayan, Rouphael, Walsh et al. (2023) "Efficacy of a bivalent (D614 + B.1.351) SARS-CoV-2 recombinant protein vaccine with AS03 adjuvant in adults: a phase 3, parallel, randomised, modified double-blind, placebo-controlled trial" *Lancet Respir Med*
25. Lu, Hogenesch (2013) "Kinetics of the inflammatory response following intramuscular injection of aluminum adjuvant" *Vaccine (Auckl)*
26. Ebenig, Muraleedharan, Kazmierski et al. (2022) "Vaccine-associated enhanced respiratory pathology in COVID-19 hamsters after T H 2-biased immunization" *Cell Rep*
27. O'hagan, Lodaya, Lofano (2020) "The continued advance of vaccine adjuvants -"we can work it out" *Semin Immunol*
28. Aucouturier, Dupuis, Deville et al. (2002) "Montanide ISA 720 and 51: a new generation of water in oil emulsions as adjuvants for human vaccines" *Expert Rev Vaccines*
29. Tabynov, Turebekov, Babayeva et al. (2022) "An adjuvanted subunit SARS-CoV-2 spike protein vaccine provides protection against Covid-19 infection and transmission" *NPJ Vaccines*
31. Feitsma, Janssen, Boersma et al. (2023) "A randomized phase I/II safety and immunogenicity study of the Montanide-adjuvanted SARS-CoV-2 spike protein-RBD-Fc vaccine, AKS-452" *Vaccine (Auckl)*
32. Heitmann, Bilich, Tandler et al. (2022) "A COVID-19 peptide vaccine for the induction of SARS-CoV-2 T cell immunity" *Nature*
33. Heitmann, Tandler, Marconato et al. (2023) "Phase I/II trial of a peptide-based COVID-19 T-cell activator in patients with B-cell deficiency" *Nat Commun*
34. Gerdts (2024) "A randomized, observer-blind, dose-escalation phase 2 clinical trial of COVAC-2 in generally healthy adults. NCT05209009. Clinical trial registration"
35. Peng, Ding, Jiang et al. (2022) "Discovery of potential anti-SARS-CoV-2 drugs based on large-scale screening in vitro and effect evaluation in vivo" *Sci China Life Sci*
36. Flurkey, Currerj, Harrison et al. (2007) "Chapter 20 -Mouse models in aging research"
37. Cao, Wang, Peng et al. (2022) "A single vaccine protects against SARS-CoV-2 and influenza virus in mice" *J Virol*
38. Keeton, Tincho, Ngomti et al. (2022) "T cell responses to SARS-CoV-2 spike cross-recognize Omicron" *Nature*
39. Song, Zhang, Israelow et al. (2021) "Neuroinvasion of SARS-CoV-2 in human and mouse brain" *J Exp Med*
40. Cavazzoni, Hanson, Podestà et al. (2022) "Follicular T cells optimize the germinal center response to SARS-CoV-2 protein vaccination in mice" *Cell Rep*
41. Tangye, Burnett, Bull (2022) "Getting to the (germinal) center of humoral immune responses to SARS-CoV-2" *Cell*
42. Dillard, Taft-Benz, Knight et al. (2024) "Adjuvantdependent impact of inactivated SARS-CoV-2 vaccines during heterologous infection by a SARS-related coronavirus" *Nat Commun*
43. Grigoryan, Feng, Bellusci et al. (2024) "AS03 adjuvant enhances the magnitude, persistence, and clonal breadth of memory B cell responses to a plant-based COVID-19 vaccine in humans" *Sci Immunol*
44. Israelow, Mao, Klein et al. (2021) "Adaptive immune determinants of viral clearance and protection in mouse models of SARS-CoV-2" *Sci Immunol*
45. Ishii, Nomura, Yamamoto et al. (2022) "Neutralizing-antibody-independent SARS-CoV-2 control correlated with intranasal-vaccine-induced CD8 + T cell responses" *Cell Rep Med*
46. Alleva, Delpero, Scully et al. (2021) "Development of an IgG-Fc fusion COVID-19 subunit vaccine, AKS-452" *Vaccine (Auckl)*
47. Janssen, Feitsma, Boersma et al. (2022) "Phase I interim results of a phase I/II study of the IgG-Fc fusion COVID-19 subunit vaccine, AKS-452" *Vaccine (Auckl)*
48. Dallan, Proietto, Laurentis et al. (2024) "Age differentially impacts adaptive immune responses induced by adenoviral versus mRNA vaccines against COVID-19" *Nat Aging*
49. Collier, Ferreira, Kotagiri et al. (2021) "Age-related immune response heterogeneity to SARS-CoV-2 vaccine BNT162b2" *Nature*
50. Ansari, Ahmad, Alam et al. (2024) "Virus-induced host chemokine CCL2 in COVID-19 pathogenesis: Potential prognostic marker and target of antiinflammatory strategy" *Rev Med Virol*
51. Zuo, He, Liang et al. (2024) "The persistence of SARS-CoV-2 in tissues and its association with long COVID symptoms: a cross-sectional cohort study in China" *Lancet Infect Dis*
52. Schultheiß, Willscher, Paschold et al. (2022) "The IL-1β, IL-6, and TNF cytokine triad is associated with post-acute sequelae of COVID-19" *Cell Rep Med*
53. Lacasse, Dubuc, Gudimard et al. (2025) "Delayed viral clearance and altered inflammatory responses affect severity of SARS-CoV-2 infection in aged mice" *Immun Ageing*
54. Subramaniam, Kenney, Jayaraman et al. (2024) "Aging is associated with an insufficient early inflammatory response of lung endothelial cells in SARS-CoV-2 infection" *Front Immunol*
55. Hoskins, Davies, Smith et al. (1979) "Assessment of inactivated influenzaA vaccine after three outbreaks of influenza A at Christ's Hospital" *Lancet*
56. Jones-Gray, Robinson, Kucharski et al. (2023) "Does repeated influenza vaccination attenuate effectiveness? A systematic review and meta-analysis" *Lancet Respir Med*
57. Munro, Feng, Janani et al. (2022) "Safety, immunogenicity, and reactogenicity of BNT162b2 and mRNA-1273 COVID-19 vaccines given as fourth-dose boosters following two doses of ChAdOx1 nCoV-19 or BNT162b2 and a third dose of" *Lancet Infect Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12589543&blobtype=pdf | # Corrections & amendments Author Correction: A single residue in the yellow fever virus envelope protein modulates virion architecture and antigenicity
Summa Bibby, James Jung, Yu Low, Alberto Amarilla, Natalee Newton, Connor Scott, Jessica Balk, Yi Ting, Morgan Freney, Benjamin Liang, Timothy Grant, Fasséli Coulibaly, Paul Young, Roy Hall, Jody Hobson-Peters, Naphak Modhiran, Daniel Watterson, Nature Communications |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12567752&blobtype=pdf | # Transcontinental Spread of HPAI H5N1 from South America to Antarctica via Avian Vectors
Ruifeng Xu, Minhao Gao, Nailou Zhang, Zhenhua Wei, Zheng Wang, Lei Zhang, Yang Liu, Zhenhua Zheng, Liulin Chen, Haitao Ding, Wei Wang
## Abstract
During China's 41st Antarctic research expedition, samples were collected from wildlife on the Fildes Peninsula, South Shetland Islands, Antarctica. Real-time RT-PCR screening confirmed H5N1 positivity, representing the first identification of the virus in brown skuas on the Fildes Peninsula. Whole-genome sequences obtained from positive samples via next-generation sequencing were subjected to phylogenetic and phylogeographic analyses. The results revealed that these Antarctic strains are most closely related to H5N1 viruses circulating in South America, particularly from Peru and Chile, suggesting a likely introduction via avian migration routes. Furthermore, a unique 17-amino-acid deletion was identified in the stalk region of the neuraminidase (NA) gene, which is uncommon among globally sampled clade 2.3.4.4b variants. This study confirms the arrival of HPAI H5N1 in the Antarctic continent and underscores the necessity for enhanced surveillance to understand the viral ecology and potential risks within this unique ecosystem.
## 1. Introduction
Highly Pathogenic Avian Influenza (HPAI) H5N1 viruses originating from the A/Goose/Guangdong/1/96 (Gs/GD) lineage have demonstrated rapid genomic evolution and remarkable cross-species transmission capabilities. Clade 2.3.4.4b outbreaks have intensified significantly since 2020, spreading rapidly across Asia, Europe, Africa, North America, and South America, with substantial impacts on poultry and wildlife [1]. Notably, this clade has demonstrated unprecedented adaptive evolution in mammalian hosts, with confirmed infections in dairy cattle and domestic cats on farms [2], as well as mass mortality events in southern elephant seals [3]. Human infections have also been documented [4], though sustained human-to-human transmission has not been observed.
The Antarctic region encompasses the Antarctic ice shelves, surrounding waters, and all island territories south of the Antarctic Convergence (Antarctic Polar Front), a marine boundary where cold Antarctic waters meet the warmer sub-Antarctic waters [5]. This region supports unique ecosystems that serve as critical habitats for numerous avian and marine mammal species. Despite their geographical isolation, wildlife, such as brown skuas (Stercorarius antarcticus), southern giant petrels (Macronectes giganteus), the Southern Elephant Seal (Mirounga leonina), and the Antarctic Fur Seal (Arctocephalus gazella), regularly breed in Antarctica but partially migrate to the coasts of Chile and Argentina during winter. Growing evidence suggests that viruses from South America may be spreading southward to Antarctica. In 2023, HPAI H5N1 was detected in sub-Antarctic regions including South Georgia (54 • 15 ′ S, 36 • 45 ′ W) and the Falkland Islands (51 • 42 ′ S, 57 • 51 ′ W) [6]. By 2024, the virus had reached James Ross Island (64 • 10 ′ S, 57 • 45 ′ W) in Antarctica, where it was identified in brown skuas, though no signs of HPAI were observed in wildlife on the Fildes Peninsula [7].
## 2. Materials and Methods
## 2.1. Sample Collection
During the 41st Chinese Antarctic Research Expedition, field investigations were conducted across the Fildes Peninsula and surrounding areas from December 2024 to February 2025. Expedition members wearing full personal protective equipment collected influenza samples from both live animals and atypical mortality cases using standardized protocols for sterilization, disinfection, and hazardous waste disposal. The spatial distribution of sampling sites was mapped using Google Earth. A total of 11 biological samples were obtained, including oropharyngeal swabs, cloacal swabs, fecal samples, and brain tissue specimens. All samples were preserved in cryotubes containing 1 mL RNA later and stored at 4 • C after collection and preserved at -80 • C.
## 2.2. qPCR
For viral genetic testing, we utilized a portable real-time fluorescence PCR system and Influenza A Virus with H5/H7 Subtype Nucleic Acid Detection Kit (PCR-fluorescent probe method) (Zhongkeshengyi science and technology Co., Ltd., Beijing, China), following the manufacturer's instructions for avian influenza virus screening. Samples with Ct values ≤35, accompanied by a characteristic S-shaped amplification curve, were identified as avian influenza virus-positive.
## 2.3. RNA Extraction
The processed samples were subjected to RNA extraction using the MiniBEST Viral RNA/DNA Extraction Kit Ver. 5.0 (Takara, Osaka, Japan) following the manufacturer's protocol. The extracted RNA was immediately stored at -80 • C until downstream molecular analysis. For each extraction procedure, RNase-free water was included as a negative control.
## 2.4. Next-Generation Sequencing
Influenza virus-positive samples identified via PCR-fluorescent probe assay were selected for Next-generation sequencing (NGS) to analyze genomic characteristics. Total RNA libraries were constructed using the VAHTS Universal V8 RNA-seq Library Prep Kit for Illumina (Vazyme, Nanjing, China) in accordance with the manufacturer's protocol. Approximately 1 µg of total RNA was reverse-transcribed into cDNA, followed by fragmentation, adapter ligation, PCR amplification, and purification. Library quality control was performed using an Agilent 2100 Bioanalyzer. Final sequencing was conducted on an Illumina NovaSeq platform (San Diego, CA, USA).
## 2.5. Sequencing Data Assembly and Bioinformatic Analysis
According to the method described by Zhang et al. [8] and Yu et al. [9], we performed bioinformatics analysis on whole-genome sequencing data to assemble sequences of eight different influenza virus segments. Briefly, raw sequencing reads were first subjected to quality assessment and removal of adapter sequences/low-quality reads using Fastp. Quality-controlled sequences were then aligned against a host genome database using Kraken2 to eliminate host-derived sequences. De novo assembly was conducted with Megahit, followed by assembly polishing with Pilon and sequence consensus optimization using CAP3. The improved sequences served as references for reassembling high-quality reads to generate final reference-based assemblies. 514 H5N1 genomic sequences were downloaded from the GISAID database and NCBI, yielding eight datasets corresponding to the eight gene segments of H5N1. Each dataset underwent multiple sequence alignment using MUSCLE, followed by phylogenetic analysis that incorporated the H5N1 sequences obtained in our study. Maximum likelihood phylogenetic trees were constructed using FastTree, with subsequent visualization performed using the ggtree package. For spatiotemporal phylogenetic analysis, we employed Nextstrain to infer the viral evolutionary origin, mutation rate, and transmission patterns.
## 3. Results
## 3.1. Case Description and Virus Detection
During China's 41st Antarctic Scientific Expedition, we collected samples in the vicinity of the Great Wall Station (62 • 12 ′ 59 ′′ S, 58 • 57 ′ 52 ′′ W) on the Fildes Peninsula, located in western King George Island of the South Shetland Islands, Antarctica.
On 7 December 2024, a critically ill, flightless thievish gull with smooth plumage but abnormal behavior was found near the Great Wall Station in China. The bird died on 10 December, and a necropsy was subsequently performed, with lung and brainstem tissues collected. Two more dead brown terns with disheveled plumage were subsequently found in the area on 25 and 26 December, and tissue samples were taken from them; a fresh fecal sample from a flying brown tern was also taken on 26 December. Real-time RT-PCR screening initially detected that these early samples were all positive for HPAI H5.
Follow-up surveillance was also conducted in 2025, and on 17 January 2025, fresh elephant seal feces were collected, and an intact submerged Antarctic fur seal carcass was found from which pharyngeal and cloacal swabs were collected. Follow-up sampling also included thieving gull feces collected on 22 January and 16 February, as well as penguin feces and a fresh fecal sample from a giant petrel collected on 13 February.
Real-time RT-PCR screening detected H5 positivity in one elephant seal fecal sample, along with H7 positivity in one fur seal sample, though subsequent retesting failed to confirm positivity in the elephant seal and fur seal specimens.
The full chronology of sampling events and all positive species are detailed in Figure 1 and Supplementary Table S1.
## 3.2. Genomic and Phylogenetic Analysis
To investigate the potential origins, genetic characteristics, and evolution of these viruses, we performed next-generation sequencing (NGS) and phylogenetic analyses, successfully obtaining complete genome sequences for all four H5N1-positive skua samples. The H5N1 viruses identified on the Fildes Peninsula, Antarctica were formally designated as A/brown skua/Fildes Peninsula/B1/2024(H5N1), A/brown skua/Fildes Peninsula/B2/2024(H5N1), A/brown skua/Fildes Peninsula/B3/2024(H5N1), and A/brown skua/Fildes Peninsula/F4/2024(H5N1), with their complete genome sequences deposited in the Global Initiative on Sharing All Influenza Data (GISAID) under accession numbers EPI_ISL_19847535, EPI_ISL_19847536, EPI_ISL_19847538, and EPI_ISL_19847539, respectively. The genome sequences of the viruses possessed high homology with each other. To further characterize the obtained viral strains, we retrieved 514 representative H5N1 sequences from the GISAID database for genotyping (Supplement Table S2). Subtyping based on hemagglutinin (HA) and neuraminidase (NA) protein sequences confirmed these isolates as H5N1 subtype, belonging to clade 2.3.4.4b genotype B3.2 (Figure S1). Additionally, we retrieved 1146 representative H5N1 clade 2.3.4.4b sequences from the National Center for Biotechnology Information (NCBI) for phylogenetic analysis (Supplement Table S3). Phylogenetic analysis revealed that viral sequences from the Fildes Peninsula clustered closely with H5N1 strains originating from South America, particularly demonstrating strong phylogenetic affinity with Peruvian and Chilean isolates across all genomic segments (Figure 2). These genetically similar viruses exhibit broad circulation among both avian and marine mammal populations, and South America is the predominant geographic reservoir for this viral lineage (Figures S2 andS3).
To investigate the potential introduction pathway of HPAI H5N1 to the Fildes Peninsula, we performed a discrete phylogeographic analysis using 90 closely related HA and NA sequences identified in our initial screening. The spatiotemporal reconstruction indicated Peru as the likely source country for the HPAI H5N1 virus detected on Fildes Peninsula, with transmission occurring through Chile before reaching Antarctica. Molecular dating traced the HA sequences to a Chilean strain sampled on 3 December 2022 (95% HPD: 13 November-10 December 2022), while the NA sequences originated from Chilean viruses circulating between 25 November 2022 (95% HPD: 10 November 2022-7 January 2023). This spatiotemporal reconstruction reveals a transmission pathway whereby the virus spilled over from South American poultry populations to brown skuas, subsequently reaching Fildes Peninsula's wildlife through the birds' natural migratory routes. (Figure S4A,B).
Comparative analysis with the reference strain A/chicken/OHiggins/241252-3/2023 revealed a 17-amino acid deletion (positions 58-74) in the stalk region of the NA protein. This deletion pattern was identical to that observed in the H5N1 strain A/brown skua/Torgersen Island/81-b82/2024 isolated from Torgersen Island but was not found in other H5N1 clade 2.3.4.4b variants. Notably, the NA gene of the Fildes Peninsula isolate harbored the substitutions E57K, V321I, and S366I, while the HA gene contained Y153H. However, continued genomic surveillance is necessary to evaluate the epidemiological significance of these findings.
## 4. Discussion
This study reports the first detection of HPAI H5N1 in brown skuas on the Fildes Peninsula, Antarctica, characterized by a unique 17-aa NA stalk deletion (58-74) currently found only in Antarctic strains. Multiple studies have documented NA stalk deletions in various influenza strains: A/Hong Kong/159/97 (H5N1) possesses a 19-aa deletion (positions 54-72) that reduces viral release capacity; A/chicken/Hubei/327/2004 (H5N1) contains a 20-aa deletion (positions 49-68); while A/Puerto Rico/8/34 (H1N1) and A/WSN/33 (H1N1) exhibit 15-aa (positions 63-77) and 16-aa (positions 57-72) deletions, respectively [10,11]. Reverse genetics studies demonstrate that NA stalk length influences viral replication kinetics; viruses with shorter stalks replicate significantly faster than their long-stalk counterparts. NA stalk length correlates with H5N1 virulence and pathogenicity; a truncated NA stalk might enhance the pathogenicity of the H5N1 subtype AIV to the mallard duck nervous system [12]. The increasing prevalence of NA stalk deletions in avian H5N1 viral isolates suggests potential adaptive advantages that may enhance cellular fitness and expand host tropism. While these mutations could significantly alter viral pathogenicity, transmissibility, and host specificity, their precise biological consequences require further experimental validation. The ecological and evolutionary significance of this Antarctic-specific deletion warrants investigation.
It should be noted that this study has limitations, including the restricted species coverage and small sample size. These constraints underscore the need for more systematic and continuous monitoring efforts to obtain more representative samples. Given the increasing human activities in Antarctica and its unique ecological significance, such surveillance is not only critical for understanding viral ecology but also essential for the early detection of potential zoonotic transmission risks.
## References
1. Peacock, Moncla, Dudas et al. "The global H5N1 influenza panzootic in mammals" *Nature*
2. Caserta, Frye, Butt et al. "Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle" *Nature*
3. Uhart, Vanstreels, Nelson et al. (2024) "Epidemiological data of an influenza A/H5N1 outbreak in elephant seals in Argentina indicates mammal-to-mammal transmission" *Nat. Commun*
4. Garg, Reinhart, Couture et al. (2025) "Highly Pathogenic Avian Influenza A(H5N1) Virus Infections in Humans" *N. Engl. J. Med*
5. Moore, Abbott, Richman (1999) "Location and dynamics of the Antarctic Polar Front from satellite sea surface temperature data" *J. Geophys. Res. Oceans*
6. Banyard, Bennison, Byrne et al. (2024) "Detection and spread of high pathogenicity avian influenza virus H5N1 in the Antarctic Region" *Nat. Commun*
7. Bennett-Laso, Berazay, Muñoz et al. (2024) "Confirmation of highly pathogenic avian influenza H5N1 in skuas" *Front. Vet. Sci*
8. Zhang, Hu, Zhang et al. (2025) "Virome landscape of wild rodents and shrews in Central China" *Microbiome*
9. Yu, Jin, Cui et al. (2014) "Influenza H7N9 and H9N2 Viruses: Coexistence in Poultry Linked to Human H7N9 Infection and Genome Characteristics"
10. Matrosovich, Zhou, Kawaoka et al. (1999) "The Surface Glycoproteins of H5 Influenza Viruses Isolated from Humans, Chickens, and Wild Aquatic Birds Have Distinguishable Properties" *J. Virol*
11. Zhou, Yu, Hu et al. (2009) "The Special Neuraminidase Stalk-Motif Responsible for Increased Virulence and Pathogenesis of H5N1 Influenza A Virus" *PLoS ONE*
12. Chen, Quan, Wang et al. "Truncation or Deglycosylation of the Neuraminidase Stalk Enhances the Pathogenicity of the H5N1 Subtype Avian Influenza Virus in Mallard Ducks"
13. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671217&blobtype=pdf | # Evaluation of the RHAM-based Pluslife Rapid SARS-CoV-2 assay and comparison with EasyNAT Rapid SARS-CoV-2 assay and Wondfo 2019-nCoV antigen assay
Guosheng Zhong, Zhiwei Zhao, Yuting He, Xuegao Yu, Tingting Yang, Huanliang Liu, Shibing Li, Cha Chen, Mingyong Luo, Bin Huang
## Abstract
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) remains a significant public health challenge. This study evaluated the Pluslife SARS-CoV-2 assay employing proprietary RNase Hybridization-Assisted amplification (RHAM) isothermal amplification technology, with comparative performance assessment against the EasyNAT Rapid SARS-CoV-2 assay and the Wondfo 2019-nCoV antigen assay. Analytical testing determined the Pluslife assay's limit of detection to be 400 copies/mL, with no cross-reactivity against common respiratory viruses and no interference from tested substances. Using RT-qPCR as the reference standard on 197 nasopharyngeal swabs, the Pluslife assay showed a positive percentage agreement (PPA) of 96.30% and a negative percentage agreement (NPA) of 99.30% (κ = 0.962, P < 0.001). The EasyNAT assay demonstrated a PPA of 96.30% and NPA of 98.60% (κ = 0.949, P < 0.001). The Wondfo antigen assay showed a PPA of 81.48% and NPA of 100% (κ = 0.865, P < 0.001). The Pluslife assay demonstrates comparable diagnostic accuracy to EasyNAT for SARS-CoV-2 detection, while offering advantages in portability, speed, and cost. Its performance significantly exceeded that of the Wondfo antigen assay, particularly in reducing false negatives. The Pluslife Mini Dock platform shows potential for rapid SARS-CoV-2 screening in diverse settings.
IMPORTANCEThe newly developed Pluslife Rapid severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) assay is a next-generation molecular point-of-care test ing (POCT) method, based on RNase Hybridization-Assisted amplification (RHAM), a self-developed isothermal amplification molecular detection technology of Pluslife Biotech. The purpose of this study is to evaluate the analytical and clinical performance of this new SARS-CoV-2 POCT method and compare the clinical performance with the EasyNAT Rapid SARS-CoV-2 assay and the Wondfo 2019-nCoV antigen assay. The results of this study are of great significance to clinical practice.
and other processes (3). Therefore, RT-qPCR is not suitable for point-of-care testing (POCT) due to its time-consuming nature.
In order to achieve rapid on-site screening of SARS-CoV-2, a variety of technologies based on nucleic acid isothermal amplification have emerged, including loop-mediated amplification (LAMP) (4), cross-priming amplification (CPA) (5,6), rolling circle amplifi cation (7,8), and recombinase polymerase amplification (RPA) (9). These thermostatic amplification techniques do not involve denaturation, annealing, or other steps and are fast and easy to operate. They are a good choice for clinical POCT implementa tion. In addition, SARS-CoV-2 testing POCT products based on isothermal amplification technology combined with other technologies have emerged (10)(11)(12). The Pluslife Rapid SARS-CoV-2 assay is a new generation of molecular POCT product, utilizing Pluslife's proprietary RNase Hybridization-Assisted amplification (RHAM) technology (13,14). The Pluslife Rapid SARS-CoV-2 assay integrates LAMP and RNase HII reporter signal visualiza tion technology and has some advantages, such as low cost and fast results, compared with the traditional POCT products.
The EasyNAT Rapid SARS-CoV-2 assay is a newly developed CPA kit by Ustar Bio technologies designed to be used with Ustar's nucleic acid amplification and detec tion analyzer. The EasyNAT Rapid SARS-CoV-2 assay integrates RNA extraction, RNA purification, and target gene amplification within a single cartridge. It demonstrates high sensitivity and specificity and has been validated for clinical rapid screening of SARS-CoV-2 (15). But the Ustar's nucleic acid amplification and detection analyzer is neither compact nor portable, limiting its suitability for resource-limited or grassroots settings. The Wondfo 2019-nCoV antigen assay, based on a colloidal gold immunochro matography method, is used to qualitatively detect 2019-nCoV antigen in nasopharyng eal swabs (16). The Wondfo 2019-nCoV antigen assay is simple and fast to operate, making it suitable for home self-tests and large-scale field screening. However, because it detects only surface proteins of the virus, the assay is susceptible to producing false-negative results (17). As a next-generation molecular POCT platform, the Pluslife Mini Dock provides a new molecular detection solution for primary medical institutions and patients with its compact size, rapid processing speed, high accuracy, and costeffec tiveness. Currently, there are no publicly available data comparing the Pluslife Rapid SARS-CoV-2 assay, the EasyNAT Rapid SARS-CoV-2 assay, and the Wondfo 2019-nCoV antigen assay. The objective of this study is to determine the performance characteristics of the Pluslife Rapid SARS-CoV-2 assay and evaluate the consistency of clinical diagnosis with the RT-qPCR. Furthermore, the clinical utility of the Pluslife SARS-CoV-2 assay is evaluated by comparing its performance with the EasyNAT Rapid SARS-CoV-2 assay and the Wondfo 2019-nCoV antigen assay.
## MATERIALS AND METHODS
## The Pluslife Rapid SARS-CoV-2 assay used in this study
The Pluslife Rapid SARS-CoV-2 Card (Pluslife Biotech, Guangzhou, China) was performed on the integrated nucleic acid testing device Pluslife Mini Dock (Pluslife Biotech, Guangzhou, China). The Pluslife Rapid SARS-CoV-2 assay is based on an isother mal amplification method and enzyme-digested probe technology, enabling specific detection of the SARS-CoV-2 N gene and ORF1ab gene. During amplification in the isothermal system, a large number of target gene sequence copies are generated. When the fluorescencelabeled probe hybridizes to the complementary sequences, it is cleaved, resulting in the emission of fluorescent signals. The Pluslife Mini Dock detects and analyzes fluorescence signals, automatically reporting positive, negative, or invalid results.
In this study, the Pluslife Rapid SARS-CoV-2 assay was conducted following the manufacturer's instructions. First, the disposable sampling swab was inserted into the nucleic acid-releasing agent to release SARS-CoV-2 nucleic acid. Then, 400-450 µL of SARS-CoV-2 nucleic acid release solution was dropped into the SARS-CoV-2 Reaction Card sample tube. After securely screwing the cap of the SARS-CoV-2 Reaction Card sample tube, the air button on top of the card cap was pressed to dissolve the beads in the reaction chip, followed by shaking the card vertically 10 times (within approx imately 5 sec). Subsequently, the card was inserted into the integrated nucleic acid testing device, which initiated the testing protocol. The detection process typically took approximately 15-35 min.
## Analytical performance evaluation of the Pluslife Rapid SARS-CoV-2 assay
The limit of detection (LOD) of the Pluslife Rapid SARS-CoV-2 assay was determined using the SARS-CoV-2 nucleic acid quantitative detection reference (BDS Biotechnolo gies, Guangzhou, China) for absolute quantification. Seven concentrations (800, 600, 500, 400, 300, 200, and 100 copies/mL) were obtained by diluting the SARS-CoV-2 reference (1.0×10 10 copies/mL). Each concentration was tested in 20 replicates. The LOD was defined as the lowest concentration achieving ≥95% positive detection of SARS-CoV-2 target genes (18).
The cross-reactivity profile of the Pluslife Rapid SARS-CoV-2 assay was systematically evaluated against common respiratory pathogens, including HCoV-229E, HCoV-OC43, Influenza A Virus (IAV), Influenza B Virus (IBV), Epstein-Barr Virus (EBV), Respiratory Syncytial Virus (RSV), Human Metapneumovirus (HMPV), Human Parainfluenza Viruses (HPIVs), and Human Cytomegalovirus (HCMV). Sequenceconfirmed nucleic acid-positive clinical specimens for HCoV-229E, HCoV-OC43, IAV, IBV, EBV, RSV, HMPV, HPIVs, and HCMV were obtained from our laboratory archives. Each viral specimen underwent replicate testing, with consistently negative results interpreted as demonstrating the assay's specificity.
Interference testing for the Pluslife Rapid SARS-CoV-2 assay was assessed in the presence of a low SARS-CoV-2 concentration. Exogenous interfering substances, including oseltamivir (1.875 mg/L), ceftriaxone (1.125 g/L), ribavirin (16.5 mg/L), meropenem (235 mg/mL), tobramycin (30 mg/L), and endogenous interfering substan ces such as human blood (0.16% vol/vol) were evaluated. All test results were positive, indicating no interference from any of the tested substances.
## Evaluation of the clinical performance of the Pluslife Rapid SARS-CoV-2 assay
## Clinical samples
This study analyzed retrospectively collected nasopharyngeal swab samples from 197 patients suspected of COVID-19 infection in the outpatient and emergency department of the First Affiliated Hospital of Sun Yat-Sen University between September 2023 and November 2023. These samples were stored at -80℃ for subsequent analysis and subjected to SARS-CoV-2 testing using a commercial RT-qPCR kit (DAAN GENE, Guangz hou, China). RT-qPCR served as the gold standard for SARS-CoV-2 detection. Positive and negative samples were first identified based on RT-qPCR results. Each sample was then tested using three rapid SARS-CoV-2 assays: the Pluslife Rapid SARS-CoV-2 assay, the EasyNAT Rapid SARS-CoV-2 assay (Ustar Biotechnologies, Hangzhou, China), and the Wondfo Novel Coronavirus (2019-nCoV) antigen assay (Wondfo Biotech, Guangzhou, China). The clinical performance of these assays was compared.
This study utilized residual biological specimens obtained from clinical settings. Patients with medical orders for COVID-19 testing were invited to participate, and verbal informed consent was obtained from all participants. The study protocol was approved by the Medical Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University.
## SARS-CoV-2 detection
All assays were performed in accordance with the manufacturer's instructions. For the RT-qPCR assay serving as the gold standard, commercial kits (DAAN GENE, Guangzhou, China) for SARS-CoV-2 detection were employed. The entire process included nucleic acid extraction, reverse transcription of RNA into complementary DNA, and real-time quantitative PCR amplification using the ABI7500 system (Applied Biosystems, CA, USA). Two hundred microliters of the UTM samples were processed for nucleic acid extraction using the Nucleic Acid Isolation or Purification Kit (DAAN GENE, Guangzhou, China). 5 µL of the nucleic acid elution was used for each RT-PCR reaction. A test result was defined as positive if the cycle threshold (Ct) value for either the ORF1ab or N gene was ≤40.
The EasyNAT Rapid SARS-CoV-2 assay was conducted using Ustar's nucleic acid amplification and detection analyzer (Ustar Biotechnologies, Hangzhou, China). A 1 mL aliquot of vortex-mixed 2019-nCoV-RNA extraction solution (Ustar Biotechnologies, Hangzhou, China) and 0.5 mL of the sample were added into the sample chamber. The cartridge lid was sealed, and the cartridge was gently shaken to ensure thorough mixing of the solution and sample. Finally, the cartridge was then inserted into the detection analyzer for analysis.
The Wondfo 2019-nCoV antigen assay was performed using the Wondfo antigen test card (Wondfo Biotech, Guangzhou, China). The nasopharyngeal swab sample was inserted into the sample extraction solution and rotated 10 times to ensure optimal dissolution. Then 3-4 drops (80 µL) of the mixture were added to the sample well of the test card. Results were interpreted within a 15-20 min period.
## Statistical analysis
The test results of the Pluslife Rapid SARS-CoV-2 assay, the EasyNAT Rapid SARS-CoV-2 assay, and the Wondfo Novel Coronavirus (2019-nCoV) antigen assay were compared to those of RT-qPCR using positive percentage agreement (PPA), negative percentage agreement (NPA), and overall rate of agreement (ORA). In addition, Cohen's kappa (ĸ) was calculated to assess the overall consistency between each assay and RT-qPCR. Confidence intervals (CIs) were computed using the Wilson score method. Statistical analyses were performed using IBM SPSS Statistics 25. P < 0.05 was considered statisti cally significant.
## RESULTS
## Analytical performance of the Pluslife assay
To determine the LOD of Pluslife Rapid SARS-CoV-2 assay, serial dilutions of the standard were prepared at varying concentrations and tested according to the manufacturer's instructions. The LOD was confirmed as 400 copies/mL (Table 1). To further assess cross-reactivity, the assay was evaluated against other respiratory viruses, including HCoV-229E, HCoV-OC43, IAV, IBV, EBV, RSV, HMPV, HPIVs, and HCMV. The results showed no cross-reactivity with these respiratory viruses, indicating the high specificity for SARS-CoV-2 (Table 2). Interfering substances at concentrations up to those listed in Table 3 showed no interference with the assay performance.
## Clinical performance of three Rapid SARS-CoV-2 assays
A total of 197 nasopharyngeal swab samples were tested for SARS-CoV-2 by RT-qPCR and three rapid SARS-CoV-2 assays: the Pluslife Rapid SARS-CoV-2 assay, the EasyNAT Rapid SARS-CoV-2 assay, and the Wondfo 2019-nCoV antigen assay. Comprehensive Compared with RT-qPCR, the Pluslife Rapid SARS-CoV-2 assay demonstrated a PPA of 96.30% and an NPA of 99.30%. The ORA was 98.48%, and the two assays demonstrated a high level of consistency (ĸ = 0.962, P < 0.001) (Table 4). The Pluslife assay achieved a 100% positive detection rate for samples with Ct values <30, and a rate >93% for samples with Ct values of 30-40 (Table 5). Three samples exhibited discordant results: two RT-qPCR positive samples tested negative by Pluslife (with Ct values for the N gene or ORF1ab gene exceeding 35), and one RT-qPCR negative sample tested positive by Pluslife (Table 6).
Similarly, compared with RT-qPCR, the PPA and NPA of the EasyNAT Rapid SARS-CoV-2 assay were 96.30% and 98.60%. The ORA was 97.97%, and the results were highly consistent (k = 0.949, P < 0.001) (Table 4). The EasyNAT assay achieved a 100% detection rate for samples with Ct values <30, 93.10% for samples with Ct values of 30-35, and 100% for samples with Ct values ≥35 (Table 5). Four samples exhibited discordant results: two RT-qPCR-positive samples tested negative by EasyNAT (with Ct values for the N gene or ORF1ab gene >35), and two RT-qPCR-negative samples tested positive by EasyNAT (Table 6).
For the Wondfo 2019-nCoV antigen assay, comparison with RT-qPCR showed a PPA of 81.48% and an NPA of 100.00%. The ORA value was 94.92%, showing good consistency (ĸ = 0.865, P < 0.001) (Table 4). The detection rate of the Wondfo assay was 100.00% for Ct values <25; 93.75% for Ct values of 25-30; 86.21% for Ct values of 30-35; and 28.57% for Ct values ≥35 (Table 5). Ten samples showed discordant results between the Wondfo assay and RT-qPCR, with Ct values exceeding 33 for nine of these samples (Table 6).
## Comparison of the three Rapid SARS-CoV-2 assays
Clinical performance comparison of the three rapid SARS-CoV-2 assays on the 54 RT-qPCR-positive samples is detailed in Table 7. Among these positives, the three methods yielded positive results in 52 (96.30%), 52 (96.30%), and 44 (81.48%) cases, respectively. The results showed significant discrepancies (χ 2 = 10.008, P < 0.05). The positive rate of the Pluslife assay matched that of the EasyNAT assay. Pairwise comparison revealed statistically significant disparities in detection rates between the molecular assays (Pluslife and EasyNAT) and the antigen-based assay (Wondfo), with both molecular assays exhibiting significantly higher positive rates.
## DISCUSSION
Coronavirus disease 2019 (COVID-19) remains one of the leading causes of epidemic respiratory infections. Currently, the most widely accepted diagnostic test for SARS-CoV-2 is still based on the gold standard method RT-qPCR (19). However, laboratory RT-qPCR testing remains inaccessible in lots of countries and cannot be widely implemented in resource-limited settings (20). Although rapid antigen tests offer convenience, their sensitivity is inferior to molecular assays (21,22). In recent years, emerging isothermal amplification technologies (eg., CPA, RPA) have enabled portable molecular diagnostics, yet most commercial systems remain cost-prohibitive for point-of-care use. This study evaluated the performance of the novel Pluslife Rapid SARS-CoV-2 assay, a palm-sized POCT platform, using RT-qPCR as the reference method, with comparative analysis against established assays: the EasyNAT Rapid SARS-CoV-2 molecular POCT and the Wondfo 2019-nCoV GICA.
The Pluslife Rapid SARS-CoV-2 assay exhibited a low LOD of 400 copies/mL. It showed no cross-reactivity with other common respiratory viruses and no interference from tested substances, demonstrating a good diagnostic performance. For clinical samples, there was a good consistency (ĸ = 0.962, P <0.001) between the Pluslife Rapid SARS-CoV-2 assay and RT-qPCR. Using the Pluslife Rapid SARS-CoV-2 assay, the positive samples can be determined as fast as 7 min, while negative samples can be confirmed in about 35 min (Table 8). Its rapid, accurate, user-friendly, and costeffective features align with POCT criteria (23,24), making it suitable for primary care and resource-limited settings. In addition, this study represents the first comparative analysis of the Pluslife Rapid SARS-CoV-2 assay, the EasyNAT Rapid SARS-CoV-2 assay, and the Wondfo 2019-nCoV antigen assay. Using RT-qPCR as the gold standard, the PPA of the Pluslife assay achieved 96.30%, which was in line with 96.30% of the EasyNAT assay and superior to 81.48% of the Wondfo assay. For NPA, the Pluslife assay demonstrated an accuracy of 99.30%, surpassing the EasyNAT assay (98.60%). It is worth mentioning that among the 197 clinical samples, the ORA of the Pluslife assay was 98.48%, exceeding both the EasyNAT assay (97.97%) and the Wondfo assay (94.92%), indicating superior clinical reliability. For RT-qPCR positive samples, statistical analysis revealed significant differences among the three assays (χ 2 = 10.008, P < 0.05). Pairwise comparisons revealed significant discrepan cies between both molecular assays (Pluslife, EasyNAT) and the Wondfo antigen assay.
Notably, while the Pluslife and EasyNAT assays demonstrated equivalent efficacy in detecting positive samples, the EasyNAT assay requires an expensive benchtop instrument and sophisticated laboratory infrastructure. In contrast, the Pluslife Mini Dock is extensively smaller and more economical, with faster results (positives: ~7 min vs 40 min). Moreover, the Pluslife assay demonstrates clear economic superiority in per-device and per-test costs (Table 8), making it more portable, faster, and costeffec tive. Besides, the difference between the Pluslife assay and the Wondfo assay was statistically significant. The PPA of the Pluslife assay was about 15% higher than that of the Wondfo assay. When Ct values for the N gene or ORF1ab gene exceeded 35, the Wondfo assay was more likely to miss the positive samples.
The emergence of molecular POCT plays a critical role in the rapid diagnosis and differentiation of respiratory diseases (25)(26)(27). Between the two rapid SARS-CoV-2 molecular POCT assays compared in this study, the Pluslife assay showed better performance than the EasyNAT assay, featuring good detection performance, compact size, low cost, simple operation, and rapid result delivery. Moreover, unlike the Wondfo assay, which detects viral surface proteins (28,29), the Pluslife Mini Dock amplifies SARS-CoV-2 RNA exponentially, significantly reducing the likelihood of false-negative results within minutes.
There are several limitations of this study. First, RT-qPCR results-rather than clinical disease diagnosis-were used as the gold standard. RT-qPCR reliability may be influenced by incomplete clinical follow-up data. RT-qPCR results might not accurately reflect the SARS-CoV-2 carrying status of samples (30). PPA and NPA are inherently better suited for consistency assessment between two diagnostic methods. Therefore, instead of sensitivity and specificity, the PPA and NPA were used to characterize the clinical performance of the rapid SARS-CoV-2 assay compared with RT-qPCR (25). Additionally, viral genome sequencing was not performed, precluding assessment of variant impacts. Moreover, the sample size was small, and most of the Ct values for positive samples exceeded 35. This may explain why the PPA value was not as high as the 99% PPA value provided by Laura Herrmann (31). Finally, clinical samples were not tested by all four methods concurrently, and SARS-CoV-2 in samples stored at -80℃ may have degraded during repeated freeze-thaw processes.
In summary, the Pluslife assay shows good consistency with RT-qPCR. Compared with the EasyNAT assay, the Pluslife assay exhibits the same excellent detection efficiency of SARS-CoV-2 positive samples, with smaller size, lower cost, and faster detection speed. Compared with the Wondfo assay, the clinical performance of the Pluslife assay is more prominent, with a lower false-negative rate. The Pluslife Mini Dock's compact design, portability, simple operation, rapid detection, capacity, and costeffectiveness support its use for SARS-CoV-2 screening in resource-limited settings.
## References
1. Wu, Zhao, Yu et al. (2020) "A new coronavirus associated with human respiratory disease in China" *Nature*
2. Bwire, Majigo, Njiro et al. (2021) "Detection profile of SARS-CoV-2 using RT-PCR in different types of clinical specimens: a systematic review and meta-analysis" *J Med Virol*
3. Vila Muntadas, Sunyer, Garcia-Navarro (2021) "COVID-19 diagnostic tests: importance of the clinical context" *Med Clin (Engl Ed)*
4. Zhang, Chen, Liu et al. (2021) "Sensitive and rapid on-site detection of SARS-CoV-2 using a gold nanoparticle-based high-throughput platform coupled with CRISPR/Cas12-assisted RT-LAMP" *Sens Actuators B Chem*
5. Wu, Shi, Chen et al. (2021) "Comparative evaluation of six nucleic acid amplification kits for SARS-CoV-2 RNA detection" *Ann Clin Microbiol Antimicrob*
6. Lu, Huang, Dong et al. (2022) "Automatic system for high-throughput and high-sensitivity diagnosis of SARS-CoV-2" *Bioprocess Biosyst Eng*
7. Chaibun, Puenpa, Ngamdee et al. (2021) "Rapid electrochemical detection of coronavirus SARS-CoV-2" *Nat Commun*
8. Zhu, Guo, Wang et al. (2023) "An ultra-sensitive one-pot RNA-templated DNA ligation rolling circle amplificationassisted CRISPR/Cas12a detector assay for rapid detection of SARS-CoV-2" *Biosensors and Bioelectronics*
9. Lu, Tong, Han et al. (2022) "Fast and sensitive detection of SARS-CoV-2 RNA using suboptimal protospacer adjacent motifs for Cas12a" *Nat Biomed Eng*
10. Zingg, Yang, Seely et al. (2023) "Rapid isothermal point-of-care test for screening of SARS-CoV-2 (COVID-19)" *Asp Mol Med*
11. Zhang, Zhao, Chen et al. (2022) "Fully automated CRISPR-LAMP platform for SARS-CoV-2 delta and omicron variants" *Anal Chem*
12. Srivastava, Prasad (2023) "Isothermal nucleic acid amplification and its uses in modern diagnostic technologies"
13. Xiao, Liu, Kang et al. "Rapid and accurate detection of SARS-CoV-2 using the RHAM technology" *Sci Rep*
14. Dobosy, Rose, Beltz et al. (2011) "RNase H-dependent PCR (rhPCR): improved specificity and single nucleotide polymorphism detection using blocked cleavable primers" *BMC Biotechnol*
15. Fernández-Sánchez, Martín-Bautista, Rivas-Ruiz et al. (2024) "Evaluation of the EasyNAT SARS-CoV-2 assay PCR test for the diagnosis of SARS-CoV-2 infection" *J Virol Methods*
16. Wang, Lu, Li et al. (2023) "Application of ultrasensitive assay for SARS-CoV-2 antigen in nasophar ynx in the management of COVID-19 patients with comorbidities during the peak of 2022 Shanghai epidemics in a tertiary hospital" *Clinical Chemistry and Laboratory Medicine (CCLM)*
17. Kevadiya, Machhi, Herskovitz et al. (2021) "Diagnostics for SARS-CoV-2 infections"
18. Li, Jiao, Wang et al. (2021) "Key points of technical review for the registration of SARS-Cov-2 nucleic acid tests in China" *Bioanalysis*
19. Rai, Kumar, Deekshit et al. (2021) "Detection technologies and recent developments in the diagnosis of COVID-19 infection" *Appl Microbiol Biotechnol*
20. Eibinger, Kessler, Stelzl et al. (2022) "SARS-CoV-2 RNA testing using different assaysimpact on testing strategies in a clinical setting" *Int J Mol Sci*
21. Yüce, Filiztekin, Özkaya (2021) "COVID-19 diagnosis -a review of current methods" *Biosens Bioelectron*
22. Chaimayo, Kaewnaphan, Tanlieng et al. (2020) "Rapid SARS-CoV-2 antigen detection assay in comparison with real-time RT-PCR assay for laboratory diagnosis of COVID-19 in Thailand" *Virol J*
23. Kosack, Page, Klatser (2017) "A guide to aid the selection of diagnostic tests" *Bull World Health Organ*
24. Mitra, Sharma (2021) "POCT in developing countries" *EJIFCC*
25. Brendish, Schiff, Clark (2015) "Point-of-care testing for respiratory viruses in adults: the current landscape and future potential" *J Infect*
26. Zhang, Ma, Ahmed et al. (2022) "Advanced point-of-care testing technologies for human acute respiratory virus detection" *Adv Mater Weinheim*
27. Teixeira, Pallás-Tamarit, Juste-Dolz et al. (2022) "An all-in-one point-of-care testing device for multiplexed detection of respiratory infections" *Biosens Bioelectron*
28. Aguilar-Shea, Vera-García, Güerri-Fernández (2021) "Rapid antigen tests for the detection of SARS-CoV-2: A narrative review" *Aten Primaria*
29. Weitzel, Pérez, Tapia et al. (2021) "SARS-CoV-2 rapid antigen detection tests" *Lancet Infect Dis*
30. Tali, Leblanc, Sadiq et al. (2021) "Tools and techniques for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)/COVID-19 detection" *Clin Microbiol Rev*
31. Herrmann, Breuer, Duc et al. (2024) "Comparison of the diagnostic accuracy of the pluslife mini dock RHAM technology with abbott ID now and cepheid genXpert: a retrospective evaluation study" *Sci Rep* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12757585&blobtype=pdf | # Contemporary Highly Pathogenic Avian Influenza (H5N1) Viruses Retain Neurotropism in Human Cerebral Organoids
Kerry Goldin, Meaghan Flagg, Tessa Lutterman, Bridget Brackney, Katie Williams, Cathryn Haigh, Emmie De Wit
## Abstract
Background. Clade 2.3.4.4b highly pathogenic avian influenza (HPAI) H5N1 viruses are widely circulating in North America with unprecedented transmission into novel host species. A high incidence of neurologic disease is observed in carnivores infected with clade 2.3.4.4b HPAI H5N1 viruses, and historical outbreaks of HPAI H5N1 in humans are also associated with neurologic complications, raising concerns about neurotropism and neurovirulence of clade 2.3.4.4b HPAI H5N1 viruses.Methods. We analyzed virus replication kinetics, cellular tropism, and host responses to infection in human cerebral organoids (hCOs) inoculated with clade 2.3.4.4b HPAI H5N1 viruses compared to a historical clade 1 HPAI H5N1 virus and a 2007 seasonal influenza A virus.Results. HPAI H5N1 viruses replicated to high titers in hCOs, but replication of the seasonal influenza A virus was not detected. Viral antigen and RNA were detected primarily in neuron-and astrocyte-like cells. Interferon responses to infection with HPAI H5N1 viruses were observed in a small population of bystander cells. Higher levels of cell death and proinflammatory cytokines and chemokines were observed in organoids inoculated with the historical HPAI H5N1 isolate.Conclusions. Clade 2.3.4.4b HPAI H5N1 viruses exhibit similar neurotropism compared to a historical clade 1 HPAI H5N1 virus. Lower levels of cell death and inflammatory cytokine production induced by clade 2.3.4.4b viruses may indicate reduced neuropathogenic potential of these viruses in humans.
Influenza A viruses (IAVs) can cause severe respiratory and extra-respiratory disease. The most common extra-respiratory complication occurs in the central nervous system [1]. Neurologic manifestations have been documented in cases of infection by highly pathogenic avian influenza (HPAI) H5N1 [2] and seasonal IAV [3] and are most common in children [1,3]. Neurological syndromes associated with IAV can be broadly classified into influenza-associated encephalopathy or encephalitis and post-influenza syndromes, including autoimmune and neurodegenerative conditions [1]. Human autopsy studies of IAV infections with thorough characterization of the brain are severely limited. For HPAI H5N1 cases, the few publications that examine the brain have reported IAV antigen and RNA in neurons and glia [4][5][6]. In autopsy studies performed in cases of fatal H1N1 infection, even individuals who exhibited neurologic signs, virus was typically not found in cerebral spinal fluid or brain tissue [4,7,8]. Unlike H5N1, where neuropathology coincides with virus infection in the brain, H1N1 appears to cause neuropathology indirectly as the result of the immune response to infection elsewhere [7].
In North America, spillover events of HPAI H5N1 viruses of clade 2.3.4.4b into mammalian species have occurred frequently in the past few years, particularly in carnivores [9][10][11][12]. In March 2024, HPAI H5N1 clade 2.3.4.4b was found in dairy cattle in the US dairies. Subsequently, the virus spread rapidly throughout dairies in the United States, and transmission events between cattle, between cattle and cats [13], and between cattle and humans [14] have been documented. This mammal-to-mammal transmission is concerning, as it may indicate adaption to mammalian hosts. Interestingly, encephalitis is a common finding in carnivores infected with this clade of H5N1 [12,13,15]. The increase in mammal-to-mammal transmission and the potential neurotropism warrants investigation into the neurotropic potential of contemporary HPAI H5N1 viruses in humans.
Neurologic manifestations have been observed in HPAIinoculated laboratory animals, including mice [16] and ferrets [17]. Importantly, the frequency of these manifestations is greatly influenced by mouse strain [18,19] and inoculation route [20]. This can make direct assessment of neurotropism and neuropathology between laboratory animal models and humans difficult. Human cerebral organoids (hCOs) are a sophisticated 3-dimensional model of the developing human brain derived from induced pluripotent stem cells [21,22]. Human cerebral organoids contain multiple neural cell types including neurons, glial cells, and progenitors.
Here, we compared the neurotropism and neuropathogenic potential of 2 contemporary clade 2.3.4.4b HPAI H5N1 viruses, 1 historical HPAI H5N1 virus and 1 seasonal influenza A H1N1 virus in hCOs. We found that the seasonal H1N1 virus was unable to infect hCOs, unlike the 3 HPAI H5N1 isolates. The 2 contemporary HPAI H5N1 isolates induced less cytotoxicity and proinflammatory cytokine production than the historical HPAI H5N1 isolate, suggesting that while the contemporary viruses retain neurotropism, they may be less neuropathogenic.
## METHODS
## Biosafety
The Institutional Biosafety Committee approved work with HPAI H5N1 virus at Biosafety Level 3 conditions. Sample inactivation was conducted according to IBC-approved standard operating procedures for removal of specimens from high containment.
## Virus and Cells
The following wild-type viruses were used: clade 2.3.4.4b, genotype B3.13 HPAI H5N1 isolate A/Bovine/Ohio/B24osu-342/ 2024 (EPI_ISL_678615; hereafter bOH2024); clade 1 HPAI H5N1 isolate A/Vietnam/1203/2004 (NCBI taxid:284218; hereafter VN2004); clade 2.3.4.4b, genotype B3.6 HPAI H5N1 isolate A/mountain lion/Montana/2/2024 (EPI_ISL_19880418; hereafter mMT2024), isolated from brain tissue from a deceased mountain lion, and H1N1 isolate A/Brisbane/59/07 (NCBI taxid:504904; hereafter BR2007). All viruses were propagated once in Madin-Darby canine kidney (MDCK) cells in infection media: minimum essential media (MEM) supplemented with 1 mM L-glutamine, 50 U/mL penicillin, 50 μg/ mL streptomycin, 1 × NEAA, 20 mM HEPES, and 4 µg/mL TPCK trypsin. Madin-Darby canine kidney cells were maintained in MEM supplemented with 10% fetal bovine serum, 1 mM L-glutamine, 50 U/mL penicillin, and 50 μg/mL streptomycin, 1 × NEAA, 20 mM HEPES. Mycoplasma testing is performed monthly; cells remained negative throughout the study. Virus stocks were sequenced and found not to contain SNPs compared to reference sequences.
## Cerebral Organoid Generation, Culture, and Infection
Generation and maintenance of hCOs were performed as described previously [23]. The hCOs were derived from an induced human pluripotent stem cell line (ATCC-1023). The human samples used in this study were obtained from a commercial source (ATCC, #KYOU-DXR0109B) with no identifying donor information. Thus, the NIH Office of Human Subjects Research Protections determined these samples to be exempt from NIH Institutional Review Board review. The hCOs used in this experiment were 6-8-months-old.
For each virus, 2 organoids were incubated with 1000 TCID50 per well in a 6-well plate in STEMCELL Technologies cerebral organoid basal media on an orbital shaker set to 65 rpm at 37°C and 5% CO 2 . Accurate cell counts could not be determined, as dissociating the organoid would destroy it and an estimate of cell counts based on diameter is likely to be inaccurate due to variably sized necrotic cores. For this reason, 1000 TCID50 was used as the inoculation dose, rather than MOI. Three replicate wells were analyzed. After 1 hour, the inoculum was removed, and organoids were washed by adding fresh media (STEMCELL Technologies cerebral organoid basal media) and being placed on the orbital shaker in the incubator for 10 minutes. After 10 minutes, media was removed and replaced with 4 mL media (STEMCELL Technologies cerebral organoid basal media + supplement E) and organoids were placed in the incubator, on the orbital shaker set to 65 rpm, at 37°C and 5% CO 2 .
## Virus Titrations
Endpoint titration was performed in MDCK cells seeded in 96-well plates. Samples were serially diluted 10-fold in infection media. Cells were washed 2× with PBS prior to the addition of culture supernatant. Cells were incubated at 37°C and 5% CO 2 . After 3 days, hemagglutination assays were performed to determine the presence of infectious IAV. Titers were calculated using the Spearman-Karber method.
## Cytotoxicity assay
For each inoculation group, culture supernatant was collected daily from 3 replicate wells. Lactate dehydrogenase (LDH) release into supernatant was measured using the LDH-Glo Cytotoxicity Assay (Promega) according to the manufacturer's instructions. Lactate dehydrogenase standards were run in parallel to calculate concentration.
## Immunofluorescence and Immunohistochemistry
Detailed description of hCO immunofluorescence and immunohistochemistry is available in the Supplementary Data.
## Single-Cell RNA Sequencing
Detailed description of hCO single-cell RNA sequencing (scRNA-seq) library preparation and analysis is available in the Supplementary Data.
## Cytokine Analysis
The Luminex Discovery Assay Human Premixed Multi-Analyte Kit (LXSAHM) was used to quantify the cytokine and chemokine response for 17 analytes: CXCL2, IFN-alpha, IFN-gamma, IFN-beta, IL-1 alpha, IL-1 beta, IL-4, IL-8, M-CSF, MMP-12, VEGF-A, CXCL10, IL-2, IL-6, IL-10, MMP-9, and TNF-alpha. Samples were prepared according to manufacturer's instructions. Plates were read within 1 hour using Bioplex Luminex 200 instrument (Bio-Rad).
## Statistical Analyses
Statistical analysis was performed in GraphPad Prism (v10) or python. Luminex data were analyzed using Bioplex (Bio-Rad) software. P-values < .05 were considered statistically significant.
## Data Availability
Data included in this manuscript have been deposited in Figshare at 10.6084/m9.figshare.29275193; scRNA-seq data are available in NCBI GEO under accession number GSE301348.
## RESULTS
## HPAI H5N1 Rapidly Replicates and Causes Cytotoxicity in hCOs
We compared virus replication and cytotoxicity of 3 HPAI H5N1 isolates and 1 H1N1 seasonal IAV isolate in the hCOs model. The 3 HPAI H5N1 isolates, VN2004, bOH2024, and mMT2024, rapidly replicated and released infectious virus into the culture supernatant as early as 1 day postinoculation (dpi) (Figure 1A). bOH2024-inoculated organoids continued to produce infectious virus throughout the experiment; however, VN2004 and mMT2024 virus production decreased over time, starting at 3 dpi. No evidence of virus replication was observed for seasonal H1N1 virus BR2007. All H5N1 isolates produced similar quantities of infectious virus over the course of the study (Figure 1B). Cytotoxicity was determined using a commercially available LDH assay. Elevated cytotoxicity was observed in organoids inoculated with HPAI H5N1 viruses (Figure 1C). Lactate dehydrogenase elevations in bOH2024and mMT2024-inoculated organoids were less rapid and peaked at lower levels compared to VN2004. Minimal cytotoxicity was observed in BR2007-inoculated organoids. Infection of hCOs by the HPAI H5N1 isolates was confirmed by immunohistochemistry at 7 dpi (Figure 1D). Abundant IAV nucleoprotein (NP) antigen was observed in the 3 HPAI H5N1-inoculated groups, but was not observed in BR2007or mock-inoculated organoids.
## IAV Antigen in hCOs Is Observed in Areas Immunoreactive for Markers of Neuronal and Astrocyte Cell Types
To confirm the presence of neural-like cell populations within the hCOs and determine the cellular tropism of the HPAI H5N1 virus isolates, serial formalin-fixed paraffin-embedded sections were labeled with canonical markers for neural progenitors, neurons, and astrocytes (Figure 2; Supplementary Fig. 1). All organoids expressed small focal regions of PAX6, the neural progenitor marker (Figure 2). All organoids had moderate to abundant MAP2 expression, representing neuron-like cells (Figure 2). These MAP2-positive cells often formed a dense meshwork of neuropil-like processes. Similarly, all organoids had moderate GFAP expression, representing astrocytelike cells, also seen forming long processes near areas of MAP2 expressing cells (Figure 2).
Additionally, virus distribution throughout the organoids was visualized. While co-localization could not be confirmed definitively, due to labeling being performed on serial sections, the general pattern suggests viral antigen in areas enriched for neuronal and potentially astrocyte-like cells in the organoids inoculated with the 3 HPAI H5N1 isolates (Figure 2), in line with detection of influenza virus antigen in neuronal and glial cells in the CNS of deceased patients [4][5][6]. In line with the virus titration and immunohistochemistry data, BR2007-and mock-inoculated organoids were negative for IAV antigen. Notably, different primary antibodies and antigen retrieval methods were used for virus antigen immunohistochemistry and immunofluorescence, which likely explains the difference in immunoreactivity between the 2 techniques.
Of note, viral antigen and neural markers were not detected in the core of some of the hCOs (Supplementary Fig. 1), likely due to the presence of necrotic cores, which are a common artifact in older organoids that are too large for oxygen and nutrients to diffuse to the center [21].
## scRNA-Seq of Infected Cerebral Organoids Reveals Antiviral Response in Bystander Cells
To evaluate host responses to infection, we conducted scRNA-seq on IAV-or mock-inoculated hCOs at 1 and 3 dpi. Reads were aligned to a combined human and IAV reference in order to identify both host and viral transcripts. After batch integration and dimensionality reduction, cells visualized by UMAP were strongly separated according to the presence of viral counts (Figure 3A). Cells derived from samples where productive virus replication was observed (VN2004, bOH2024, mMT2024) clustered separately from uninfected and mock-infected cells, as well as cells derived from samples where virus replication was not observed (BR2007). Within these clusters, samples inoculated with different H5N1 viruses were well-integrated (Figure 3B). We observed a high fraction of counts aligned to IAV transcripts which was most pronounced at 3 dpi, accounting for a large fraction of the transcriptional variation within the dataset (Figure 3C andD). In some clusters of infected cells, viral reads accounted for 50%-80% of total counts (Figure 3E). Higher percentages of viral counts were observed in samples inoculated with contemporary HPAI H5N1 isolates compared to VN2004 (Figure 3E), despite no statistically significant difference in titers of infectious virus between HPAI H5N1-inoculated samples (Figure 1A andB). Viral reads were not detected in BR2007-inoculated hCOs (Figure 3E), in accordance with no detectable virus replication (Figure 1A). Due to the significantly reduced capture of host transcripts, cells with >20% viral counts were excluded from further analysis, and nearest neighbor computation and clustering was repeated.
We utilized unsupervised clustering to identify distinct cell populations in order to characterize virus infection and host responses. Analysis of host gene expression revealed populations of neurons, astrocytes, and progenitors at various stages of neural differentiation (Figure 4A; Supplementary Fig. 2). HPAI H5N1-infected cells, indicated by the presence of >5% viral counts, were found in clusters 3, 4, and 11 (Figure 4B andC). While these clusters were somewhat heterogeneous, they predominantly expressed neuronal cell marker genes (Figure 4D), indicating HPAI H5N1 primarily replicates in neurons.
To evaluate host responses to infection at a functional level, we conducted over-representation analysis (ORA) at the single cell level with hallmark gene sets obtained from the molecular signatures database (MSigDB). To identify signaling pathways perturbed by virus infection, we ranked the pathways in each cluster that were most differentially expressed compared to mock-inoculated cells. We utilized the entire dataset of mockinoculated cells since infected cell clusters contained mixed cellular identity and thus direct comparison of infected versus uninfected cells in the same population was not possible. Additionally, comparison to mock-inoculated cells as opposed to uninfected clusters allowed identification of host responses in bystander cells. The top 5 pathways identified in each cluster are shown in Supplementary Fig. 3. In most clusters, we did not observe major differences in pathway enrichment in HPAI H5N1-inoculated samples that differed from BR2007-or mock-inoculated samples (Supplementary Fig. 3), indicating the lack of a coherent response to infection in the majority of cells. Surprisingly, interferon (IFN) responses were only observed in a minority of cells, primarily in cluster 1 (Figure 4E). The peak IFN response in cluster 1 occurred at 3 dpi and was only detected in samples where productive virus replication was also observed, indicating that the lack of virus replication in BR2007-inoculated organoids was not due to induction of antiviral innate immune responses (Figure 4F; Supplementary Fig. 4). The magnitude of the IFN response in cluster 1 was similar between H5N1 isolates (Figure 4F). However, we observed several clusters where IFN responses were only detected in VN2004-inoculated samples, indicating this isolate may drive proinflammatory responses in a broader range of cells (Supplementary Fig. 4). Cells with high IFN response scores had low percentages of viral counts, indicating that IFN responses were occurring primarily in bystander cells (Figure 4G). This pattern was observed for all H5N1 isolates, suggesting similar innate immune evasion capacities.
## VN2004 Induced Proinflammatory Response in hCOs
Production of proinflammatory cytokines and chemokines and release into culture supernatant was quantified (Figure 5). Cytokine and chemokine release was first observed at 5 dpi, corresponding with the increases in cytotoxicity in the H5N1-inoculated organoids (Figure 1B), suggesting that inflammatory responses may contribute to virus-induced cytotoxicity. VN2004 induced the most robust proinflammatory response in hCOs, with trends toward elevations of IL-8, M-CSF, CXCL10, IFN-beta, IL-1 alpha, and IL-6 (Figure 5). BR2007-inoculated organoids had mild elevations in IL-4 and VEGF-A, although this was primarily driven by a single replicate and therefore difficult to interpret.
## DISCUSSION
Influenza A virus infections in humans are linked to neurologic disease [1], with HPAI H5N1 viruses associated with a higher incidence of neurologic disease compared to seasonal viruses [24]. Contemporary clade 2.3.4.4b HPAI H5N1 viruses have become widespread in North America. Their expansion into cattle and unprecedented ability to transmit between species has raised concerns about the consequences of spillover into humans. Seventy human cases have been reported in the United States thus far, generally resulting in mild disease with no reports of neurologic complications [14,25]. This is in stark contrast to historical HPAI H5N1 outbreaks in humans, where high mortality rates were observed [26], and cases of HPAI H5N1-associated encephalitis were reported in children [2]. In our study, all 3 HPAI H5N1 isolates replicated equally efficient, while the seasonal IAV isolate did not replicate. This result aligns with observations of higher rates of neurologic complications following HPAI compared to seasonal IAV virus infections in humans and laboratory animals. This difference may be explained by several factors. HPAI viruses contain a multibasic cleavage site in HA which has been shown to increase virus replication in neural cells [27]. Additionally, differential sialic acid receptor usage between human-adapted seasonal IAV viruses and HPAI viruses may impact neurotropism [28].
Since the neurologic complications of IAV infection are associated with inflammation [1], we sought to characterize host responses to virus replication in neural cells. By using the hCO model, we are able to characterize the neural response in isolation, without the potential proinflammatory effect of microglia, endothelial cells, or circulating leukocytes. Single-cell RNA sequencing analysis identified an IFN response to infection with HPAI H5N1 viruses but not BR2007, in accordance with the presence of virus replication. While all HPAI H5N1 isolates induced an IFN response, VN2004 induced responses in a broader range of cell populations. Additionally, production of proinflammatory cytokines and chemokines was observed only in organoids inoculated with VN2004. These data suggest that this isolate may drive more inflammation in the CNS, which may be associated with the greater incidence of neurological disease in historical versus contemporary HPAI H5N1 outbreaks in humans. We detected increased levels of IL-6 in the supernatant of organoids inoculated with VN2004. Elevated IL-6 in the serum or cerebral spinal fluid of human patients has been associated with the development of influenza-associated encephalopathy and may be a negative prognostic indicator, highlighting the ability of hCOs to model clinically relevant aspects of human disease. Notably, IFN responses were only detected in a small fraction of sequenced cells, and these cells were either uninfected or had low frequency of viral counts. Combined with the delayed and limited induction of proinflammatory cytokines, this suggests that HPAI H5N1 viruses can efficiently counteract innate immune responses in neural cells. Given the limited proinflammatory responses observed in HPAI H5N1-inoculated hCOs, it is likely that additional cell types play a dominant role in IAV-associated encephalitis and encephalopathy. It is also possible that the limited detection of proinflammatory cytokines is related to the low frequency of cells responding to HPAI H5N1 infection. Analyses of larger number of hCOs could overcome this limitation. The high fraction of viral transcripts (50%-80% of total reads in some cells) in the HPAI H5N1-inoculated organoids at 3 dpi was very striking but in alignment with the widespread detection of viral antigen. As this was the main driver of variation in our dataset, it made analysis of the host response, and even identification of cell type, in these HPAI H5N1-inoculated 3 dpi samples challenging. Influenza A virus can selectively shut off transcription of host genes via endoribonuclease-mediated RNA degradation of host RNA; however, the magnitude of host transcript dropout was unexpected in this study. This may explain the rapid disease progression and lack of inflammation that has been observed in some mouse studies where IAV infection in the CNS was robust, rapid, and not associated with infiltrating inflammatory cells [19].
Another goal of this study was to determine which neural cell types were susceptible to IAV infection and whether clade 2.3.4.4b HPAI H5N1 viruses exhibit altered cellular tropism compared to historical isolates since there are reports of broadened sialic acid receptor usage for these viruses [29]. Our results from immunofluorescence and scRNA-seq suggest that neuronand astrocyte-like cells are infected in the HPAI H5N1inoculated organoids, and neural cellular tropism does not differ between historical and contemporary H5N1 viruses. This is in accordance with previous studies in mice [16,19,30] and ferrets [17] inoculated with HPAI H5N1 virus where viral antigen was present in neurons and astrocytes.
Our study provides valuable information on the neurotropic and neuropathogenic potential of currently circulating HPAI H5N1 viruses by demonstrating similar replicative fitness in neural cells compared to a historical virus with known neurological manifestations, but possible reduced neuropathogenicity indicated by reduced cytotoxicity and proinflammatory cytokine production. The hCO model offers a complex, human-derived system in which to study virus neurotropism that can help maximize translational rigor and improve the predictive power for the development of human neurologic disease.
## References
1. Sellers, Hagan, Hayden et al. (2017) "The hidden burden of influenza: a review of the extra-pulmonary complications of influenza infection" *Influenza Other Respir Viruses*
2. De Jong, Bach, Phan (2005) "Fatal avian influenza A (H5N1) in a child presenting with diarrhea followed by coma" *N Engl J Med*
3. Goenka, Michael, Ledger (2014) "Neurological manifestations of influenza infection in children and adults: results of a national British surveillance study" *Clin Infect Dis*
4. Akins, Belko, Uyeki et al. (2010) "H1n1 encephalitis with malignant edema and review of neurologic complications from influenza" *Neurocrit Care*
5. Gao, Dong, Dong (2010) "A systematic molecular pathology study of a laboratory confirmed H5N1 human case" *PLoS One*
6. Gu, Xie, Gao (2007) "H5n1 infection of the respiratory tract and beyond: a molecular pathology study" *Lancet*
7. Lee, Wong, Chan (2010) "Acute encephalopathy associated with influenza A infection in adults" *Emerg Infect Dis J*
8. Lee, Smith, Rao (2011) "Fatal H1N1-related acute necrotizing encephalopathy in an adult" *Case Rep Crit Care*
9. Jakobek, Berhane, Nadeau (2023) "Influenza A (H5N1) virus infections in 2 free-ranging black bears (Ursus americanus) Quebec" *Canada. Emerg Infect Dis*
10. Elsmo, Wünschmann, Beckmen (2022) "Highly pathogenic avian influenza A (H5N1) virus clade 2.3. 4.4 b infections in wild terrestrial mammals, United States" *Emerg Infect Dis*
11. Cronk, Caserta, Laverack (2023) "Infection and tissue distribution of highly pathogenic avian influenza A type H5N1 (clade 2.3. 4.4 b) in red fox kits (Vulpes vulpes)" *Emerg Microbes Infect*
12. Puryear, Sawatzki, Hill (2023) "Highly pathogenic avian influenza A (H5N1) virus outbreak in new England seals, United States" *Emerg Infect Dis*
13. Burrough, Magstadt, Petersen (2024) "Highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b virus infection in domestic dairy cattle and cats, United States" *Emerg Infect Dis*
14. Uyeki, Milton, Hamid (2024) "Highly pathogenic avian influenza A(H5N1) virus infection in a dairy farm worker" *N Engl J Med*
15. Vreman, Kik, Germeraad (2023) "Zoonotic mutation of highly pathogenic avian influenza H5N1 virus identified in the brain of multiple wild carnivore Species" *Pathogens*
16. Bissel, Giles, Wang et al. (2012) "Acute murine H5N1 influenza A encephalitis" *Brain Pathol*
17. Peng, Yun, Chumakova (2012) "Neuropathology of H5N1 virus infection in ferrets" *Vet Microbiol*
18. Otte, Sauter, Alleva et al. (2011) "Differential host determinants contribute to the pathogenesis of 2009 pandemic H1N1 and human H5N1 influenza A viruses in experimental mouse models" *Am J Pathol*
19. Goldin, Van Tol, Johnson (2025) "Enhanced neurotropism of bovine H5N1 compared to the Vietnam H5N1 isolate in C57BL/6J mice" *Viruses*
20. Eisfeld, Biswas, Guan (2024) "Pathogenicity and transmissibility of bovine H5N1 influenza virus" *Nature*
21. Lancaster, Knoblich (2014) "Generation of cerebral organoids from human pluripotent stem cells" *Nat Protoc*
22. Lancaster, Renner, Martin (2013) "Cerebral organoids model human brain development and microcephaly" *Nature*
23. Williams, Foliaki, Race (2023) "Neural cell engraftment therapy for sporadic Creutzfeldt-Jakob disease restores neuroelectrophysiological parameters in a cerebral organoid model" *Stem Cell Res Ther*
24. Bauer, Benavides (2023) "Veldhuis Kroeze EJB, de WE, van Riel D. The neuropathogenesis of highly pathogenic avian influenza H5Nx viruses in mammalian species including humans" *Trends Neurosci*
25. Morse, Coyle, Mikesell (2024) "Influenza A(H5N1) virus infection in two dairy farm workers in Michigan" *N Engl J Med*
26. Tran, Nguyen, Nguyen (2004) "Avian influenza A (H5N1) in 10 patients in Vietnam" *N Engl J Med*
27. Siegers, Van De Bildt, Lin (2019) "Viral factors important for efficient replication of influenza A viruses in cells of the central nervous system" *J Virol*
28. Zhang, Pu, Sun (2021) "Neurovirulence of avian influenza virus is dependent on the interaction of viral NP protein with host factor FMRP in the murine brain" *J Virol*
29. Good, Fernandez-Quintero (2024) "A single mutation in dairy cowassociated H5N1 viruses increases receptor binding breadth" *Nat Commun*
30. Jang, Boltz, Sturm-Ramirez (2009) "Highly pathogenic H5N1 influenza virus can enter the central nervous system and induce neuroinflammation and neurodegeneration" *Proc Natl Acad Sci U S A* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12841672&blobtype=pdf | # Further Evidence for the Immunosuppressive Activity of Transmembrane Envelope Protein p15E of Porcine Endogenous Retrovirus
Joachim Denner, Reinhard Schwinzer, Claudia Pokoyski, Benedikt Kaufer, Björn Dierkes, Jinzhao Ban, Lovlesh Lovlesh
## Abstract
Retroviruses are immunosuppressive, and there is evidence that a highly conserved immunosuppressive domain (isu domain) in their transmembrane envelope protein contributes to this activity. Studies have shown that inactivated retroviruses, their purified transmembrane envelope proteins, and synthetic peptides corresponding to the isu domain inhibit mitogen-triggered proliferation of peripheral blood mononuclear cells (PBMCs) and modulate their cytokine and gene expression. This has been demonstrated for human immunodeficiency virus type 1 (HIV-1), as well as for beta-and gammaretroviruses and for both exogenous and endogenous retroviruses, including syncytins. In the case of HIV-1, homopolymers of its isu peptide stimulated an increased release of IL-10, IL-6, and other cytokines from human PBMCs. Up-regulated genes included IL-6, IL-8, and IL-10, as well as MMP-1, TREM-1, and IL-1β. In vivo, in a mouse tumor model, tumor cells that were unable to induce tumors in immunocompetent animals gained the ability to do so when expressing the transmembrane envelope protein or the isu domain of various retroviruses on their surface. Here, we demonstrate that the transmembrane envelope protein p15E of PERV can modulate cytokine expression in human PBMCs. Human 293 cells were transfected with four constructs that express a portion of p15E, including the isu domain, and were cultured in the presence of a selection medium containing hygromycin. The p15E-expressing cells were co-cultured with human PBMCs, leading to the release of IL-6 and IL-10 protein and the modulation of multiple cytokines and other markers, including IL-6, IL-10, IFN-α, TNF-α, MMP1, and SEPP1. Similar, but more pronounced, effects were observed when PERV-producing 293 and pig cells were used in parallel; both expressed higher levels of p15E. Additionally, p15E expression reduced MHC class I expression, and preliminary data indicate that p15E expression could have a protective effect against cellular cytotoxicity. This finding underscores the need for further research to elucidate the dynamics of p15E expression and its immunosuppressive activity. It also contributes to the understanding of the immunosuppressive properties of pathogenic retroviruses. Furthermore, expressing the immunosuppressive p15E of PERV on the surface of a pig xenotransplant may reduce the need for pharmaceutical immunosuppressants.
## 1. Introduction
Retroviruses are immunosuppressive. This is not only true for immunodeficiency viruses such as human immunodeficiency virus 1 and 2 (HIV-1 and -2) but also for most other retroviruses, including gammaretroviruses (for a review, see [1,2]). Feline leukemia viruses (FeLVs), murine leukemia viruses (MuLVs) and koala retroviruses (KoRVs)gammaretroviruses closely related to porcine endogenous retroviruses (PERVs)-induce in the infected host not only leukemia and lymphoma but also a severe immunodeficiency which usually precedes tumor development. Immunosuppression without tumor development has also been observed.
The mechanism of how retroviruses induce immunosuppression is not well studied. The observation that many retroviruses cause immunodeficiency suggests the existence of shared mechanisms [1]. In the case of lentiviruses, however, additional viral factors clearly contribute to immune dysfunction. For example, the HIV-1 surface envelope protein gp120 binds CD4 on lymphocytes and perturbs immune signaling [3], and the viral transactivator Tat has also been implicated in immunomodulation [4]. Moreover, the accessory protein Nef from both HIV-1 and SIV contributes to AIDS pathogenesis by impairing the humoral and cellular arms of innate and adaptive immunity [5]. Nef down-regulates CD4 and MHC-1. However, gp41 and gp120 are the first viral proteins that interfere with the immune system during the initial step of HIV-1 infection. Nef and Tat are produced only after infection and may contribute to immunosuppression later.
It has been shown that inactivated retroviruses, their transmembrane envelope proteins, or synthetic peptides corresponding to a highly conserved domain in their transmembrane envelope protein, called the immunosuppressive (isu) domain, inhibit different in vitro activities of immune cells, including down-regulating the Th1-type cytokines (interferon-gamma) and up-regulating the Th2-type cytokines (interleukin 10) [1,2]. It is important to note that Nef of HIV-1 contains a sequence homology with the immunosuppressive domain of gammaretroviruses [6]. The isu domain of the transmembrane envelope protein is highly conserved among all retroviruses [1]. The isu motif is not only highly conserved among retroviruses; it is also ancient. The sequence is found in human syncytin-1 and syncytin-2, which are >25 million and >40 million years old, respectively; it is also found in HEMO, another Env protein coded by an endogenous retrovirus that is >100 million years old, and in the percomorf locus of ray-finned fish, which is also >100 million years old [7][8][9][10]. To investigate the function of the isu domain, several mutants with single amino acid substitutions at conserved positions in the isu domain of MuLV were produced [11]. The majority of these mutants abolished infectivity, indicating that such mutated viruses cannot be investigated in vivo [11,12].
The assays used to demonstrate that the retroviral transmembrane envelope proteins and the corresponding isu peptides are immunosuppressive include mitogen-triggered proliferation of PBMCs, mixed lymphocyte reaction, IL-2-stimulated proliferation of T cells, mitogen-triggered proliferation of B cells, neutrophilic and erythroid cell function, receptor motility on the cell surface of immune cells, measurement of cytokine release, and measurement of cytokine and general gene expression (for a review, see [1,2]). A synthetic peptide corresponding to the isu domain of p15E of gammaretroviruses has been shown to inhibit cytotoxic T cell activity, natural killer cell activity, and B cell activation. This peptide inhibits human IFN gamma production and TNF alpha expression. Furthermore, it activates mitogen-activated protein kinases, induces intracellular cAMP, phosphorylates protein kinase D, and inactivates protein kinase C (for a review, see [13,14]). A synthetic peptide with sequence identity to the isu domain of the transmembrane protein gp41 of HIV-1 has also been shown to inhibit protein kinase C [15].
In studies using the recombinant transmembrane envelope protein gp41 of HIV-1 [16][17][18][19][20], it may be assumed in retrospect that the retroviral materials were contaminated with traces of endotoxin, which is also able to induce IL-10 and other cytokines [21,22].
Most importantly, the retroviral transmembrane envelope proteins are also immunosuppressive in vivo. The most convincing in vivo results came from a tumor rejection assay: expression of different retroviral transmembrane envelope proteins on mouse tumor cells, which did not grow in immunocompetent mice, allowed them to produce tumors in immunocompetent animals by suppression of their immune system. This was shown for p15E of MoMuLV [23], the transmembrane envelope proteins of Mason-Pfizer monkey virus [24], human endogenous retrovirus-H (HERV-H) [25], FeLV [26], and one of two murine and one of two human syncytins [27]. Syncytins are envelope proteins of endogenous retroviruses expressed in the placenta [27]. Experiments deleting parts of the isu sequence of syncytin 2 showed that this domain is the sequence responsible for immunosuppressive activity [27]. Only one murine and one human syncytin were immunosuppressive: human syncytin-2 (HERV-FRD) and mouse syncytin-B. In contrast, human syncytin 1 (HERV-W) and murine syncytin-A were not immunosuppressive. Mutations of relevant amino acids in the isu domain allowed for switching from an immunosuppressive syncytin into a non-immunosuppressive and vice versa, as measured in the mouse tumor model [27].
Furthermore, immunization with the non-immunosuppressive form (wild-type syncytin-1 and mutated syncytin-2) induced immunoglobulin G titers 10-to 30-fold higher than the corresponding immunosuppressive form (mutant syncytin-1 and wild-type syncytin-2) [27]. This indicates that the immunosuppressive activity acts not only locally, on the surface of the tumor cells, but is generalized, also influencing antibody production.
Retrovirus infections modulate the cytokine release in the infected individuals, for example, in AIDS patients [28]. The transmembrane envelope proteins and synthetic peptides corresponding to the immunosuppressive domain have also been shown to modulate cytokine mRNA expression and release in human PBMCs. Using cytokine arrays, it was shown that the transmembrane envelope proteins of HIV-1, PERV, HERV-K, and the corresponding isu peptides induced the release of the following cytokines: IL1-α, IL-10, IL-6, IL-8, monocyte chemoattractant protein (MCP)-1, MCP-2, tumor necrosis factor (TNF-α), macrophage inflammatory protein (MIP)-1α, and MIP-3 [29][30][31]. In contrast, the expression of IL-2 and the chemokine (C-X-C motif) ligand CXCL-9 was decreased. Microarray analysis of the expression of more than 25,000 genes in human PBMCs treated with the homopolymer of the HIV-1 isu peptide or with the recombinant transmembrane envelope protein of HERV-K confirmed the cytokine data and showed up-regulation and down-regulation of more than 300 genes [30,31]. Among the genes with the highest up-regulation were IL-6, matrix metalloproteinase 1 (MMP-1), and triggering receptor expressed on myeloid cells 1 (TREM-1). Among the down-regulated genes were ficolin-1 (FCN1), selenoprotein P, plasma, 1 (SEPP1), triggering receptor expressed on myeloid cells 2 (TREM-2), and CXCL-10; all of these proteins are involved in innate immunity [30,31].
The knowledge of the mechanisms of the immunosuppressive activity of retroviruses may have importance for vaccine development against retroviruses: the mutation of the isu domain significantly increased the efficacy of a vaccine against FeLV [26] and against simianhuman immunodeficiency virus (SHIV) [32]. Cynomolgus macaques were vaccinated with measles virus replicative vectors expressing antigens of SHIV. Antigens were either the wild type or mutated in the isu domain of the envelope protein. The inactivation of the isu domain led to the induction of significantly enhanced cellular immune responses and reduced proviral loads after challenge of the vaccinees [32]. A mutation in the isu domain of gp41 of HIV-1 increased antibody production when rats were immunized with the mutated protein, in contrast to the unmutated protein [29].
Here, we describe a novel and endotoxin-free system for testing the immunosuppressive properties of a retroviral transmembrane envelope protein. For this, four constructs encoding a part of the transmembrane envelope protein p15E of PERV-A, which is a gammaretrovirus closely related to MuLV, FeLV, and KoRV-all three viruses inducing severe immunodeficiencies in their infected hosts-were transfected into human 293 cells. The transfected cells were continuously cultured in the presence of hygromycin-containing selection medium. Their expression was controlled by immunofluorescence and flow cytometry. These cells were incubated with human PBMCs, and the changes in cytokine release and cytokine expression of the PBMCs in this endotoxin-free system were analyzed in comparison to cells not expressing p15E. In addition, human PBMCs were incubated with pig kidney cells, PK15 cells, releasing PERV, and human 293 cells infected with and producing PERV-A/C [33,34]. Furthermore, p15E-expressing cells were used to study the impact of the expression of p15E on human cytotoxic cells and the expression of the major histocompatibility complex (MHC) class 1 molecules.
## 2. Results
## 2.1. Sequence Homology of the Retroviral Isu Domain
A sequence alignment of transmembrane envelope proteins of numerous retroviruses shows that the isu domain is highly conserved among all retroviruses (Figure 1). It is an α-helical structure in immediate neighborhood of a Cys-Cys loop containing three cysteines in the case of gammaretroviruses (CX 6 CC) and only two cysteines in the case of immunodeficiency viruses, including HIV-1 (Figure 1). Within the isu domain, residues L1, Q2, N3, R4, L7, D8, and L10 show the highest conservation. The main domains of the TM protein are the fusion peptide (FP), the immunosuppressive domain (isu), the cysteine loop (C-CC in the case of HIV-1), and the membranespanning domain (MSD). (B) A sequence alignment of the following retroviral sequences was performed: MuLV, murine leukemia virus; FeLV, feline leukemia virus; PERV, porcine endogenous retrovirus; KoRV, koala retrovirus; GaLV, gibbon ape leukemia virus; CKS-17 consensus sequence of the gammaretroviruses; HERV-W, HERV-FRD, human endogenous retrovirus W, FRD; BERV, bovine endogenous retrovirus; MPMV, Mason Pfizer monkey virus; BERV-P, bovine endogenous retrovirus; BaEV, baboon endogenous retrovirus; HERV4-1, human endogenous retrovirus 4-1; ERV-3, endogenous retrovirus 3; Mab-Env1-4, syncytins of Marubya lizards; HIV-1, -2, human immunodeficiency virus -1, -2; SIV cpz, simian immunodeficiency virus chimpanzee; SIV sm, SIV sooty mangabey; SIVagm, SIV African green monkeys; MMTV, mouse mammary tumor virus; HERV-K, human endogenous retrovirus-K. Identical amino acids and conservative exchanges (L = V = I) in all virus groups or in a single group are stained. The colors highlight identical amino acids at a given position, illustrating the degree of conservation.
## 2.2. Cloning and Transfection of p15E of PERV
To establish a cellular and endotoxin-free system for studying the immunosuppressive properties of p15E of PERV, two distinct expression constructs were designed according to the PERV-A/C sequence AY570980 and inserted in a vector. Both constructs contained the ectodomain of p15E, including the immunosuppressive domain, the membrane-spanning domain (MSD), the env signal peptide (SP), the furin peptidase cleavage site, and a short sequence derived from gp70 (Figure 2A,B). Neither construct contained main parts of the fusion peptide (FP) of p15E to avoid unwanted membrane fusions. All constructs were based on the crystal structure of the p15E ectodomain, which forms a homotrimer [35]. One construct, p15E-link, carried a mutation at position 1652, resulting in a cysteine-toserin substitution, which removed one of the cysteines in the immunodominant region of p15E (Figure 2D) to avoid unwanted intermolecular Cys-Cys interactions. This construct contained a longer portion of the N-terminal part of p15E, referred to as the linker region, to study whether the unstructured region in front of the N-terminal alpha helix of p15E plays a role. Both constructs were similar to the MoMuLV env used to study the immunosuppressive properties of this murine retrovirus in vivo, in a tumor system in mice [23] (Figure 2C). p15E was cloned into pVitro2-EGFP [36]. Following transfection using polyethylenimine (PEI) [37], cells were selected with 500 µg/mL hygromycin and maintained under continuous culture in selection medium.
Since, as shown below, cell surface expression of p15E was very low for these constructs, a second set of constructs was generated. In addition to the p15E-link-His construct, a mutant containing a substitution in the immunosuppressive (isu) domain was produced, in which the sequence LQNR was replaced by AAAA (Figure 2E,F). Previous studies have shown that mutations in this highly conserved region of the isu domain abrogate the ability to induce IL-10 and IL-6 [29].
Furthermore, two additional constructs expressing p15E were generated. One construct expressed human albumin as a protein carrier together with p15E, a FLAG tag, and an infrared fluorescent protein (iRFP) (Figure 2G). The second construct expressed a fragment of the PERV surface envelope protein gp70 together with p15E, a FLAG tag, and iRFP (Figure 2H). The presence of iRFP enabled easy detection of expressing cells and facilitated the selection of positive cells by fluorescence-activated cell sorting (FACS). The same construct as in (E), but with a mutation in the immunosuppressive domain, substituting LQNR into AAAA. (G) A new construct, combining human albumin with p15E, flag-tag, and infrared fluorescent protein (iRFP), (H) A new construct combining a fragment of the surface envelope protein gp70 of PERV with p15E, flag-tag, and iRFP. iRFP, infrared fluorescent protein; N-helix, N-terminal helix; C-helix, C-terminal helix; TM, transmembrane domain; ISU, immunosuppressive domain; furin, furine peptidase cut; 2AP, 2A peptide.
## 2.3. Analysis of p15E of PERV Expression in Transfected and Virus-Producing Human Cells
The expression of p15E on the surface of transfected and virus-producing cells was analyzed using two methods: immunofluorescence and flow cytometry analysis. In both cases, a specific antiserum against p15E of PERV was employed. This antiserum (#355) had previously been shown to react with recombinant and viral p15E in Western blot assays, and the epitopes of the antibody binding had been defined using overlapping peptides. GPQQLEK/T is the minor epitope in the fusion peptide proximal region (FPPR) of the N-terminal helix and FEGWFN the major epitope in the membrane proximal external region (MPER) of p15E [38,39]. Low intracellular expression of p15E was observed in the transfected cells when the immunofluorescence was performed with cell membrane permeabilization (Figure 3A). The expression of p15E on the cell surface was even lower. The intracellular expression of p15E was much stronger in virus-producing 293T cells and much stronger on the cell surface compared with the transfected cells (Figure 3). In another experiment assessing p15E expression using the same goat anti-p15E serum (#355) but an Alexa 568-conjugated anti-goat IgG, again, only low levels of p15E were detected, not only in the transfected 293 cells but also in the PERV-producing 293 cells (Figure 3C), indicating that not only does the expression of p15E fluctuate in 293T cells but so does the expression of PERV in PERV-producing 193T cells. It is important to note that the conformation of the protein in both the cytoplasm and on the cell surface remains unknown. p15E on the cell surface might trimerize, and the epitopes recognized by the goat serum may be partially masked.
Flow cytometry studies confirmed significant differences in p15E expression between the cell surface and the intracellular compartment (Figure 3D). Cell surface staining of 293T wt cells with the anti-p15E antiserum revealed a small shift in fluorescence intensity (nine arbitrary units, solid histograms) as compared to incubation of cells with the secondary reagents alone (broken histograms). This shift could be due to some unspecific binding of the antiserum to 293T cells. Fluorescence intensity was not enhanced after staining of 293T-p15E-NHR-His (eight units) or 293T-p15E-link-His cells (nine units), suggesting that p15E was not expressed on the cell surface or with very low density which was below detection level. A similar issue as described above arises here: the conformation of the protein on the cell surface remains unknown. However, clear-cut binding of the anti-p15E antiserum was demonstrated in permeabilized transfectants. Thus, mean fluorescence intensity of 16 units in 293T wt cells significantly increased to 283 and 502 units after staining of 293T-p15E-NHR-His and 293T-p15E-link-His cells, respectively. Thus, the p15E transgene was expressed in this cell model, and the protein was detected intracellularly but was barely detected on the cell surface. 63 times. (C) Immunofluorescence analysis using a p15E-specific goat antiserum (#355) and Alexa 568-conjugated anti-goat IgG of 293 cells transfected with p15E-link and p15E-NHR of wild-type 293 cells (control) and of 293 cells infected with and producing PERV. The cells were not permeabilized to study cell surface expression. Magnification: 63 times; exposure: 3 s; laser: 555 nm. (D) Flow cytometry analysis of cell surface and intracellular expression of p15E in wild-type (wt) 293T cells and 293T cells transfected with p15E-NHR-His and 293T-p15E-link-His. Cells were incubated with anti-p15E goat serum #355, followed by incubations with biotinylated bovine anti-goat Ig and APCconjugated streptavidin. Histograms show anti-p15E binding (solid lines); the numbers represent mean fluorescence intensity and the reactivity of secondary reagents alone (dotted lines). (E) Cell surface expression of p15E on 293T-p15E-NHR-His cells (cyan), on 293T-p15E-link-His cells (orange), and PERV-A/C-producing 293 cells (magenta) in comparison with wt 293 cells (red). Flow cytometry analysis was performed using the same antiserum #355 against p15E of PERV and a donkey anti-goat FITC-labeled secondary antibody.
In a second experiment, both 293T-p15E-NHR-His and 293T-p15E-link-His cells, as well as PERV-A/C-producing 293 cells, were stained with the same antiserum against p15E. A donkey anti-goat FITC-labeled secondary antibody was used, along with FACS buffer containing 1% horse serum. In this experiment, clear surface expression of p15E was detected on the transfected 293T cells, with very strong expression observed on the surface of PERV-A/C-producing cells (Figure 3E).
These results indicate that p15E expression can vary, but in some instances, a distinct and detectable expression is evident.
## 2.4. Effect of p15E of PERV on Cytokine Expression in Human PBMCs
To investigate whether 293T cells expressing p15E can induce IL-10 secretion in human PBMCs, similar to the synthetic isu peptides, recombinant transmembrane envelope proteins, and virus preparations of HIV-1 and HERV-K [30,31], p15E-expressing cells were co-incubated with purified human PBMCs. Six experiments were performed. After 24 h, IL-10 levels in the supernatant were quantified using an ELISA. An increase in IL-10 release was observed when PBMCs were incubated with 293T cells expressing both p15E constructs compared to wild-type 293T cells (Figure 4A-C). However, in two of six experiments, no increased expression was observed. Since the expression of p15E on the cell surface was shown to be variable (Figure 3D,E), a correlation with IL-10 release may be anticipated. IL-10 release from wild-type 293T was also tested and, as expected, was undetectable (Figure 4). Human PBMCs incubated with porcine embryonic kidney PK15 cells producing PERV showed a significantly increased IL-10 release (Figure 4B). Notably, a much higher release of IL-10 was observed when PBMCs were incubated with 293T cells producing PERV-A/C (Figure 4C). These cells produced the virus, as demonstrated by measuring viral RNA using a real-time PCR in the supernatant. RNA was isolated from the cell supernatant, and a real-time RT-PCR was performed using the pol primers (Table 1); ct values between 17 and 24 were measured, corresponding to 10 4.7 to 10 6.4 copies per 20 µL supernatant. The presence of cellular DNA was excluded performing the assays without RT. Immunofluorescence analysis revealed higher surface expression of p15E on PERV-producing cells compared to that observed in the transfected p15E-expressing cells (Figure 4B). Differences in induced IL-10 levels between the p15E-link-His and p15E-NHR-His constructs observed in some experiments (e.g., Figure 4A) but not in others (e.g., Figure 4B,C), along with the absence of IL-10 induction in two other cases, suggest variability in the expression of active p15E on the surface of transfected cells. In an additional experiment, the cells analyzed in Figure 3E were co-cultured with human PBMCs, and IL-6 release by the PBMCs was assessed. In contrast to the cells shown in Figure 3D, which exhibited only minimal p15E expression on the cell surface, the p15Elink-His-and p15E-NHR-His-expressing cells displayed low but detectable levels of p15E (Figure 3E). Strong p15E expression was observed on the surface of PERV-A/C-producing cells (Figure 3E). These PERV-producing cells induced robust IL-6 release (Figure 4D), whereas the p15E-link-His-expressing cells and the p15E-NHR-His-expressing cells induced a medium amount of IL-6 (Figure 4E). These findings correlate with the expression levels determined by FACS analysis (Figure 3E) and confirm that the extent of IL-10 release correlated with the extent of p15E on the cell surface.
After demonstrating that incubation with p15E-expressing cells increased IL-10 and IL-6 protein release by human PBMCs, the effect of p15E on the expression of additional cytokines and markers at the mRNA level was subsequently evaluated. Real-time RT-PCRs specific for the mRNA of IL-6, IL-10, INF-γ, TNF-α, and SEPP1 were established, and the expression was measured after 4, 6, 8, and 10 h of incubation with 293T cells expressing p15E-link-His and 293 cells producing PERV-A/C (Figure 5A). Screening of these cytokines and SEPP1 was performed based on our previous cytokine array and microarray analyses, which showed that these markers were up-regulated in human PBMCs from healthy donors following exposure to the isu peptide homopolymer of HIV-1, as well as the recombinant transmembrane envelope proteins of HERV-K [30,31]. In this experiment, expression of IL-6, IL-10, INF-γ, TNF-α, and SEPP1 mRNA increased, either steadily increasing, as in the case of IL-10, or peaking at 8 h, as in the case of IL-6, TNF-α, and INF-γ. Please note the strong expression of all mRNA in the co-incubation experiment with PERV-infected 293 cells, which showed a higher expression of p15E compared with the transfected p15E-expressing cells. However, when MMP-1, TNF-α, IL-8, and IL-6 were analyzed in a second experiment, an increase in expression of the mRNA of these molecules was only observed for PK15 cells, but not for the transfected cells, with the exception of a slight increase in MMP-1 and IL-6 in cells expressing p15E-link (Figure 5B). Similarly, in this context, the expression of p15E on the cell surface also varies, and a correlation between p15E expression and cytokine expression can be anticipated. Please note again the higher expression of the mRNA in the co-incubation experiment with PERV-producing pig PK15 cells, which showed a higher expression of p15E compared with the transfected and selected p15E-expressing cells. MMP-1 was included because, in microarray experiments with the isu peptide homopolymer of HIV-1 and the recombinant transmembrane envelope proteins of HERV-K, it was identified as the most highly up-regulated gene in human PBMCs [30].
$$SEPP1 rev ′ -TCGACAGAGCTTCTTTTG-3 ′ 954-972 SEPP1 probe ′ -6-FAM-AGAATCAGCAACCAGGAGCA-BHQ-1-3 ′ 721-740 TNF-α fw ′ -GAGAAGCAACTACAGACCCC-3 ′ this manuscript NM 000594.4 48-67 TNF-α rev ′ -CATGCTTTCAGTGCTCATGG 176-195 TNF-α probe ′ -6-FAM-ACAACCCTCAGACGCCACATCC-BHQ-1-3 ′ 76-97$$
## 2.5. Effect of p15E of PERV on Human Cytotoxic Effector Cells
A specific in vitro assay was applied to evaluate the immunosuppressive effect of p15E of PERV on cytotoxicity of effector cells. Thus, human PBMCs were cultivated for 5 days with IL-2 to induce cytotoxic activity and then co-cultivated for two hours with wild-type 293 cells and 293 cells expressing p15E as p15E-link-His (Figure 6). In a series of experiments using effector populations from different blood donors, 3 to 7% of gated CD56 + CD45 + cells (effector population) expressed CD107a. An increased proportion of CD107a + cells (11 to 34%) was observed in co-cultures with wild-type 293T cells, indicating degranulation of the effector cells by contact with 293T cells. In two experiments (Exp. 1 and 2), we observed a slight reduction in CD107a expression when p15E-expressing transfectants were used as targets. However, no reduction was seen in the other two experiments. This data indicates that under certain conditions expression of p15E on target cells triggers reduced levels of CD107a, pointing to diminished cytotoxic activity of effector cells. This obviously correlates with the level of expression of p15E. ). IL-2-activated PBMCs were cultured alone (gray columns) or with 293T wt cells (black columns) or 293T cells expressing p15E-link-His (dashed columns). Expression of CD107a was monitored after 2 h on gated CD56 + CD45 + cells.
## 2.6. Effect of p15E of PERV on MHC Class I Expression
Retroviruses are known to down-regulate MHC molecules at the cell surface. For example, HIV-1 reduces the expression of MHC class I A and B molecules, thereby shielding infected cells from cytotoxic T lymphocyte (CTL)-mediated killing [42]. A similar effect has been reported for a murine gammaretrovirus closely related to PERV [43]. To examine whether p15E expression affects MHC class I (HLA-ABC) levels, 293T wild-type cells and cells expressing p15E either as p15E-NHR-His or p15E-link-His were stained with a monoclonal antibody against human MHC class I molecules. A significant reduction of 16-20% in MHC class I expression was observed (Figure 7).
## 2.7. New Set of Expression Constructs, Transfection, and Co-Cultivation with Human PBMCs
Because p15E expression was very low when the constructs p15E NHR His and p15E Link were used (Figure 4), new expression constructs were generated. First, a mutation was introduced into the isu domain (LQNR → AAAA), which in HIV-1 is known to abolish immunosuppressive activity [29] (Figure 2F). Second, p15E was co-expressed with an infrared fluorescent protein (iRFP): in one construct together with human serum albumin and in another with a larger fragment of the PERV surface envelope protein gp70 (Figure 2G,H). These constructs were transfected into 293T cells using polyethylenimine (PEI), and cells were selected by FACS. However, the p15E expression in this set of constructs also remained very low. Nonetheless, incubation of these p15E-expressing 293T cells with normal human PBMCs resulted in an increased IL-6 and IL-10 release compared with untransfected 293T cells (Figure 8). Importantly, the isu domain mutation reduced the amount of IL-10 released (Figure 8). Interestingly, IL-10 induction by cells producing PERV-A/C was very low in this experiment, consistent with the low level of PERV expression observed in some assays (Figure 3C). This indicates that the expression of p15E and PERV fluctuated in human 293T cells for reasons that remain unknown. The extent of cytokine induction depended on the level of p15E expression in both p15E-expressing and PERV-producing cells.
## 3. Discussion
To gain further evidence for the immunosuppressive properties of the transmembrane envelope protein p15E of PERV, the main part of this molecule, including the isu domain, was expressed in human 293T cells, and its effect on human PBMCs was investigated. Unfortunately, the protein expression on the cell surface was very low and fluctuating, leading to variability in its effects across experiments. Since the expression of p15E was the only parameter fluctuating in the experiments, the modulation of the IL-10 and IL-6 release and cytokine expression found in some experiments must be associated with this molecule. It is important to note that the release of IL-10 and IL-6 (Figure 4) and the expression of cytokine mRNA (Figure 5) were high in the co-incubation experiments with PERV-infected 293 cells and PERV-producing pig PK15 cells, which showed a higher expression of p15E compared with the transfected p15E-expressing cells (Figure 3).
It remains unclear why the expression of p15E, especially on the cell surface of the transfected and selected p15E-expressing 293 cells, is so low. Surprisingly, an arginine repeat was found in the protein sequence of p15E of PERV, while it was absent in the sequence of p15E of MuLV [44]. This short arginine repeat suggests that the PERV protein could be retained in the cell, in contrast to the MuLV protein p15E [44]. Arginine/serinerich proteins are mainly localized in the cytoplasm and are targeted to the nucleus [45,46]. Future investigations will focus on optimizing the expression system to achieve higher and more stable levels of p15E expression. This will include introducing mutations in the arginine repeats, testing alternative vectors, promoters, and tagged constructs, as well as evaluating trimeric p15E and different cell lines to improve protein expression, stability, and conformation. Nevertheless, few p15E molecules can be found at the surface of the transfected cells by immunofluorescence (Figure 3). Expression of p15E was also found by flow cytometry (Figure 3E). Therefore, two methods, immunofluorescence and flow cytometry, independently showed low expression in the cytoplasm of human 293 cells and a lower but detectable expression at the cell surface. Unfortunately, the expression of p15E on the cell surface fluctuates (see Figure 3D,E); the reasons for this are unclear.
Despite the low expression of p15E, cytokine expression and cytokine release from PBMCs of healthy humans were modulated in a manner consistent with previous observations for synthetic peptides corresponding to the isu domain of PERV p15E and purified PERV particles. Furthermore, previous findings showed that peptides corresponding to the isu domain of PERV inhibited mitogen-triggered proliferation [47,48]. The sequence of the isu domain of PERV is identical to the isu domains of related gammaretroviruses, such as MuLV, FeLV, and KoRV (Figure 1). Therefore, theoretically, evidence of immunosuppressive properties in synthetic peptides, viral or recombinant p15E, or virus particles from MuLV, FeLV, and KoRV inherently extends to the isu domain of PERV, and vice versa. The immunosuppressive properties of MuLV, FeLV, and KoRV, as well as of human endogenous retroviruses such as HERV-K, are well studied in vitro and in vivo (for a review, see [1,2,49]). The envelope proteins of endogenous retroviruses, called syncytins, not only play a role in the placentogenesis but may also immunoprotect the embryo (for a review, see [50]). However, an involvement of PERV in pig placentogenesis has not yet been demonstrated.
Two main reasons may account for the limited effect of the p15E constructs used in this work. The first is its low level of expression, and the second is the possibility that it does not adopt the correct conformation required to bind to the putative receptor for the immunosuppressive domain. Our studies on peptides suggested that a certain degree of multimerization is necessary, as monomeric peptides were inactive, while only polymeric forms exhibited activity [51]. When we displayed the isu domains of HIV, PERV, and MuLV on the surface of human cells using a tetraspanin-anchored construct, no modulation of cytokine release was observed in co-cultured human PBMCs [44], likely due to an unfavorable conformation caused by proximity to the cell membrane. Similarly, when the same retroviral proteins containing the isu domains were expressed in an alternative system and released into the culture medium, no cytokine response was detected, possibly due to insufficient protein levels or lack of multimerization.
Immunosuppression is a general property of all retroviruses, and immunodeficiency viruses such as human immunodeficiency viruses HIV-1 and HIV-2 are well studied examples (for a review, see [1]). The changes in cytokine expression observed here are in agreement with the changes in cytokine expression observed when human PBMCs were incubated with polymers of synthetic peptides corresponding to the isu domain of HIV [31] or with HERV-K particles released from a human teratocarcinoma cell line, with a recombinant transmembrane envelope protein of HERV-K or with peptides corresponding to the isu domain of HERV-K [30]. Modulation of cytokine expression was also observed when FeLV was analyzed (for a review, see [2]).
Analysis of gene expression in human PBMCs treated with the HIV isu peptide or the recombinant transmembrane envelope protein of HERV-K revealed significant changes, with over 300 genes either up-or down-regulated [30,31]. Notably, IL-6, IL-10, MMP-1, and SEPP1 were among the most highly up-regulated genes-a finding that is consistent with the elevated expression levels observed in this study (Figure 4). MMP-1 is a zinc-dependent protease essential for the breakdown of extracellular matrix expressed on monocytes and macrophages [52], and SEPP1 plays an important role in innate immune responses [53,54]. In hepatitis C virus infection, SEPP1 mRNA inhibits type I interferon responses by limiting the function of retinoic-acid-inducible gene I (RIG-I), a sensor of viral RNA [54].
One advantage of the established system is the absence of endotoxin. Endotoxin can induce cytokine modulation [55], and in experiments with synthetic peptides or recombinant proteins produced in bacteria, endotoxin contamination below the detection limit of the assay used (EndoLISA System, Hyglos, Bernried, Germany) could not be ruled out. Endotoxin is a lipopolysaccharide (LPS) of the outer membrane of most Gram-negative bacteria; it binds first to the LPS-binding protein (LBP) and is transferred to cluster of differentiation 14 (CD14), where myeloid differentiation-2 protein (MD-2) and the Toll-like receptor 4 (TLR4) re-associate. Receptor binding leads to a signal transduction involving activation of the nuclear factor-kappa B (NF-κB) transcription factor, resulting in the re-lease of cytokines [55]. To avoid endotoxin contamination, in our later experiments, gp41 produced in human 293 cells was used [56]. The secreted and purified to homogeneity recombinant gp41 produced in 293 cells was soluble, glycosylated, and assembled into trimers. The protein bound to monocytes and to a lesser extent to lymphocytes and triggered the production of specific cytokines when added to normal PBMCs [56], confirming that endotoxin was not involved in the effects of gp41 of HIV-1.
In addition, the immunosuppressive properties of the transmembrane envelope protein gp41 of HIV-1 was studied in a cellular system which was endotoxin-free. For this, murine cTRAMP prostate cancer cells were transfected with a gp41-expressing vector, and gp41 expression on the cell surface was demonstrated by FACS analysis, and the cells released gp41 into the cell supernatant [56]. These cells were pulsed with the ovalbuminderived MHC-I peptide SIINFEKL and co-cultured with naïve CD8+ T cells from OT-1 mice, which carry the corresponding SIINFEKL T cell receptor. The gp41-expressing cells, but not the vector control cells, strongly inhibited IFNγ production and reduced CD25 (IL-2 receptor) expression. These findings indicated that gp41 impairs the antigen-specific response of murine CD8+ T cells by drastically suppressing IFNγ production. Furthermore, this result corroborates previous findings that retroviral transmembrane proteins or peptides corresponding to their isu domain exhibit interspecies reactivity by modulating immune cells across species (for a review, see [1]).
Since the modulation of cytokine expression by p15E of PERV observed in our experiments is identical to that reported for FeLV, which contains the same isu domain [26], and comparable to that induced by the transmembrane envelope protein gp41 of HIV-1 [16-19], which contains a related isu domain, this can be considered a common property of retroviral transmembrane envelope proteins. This conclusion is further supported by findings that synthetic peptides corresponding to the isu domains of PERV, MuLV, FeLV, and KoRV, as well as peptides derived from HIV-1, inhibit mitogen-induced proliferation of human PBMCs and modulate cytokine and gene expression in a similar manner [1,2].
The results presented here, together with previous publications, not only contribute to the understanding of the mechanisms by which pathogenic exogenous retroviruses such as MuLV, KoRV, FeLV, and HIV and endogenous retroviruses such as HERV-K and syncytins exert their immunosuppressive effects but may also have important implications for xenotransplantation. Xenotransplantation using pig cells or organs has achieved remarkable progress in recent years. Notably, the first human patients have received encapsulated pig islet cells for the treatment of diabetes [57], as well as pig hearts [58,59] and kidneys [60] for the treatment of organ failure. Although the pig organs were genetically modified to prevent hyperacute rejection and to reduce cell and antibody mediated rejection, they still required intensive pharmacological immunosuppression to prevent rejection. The expression of an immunosuppressive protein on the surface of the pig transplant may help prevent rejection, similar to what has been demonstrated in mouse tumor models [23][24][25][26][27]61]. In contrast to the mouse tumor model, where expression of the immunosuppressive retroviral transmembrane protein prevents rejection of tumor cells, in xenotransplantation, this mechanism aims to prevent rejection of a healthy transplanted organ. This approach could significantly reduce the need for pharmacological immunosuppression. The p15E of PERV is the best candidate since it is present in the genome of all pigs, and pigs are tolerant and do not produce antibodies against p15E [62]. The expression of the immunosuppressive molecule p15E on the surface of pig xenotransplants could revolutionize the field of transplantation. Despite the xenogeneic origin of these organs, the need for immunosuppressive drugs may be significantly reduced. Additionally, pig organs offer the advantage of being available in unlimited supply and present a higher level of virological safety compared to allogeneic human organs. Whereas donor pigs for xenotransplantation can be screened carefully before transplantation, historically, several human viruses-including HIV, human cytomegalovirus, and even rabies virus-have been transmitted through human organ transplants [63]. Human kidney epithelial 293T cells, 293T cells infected with and producing PERV-A/C, and porcine embryonic kidney pig PK15 cells were grown in Dulbecco Eagle Medium (DMEM) with 10% fetal bovine serum (FCS, PAN Biotech, Aidenbach, Germany, Lot P160616) and 1% penicillin-streptomycin (DMEM culture medium). Cells were maintained at 37 • C in a humidified chamber with 5% CO 2 . PK15 cells harbor PERV-A and PERV-B but not PERV-C proviruses in their genome; they release infectious virus particles. These cells were obtained from Leibniz-Institute DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany). The PERV-A/C produced by 293T cells is the result of passaging cell-free virus on human 293T cells and is characterized by multimerized repeats containing transcription factor binding sites in the long-terminal repeat (LTR) [33,34]. Thus, a higher number of repeats results in a longer LTR and, consequently, a higher viral replication rate [34]. Since the number of repeats in the LTR changes during cultivation [33], a PCR was performed using LTR-specific primers in order to characterize the virus used in the present experiments. The length of the amplicon was higher compared with that from PERV-3 • and lower compared with PERV-5 • [33] (Figure 9), indicating that a highly replicating virus is produced. The virus load was determined using a real-time PCR (see below). 293T cells were split twice a week in a 1:3 ratio; PK15 cells were split in a 1:2 ratio every 3 days after washing with phosphate-buffered solution (PBS) and trypsinization using 0.25% trypsin/0.02% EDTA (PAN Biotech, Aidenbach, Germany).
## 4. Materials and Methods
## 4.2. Cloning of p15E
Synthetic p15E constructs were produced based on the env gene of PERV-A/C (AY570980) as gene blocks (gBlock) (Integrated DNA Technologies IDT, Coralville, IA, USA) (Figure 2). All constructs contained the signal peptide of the env gene and a linker part of the gp70 gene coding for its first 23 amino acids, followed by the sequence for the furin cleavage site and a modified sequence of the p15E gene. All constructs did not contain a major part of the fusion peptide (position 1390-1431). The constructs of p15E were based on the crystal structure the p15E ectodomain which form a homotrimer, pdb_00007s94 [35]. The reported trimer formation was achieved on the basis of the overexpressed and purified ectodomain, expressed in a bacterial expression system. In our mammalian expression system, the construct is targeted to the plasma membrane and has to be processed by the plasma membrane-located peptidase furin to obtain a final product which is similar to the p15E protein expressed when PERV is produced. Expecting that the length of the N-terminal part might be important for the correct conformation, two versions of difference length were produced. The fusion peptide was removed in both constructs to avoid unwanted membrane fusion.
The construct designated p15E-link contained the following modifications: a nucleotide exchange at position 1652 from g to c (leading to a cysteine-to-serin conversion to avoid unwanted intermolecular Cys-Cys interactions). The construct designated p15E-NHR did not contain the sequence coding for the unstructured N-terminal part, including the fusion peptide, but started with the N-terminal helix (NHR) (nucleotide position 1456) of p15E; the nucleotide at position 1652 was not changed. For cloning purposes, an Eco-R1 restriction site was added to the 5 ′ end, including a Kozak sequence for optimal translation initiation. The 3 ′ end contained a sequence coding for a 6× histidine tag (p15E-link-His, p15E-NHR-His) and a stop codon followed by a Nhe-1 cutting site. The p15E gBlock with Eco-R1 and Nhe-1 site were cloned into pVitro2-EGFP [36] using the same two enzymes replacing the EGFP gene in the plasmid. All plasmids were sequenced before transfection into 293T cells.
The constructs were similar to the construct of the envelope of the Moloney MuLV (MoMuLV) prepared by Mangeney and Heidman [23] (Figure 2C). The immunosuppressive domain of PERV and MoMuLV are identical (Figure 1). The MoMuLV env construct was expressed in mouse cells that do not produce tumors in immunocompetent mice. However, when Env was expressed on the cell surface, tumors developed, indicating that the retroviral protein induced immunosuppression [23].
In addition, a second set of p15E-expressing constructs was generated. Besides the p15E-link-His construct, a mutant carrying a substitution in the isu domain was produced, in which the sequence LQNR was replaced by AAAA (Figure 2E,F). Previous studies have shown that mutations in this highly conserved region of the isu domain abrogate the ability to induce IL-10 and IL-6 [29].
Furthermore, two additional constructs expressing p15E were generated. One construct expressed human albumin as a protein carrier together with p15E, a FLAG tag, and an infrared fluorescent protein (iRFP) (Figure 2G). The other construct expressed a fragment of the PERV surface envelope protein gp70 together with p15E, a FLAG tag, and iRFP (Figure 2H). The presence of iRFP enabled easy detection of expressing cells and facilitated the selection of positive cells by FACS.
## 4.3. Transfection of p15E and Establishment of Transfected p15E-Expressing Cells
For plasmid transfection, 10 5 293T cells were seeded in a 12-well plate the day before transfection. On the next day, the medium was changed to DMEM with 5% FCS. For each transfection, 3 µL of polyethylenimine (PEI) solution (1 mg/mL) was added to 50 µL PBS; in parallel, 1 µg plasmid DNA was added to 50 µL PBS [37]. Both solutions were vortexed at high speed for 1 min. After 10 min rest at room temperature, the PEI and DNA solution were gently mixed and incubated for 3 min at room temperature. The transfection solution was added dropwise to the cells. After 3 h, the medium was changed to DMEM culture medium. Selection was started 2 days after transfection with 500 µg/mL hygromycin, and the cells were maintained under continuous culture in the selection medium containing hygromycin.
The iRFP-expressing cells were selected by FACS.
## 4.4. Peripheral Blood Mononuclear Cells (PBMCs)
At the Institute of Virology in Berlin, PBMCs were isolated from buffy coats from human blood from an anonymous donor using Ficoll-Hypaque density centrifugation with the use of 50 mL Leucosep Tubes (Greiner Bio-One, Kremsmünster, Austria) according to the instructions of the manufacturer (Greiner Bio-One). The buffy coat was diluted in a 1:2 ratio with PBS beforehand. Leucosep tubes were filled with 15 mL of Ficoll-Hypaque and centrifuged for 30 s at 1000× g at room temperature to move the Ficoll-Hypaque below the porous barrier. A volume of 30 mL of the diluted buffy coat was layered on top of the porous barrier and centrifuged at 1000× g for 10 min at room temperature without brakes. After centrifugation, the following layers were observed: plasma, enriched cell fraction PBMCs, granulocytes, and erythrocytes. The fraction containing PBMCs was harvested using a Pasteur pipette. The porous barrier effectively avoids recontamination with pelleted erythrocytes and granulocytes. Harvested PBMCs were washed twice with 10 mL of PBS and subsequently centrifuged for 10 min at 250× g. The PBMC pellet was resuspended in cell culture medium. Resuspended PBMCs were counted using a Neubauer Chamber, and 1 × 10 8 PBMCs were frozen in cryopreserved tubes and stored in liquid nitrogen in a freezing medium containing 70% DMEM, 20% FCS, and 10% dimethyl sulfoxide (DMSO). Freshly isolated PBMCs were used for co-culture experiments. The use of human blood has been approved by the ethical commission at the Medical Faculty of Humboldt University Berlin. At the Transplant Laboratory in Hannover, PBMCs were isolated from discarded material of normal routine apheresis samples from anonymized donors.
## 4.5. Ethics Declarations
At the Institute of Virology in Berlin, buffy coats from human blood from an anonymous donor were provided by Deutsches Rotes Kreuz, Blutspendedienst Nord-Ost, Berlin. The use of human blood has been approved by the ethical commission at the Medical Faculty of the Humboldt University Berlin. At the Transplant Laboratory in Hannover, PBMCs were isolated from discarded material of normal routine apheresis samples obtained from the Department of Transfusion Medicine of the Hannover Medical School. Samples were anonymized and could not be assigned to an individual donor. The local ethics committee of Hanover Medical School approved this procedure. All methods were carried out in accordance with DFG guidelines of Good Scientific Practice.
## 4.6. Co-Cultivation
Human 293T cells and porcine PK-15 cells are adherent cells and were used in the co-cultivation assay at 80 to 90% confluence. Cells were washed with PBS and incubated with 0.25% trypsin/0.02% EDTA at 37 • C for 30 s. A volume of 10 mL of culture medium containing FCS was added; cells were collected in a 15 mL falcon tube and centrifuged at 300× g for 5 min. Pellets were resuspended in 10 mL culture medium. Cells were counted twice using a Neubauer Chamber and centrifuged at 300× g for 5 min, and culture medium was added to the cell pellet to obtain 3 × 10 5 cells/100 µL. A total of 100 µL of cells was added to each well of a 96-well plate, and 100 µL of culture medium without hygromycin was added into the wells. Cells were incubated overnight at 37 • C in a humidified chamber with 5% CO 2 . The next day, 100 µL of media was removed, as 293T and PK-15 adhere to the surface of the plate, and 3 × 10 5 PBMCs in a volume of 100 µL RPMI with 15% FCS, 2 mM glutamine, and 1 mM sodium pyruvate were added for co-incubation and left overnight at 37 • C in a humidified chamber with 5% CO 2 . For PCR expression studies, pooled batches from wells with 2.5 × 10 4 /100 µL 293T cells and 7.5 × 10 4 /100 µL PBMCs were used.
## 4.7. Fluorescence Analysis
293T cells were gently dislodged from cell culture flasks with ice-cold PBS and counted with a Neubauer Chamber after trypan blue staining to detect dead cells. The cell suspension was adjusted to 250 × 10 3 cells/mL. For each sample, 50 × 10 3 cells in 200 µL PBS were transferred to a glass slide using a Cellspin 1 device (Tharmac, Limburg/Lahn, Germany) at 8000 rpm for 10 min according to the manufacturer's instructions. Slides were kept at room temperature until completely dry. Cells were fixed in 4% formaldehyde in PBS for 15 min followed by washing in PBS three times for 10 min each. Perforation of the cell membrane was achieved by 15 min incubation in PBS with 0.5% Triton X-100 (Carl Roth, Karlsruhe, Germany) followed by PBS washes as described above. Slides were incubated with 3% BSA in PBS for 1 h to reduce unspecific binding followed by incubation with a 1:100 dilution of goat anti-p15E (goat #355 [38]) in 3% BSA/PBS for 1 h. After the PBS wash (3 times, 10 min) FITC-conjugated anti-goat antibody from a donkey (Merck/Sigma Aldrich, Darmstadt, Germany) was incubated for 1 h at room temperature. Cover slips were added after all liquid was removed, and a mounting dye containing 4 ′ ,6-diamidino-2-phenylindole (DAPI) was added (Carl Roth, Karlsruhe, Germany).
Samples were analyzed with a Zeiss Axio fluorescence microscope equipped with an Axiocom 503 mono camera and a Colibri 7 LED light source using ZEN 2.3 software (all Zeiss, Oberkochen, Germany). The SMART set up provided by the software was used to adjust the fluorescence signals. Exposure time for the FITC channel was set to 5 s (DAPI 50 ms). Non-specific background signals in the FITC channel were subtracted by setting the threshold in the negative control (untransfected 293T cells) to zero/black. These setting were used for all samples. DAPI staining was adjusted to highest contrast.
## 4.8. Antibodies and Flow Cytometry
A goat anti-p15E serum was used to monitor p15E expression after transfection of human 293T. Generation and characterization of the goat anti-p15E serum #355 had been described previously [38]. Cells were incubated with serum (30 min, 1:40 dilution), followed by two additional incubation steps using biotinylated bovine anti-goat Ig (Dianova, Hamburg, Germany) and allophycocyanin (APC)-conjugated streptavidin (BD Biosciences, San Jose, CA, USA). To monitor intracellular levels of p15E, transfected and control cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with saponin (0.2%) for 10 min at room temperature, and then incubated with anti-p15E antiserum. Expression of MHC class I molecules (HLA-ABC) on 293T cells was detected by indirect staining using monoclonal antibody (mAb) W6/32 (American Type Culture Collection, ATCC, Manassas, VA, USA) and phycoerythrin (PE)-conjugated rat anti-mouse kappa light chain (BD Biosciences). Directly labeled mAb CD56-APC (B159; BD Biosciences) in combination with CD107a-PE (H4A3; BD Biosciences) were used to study degranulation of human natural killer (NK) cells in response to 293T cells. An antibody against CD45 was used: CD45-FITC (HI30, BD Biosciences). Analyses were performed on a FACSCalibur flow cytometer (Becton Dickinson, San Jose, CA, USA), and data were processed by using FCS Express 7 (De Novo software, Pasadena, CA, USA). These analyses were performed at the Medical School, Hannover.
An additional assay was performed at the Free University Berlin with the same antiserum against p15E and a donkey anti-goat FITC-labeled secondary antibody, using a special buffer composed of phosphate-buffered saline (PBS), 2 mM ethylenediaminetetraacetic acid (EDTA), and 1% horse serum. Analyses were performed on a Cytoflex S Flow Cytometer (Beckman Coulter, Krefeld, Germany) and data were processed by using Software CytExpert 2.1.
## 4.9. CD107a Assay
Cytotoxic effector cells were generated by culturing PBMCs for 5 to 7 days in the presence of 50 ng/mL IL-2. Cells (2 × 10 5 ) were co-cultured for two hours with 2 × 10 5 293T wt cells or 293T cells transfected to express p15E (293T-p15E-NHR-His or 293T-p15Elink-His). Cytotoxic activity against 293T target cells was monitored by assessing CD107a expression (degranulation) on gated CD56 + CD45 + NK cells.
## 4.10. Statistical Analysis
Statistical analyses at the Transplant Laboratory in Hannover were performed using Student's t-test, with levels of significance reported as p-values. At the Institute of Virology in Berlin, statistical analysis was performed using one-way ANOVA and using GraphPad Prism (Prism 10.2.2).
## 4.11. DNA and RNA Extraction
In order to demonstrate the presence of the sequence encoding p15E with its isu domain, the cell lines wt 293T, 293T-p15E-NHR-His and 293T-p15E-Link-His were tested using conventional PCR. Genomic DNA was isolated from transfected and non-transfected cells using DNA Easy Blood and Tissue Kit (Qiagen, Hilden, Germany). A total of 5 × 10 6 cultured cells of each cell line were centrifuged for 5 min at 300× g and resuspended in 200 µL PBS. Further steps of DNA extraction were performed using the Instruction manual provided by Qiagen to purify total DNA from animal blood or cells (DNeasy Blood and Tissue Handbook, Spin-Column protocol). All centrifugation steps were performed at room temperature. DNA concentration was quantified twice using a nanodrop 1000 spectrophotometer (PeqLab, Erlangen, Germany). Purified plasmids of p15E-NHR-His and p15E-Link-GFP (1:100 and 1:1000) were used as positive controls. Wt 293T cells were used as a negative control. DNA concentrations were standardized using nuclease free water and 5 µL of DNA sample was added to each reaction mix. DNA from virus producing 293T cells was isolated using the DNeasy Blood and Tissue kit (Qiagen). In order to analyze the expression of cytokines, RNA was isolated using the RNeasy Kit (Qiagen) and DNAse (New England Biolabs, Frankfurt am Main, Germany) treatment to remove cellular DNA.
## 4.12. Polymerase Chain Reaction (PCR), Real-Time PCR and Real-Time RT-PCR
In order to characterize PERV proviruses in virus-producing 293T cells, PCRs using specific primers for the pol and LTR region of PERV (Table 1) were performed using the AmpliTaq DNA polymerase (Applied Biosystems, Waltham, MA, USA): 5 min denaturation at 95 • C, and 45 cycles (15 s 95 • C, 30 s 62 • C, 30 s 72 • C). PCR was performed using a Biometra Thermocycler (Analytik Jena, Jena, Germany). Electrophoresis was performed in a 1.3% agarose gel including a 1 kbp DNA ladder (GeneRuler, Thermo Scientific, Waltham, MA, USA). DNA from a PERV-positive pig was used as a positive control for PCR targeting the pol sequence of PERV.
Additionally, two plasmids containing the LTR sequences of PERV-3 • and PERV-5 • served as positive controls for the characterization of the LTR regions. PERV proviruses in the virus-producing 293T cells were also determined by real-time PCR using specific primers for the pol of PERV (Table 1 and RT-qPCR were performed using qTOWER 3 G qPCR-Thermocycler (Analytik Jena, Jena, Germany).
For quantification, a standard curve was built using a synthetic gene block with a partial sequence of PERV pol.
## 4.13. Detection of Cytokine Expression by RT-PCR
In order to analyze expression of IL-6, IL-10, INF-γ, TNF-α, and SEPP1, a RT-PCR for each of the factors was established, using primers and probes as shown in Table 1. RNA was isolated from pooled culture wells, treated with DNAse I, and analyzed using human GAPDH as control: NO-ROX one step kit, 30 min reverse transcription at 50 • C, 5 min denaturation at 95 • C, 45 cycles (10 s 95 • C, 20 s 54 • C, 15 s 72 • C). A qTOWER 3 qPCR-Thermocycler was used in all experiments. Gene expression was calculated according the 2 -∆∆Ct method [64].
## 4.14. Detection of Cytokine Release by ELISA
To detect IL-10, an ELISA for human IL-10 (R & D Biosciences, Minneapolis, MN, USA) was used. Supernatants obtained from co-culture of human PBMCs with transfected and non-transfected 293T cells were used for the assay as described in Section 4.6. Supernatants were collected after 24 h of incubation by centrifuging at 2000× g for 10 min. ELISAs were performed in duplicates according to protocols of the supplier: 200 µL of standard, control, and sample supernatant were added per well of the ELISA plate, covered with an adhesive strip, incubated for 2 h at room temperature, and washed with 400 µL of wash buffer using a squirt bottle for a total of 4 washes. A total of 200 µL of human IL-10 antibody conjugate was added to each well and incubated for an hour at room temperature. Washing was repeated, and subsequently, 200 µL of substrate solution was added to each well and incubated for 30 min at room temperature.
Subsequently, 50 µL of stop solution was added to each well. Optical density was determined within 30 min using a microplate reader set to 450 nm. To detect IL-6, an ELISA for human IL-6 (Sigma Aldrich, St. Louis, MI, USA) was used. Supernatants containing proteins were collected after 24 h of incubation by centrifuging at 2000× g for 10 min. ELISAs were performed in duplicates according to protocols of the supplier. Standards, buffer solutions, and detection antibodies were prepared as mentioned in manufacturers manual. A volume of 100 µL of supernatant from co-culture of different samples was added in the wells along with the standard. Samples and the standard were incubated overnight at 4 • C with gentle shaking. The next day, solutions were discarded, and wells were washed with 300 µL wash buffer 4 times; 100 µL of biotinylated antibody was added to each well and incubated at room temperature for 60 min with gentle shaking. The biotinylated antibody was removed, wells were washed, and 100 µL of streptavidin solution was added to each well and incubated for 45 min with gentle shaking. Streptavidin was removed from wells, and the plate was washed and incubated with 100 µL TMB one-step substrate reagent for 30 min in the dark with gentle shaking; finally, 50 µL of stop solution was added and optical density was measured immediately at 450 nm on a microplate reader.
## 4.15. Endotoxin Measurement
Although all components to be used for cell culture are endotoxin-free, a measurement of endotoxin was performed using the Pierce chromogenic endotoxin quant kit (Thermo Fisher, Waltham, MA, USA). All tested materials were below the detection limit.
## 5. Conclusions
Using a novel, endotoxin-free cellular system to express the transmembrane envelope protein p15E of PERV, we confirmed the immunosuppressive properties of this molecule.
## References
1. Denner (2014) "The transmembrane proteins contribute to immunodeficiencies induced by HIV-1 and other retroviruses" *AIDS*
2. Oostendorp, Meijer, Scheper (1993) "Immunosuppression by retroviral-envelope-related proteins, and their role in nonretroviral human disease" *Crit. Rev. Oncol. Hematol*
3. Borghi, Fantuzzi, Varano et al. (1995) "Induction of interleukin-10 by human immunodeficiency virus type 1 and its gp120 protein in human monocytes/macrophages" *J. Virol*
4. Planès, Benmohamed, Leghmari et al. (2014) "HIV-1 Tat protein induces PD-L1 (B7-H1) expression on dendritic cells through tumor necrosis factor alpha-and toll-like receptor 4-mediated mechanisms" *J. Virol*
5. Quaranta, Mattioli, Giordani et al. (2009) "Immunoregulatory effects of HIV-1 Nef protein" *Biofactors*
6. Collette, Dutartre, Benziane et al. (1996) "Similarity between Nef of primate lentiviruses and p15E of murine and feline leukaemia viruses" *AIDS*
7. Mallet, Bouton, Prudhomme et al. (2004) "The endogenous retroviral locus ERVWE1 is a bona fide gene involved in hominoid placental physiology" *Proc. Natl. Acad. Sci*
8. Blaise, De Parseval, Bénit et al. (2003) "Genomewide screening for fusogenic human endogenous retrovirus envelopes identifies syncytin 2, a gene conserved on primate evolution" *Proc. Natl. Acad. Sci*
9. Heidmann, Béguin, Paternina et al. (2017) "HEMO, an ancestral endogenous retroviral envelope protein shed in the blood of pregnant women and expressed in pluripotent stem cells and tumors"
10. Henzy, Gifford, Kenaley et al. (2017) "An intact retroviral gene conserved in spiny-rayed fishes for over 100 My" *Mol. Biol. Evol*
11. Hogan, Harmon, Cid-Rosas et al. "Conserved residues of the immunosuppressive domain of MLV are essential for regulating the fusion-critical SU-TM disulfide bond" *J*
12. Eksmond, Jenkins, Merkenschlager et al. (2017) "Mutation of the putative immunosuppressive domain of the retroviral envelope glycoprotein compromises infectivity" *J. Virol*
13. Good, Ogasawara, Liu et al. (1990) "Immunosuppressive actions of retroviruses" *Lymphology*
14. Takahashi, Day, Luangwedchakarn et al. (2001) "A retroviral-derived immunosuppressive peptide activates mitogen-activated protein kinases" *J. Immunol*
15. Ruegg, Strand (1991) "A synthetic peptide with sequence identity to the transmembrane protein GP41 of HIV-1 inhibits distinct lymphocyte activation pathways dependent on protein kinase C and intracellular calcium influx" *Cell. Immunol*
16. Koutsonikolis, Haraguchi, Brigino et al. (1997) "HIV-1 recombinant gp41 induces IL-10 expression and production in peripheral blood monocytes but not in T-lymphocytes" *Immunol. Lett*
17. Takeshita, Breen, Ivashchenko et al. (1999) "Induction of IL-6 and IL-10 production by recombinant HIV-1 envelope glycoprotein 41 (gp41) in the THP-1 human monocytic cell line" *Cell. Immunol*
18. Speth, Joebstl, Barcova et al. (2000) "HIV-1 envelope protein gp41 modulates expression of interleukin-10 and chemokine receptors on monocytes, astrocytes and neurones" *AIDS*
19. Barcova, Kacani, Speth et al. (1998) "Gp41 envelope protein of human immunodeficiency virus induces interleukin (IL)-10 in monocytes, but not in B, T, or NK cells, leading to reduced IL-2 and interferon-γ production" *J. Infect. Dis*
20. Kim, Kim, Choi et al. (1999) "Soluble expression in Escherichia coli of murine endogenous retroviral transmembrane envelope protein having immunosuppressive activity" *Protein Expr. Purif*
21. Pengal, Ganesan, Wei et al. (2006) "Lipopolysaccharide-induced production of interleukin-10 is promoted by the serine/threonine kinase Akt" *Mol. Immunol*
22. Ivanusic, Denner (2023) "Sensitive detection of lipopolysaccharide (LPS) by monitoring of IL-10 secretion from PBMCs" *MicroPubl. Biol*
23. Mangeney, Heidmann (1998) "Tumor cells expressing a retroviral envelope escape immune rejection in vivo"
24. Blaise, Mangeney, Heidmann (2001) "The envelope of Mason-Pfizer monkey virus has immunosuppressive properties" *J. Gen. Virol*
25. Mangeney, De Parseval, Thomas et al. (2001) "The full-length envelope of an HERV-H human endogenous retrovirus has immunosuppressive properties" *J. Gen. Virol*
26. Schlecht-Louf, Mangeney, El-Garch et al. (2013) "A targeted mutation within the FeLV envelope protein immunosuppressive domain to improve a canarypox-based FeLV vaccine" *J. Virol*
27. Mangeney, Renard, Schlecht-Louf et al. (2007) "Placental syncytins: Genetic disjunction between the fusogenic and immunosuppressive activity of retroviral envelope proteins" *Proc. Natl. Acad. Sci*
28. Kedzierska, Crowe (2001) "Cytokines and HIV-1: Interactions and clinical implications" *Antivir. Chem. Chemother*
29. Morozov, Morozov, Semaan et al. (2012) "Single mutations in the transmembrane envelope protein abrogate the immunosuppressive property of HIV-1" *Retrovirology*
30. Morozov, Dao Thi, Denner (2013) "The transmembrane protein of the human endogenous retrovirus-K (HERV-K) modulates cytokine release and gene expression" *PLoS ONE*
31. Denner, Eschricht, Lauck et al. (2013) "Modulation of cytokine release and gene expression by the immunosuppressive domain of gp41 of HIV-1" *PLoS ONE*
32. Nzounza, Martin, Dereuddre-Bosquet et al. (2021) "A recombinant measles virus vaccine strongly reduces SHIV viremia and virus reservoir establishment in macaques" *NPJ*
33. Karlas, Irgang, Votteler et al. (2010) "Characterisation of a human cell-adapted porcine endogenous retrovirus PERV-A/C" *Ann. Transplant*
34. Denner, Specke, Thiesen et al. (2003) "Genetic alterations of the long terminal repeat of an ecotropic porcine endogenous retrovirus during passage in human cells" *Virology*
35. Dean, Serrão, Lee "Structure of the Core Postfusion Porcine Endogenous Retrovirus Fusion Protein. mBio 2022, 13, e0292021"
36. Schippers (2015) "Functional Characterization of the Potential Immune Evasion Proteins pUL49.5 and p012 of Marek's Disease Virus (MDV)"
37. Yang, Zhou, Li et al. (2017) "Optimized PEI-based Transfection Method for Transient Transfection and Lentiviral Production" *Curr. Protoc. Chem. Biol*
38. Kaulitz, Fiebig, Eschricht et al. (2011) "Generation of neutralising antibodies against porcine endogenous retroviruses (PERVs)" *Virology*
39. Fiebig, Stephan, Kurth et al. (2003) "Neutralizing antibodies against conserved domains of p15E of porcine endogenous retroviruses: Basis for a vaccine for xenotransplantation?" *Virology*
40. Czauderna, Fischer, Boller et al. (2000) "Establishment and characterization of molecular clones of porcine endogenous retroviruses replicating on human cells" *J. Virol*
41. Yang, Güell, Niu et al. (2015) "Genome-wide inactivation of porcine endogenous retroviruses (PERVs)" *Science*
42. Kamp, Breij, Nottet et al. (2001) "Interactions between major histocompatibility complex class II surface expression and HIV: Implications for pathogenesis" *Eur. J. Clin. Investig*
43. Jones, Moors, Ryan et al. (1992) "Altered macrophage antigen-presenting cell function following Friend leukemia virus infection" *Viral Immunol*
44. Ivanusic, Pietsch, König et al. (2018) "Absence of IL-10 production by human PBMCs co-cultivated with human cells expressing or secreting retroviral immunosuppressive domains" *PLoS ONE*
45. Isshiki, Tsumoto, Shimamoto (2006) "The serine/arginine-rich protein family in rice plays important roles in constitutive and alternative splicing of pre-mRNA" *Plant Cell*
46. Manley, Krainer (2010) "A rational nomenclature for serine/arginine-rich protein splicing factors (SR proteins)" *Genes Dev*
47. Denner (1998) "Immunosuppression by Retroviruses: Implications for Xenotransplantation" *Ann. N. Y. Acad. Sci*
48. Tacke, Kurth, Denner (2000) "Porcine endogenous retroviruses inhibit human immune cell function: Risk for xenotransplantation?" *Virology*
49. Snyderman, Cianciolo (1984) "Immunosuppressive activity of the retroviral envelope protein P 15E and its possible relationship to neoplasia" *Immunol. Today*
50. Denner (2016) "Expression and function of endogenous retroviruses in the placenta" *APMIS*
51. Denner, Norley, Kurth (1994) "The immunosuppressive peptide of HIV-1: Functional domains and immune response in AIDS patients" *AIDS*
52. Woessner, Nagase (2000) "Matrix Metalloproteinases and TIMPs"
53. Hurwitz, Klaus, Llabre et al. (2007) "Suppression of human immunodeficiency virus type 1 viral load with selenium supplementation: A randomized controlled trial" *Arch. Intern. Med*
54. Murai, Honda, Shirasaki et al. (2019) "Induction of Selenoprotein P mRNA during Hepatitis C Virus Infection Inhibits RIG-I-Mediated Antiviral Immunity" *Cell Host Microbe*
55. Wang, Quinn, Endotoxins (2010) "Lipopolysaccharides of gram-negative bacteria" *Subcell. Biochem*
56. Mühle, Lehmann, Hoffmann et al. (2017) "Antigenic and immunosuppressive properties of a trimeric recombinant transmembrane envelope protein gp41 of HIV-1" *PLoS ONE*
57. Cooper, Matsumoto, Abalovich et al. (2016) "Progress in Clinical Encapsulated Islet Xenotransplantation"
58. Griffith, Goerlich, Singh et al. (2022) "Genetically Modified Porcine-to-Human Cardiac Xenotransplantation" *N. Engl. J. Med*
59. Mohiuddin, Singh, Scobie et al. (2023) "Graft dysfunction in compassionate use of genetically engineered pig-to-human cardiac xenotransplantation: A case report" *Lancet*
60. Montgomery, Stern, Lonze et al. (2022) "Results of Two Cases of Pig-to-Human Kidney Xenotransplantation" *N. Engl. J. Med*
61. Schlecht-Louf, Renard, Mangeney et al. (2010) "Retroviral infection in vivo requires an immune escape virulence factor encrypted in the envelope protein of oncoretroviruses" *Proc. Natl. Acad. Sci*
62. Keller, Petersen, Niemann et al. (2014) "Lack of antibody response in pigs immunized with the transmembrane envelope protein of porcine endogenous retroviruses" *J. Gen. Virol*
63. Nellore, Fishman (2018) "Donor-derived infections and infectious risk in xenotransplantation and allotransplantation. Xenotransplantation"
64. Livak, Schmittgen (2001) "Analysis of relative gene expression data using real-time quantitative PCR and the 2 -∆∆CT Method" *Methods*
65. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607705&blobtype=pdf | # Distinct immune properties of the N-and C-termini of the immunosuppressive domain of Ebola virus glycoprotein
Mathieu Iampietro, Sivakumar Periasamy, Philipp Ilinykh, Yuan Qiu, Yakun Liu, Abhinit Nagar, Bin Gong, Alexander Bukreyev
## Abstract
Ebola virus (EBOV) causes a severe human disease with high lethality. Pathogenesis of EBOV disease is characterized by a paradoxical combination of hyperinflammation and immunosuppression. EBOV has a single envelope glycoprotein (GP), which is a type I transmembrane protein with strong immunomodulatory effects. GP contains a conserved immunosuppressive domain (ISD) with a high similarity to ISDs of envelope proteins retroviruses. To investigate the effects of ISD, a set of 17 EBOV viral-like particle (VLP) constructs containing the entire EBOV nucleoprotein, VP40, and GP with single alanine or glycine substitutions in each position of ISD was generated and tested in human peripheral blood mononuclear cells (PBMCs). Wild-type VLPs induced inflammatory responses; however, when added to pre-stimulated cells, they reduced inflammation, thus exerting immunosuppressive properties. Substitution of lysine at ISD position 5 (Lys-5) increased anti-inflammatory properties by reducing proliferative responses of VLPs and also reducing nuclear factor of activated T-cells 1 (NFAT1) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) activation. In contrast, substitution of tryptophan at position 14 (Trp-14) increased both the proinflammatory and the proliferative responses and the adhesion of VLP-infected human monocytes to microvascular endothelial cells. Thus, the ISD N-and C-termini have proand anti-inflammatory properties, respectively, suggesting their unique implications in the EBOV pathogenesis. Furthermore, the immunomodulating effects of ISD were also mediated by shed GP, which is abundant in the medium. These data may be useful for the development of treatments for diseases caused by EBOV by targeting the ISD. IMPORTANCE Our data suggest that the ISD N-terminus plays a role in activating immune cells and pro-inflammatory response. In contrast, the C-terminus of ISD downregulates the pro-inflammatory response through the reduction of NF-kB and NFAT activities. The data also show that EBOV GP increases the adhesion of monocytes to endothelial cells, and the effect is inhibited by the ISD C-terminus. Moreover, the data demonstrate that the immunomodulating effects of ISD are mediated not only by the virus-associated GP but also by the shed GP, which is abundant in the medium. Pathogenesis of the disease caused by EBOV is characterized by hyperinflammation and some features of immunosuppression, which could in part be affected by the complex effects of the ISD. These data indicate that targeting the ISD may be considered for the development of treatments for the disease caused by EBOV.
EBOV belongs to the Filoviridae family and has a single-stranded non-segmented negative-sense RNA genome encoding seven major genes located in a linear order (5). The gene number four produces two major proteins: a 676-amino acid-long type 1 transmembrane glycoprotein (GP), which is anchored in the viral membrane, and a 364-amino acid-long soluble glycoprotein (sGP). GP is produced from mRNA undergoing co-transcriptional editing, whereas sGP is produced from its unedited version (6)(7)(8). GP is post-translationally cleaved by the host cellular protease furin and is displayed on the virion's surface as two subunits, GP1 and GP2, linked by disulfide bonds (9). EBOV GP1 is highly glycosylated with N-linked and O-linked glycans, whereas GP2 contains mostly N-linked glycans (10). In addition, shed GP, which represents soluble full-length GP ectodomain resulting from proteolysis of GP by cellular tumor necrosis factor α-convert ing enzyme (TACE), is also present in the extracellular compartment (11).
EBOV GP mediates viral attachment, fusion to the membrane, and subsequent cellular entry (4,11,12). Furthermore, GP activates dendritic cells, macrophages, and T-cells through binding to toll-like receptor 4 (TLR4) (12)(13)(14). In addition, GP is implicated as a main factor of virulence contributing to vascular cell toxicity and injury that under lie the phenomenon of hemorrhagic fever occurring during EBOV infection (15,16). Importantly, the interaction of viral GP with immune cell populations leads to dysregu lation of the immune response during EBOV disease (4,14). One of the viral factors involved in modulating the immune response could be the immunosuppressive domain (ISD) located in GP2 (17). EBOV GP ISD represents a sequence of 17 amino acids (a.a.) initially identified in the ectodomains of retroviral envelope proteins (18) and demonstra ted to possess immunosuppressive activity in multiple experimental systems (18)(19)(20)(21)(22). Subsequently, a similar ISD was identified in the GP ectodomain of Marburg virus, which is another filovirus (23). However, the biological effects of EBOV and Marburg virus ISD remain obscure. One report studying the GP crystal structure suggested that the EBOV ISD is unlikely to have an immunosuppressive effect (24). Another study reported that incubation of human peripheral blood mononuclear cells (PBMCs) with a 17-mer peptide mimicking ISD of EBOV or other filoviruses inhibits activation of CD4 + and CD8 + T-cells (25); however, peptides may not necessarily reproduce biological properties of ISD in the context of the full-length protein. Here, to test the effects of ISD with a more authentic system, we generated a set of EBOV viral-like particles (VLPs) and recombinant shed GP with single point mutations in amino acid residues constituting the ISD. We exposed human PBMCs to the mutated and non-mutated VLPs. We found that a mutation in Lys-5 (N-terminus) enhances the release of anti-inflammatory factors, whereas the mutation in Trp-14 (C-terminus) has opposite effects. Overall, the strikingly opposite effects of the Nand C-termini of ISD may have implications in EBOV pathogenesis.
## RESULTS
## EBOV GP triggers inflammation in non-stimulated cells but has immunosup pressive effects in pre-stimulated cells
The EBOV GP ISD has various levels of amino acid similarity to ISD of other ebolaviruses and retroviruses (Fig. 1A). To verify that the ISD has a capacity to reduce inflammation, we generated VLPs by transfecting 293T cells with plasmids encoding VP40, NP, and WT GP. After 72 h post-transfection, the supernatants were harvested, and VLPs were purified by sucrose gradient centrifugation and analyzed for the incorporation of the EBOV proteins. Analysis of VLPs by western blotting demonstrated the presence of expected proteins (Fig. 1B). To evaluate the effect of ISD on inflammation, we tested the activities of key transcription factors, NF-κB and activator protein 1 (AP-1), which are involved in multiple biological and pathological processes, including inflammation (26), using THP-1 XBlue cells that allow monitoring their activation by expression of both NF-κB-and AP-1-indu cible SEAP reporter gene. We exposed THP-1 XBlue cells to VLPs or medium from mock-transfected cells or pre-stimulated with 12-O-tetradecanoylphorbol-13-acetate (TPA) and ionomycin to induce a pro-inflammatory state (Fig. 1C). Although forskolin, used as a control for anti-inflammatory activity, inhibited inflammation in both culture environments, WT VLPs increased NF-κB and AP-1 activity in mock-treated cells (Fig. 1C left panel), whereas it impaired their activities when cells were previously treated with TPA/ionomycin (Fig. 1C right panel). We next repeated our experimental procedures using primary human PBMCs and found that both IL-2 (Fig. 1D) and TNFα levels (Fig. 1E) were slightly increased in mock-treated cells (Fig. 1D and E left panels), whereas their levels reduced following culture of WT VLPs in pre-activated cells (Fig. 1D and E right panels), confirming our previous results. Interestingly, these data describe complex effects of EBOV GP and the potential role of its ISD during EBOV immunopathogenesis, as it triggered inflammatory signatures in non-stimulated cells while significantly reducing inflammatory response in pre-activated cells.
## The N-and C-terminal amino acids of EBOV GP ISD have opposite immuno modulating effects
To better understand the immunomodulating effects of the EBOV GP ISD, we designed a set of VLPs including WT and 17 ISD single mutants, with an alanine (Ala) substitution in each residue of the 17-mer ISD sequence, except Ala in position 6, which was replaced with glycine (Gln) (Fig. S1A). This was followed by the generation of VLPs as described above. The mutant VLPs showed comparable expressions of GP, NP, and VP40 to that of WT VLP (Fig. S1B).
To test the effects of the ISD mutations on the expression of cytokines, human PBMCs were exposed to WT or mutant VLPs at 10 µg/mL for 24 h, and the percentages of cells positive for IL-2 and IL-12, as markers of activation, were determined by flow cytometry (Fig. 2A andB). Compared with mock-treated cells, treatment of cells with WT VLPs resulted in an increase in percentages of IL-2 + cells from 0.06% to 1.81% (Fig. 2A) and IL-12 + cells from 3.16% to 7.33% (Fig. 2B), supporting the previous observations (Fig. 1C to E). When cells were exposed to mut 5 (which corresponds to GP K588A) VLPs, a reduction in the percentages of IL-2 + cells (1.04%) (Fig. 2A) and IL-12 + cells (3.76%) (Fig. 2B) was observed when compared with WT VLP. In contrast, mut 14 (which corresponds to GP W597A) induced the highest percentages of IL-2 + cells (5.67%) (Fig. 2A) and IL-12 + cells (14.56%) (Fig. 2B). In parallel, we evaluated the ISD activity on the expression of known anti-inflammatory mediators IL-10 (Fig. 2C) and cyclic AMP (cAMP) (Fig. 2D). Although the anti-inflammatory activity of IL-10 is well known (27), cAMP was demonstrated to be implicated for the immunosuppressive activity of ISD in retroviruses (21,28,29). cAMP is also known to inhibit cell activation and pro-inflammatory cytokine production (30)(31)(32). As such, we tested the amounts of IL-10 + and cAMP + cells cultured for the ISD mutants. Strikingly, the mut 5 VLPs demonstrated the highest percentages of IL-10 + (7.28%) compared with WT VLPs (5.71%) (Fig. 2C) and cAMP levels (60% increase over WT) (Fig. 2D). In contrast, mut 14 VLPs demonstrated the lowest percentage (3.94%) of IL-10 + cells compared with WT and a 35% reduction of cAMP levels (Fig. 2C andD), consistent with an opposite trend in the cytokine responses. To determine the effects of the ISD amino acids 5 and 14 in the context of live EBOV infection, we introduced each mutation in the full-length clone encoding EBOV expressing GFP (28) (EBOV-GFP) and attempted to recover the associated virus, along with the control clone without the mutation, as previously described (29). Two attempts resulted in no viable virus recovered from the mutated constructs, whereas the WT control virus without a mutation was easily recovered. These data suggest that K588 and W597 in the ISD are critical for the viability of EBOV. Overall, our data further support the identification of K588 and W597 within the ISD as the residues responsible for its immunomodulating properties.
## The ISD Lys-5 and Trp-14 residues have opposite effects on the activation of PBMCs
Computational analysis of the 3D structure of GP in relation to immunosuppressive domain (ISD) with greater detail for K588 and W597 is shown in Fig. 3A through C. The sequences of GP and the ISD mutants were analyzed using PyMOL software, an open-source molecular visualization system, for amino acid interactions. In the α-helix of ISD, the hydrogen bonding of K (lysine) contributes to the ability of α-helix 4 to interact with the N-terminal half of the protein, whereas W (tryptophan) maintains interaction with α-helix 5. The mutation to either K or W could maintain an intact interaction with the respective regions. As noted above, EBOV GP mediates viral attachment and entry into the host cells. We therefore tested whether the mutations have any impact on attachment, which could affect the cytokine response. First, we tested cell binding using human PBMCs. For that, human PBMCs were isolated and cultured with WT or mut 5 or mut 14 EBOV VLPs for 2 h at 4°C. Then, the cells were harvested, and percentages of cells with bound WT, mut 5, and mut 14 EBOV VLPs were quantified by flow cytometry. Equivalent levels of binding of WT and mutated VLPs were detected (Fig. 3D). Moreover, these data were confirmed when our analysis focused specifically on isolated B-and T-cell populations (Fig. S2A andB). These data suggest that mutations K588A and W597A do not have any strong effects on VLP-cell interactions. Next, we tested how the mutations impact the expression of inflammatory cytokines or activation markers, focusing on the T-cell population. Human PBMCs were mock-treated or cultured with WT or mut 5 or mut 14 VLPs for 48 h, and expressions of TNFα, IFNγ, IL-4 (Fig. 3E), CD25, CD69, and CD2 (Fig. 3F) were evaluated by flow cytometry. Consistent with the previous data, WT VLPs induced an increase in all these molecules compared with mock-treated cells. Mut 5 VLPs induced a reduced level of TNFα + (33.4% compared with WT VLPs), IFNγ + (21.7%), CD25 + (19.7%), CD69 + (22.4%), and CD2 + (34.4%) cells compared with WT VLPs. In contrast, mut 14 VLPs induced increased levels of TNFα + (34.3%), IFNγ + (31%), CD25 + (21%), CD69 + (17.3%), and CD2 + cells (47.3%) compared with WT VLPs. Interestingly, in comparison to WT VLPs, both mut 5 and mut 14 VLPs induced an increase in levels of IL-4 (57.3% and 45.3%, respectively), which is a pivotal cytokine involved in multiple immune mechanisms in either pro-or anti-inflammatory environments (30,31). As the mutations did not affect the binding of VLPs to cells (Fig. 3D), the observed changes in the expression of cytokines and activation markers associated with the mutations are due to the cell-intrinsic processes in response to exposure to VLPs.
## The ISD Trp-14 residue reduces GP-induced inflammation mediated by the ISD Lys-5 residue
It is well established that viral ISDs inhibit inflammatory responses (18-22, 25, 32). Consistent with our previous observations, the C-terminal mut 14 appeared to have a pro-inflammatory effect, suggesting an anti-inflammatory effect of the Trp-14 residue. Unexpectedly, our analysis demonstrated that the mut 5 ISD has an anti-inflammatory effect compared with WT, suggesting a pro-inflammatory effect of the specific Lys-5 residue in the WT sequence, contrasting the expected role of ISD. We therefore tested the effects of both mutations on the activity of the transcription factors involved in inflammatory responses (Fig. 4). Indeed, NF-κB and NFAT1 are major transcription factors that play important roles in the induction of cytokine gene expressions, cell activation, and the implementation of an inflammatory environment (33)(34)(35)(36)(37)(38). First, luciferase assays demonstrated modulations in the profiles of NFAT1 expression (Fig. 4A) similar to the ones elicited for cytokines and activation markers previously tested in PBMCs (Fig. 3E andF). Specifically, mut 5 VLPs had reduced the activity of NFAT1 to levels comparable with the effect of cyclosporine A (CsA) and reduced the activity of NF-κB. In contrast, mut 14 VLPs demonstrated elevated activities compared with WT VLP (Fig. 4A andB).
Next, imaging flow cytometry experiments were performed to analyze the nuclear localization of NFAT1-GFP in monocytic THP-1 cells. Unlike mock-treated cells, cells cultured with WT VLP displayed the presence of NFAT1-GFP in the nucleus (Fig. 4C), thus representing its active profile. Cells stimulated with ionomycin (positive control) or mut 14 VLPs demonstrated increased nuclear localization of NFAT1-GFP compared with WT VLP. In contrast, cells exposed to mut 5 VLPs demonstrated reduced nuclear localization of NFAT1-GFP (Fig. 4C). These data paralleled the effects of the mutations on cytokine expression (Fig. 1 and2). We hypothesized that the pro-inflammatory effect of the ISD N-terminus is associated with its induction of the active mono-phosphorylated form of NFAT1 (39). To test the hypothesis, human PBMCs were cultured with WT, mut 5, or mut 14 VLPs for 24 h, lysed, and analyzed by western blotting for the transcriptionally active mono-phosphorylated NFAT1 and its inactive multi-phosphorylated form. Ionomycin alone, which is a potent stimulator of cell proliferation, induced mono-phosphorylated NFAT1, whereas CsA completely inhibited its activity. Importantly, WT VLPs elevated the levels of mono-phosphorylated NFAT1 in PBMCs, concomitantly with previous observa tions. Moreover, PBMCs cultured with mut 5 VLPs did not elicit any significant activation of NFAT1 as seen by high levels of multi-phosphorylated form, but cells cultured with mut 14 VLPs displayed high levels of mono-phosphorylated NFAT1 equivalent to that induced by ionomycin (Fig. 4D).
To further investigate the properties of EBOV GP ISD on anti-inflammatory-related signaling, we tested the activation of both the mammalian target of rapamycin (mTOR) protein and protein kinase B (Akt) (40). Human PBMCs were incubated with WT, mut 5, or mut 14 VLPs for 24 h, lysed, and analyzed by western blotting for the expression of active p-Akt and p-mTOR. WT VLPs cultured with PBMCs increased the levels of p-mTOR and p-Akt compared with the mock-treated condition (Fig. 4E). Contrary to what was described when investigating pro-inflammatory factors NFAT1 and NF-κB (Fig. 4A, B, andC), the levels of anti-inflammatory factors p-Akt and p-mTOR were increased in the presence of mut 5 VLPs while decreased in the presence of mut 14 VLPs when compared with that of WT VLPs (Fig. 5E). Overall, these data suggest that the N-terminal part of EBOV GP ISD induces pro-inflammatory pathways, whereas the C-terminal part exerts an anti-inflammatory signaling.
## EBOV GP ISD N-and C-termini have opposite effects on cell proliferation and viral replication
Next, we tested whether the mutations in ISD affect cell proliferation. To measure the effects of the mutations on cell division, human PBMCs were pre-labeled with carboxy fluorescein succinimidyl ester (CFSE), incubated with the different VLP constructs for 48 h, and the dilution of CFSE as a measure of cell proliferation was quantified by flow cytometry. Cells treated with mut 5 VLP demonstrated a reduction in cell prolifera tion, compared with WT VLPs, whereas mut 14 VLP-treated cells elicited an increase in proliferation (Fig. 5A andB). The immunosuppressive effect of the ISD C-terminus, based on the mut 14 response, is consistent with the suppression of cell proliferation previously described for the EBOV GP ISD-specific 17-mer peptide (27). These data suggest that the ISD C-terminus, which has an immunosuppressive effect, also inhibits cell proliferation.
Our previous studies have demonstrated that the direct binding of EBOV GP to TLR4 on monocytes promoted their differentiation, thus leading to their increased susceptibil ity to EBOV infection (13,41). Therefore, we evaluated the effects of the mutations on viral replication in THP-1 monocytic cells. THP-1 cells were incubated with WT or the mutated VLPs for 24 h and infected with EBOV-GFP for 48 h. Next, viral replication was assessed by flow cytometry by quantitation of the percentages of infected (GFP+) cells and the mean fluorescent intensity (MFI) (Fig. 5C to E). Pre-treatment of THP-1 cells with WT VLPs favored an increased EBOV infection compared with mock-treated cells based on both percentages of GFP + cells and the MFI, thus confirming our previous results (13) (Fig. 5C andD). Furthermore, pre-culture with mut 5 VLP increased the infection compared with WT VLP, whereas pre-treatment with mut 14 VLP reduced it (Fig. 5C to E). These data show opposite effects of the N-and C-termini on the susceptibility of THP-1 cells to EBOV infection, which are consistent with the results on cytokine expression (Fig. 1D andE 2, and3) and the activation of different signaling pathways (Fig. 1C and4). Taken together, these results demonstrate that the ISD N-terminus induces a pro-inflammatory cellular environment, promotes cell proliferation, but reduces EBOV infection. In contrast, the ISD C-terminus promotes an anti-inflammatory environment, limits cell proliferation, but increases EBOV infection.
## EBOV GP increases the adhesion of monocytes to endothelial cells, and the effect is inhibited by the ISD C-terminus
As demonstrated previously, mut 14 VLPs enhanced the expression of pro-inflammatory cytokines and decreased the cAMP levels compared with WT VLPs, thus suggesting EBOV GP ISD Trp-14 anti-inflammatory property (Fig. 2). The anti-inflammatory properties of cAMP are mediated by exchange protein directly activated by cAMP type 1 (EPAC-1), which is an intracellular cAMP receptor (42)(43)(44). The cAMP-EPAC1 axis suppresses the inflammatory process by stabilizing microvascular endothelial cells in inflammation (45). The adhesion of leukocytes to the microvascular endothelium, prior to their infiltration into the perivascular space, is a key characteristic of focal inflammation (46). In the bloodstream, leukocytes must strongly attach to the endothelial surface to overcome detachment by shear stress from blood flow. Therefore, leukocyte adhesion can be investigated by measuring real-time nanoscale binding force between living leukocytes and microvascular endothelial cells. Fluidic AFM technology is particularly well suited to achieve this goal (47). Using a conventional monocyte adhesion assay (48) and fluidic AFM, we tested whether the ISD and its mutations affect the monocyte-endothelial cell adhesion as part of the inflammatory process. For that, THP-1 monocytes were labeled with the fluorescent dye Calcein-AM and added atop a human brain microvascular endothelial cell (HBMVEC) monolayer, and the co-cultures were exposed to WT or mutant VLPs (0.3 µg/mL) followed by incubation for 72 h. Next, the cells were washed with phosphate-buffered saline (PBS), and the relative fluorescence intensities for the number of Calcein-AM-labeled THP-1 cells adhered to the monolayer of HBMVEC were scored using relative fluorescence units (rfu) (49). We found that WT VLPs increased THP-1 adherence to HBMVEC cells compared to mock-treated cells, and mut 5 demonstrated only a marginal reduction of the adherence. In contrast, mut 14 demonstrated further increased adherence (Fig. 6A). Therefore, mut 14 was further characterized by measuring the adhesion force between single-live THP-1 and single-live HBMVEC cells by fluidic atomic force microscopy (fluidic AFM). Again, we observed increased adhesion forces for mut 14, compared with WT VLPs (Fig. 6B). Thus, the anti-inflammatory effect of Trp-14 in WT ISD is accompanied by reduced monocyte adhesion.
We hypothesized that the higher monocyte-endothelial cell adhesion by mut 14 can be associated with a reduction of cAMP, resulting in a decrease of EPAC-1 levels (45). To test this hypothesis, we incubated THP-1-HBMVEC cocultures with mut 14 VLP for 6 h and then treated them with ESA I942, a specific activator of EPAC-1 expression, to measure cell-cell adhesion. As expected, the activation of EPAC-1 significantly reduced monocyteendothelial cell adhesion as measured by lower rfu of THP-1 cells (Fig. 6C, E, andF). Similarly, the fluidic AFM analysis confirmed that the activation of EPAC-1 by ESA I942 reduced the adhesion force between THP-1 and HBMVEC (Fig. 6D). Thus, mut 5 demon strated an intact cAMP-EPAC-1 axis, which protects barrier function, contributing to reduced inflammatory response. In contrast, mut 14 demonstrated a compromised cAMP-EPAC-1 axis contributing to the pro-inflammatory response. Overall, these data demonstrate that ISD facilitates the adhesion of monocytes to endothelial cells mediated by EPAC-1, presumably as a consequence of its overall anti-inflammatory effect mediated by its C-terminus.
## The N-and C-termini of shed GP ISD trigger expression of pro-and antiinflammatory cytokines, respectively
We have characterized the effects of ISD on EBOV GP exposed on the surface of VLPs, thus mimicking the effects of GP on the surface of EBOV particles. However, an alterna tive form of GP is released into the extracellular medium from infected cells (11). Specifically, during EBOV infection, the GP ectodomain is cleaved off by TACE, shed extensively, and triggers immune cell activation and cell death (13,50). We hypothesized that ISD in shed GP contributes to immune modulation. To test this hypothesis and investigate the role of the N-and C-termini of the ISD in the context of shed GP alone without any other viral factor, we generated WT, mut 5, and mut 14 shed GP proteins as described previously (13) and incubated them in PBMC cultures for 24 h and 96 h. The supernatants were harvested and analyzed by bead-based multiplex assay to measure the levels of secreted pro-and anti-inflammatory cytokines. Globally, WT, mut 5, and mut 14 shed GPs increased both pro-and anti-inflammatory cytokine levels in comparison to mock-treated PBMCs on both day 1 and day 4 (Fig. S3A andB), thus confirming the previous results obtained with VLPs. A more detailed analysis demonstrated that in comparison to WT shed GP, mut 5 shed GP increased mostly anti-inflammatory cytokines, specifically IL-10, IL-5, IL-13, and IL-4, particularly on day 4 (Fig. 7). In contrast, mut 14 shed GP mostly increased pro-inflammatory cytokines, including GM-CSF, IFNγ, IL-8, IL-12 (p70), IL-6, TNFα, and IL-4. Induction of IL-4 by both mutants is consistent with the observations showing induction of IL-4 by both mut 5 and mut 14 VLPs (Fig. 3E). Overall, the pattern of cytokine expression by shed GP mutants was similar to that of the VLP mutants (Fig. 1C through E left panels, Fig. 2 and3E). Altogether, these data demonstrate that the ISD present in shed GP is capable of modulating the immune response with its N-and C-termini having pro-and anti-inflammatory properties, respectively.
## EBOV shed GP is internalized through interaction with TLR4 and modulates its downstream signaling pathway
We previously reported that shed GP activates immune cells in a TLR4-dependent manner to promote cellular differentiation and viral replication (13). Here, we tested the role of shed GP ISD in the activation of cells through TLR4. First, to evaluate binding of each shed GP version (Fig. 8A, 1st and 3rd column), HEK 293 cells stably expressing TLR4 (293-TLR4), or THP-1 cells were mock-treated or treated with recombinant TLR4 (rTLR4) and exposed to WT or both mutated shed GPs for 2 h on ice to inhibit internalization. Then, to further investigate internalization (Fig. 8A, 2nd and 4th column), following the same procedure, the cells were washed and then either trypsinized for 10 min or mock-treated, and shifted to 37°C for 1 h to promote internalization. Thereafter, the cells were lysed, and the lysates were analyzed by western blotting with antibodies specific to GP or TLR-4. Exposure of cells to WT, mut 5, and mut 14 shed GP demonstrated no difference in cellular binding and internalization, as the GP signal remained unchanged (Fig. 8A, 2nd line), thus corroborating previous observations with VLPs (Fig. 3D; Fig. S2). However, pre-treatment with rTLR4 significantly reduced EBOV GP binding and its internalization (Fig. 8A, 3rd line). Treatment of cells with trypsin has stripped off cell membrane-bound GP, but not the internalized GP (Fig. 8A, 4th line). Together, these data demonstrate that TLR4 plays an important role in the internalization of shed GP and that mutations at positions 5 and 14 in the ISD do not have any impact either on its binding or its internalization.
Next, we determined the capacity of ISD mutations to modulate TLR4-related transcription factor activities involved in multiple cellular processes. 293-TLR4 cells were transfected with NF-κB-Luc and NFAT-Luc plasmids for 48 h and mock-treated or treated with the specific TLR4 signaling inhibitor CLI-095 or rTLR4 with or without CsA, a specific inhibitor of the NFAT signaling pathway. Then, the cells were cultured with WT, mut 5, or mut 14 EBOV shed GP for 24 h and subjected to luciferase assays. As demonstrated previously, WT EBOV shed GP activated both NF-κB and NFAT1 pathways in mock-treated cultures (Fig. 1C, 4A andB and8B andC). Shed GP mut 5 and mut 14 displayed reduction and increase in both NF-kB and NFAT activities, respectively (Fig. 8B andC). Moreover, to characterize the role of TLR4, we pre-treated cell cultures with the specific TLR4 signaling inhibitor CLI-095 or rTLR4, which significantly reduced the activities of both transcription factors perpetrated by the shed GPs (Fig. 8B andC). Together, these data demonstrate that following TLR4 engagement with EBOV shed GP, K588 promotes cellular activation and expression of pro-inflammatory cytokines via NF-kB and NFAT pathways, whereas W597 inhibits these effects.
## DISCUSSION
The ISD, a conserved region in the retroviral envelope glycoproteins, was shown to inhibit immune responses in multiple systems (18,(20)(21)(22). Likewise, EBOV and Marburg virus have a conserved ISD in their glycoprotein (17,23,24,(51)(52)(53). An in vitro study, which used a 17-residue peptide that mimics EBOV GP ISD, has demonstrated an immunosuppressive activity by inhibiting activation of CD4 + and CD8 + T cells (25). As already noted, a linear peptide treatment may not reproduce the biological effects of ISD in a fully conformational GP. As such, we first evaluated the effects of EBOV GP ISD on human PBMCs using EBOV VLPs consisting of the NP, VP40, and GP proteins. We demonstrated that VLPs decrease the transcriptional activity and cytokine expres sion in pre-stimulated PMBCs (Fig. 1C through E). To further investigate ISD properties, we generated 17 VLPs displaying single mutations through alanine or glycine substitu tions in the full sequence of ISD. That strategy allowed us to identify two residues playing important roles in the modulation of inflammation (Fig. 2). Indeed, as mut 5 VLPs enhanced the induction of anti-inflammatory cytokines and reduced activation of T cells, the WT N-terminus sequence of ISD is likely to induce a pro-inflammatory response. In contrast, mut 14 VLPs elicited pro-inflammatory-associated cytokines and T-cell activation, thus suggesting that the WT C-terminus sequence of ISD is involved in its immunosuppressive properties. The pro-inflammatory effect of the ISD N-terminus is remarkable, as it is opposite to the expected immunosuppressive effects of the conserved ISD. These data demonstrate a balance and a possible dichotomy in the cellular response elicited by ISD.
However, it remains unclear how the mutants with opposite effects impact the immune responses. We therefore tested the mutants for cell binding and entry, which did not show any significant difference between WT, mut 5, and mut 14 VLPs (Fig. 3D) and shed GPs (Fig. 9A). We then determined that following cell entry, the ISD mutants differently trigger signaling pathways, as mut 5 VLPs favored the activation of the anti-inflammatory Akt/mTor axis, whereas mut 14 VLPs triggered the pro-inflammatory NFAT and NF-κB pathways in comparison to WT VLPs. As our initial data demonstrate global differences in the induction of pro-and anti-inflammatory responses, we next tested the ability of the VLP mutants to alter cell proliferation and also viral replication. Consistent with cytokine profiles, cells cultured with mut 5 and mut 14 VLPs displayed reduced and increased cell proliferation, respectively. The effect of the mutations in the ISD on EBOV replication was opposite: increased in the presence of mut 5 and reduced in the presence of mut 14, potentially due to the characteristics of the immune environment induced by each mutant, thus hampering or enhancing the antiviral effects, respectively.
EBOV replicates in multiple cell types, resulting in a complex pathogenesis that includes an uncontrolled and unbalanced immune response leading to high fatality rates (3). Indeed, EBOV triggers hypercytokinemia, coagulopathy, vascular damage, and immune-mediated tissue pathologies (15,54), contributing to the damage of vital organs (3). During the process of inflammation, both cytokine induction and breaks in cell-junction barriers facilitate the transmigration of pro-inflammatory cells through capillary endothelial cells (55). Vascular endothelial barrier function is maintained by several mechanisms, such as the one involving the cAMP-EPAC1 axis, which likely reduces inflammation, thus sustaining its shielding properties and minimizing transmigration and inflammation (45). We observed that mut 14 increased monocyte-endothelial cell adhesion with an elevated binding force when compared with the WT or mut 5 VLPs. As mut 14 also reduced cAMP, which acts through EPAC-1 (42,56) to maintain vascular permeability, it is concluded that mut 14 could enhance vascular permeability and transmigration, as it has been demonstrated to promote pro-inflammatory response. These results suggest that the residue W597 reduces the inflammation, contributing to "immunosuppression. " The induction of cAMP by W597 can potentially shut down the proximal T-cell activation, which was previously reported as a negative regulator of T cells (57). On the other hand, the capacity of the residue K588 to keep cAMP at low levels can trigger inflammatory processes, including T-cell activation, cell-cell adhesion, and cellular transmigration.
During EBOV infection, shed GP, which is abundantly released from infected cells, activates bystander immune cells, resulting in excessive release of inflammatory cytokines, increased vascular permeability, and dysregulated inflammatory changes (13,50). To test if this protein further amplifies the effects of GP anchored in viral particles, as demonstrated using VLPs, we generated EBOV shed GP WT, mut 5, and mut 14. Indeed, and similarly to the observations with EBOV VLPs, shed GP mutants did not show any difference in cellular binding and entry. Again, exposure of human PBMCs to the mutated shed GP resulted in differential cellular activation and cytokine response, as mut 5 promoted an anti-inflammatory response, whereas mut 14 enhanced pro-inflammatory response (Fig. S3; Fig. 7). Additional experiments demonstrated that shed GP mut 5 and mut 14 displayed reduction and increase in both NF-kB and NFAT activities, respectively (Fig. 8).
Overall, our data suggest that the ISD N-terminus plays a role in activating immune cells and pro-inflammatory response. In contrast, the C-terminus of ISD downregulates the pro-inflammatory response through the reduction of NF-kB and NFAT activities. Thus, the expected immunosuppressive and anti-inflammatory effects of the ISD are associated with its C-terminus, whereas the N-terminus has unexpected pro-inflammatory properties (Fig. 9). These data also show that EBOV GP increases the adhesion of monocytes to endothelial cells, and the effect is inhibited by the ISD C-terminus (Fig. 9). Moreover, the data show that the immunomodulating effects of ISD are mediated not only by the virus-associated GP but also shed GP, which is abundant in the medium. Although no molecular mechanism of action has ever been associated with EBOV GP ISD, evaluating its potential association with the receptors from the TYRO3 receptor tyrosine kinase family and its transduction signaling might be of interest. Indeed, TYRO3 receptors are responsible for activating the Akt/mTOR pathway, thus promoting an anti-inflammatory response (58,59). Moreover, members of the Tyro3 receptor tyrosine kinase family are involved in cell entry of EBOV (60). Pathogenesis of EBOV disease is characterized by a paradoxical combination of hyperinflammation (61)(62)(63)(64)(65)(66)(67)(68) and features of immunosuppression (29,(69)(70)(71), reviewed in ref (72), which could in part be affected by the complex effects of the ISD presented in this study. These data may be useful for the development of treatments for the disease caused by EBOV by targeting the ISD.
Our study has limitations. First, although VLPs reproduce the three-dimensional structure of EBOV particles, they cannot completely reproduce the complexity of the biological effects of ISD in the context of EBOV infection. Second, some of the effects found in this study, particularly the modulation of the adhesion of monocytes to endothelial cells, should be further investigated in the context of EBOV infection in vivo.
## MATERIALS AND METHODS
## Virus, VLPs, and shed GPs
The recombinant EBOV, strain Mayinga, expressing green fluorescent protein (EBOV-GFP) (28) was recovered from the cDNA and propagated by three passages in Vero E6 cell monolayers as previously described (29). The viral stocks were quantified by plaque titration in Vero E6 monolayers. All work with EBOV was performed in a biosafety level 4 (BSL-4) laboratory of the Galveston National Laboratory. Generation of EBOV VLPs was performed as described previously (13). The mutants coding for pcDNA3.1 mut 5 EBOV GP-eGFP and pcDNA3.1 mut 14 EBOV GP-eGFP were generated using the Q5 site-directed mutagenesis kit (New England Biolabs) according to the manu facturer's procedure. EBOV VLPs were generated using the plasmids WT, mut 5, or mut 14 pcDNA3.1 EBOV GP-eGFP, VP40 pWRG7077:64759-2010-233-1-4_VP40_optVP40 provided by Dr. Sina Bavari (U.S. Army Medical Research Institute of Infectious Diseases), and EBOV NP pCEZ-NP provided by Dr. Yoshihiro Kawaoka (University of Wisconsin) following transfection in 293T cells using TransIT-LT1 Mirus reagent (Mirus Bio) for 72 h. Transfected cells were collected and centrifuged at 10,000 × g for 10 min to remove cell debris. Then, cell supernatants were purified by pelleting through a 20% (wt/vol) sucrose cushion at 100,000 × g, 4°C for 120 min using a Beckman ultracentrifuge. The pellets were resuspended in 20 mL of PBS and further purified by ultracentrifugation at 100,000 × g, 4°C for 60 min. In parallel, VLPs were also concentrated using the Lenti-X concentrator reagent (Takara) following the manufacturer's protocol. Each suspension containing the VLPs was determined by western blotting. The total protein concentra tions of the VLP or virus preparations were determined after lysis in Nonidet P-40 detergent by using a Pierce BCA assay (Thermo Fisher Scientific). Briefly, the protein concentration in samples was determined using BCA reagent and read at 562 nm. The BSA standards were used to determine the protein concentration in samples. Recombi nant shed GPs were generated by transfection of WT, mut 5, or mut 14 pcDNA3.1 EBOV GP-eGFP plasmid in 293T cells using TransIT-LT1 Mirus reagent (Mirus Bio) for 72 h. Transfected cells were collected and centrifuged at 10,000 × g for 10 min to remove cell debris, and supernatants containing shed GP were concentrated using Centricon-Plus 70 centrifugal filter units (EMD Millipore) following the manufacturer's recommendations. Concentration of shed GPs in each sample was determined using Pierce BCA Protein Assay (ThermoFisher Scientific). To ensure that we use equal concentrations of the shed GPs WT, mut 5, and mut 14, we performed western blot analysis using an anti-EBOV GP (Integrated BioTherapeutics, #0201-020), followed by densitometry analysis (ImageJ software) for normalization. In parallel, VLPs were characterized by western blot analysis using anti-EBOV VP40 (Integrated BioTherapeutics, #0301-010), anti-EBOV NP (Integrated BioTherapeutics, #0301-012), and anti-EBOV GP (Integrated BioTherapeutics, #0201-020).
## Cells
Human monocytic cells THP-1 obtained from the American Type Culture Collection (ATCC) and THP-1 Blue NF-κB cells (InvivoGen) were cultured in RPMI 1640 medium (Thermo Fisher Scientific) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (GE Hyclone) and 1% HEPES (Corning). Vero-E6 and 293T cell lines (obtained from the ATCC) and HEK 293-TLR4 cell line (293/HTLR4a. InvivoGen) were cultured in Dulbecco's modified Eagle's medium, supplemented with 10% FBS (Thermo Fisher Scientific), 1% HEPES (Corning), 1% nonessential amino acids (Sigma-Aldrich), 1% sodium pyruvate (Sigma-Aldrich), and 2% penicillin-streptomycin mix (Thermo Fisher Scientific). Human brain microvascular endothelial cells (HBMVECs) (10HU-051, iXCells Biotechnolo gies) were grown in endothelial cell growth medium (Cell Applications) supplemented with 10% heat-inactivated FBS with humidity in 5% CO 2 at 37°C. Cells were maintained in human endothelial cell growth medium during the experiments. The temperature controller (NanoSurf ) kept the liquid environment at 37°C during all measurements.
## Isolation and culture of human PBMCs
Buffy coats were obtained from deidentified healthy adult donors according to a clinical protocol approved by the University of Texas Medical Branch at Galveston (UTMB) Institutional Review Board. PBMCs were isolated by Histopaque (Sigma-Aldrich) gradient as recommended by the manufacturer. Then, CD3 + T-cells and CD20 + B-cells were isolated from fresh PBMCs using a cell-specific negative selection enrichment kit (StemCell Technologies). Purity of the isolated lymphocytes typically ranged from 93% to 95% as determined by flow cytometry using a LSR Fortessa flow cytometer (BD Biosciences) at the UTMB Flow Cytometry Core Unit. Data were analyzed using FlowJo v8 (FlowJo, LLC).
## VLP treatment, cAMP level evaluation, and flow cytometry analysis of PBMCs
PBMCs were cultured at 10 6 per milliliter and pulsed with EBOV VLPs at 10 µg/mL and maintained in RMPI-1640 medium (Sigma-Aldrich) supplemented with 10% of heat-inactivated FBS. As a control, cells were stimulated with TPA (Sigma-Aldrich) at 25 ng/mL and 0.5 µM of ionomycin (Sigma-Aldrich) or 10 µM Forskolin (Sigma-Aldrich). Two hours following pulsing, Brefeldin-A (Sigma-Aldrich) was added at 10 µg/mL, and the cells were incubated for an additional 16 h at 37°C. Levels of cAMP in PBMCs were evaluated using a specific cyclic AMP ELISA kit (Cayman) following the manufacturer's instructions. In parallel, PBMCs were washed twice with PBS by spinning at 250 × g for 5 min at 4°C and fixed with 0.1 mL 4% paraformaldehyde (Fisher Bioreagents). PBMCs were permeabilized with permeabilization buffer (eBiosciences), stained with antibodies specific for IL-2 (BD Biosciences, #341116), IL-10 (BD Biosciences, #554707), IL-12 (BD Biosciences, #554576), TNFα (BD Biosciences, #340534), IFNγ (BD Biosciences, #554702), or IL-4 (BD Biosciences, #554485), washed with PBS, and resuspended in 500 µL of PBS. To evaluate activation markers on T-lymphocytes, cells were surface-stained with antibodies specific for CD25 (BD Biosciences, #347643), CD69 (BD Biosciences, #340560), or CD2 (BD Biosciences, #555327) for 30 min and washed. Flow cytometry was performed using a LSR Fortessa flow cytometer (BD Biosciences) at the UTMB Flow Cytometry Core Unit, and data were analyzed using FlowJo.
## Cell proliferation assay and EBOV infectivity
Proliferation assay was performed using CellTrace CFSE Cell Proliferation Kit (Thermo Fisher Scientific) following the manufacturer's instructions. Briefly, fresh PBMCs were obtained from healthy donor blood as described above. PBMCs were labeled with CFSE and incubated with 20% FBS to quench excess CFSE and washed three times with PBS. Cells were mock-treated or cultured with VLPs for 48 h at 37°C. Then, cells were harvested and analyzed by flow cytometry for CFSE dilution. To evaluate EBOV infectivity, THP-1 cells were pulsed with medium alone or with VLPs for 24 h. Then, cells were centrifuged for 5 min at 250 × g, supernatants were removed, and the cells were cultured with EBOV-GFP at an MOI of 0.1 PFU/cell for an additional 48 h. Cells were harvested, fixed with 4% formaldehyde for 24 h at 4°C, and analyzed for GFP expression by flow cytometry.
## Binding and internalization assays
Human donor PBMCs were isolated with a Histopaque (Sigma-Aldrich) gradient as recommended by the manufacturer. Then, T-lymphocytes and B-lymphocytes were isolated from PBMCs by negative selection using magnetic microbead separation kits (Miltenyi Biotec) to keep CD3 + and CD19 + cells, respectively. The cells were cultured with fresh medium alone or in the presence of WT, mut 5, or mut 14 EBOV VLP for 2 h on ice. Thereafter, cells were immunostained with rabbit antibodies raised against EBOV virus-like particles (VLP) (Integrated BioTherapeutics) and analyzed by flow cytometry. To evaluate the role of TLR4 in binding and internalization of shed GP, HEK 293-TLR4, or THP-1 cells were plated at 10 6 cells per well in U-bottom 96-well plates (Thermo Fisher Scientific), mock-treated or pre-treated with rTLR4 (RnD Systems, #1478-TR-050), and placed on ice to prevent internalization of VLPs. EBOV shed GPs were added, and cells were incubated for 2 h on ice and washed with PBS containing 2% heat-inactiva ted FBS. Then, to investigate internalization, cells were either Immediately trypsinized or incubated at 37°C for 1 h to promote internalization and then trypsinized; non-tryp sinized samples were used as a control. Thereafter, cell lysates were immunostained with the following antibodies: anti-EBOV GP (Integrated BioTherapeutics, #0201-020), anti-TLR4 (Santa Cruz Biotechnology, #sc-293072), and anti-GAPDH (Cell Signaling, #8884S).
## Analysis of signaling pathways
Isolated PBMCs were plated at a concentration of 2 × 10 6 cells per well in 24-well plates (Thermo Fisher Scientific). Then, the cells were stimulated or not with 25 ng/mL TPA (Sigma-Aldrich) and with 0.5 µM ionomycin (Sigma-Aldrich), treated or not with 1 µM of CsA (Sigma-Aldrich), or incubated with WT, mut 5, and mut 14 VLPs for 16 h at 37°C, and washed with PBS containing 2% heat-inactivated FBS. Cell lysates were collected for western blot analysis with antibodies specific for multi-p NFAT1 and mono-p NFAT1 (Novus Biologicals, #25A10.D6.D2), p-AkT (Cell Signaling, #9271), AkT (Cell Signaling, #4691), p-mTOR (Cell Signaling, #2971) and mTOR (Cell Signaling, #2972), EBOV VP40 (Integrated BioTherapeutics, #0301-010), and GAPDH (Cell Signaling, #8884S).
## Imaging flow cytometry analysis
THP-1 cells were transfected with a plasmid coding for NFAT1-GFP protein previously described (73) using Lipofectamine LTX reagent (ThermoFisher Scientific) according to the manufacturer's instructions. Then, transfected THP-1 cells were cultured at 10 6 cells per well in a 96-well plate, stimulated or not with TPA (Sigma-Aldrich) at 25 ng/mL and ionomycin at 0.5 µM (Sigma-Aldrich), cultured with medium alone or with VLPs for 16 h, harvested, and stained as described for flow cytometry experiments above. Analyses of EBOV VLP proteins and NFAT1-GFP proteins were performed using an AMNIS FlowSight Imaging flow cytometer (Sigma-Aldrich) with a minimum of 2,000 events acquired. The purity of the virions was monitored and confirmed during imaging flow cytometry experiments with gating by "aspect ratio" on the y-axis and "area" on the x-axis in the bright-field channel. Data were analyzed with the IDEAS version 2.0 software.
## Luciferase assays
293T and 293-TLR4 cells were seeded at 10 5 cells per well in 12-well plates (Sigma-Aldrich), transfected with NFκB-Luc (Addgene, #111216) or NFAT-Luc (Addgene, #17870) plasmids using TransIT LT1 transfection reagent (Mirus Bio LLC) and incubated at 37°C for 48 h. Cells were then stimulated with 25 ng/mL TPA and 0.5 µM of ionomycin, or 1 µM of CsA, 10 µg/mL of rTLR4 (RnD Systems, #1478-TR-050), or 100 ng/mL CLI-095 (InvivoGen) for 1 h. Next, cells were pulsed with medium alone or with EBOV VLPs for an additional 24 h. Then, cells were lysed with Pierce Luciferase Cell lysis buffer (Thermo Fisher Scientific), and cell lysates were assayed for luciferase activity using a luminometer (Glomax 20/20, Promega). Bicinchoninic acid (BCA) protein assays (Thermo Scientific) were used for normalization.
## Multiplex analysis of serum cytokines and chemokines
Isolated PBMCs were seeded at 10 6 cells per ml in a 12-well plate (Corning). Then, cells were cultured with medium alone or in the presence of WT shed GP, mut 5 shed GP, or mut 14 shed GP for 24 h or 96 h. Supernatants were harvested and centrifuged at 9,000 × g for 10 min at 4°C to remove cell lysates or debris. Next, supernatants were analyzed using a Multiplex magnetic bead-based assay (Eve Technologies) to evaluate cytokine levels for GM-CSF, IFNγ, IL-6, IL-8, IL-12(p70), TNFα, IL-4, IL-10, IL-5, and IL-13.
## Monocyte adhesion assay
Human brain microvascular endothelial cell (HBMVEC), passages 6 and 7, were seeded on 48-well plates and cultured at 37°C and 5% CO 2 . The traditional fluorescent microscopybased monocyte adhesion assay was processed as previously described (48). Briefly, calcein-AM-labeled THP-1 monocytes (5 × 10 4 cells per well) were added to HBMVEC monolayer (5 × 10 4 cells/well) in 48-well plates. After co-cultures for a designed time, non-adherent monocytes were gently washed off using PBS, and adherent monocytes were fixed with 4% paraformaldehyde. Monocyte adhesion was visualized under a fluorescent microscope (Olympus BX51) using a 10 × objective. The relative fluorescence intensity was calculated using ImageJ software from three different fields per well. The results are representative of at least three independent experiments.
## Fluidic AFM single-living THP-1-HBMVEC vertical binding force (VBF) measurement
For the fluidic AFM studies, the HBMVECs were co-cultured with THP-1 cells at 1:1 ratio. The fluidic AFM system coupling Nanosurf Core AFM (Nanosurf ) and Fluidic Pressure Controller (Cytosurge AG) was used for this assay. The air in the reservoir on the backside of the micropipette microchannel was removed by PBS, and the reservoir was connec ted to the Pneumatic Connector (Cytosurge AG). After subsequent connection to the Fluidic Pressure Controller, positive pressure (20 mBar) was applied to enable PBS to flow through the microchannel within the micropipette. Under a phase contrast microscope, the micropipette Fluidic AFM cantilever (Cytosurge AG) approached a single monocyte in medium, and a -800-mbar pressure was applied to capture a single THP-1 cell by adsorbing it onto the aperture of the micropipette. This micropipette was then used as a cell probe to measure the VBFs between an HBMVEC and the THP-1 cell on the micropipette using force spectroscopy as we described (45). The cell-micropipette was driven to approach the HBMVEC monolayer with 2 nN as the setpoint force and paused on the surface of the cell for 0.5 min. This defined time established the interaction on the surfaces between a single THP-1 and HBMVEC. The force spectroscopy was performed to measure the unbinding force during rupture of the interaction between the THP-1 and HBMVEC. The unbinding effort was assessed by measuring the work done (in picojoules [pJ]), which was calculated by integrating the area under the force-distance (F-D) curve (representative curves shown in Fig. 6F) using software as described previously (47,74). The Temperature Controller (Nanosurf ) kept the liquid environment at 37°C during all measurements. To avoid any cell debris affecting the microfluid channel of the canti lever from previous measurements, a new micropipette was routinely installed for a new measurement. At least five single sets of THP-1-HBMVEC were measured with five cantilevers, respectively, per group (n = 5 per group in Fig. 6B andD).
## Statistical analyses
Each independent experiment was performed in biological duplicates or triplicates. Data are presented as mean ± SEM of at least three independent experiments. Statistical methods used are mentioned in figure legends, and the statistics were calculated using GraphPad Prism 6 (GraphPad). P values of <0.05 were considered statistically significant.
## References
1. Feldmann, Geisbert (2011) "Ebola haemorrhagic fever"
2. Prescott, Marzi, Safronetz et al. (2017) "Immunobiology of Ebola and Lassa virus infections" *Nat Rev Immunol*
3. Baseler, Chertow, Johnson et al. (2017) "The pathogenesis of Ebola virus disease" *Annu Rev Pathol*
4. Mohamadzadeh, Chen, Schmaljohn (2007) "How Ebola and Marburg viruses battle the immune system" *Nat Rev Immunol*
5. Jun, Leuze, Nookaew et al. (2015) "Ebolavirus comparative genomics"
6. Sanchez, Trappier, Mahy et al. (1996) "The virion glycoproteins of Ebola viruses are encoded in two reading frames and are expressed through transcriptional editing" *Proc Natl Acad Sci*
7. Volchkova, Klenk, Volchkov (1999) "Delta-peptide is the carboxy-terminal cleavage fragment of the nonstructural small glycoprotein sGP of Ebola virus" *Virology (Auckl)*
8. Lee, Fusco, Hessell et al. (2008) "Structure of the Ebola virus glycoprotein bound to an antibody from a human survivor" *Nature*
9. Volchkov, Feldmann, Volchkova et al. (1998) "Processing of the Ebola virus glycoprotein by the proprotein convertase furin" *Proc Natl Acad Sci*
10. Jeffers, Sanders, Sanchez (2002) "Covalent modifications of the Ebola virus glycoprotein" *J Virol*
11. Dolnik, Volchkova, Garten et al. (2004) "Ectodomain shedding of the glycoprotein GP of Ebola virus" *EMBO J*
12. Okumura, Pitha, Yoshimura et al. (2010) "Interaction between Ebola virus glycoprotein and host Toll-like receptor 4 leads to induction of proinflammatory cytokines and SOCS1" *J Virol*
13. Iampietro, Santos, Lubaki et al. (2018) "Ebola virus shed glycoprotein triggers differentiation, infection, and death of monocytes through toll-like receptor 4 activation" *J Infect Dis*
14. Martinez, Tantral, Mulherkar et al. (2011) "Impact of Ebola mucin-like domain on antiglycoprotein antibody responses induced by ebola virus-like particles" *J Infect Dis*
15. Yang, Duckers, Sullivan et al. (2000) "Identification of the Ebola virus glycoprotein as the main viral determinant of vascular cell cytotoxicity and injury" *Nat Med*
16. Volchkov, Volchkova, Muhlberger et al. (2001) "Recovery of infectious Ebola virus from complementary DNA: RNA editing of the GP gene and viral cytotoxicity" *Science*
17. Volchkov, Blinov, Netesov (1992) "The envelope glycoprotein of Ebola virus contains an immunosuppressive-like domain similar to oncogenic retroviruses" *FEBS Lett*
18. Cianciolo, Copeland, Oroszlan et al. (1985) "Inhibition of lymphocyte proliferation by a synthetic peptide homologous to retroviral envelope proteins" *Science*
19. Blinov, Krasnov, Shargunov et al. (2013) "Mechanisms of retroviral immunosuppressive domain-induced immune modulation" *Mol Biol (Moskow)*
20. Eksmond, Jenkins, Merkenschlager et al. (2017) "Mutation of the putative immunosuppressive domain of the retroviral envelope glycoprotein compromises infectivity" *J Virol*
21. Denner (2014) "The transmembrane proteins contribute to immunodeficiencies induced by HIV-1 and other retroviruses" *AIDS*
22. Morozov, Morozov, Semaan et al. (2012) "Single mutations in the transmembrane envelope protein abrogate the immunosuppres sive property of HIV-1" *Retrovirology (Auckl)*
23. Bukreyev, Volchkov, Blinov et al. (1993) "The GP-protein of Marburg virus contains the region similar to the "immunosuppressive domain" of oncogenic retrovirus P15E proteins" *FEBS Lett*
24. Weissenhorn, Carfí, Lee et al. (1998) "Crystal structure of the Ebola virus membrane fusion subunit, GP2, from the envelope glycoprotein ectodomain" *Mol Cell*
25. Yaddanapudi, Palacios, Towner et al. (2006) "Implication of a retrovirus-like glycoprotein peptide in the immunopathogenesis of Ebola and Marburg viruses" *FASEB J*
26. Fujioka, Niu, Schmidt et al. (2004) "NF-kappaB and AP-1 connection: mechanism of NF-kappaB-dependent regulation of AP-1 activity" *Mol Cell Biol*
27. Saraiva, Vieira, O'garra (2020) "Biology and therapeutic potential of interleukin-10" *J Exp Med*
28. Towner, Paragas, Dover et al. (2005) "Generation of eGFP expressing recombinant zaire ebolavirus for analysis of early pathogenesis events and highthroughput antiviral drug screening" *Virology (Auckland)*
29. Lubaki, Ilinykh, Pietzsch et al. (2013) "The lack of maturation of ebola virus-infected dendritic cells results from the cooperative effect of at least two viral domains" *J Virol*
30. Anovazzi, Medeiros, Pigossi et al. (2017) "Functionality and opposite roles of two interleukin 4 haplotypes in immune cells" *Genes Immun*
31. Luzina, Keegan, Heller et al. (2012) "Regulation of inflammation by interleukin-4: a review of "alternatives" *J Leukoc Biol*
32. Schlecht-Louf, Mangeney, Letzelter et al. (2010) "Retroviral infection in vivo requires an immune escape virulence factor encrypted in the envelope protein of oncoretroviruses" *Proc Natl Acad Sci*
33. Hogan (2017) "Calcium-NFAT transcriptional signalling in T cell activation and T cell exhaustion" *Cell Calcium*
35. Macian (2005) "NFAT proteins: key regulators of T-cell development and function" *Nat Rev Immunol*
36. Hermann-Kleiter, Baier (2010) "NFAT pulls the strings during CD4+ T helper cell effector functions" *Blood*
37. Martinez, Pereira, Äijö et al. (2015) "The transcription factor NFAT promotes exhaustion of activated CD8" *Immunity*
38. Shin, Kim, Cho et al. (2019) "Coupled feedback regulation of nuclear factor of activated T-cells (NFAT) modulates activation-induced cell death of T cells" *Sci Rep*
39. Rao, Luo, Hogan (1997) "Transcription factors of the NFAT family: regulation and function" *Annu Rev Immunol*
40. Okamura, Aramburu, García-Rodríguez et al. (2000) "Concerted dephosphorylation of the transcription factor NFAT1 induces a conformational switch that regulates transcriptional activity" *Mol Cell*
41. Weichhart, Säemann (2008) "The PI3K/Akt/mTOR pathway in innate immune cells: emerging therapeutic applications" *Ann Rheum Dis*
42. Mathieu, Younan, Nishida et al. (2017) "Ebola virus glycoprotein directly triggers T lymphocyte death despite of the lack of infection" *PLoS Pathog*
43. Parnell, Smith, Palmer et al. (2012) "Regulation of the inflammatory response of vascular endothelial cells by EPAC1" *Br J Pharmacol*
44. De Rooij, Zwartkruis, Verheijen et al. (1998) "Epac is a Rap1 guanine-nucleotide-exchange factor directly activated by cyclic AMP" *Nature*
45. Kawasaki, Springett, Mochizuki et al. (1998) "A family of cAMP-binding proteins that directly activate Rap1" *Science*
46. Xiao, Zhang, Su et al. (2021) "Intracellular receptor EPAC regulates von willebrand factor secretion from endothelial cells in a PI3K-/eNOSdependent manner during inflammation" *J Biol Chem*
47. Muller (2002) "Leukocyte-endothelial cell interactions in the inflammatory response" *Lab Invest*
48. Qiu, Chien, Maroulis et al. (2022) "Extending applications of AFM to fluidic AFM in single living cell studies" *J Cell Physiol*
49. Bocquet, Tembely, Rioult et al. (2021) "Characterization of novel interactions with membrane NEU1 highlights new regulatory functions for the elastin receptor complex in monocyte interaction with endothelial cells" *Cell Biosci*
50. Liu, Zhou, Su et al. "Gong B. 2021. Endothelial exosome plays a functional role during rickettsial infection" *mBio*
51. Escudero-Pérez, Volchkova, Dolnik et al. (2014) "Shed GP of ebola virus triggers immune activation and increased vascular permeability" *PLoS Pathog*
52. Becker (1995) "Retrovirus and filovirus "immunosuppressive motif" and the evolution of virus pathogenicity in HIV-1, HIV-2, and Ebola viruses" *Virus Genes*
53. Weissenhorn, Calder, Wharton et al. (1998) "The central structural feature of the membrane fusion protein subunit from the ebola virus glycoprotein is a long triple-stranded coiled coil" *Proc Natl Acad Sci*
54. Bukreyev, Volchkov, Blinov et al. (1993) "The VP35 and VP40 proteins of filoviruses. homology between marburg and ebola viruses" *FEBS Lett*
55. Villinger, Rollin, Brar et al. (1999) "Markedly elevated levels of interferon (IFN)-gamma, IFN-alpha, interleukin (IL)-2, IL-10, and tumor necrosis factor-alpha associated with fatal Ebola virus infection" *J Infect Dis*
56. Muller (2013) "Getting leukocytes to the site of inflammation" *Vet Pathol*
57. Moy, Bodmer, Blackwell et al. (1998) "cAMP protects endothelial barrier function independent of inhibiting MLC20dependent tension development" *Am J Physiol*
58. Wehbi, Taskén (2016) "Molecular mechanisms for cAMP-mediated immunoregulation in T cells -role of anchored protein kinase A signaling units" *Front Immunol*
59. Weichhart, Haidinger, Katholnig et al. (2011) "Inhibition of mTOR blocks the anti-inflammatory effects of glucocorticoids in myeloid immune cells" *Blood*
60. Lee, Chun (2019) "Anti-inflammatory role of TAM family of receptor tyrosine kinases via modulating macrophage function" *Mol Cells*
61. Shimojima, Takada, Ebihara et al. (2006) "Tyro3 family-mediated cell entry of Ebola and Marburg viruses" *J Virol*
62. Cilloniz, Ebihara, Ni et al. (2011) "Functional genomics reveals the induction of inflammatory response and metalloproteinase gene expression during lethal Ebola virus infection" *J Virol*
63. Menicucci, Versteeg, Woolsey et al. (2017) "Transcriptome analysis of circulating immune cell subsets highlight the role of monocytes in zaire ebola virus makona pathogenesis" *Front Immunol*
64. Jankeel, Menicucci, Woolsey et al. (2020) "Early transcriptional changes within liver, adrenal gland, and lymphoid tissues significantly contribute to ebola virus pathogenesis in cynomolgus macaques" *J Virol*
65. Wiedemann, Foucat, Hocini et al. (2020) "Long-lasting severe immune dysfunction in ebola virus disease survivors" *Nat Commun*
66. Kotliar, Lin, Logue et al. (2020) "Single-cell profiling of ebola virus disease in vivo reveals viral and host dynamics" *Cell*
67. Colavita, Biava, Castilletti et al. (2014) "Inflammatory and humoral immune response during ebola virus infection in survivor and fatal cases occurred in sierra leone during the" *Viruses*
68. Patrick, Iampietro, Nishida et al. (2017) "Ebola virus binding to Tim-1 on T lymphocytes induces a cytokine storm" *mBio*
69. Younan, Ramanathan, Graber et al. (2017) "The toll-like receptor 4 antagonist eritoran protects mice from lethal filovirus challenge" *mBio*
70. Yen, Mulder, Martinez et al. (2014) "Molecular basis for ebolavirus VP35 suppression of human dendritic cell maturation" *J Virol*
71. Lubaki, Younan, Santos et al. (2016) "The ebola interferon inhibiting domains attenuate and dysregulate cell-mediated immune responses" *PLoS Pathog*
72. Younan, Santos, Ramanathan et al. (2019) "Ebola virus-mediated T-lymphocyte depletion is the result of an abortive infection" *PLoS Pathog*
73. Messaoudi, Amarasinghe, Basler (2015) "Filovirus pathogenesis and immune evasion: insights from ebola virus and marburg virus" *Nat Rev Microbiol*
74. Iampietro, Morissette, Gravel et al. (2014) "Inhibition of interleukin-2 gene expression by human herpesvirus 6B U54 tegument protein" *J Virol*
75. Bei, Miranda-Morales, Gan et al. (2023) "Circulating exosomes from alzheimer's disease suppress vascular endothelial-cadherin expression and induce barrier dysfunction in recipient brain microvascular endothelial cell" *J Alzheimers Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12451611&blobtype=pdf | # ICTV Virus Taxonomy Profile: Duplodnaviria 2025
Evelien Adriaenssens, Ryan Cook, Valerian Dolja, Eugene Koonin, Mart Krupovic, Jens Kuhn, Cédric Lood, Alejandro Reyes Muñoz, Dann Turner, Arvind Varsani, Paola Vaz, Thomas Waltzek, Yuri Wolf, Natalya Yutin, F Murilo Zerbini
## Abstract
The realm Duplodnaviria includes viruses of archaea, bacteria and eukaryotes, with linear dsDNA genomes. Duplodnavirians share a distinct morphogenetic module of four hallmark genes encoding the HK97-fold major capsid protein, a genome packaging ATPasenuclease (large terminase subunit), a portal protein and a capsid maturation protease. This is a summary of the International Committee on Taxonomy of Viruses (ICTV) Report on the realm Duplodnaviria, which is available at ictv.global/report/duplodnaviria.
## VIRION
Particles of viruses of the class Herviviricetes (phylum Peploviricota, eukaryotic hosts) are 150-200 nm in diameter and are pleomorphic, mostly spherical, with a glycoprotein-containing lipid envelope that encloses a tegument and an icosahedral capsid.
Particles of viruses of the class Caudoviricetes (phylum Uroviricota, bacterial or archaeal hosts) are head-tailed and do not have an envelope [1,2]. The head is icosahedral and may be isometric or prolate (head diameter 40-200 nm). The tail is typically 10-350 nm in length but can be around 800 nm in some bacterial viruses [3].
The protein composition of mature particles varies greatly among viruses of different families. Only herviviricete particles contain lipids. Carbohydrates have been reported for herviviricete virions [4] and certain caudoviricetes infecting mycobacteria [5].
## GENOME
All duplodnavirians have linear dsDNA genomes when packaged in the particle. Genome lengths are 108.4-322.3 kbp for herviviricetes and 11.6->660 kbp for caudoviricetes. Depending on the replication and packaging mechanisms, genomes can have defined termini with reiterated sequences (direct or inverted repeats), cohesive 3′-or 5′-overhangs, or circularly permuted genome architectures [6]. Genes with associated functions are typically clustered, with conserved, essential genes localized internally/centrally in the genome; scattered gene arrangements also occur.
## REPLICATION
Duplodnavirians with longer genomes usually encode their own DNA polymerases [7]. Herpesvirals uniformly encode family B DNA polymerases and several other replication
$$ICTV$$
## TAXONOMY
Current taxonomy: ictv.global/taxonomy. The realm Duplodnaviria was established in 2020 (Master Species List #35) (Fig. 1).
Members have a dsDNA genome encoding a morphogenetic module consisting of the major capsid protein with the HK97 structural fold, a portal protein, the terminase complex and a capsid maturation protease. The module is distinct from that encoded by other known dsDNA viruses, including those in the realm Varidnaviria.
## RESOURCES
Full ICTV Report on the realm Duplodnaviria: duplodnaviria.
## References
1. Dion, Oechslin, Moineau (2020) "Phage diversity, genomics and phylogeny" *Nat Rev Microbiol*
2. Baquero, Liu, Wang et al. (2020) "Structure and assembly of archaeal viruses" *Adv Virus Res*
3. Minakhin, Goel, Berdygulova et al. (2008) "Genome comparison and proteomic characterization of Thermus thermophilus bacteriophages P23-45 and P74-26: siphoviruses with triplex-forming sequences and the longest known tails" *J Mol Biol*
4. Johnson, Spear (1983) "O-linked oligosaccharides are acquired by herpes simplex virus glycoproteins in the Golgi apparatus" *Cell*
5. Freeman, Robotham, Parks et al. (2023) "Virion glycosylation influences mycobacteriophage immune recognition" *Cell Host Microbe*
6. Gulyaeva, Garmaeva, Kurilshikov et al. (2022) "Diversity and ecology of Caudoviricetes phages with genome terminal repeats in fecal metagenomes from four Dutch cohorts" *Viruses*
7. Kazlauskas, Krupovic, Venclovas (2016) "The logic of DNA replication in double-stranded DNA viruses: insights from global analysis of viral genomes" *Nucleic Acids Res*
8. Yutin, Tolstoy, Mutz et al. (2024) "DNA polymerase swapping in Caudoviricetes bacteriophages" *Virol J*
9. Klupp, Mettenleiter (2023) "The knowns and unknowns of herpesvirus nuclear egress" *Annu Rev Virol*
10. Schwarzer, Hackl, Oksanen et al. (2023) "Archaeal host cell recognition and viral binding of HFTV1 to its Haloferax host" *mBio* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12570470&blobtype=pdf | # Repeated thermal stress exposure in Aedes aegypti co-infected with Wolbachia and dengue virus
Suk Ser, Fhallon Ware-Gilmore, Nina Dennington, Adam Miller, Brianna Mcnulty, Makael Harris, Matthew Jones, Matthew Hall, Carla Sgrò, Katriona Shea, Elizabeth Mcgraw
## Abstract
Climate change is increasing the frequency and intensity of heatwaves, affecting the thermal tolerance of mosquitoes and potentially influencing the efficacy of the biological control agent, Wolbachia. This study investigates the impact of repeated thermal stress on Aedes aegypti mosquitoes co-infected with Wolbachia and dengue virus (DENV). We exposed infected mosquitoes (singly and in co-infection) to varying intensities, frequencies, and durations of thermal stress to assess their thermal sensitivity via a "knockdown assay" compared to uninfected controls. Our results demonstrate that co-infection with Wolbachia and DENV significantly increases thermal sensitivity, with mosquitoes exhibiting a twofold faster median knockdown time than either singly infected or uninfected controls in most cases. A comparison of mosquitoes with no prior heat exposure to those given a single exposure revealed some evidence of heat hardening, or a slight lengthening of time to knockdown. Additional exposures provided no substantial benefit, however. Extended thermal stress (60 mins) also significantly reduced DENV loads, while Wolbachia loads remained stable, indicating that prolonged heat may disrupt viral replication without affecting bacterial symbiosis. These findings suggest that heatwaves could lower vector competence and disproportionately affect DENV-infected mosquitoes in Wolbachia-release areas, with implications for biocontrol strategies. Field studies should explore how infection affects mosquitoes' ability to modulate thermal exposure behaviorally, providing insights for optimizing Wolbachiabased control efforts. IMPORTANCE Dengue virus (DENV), spread by the mosquito Aedes aegypti, is a major global health threat affecting millions of people. This study examines how repeated exposures to heat stress affect the thermal tolerance of mosquitoes infected with DENV and/or Wolbachia, a bacterium used for biological control. These repeated exposures mimic the experience of mosquitoes in the wild experiencing heatwaves of increasing frequency under climate change. Our research shows that Ae. aegypti co-infected with Wolbachia and DENV is more susceptible to thermal stress than singly infected or uninfected mosquitoes. We also demonstrate that multiple independent thermal stress exposures do not exacerbate the effect of infection. Understanding these interactions is essential for predicting how climate change may affect dengue transmission and the resilience of Wolbachia-based interventions.KEYWORDS mosquito, dengue virus, vector-biology, climate change, thermal tolerance, Wolbachia, vector control, heat wave G lobal warming is currently increasing average temperatures and is associated with more frequent and intense heatwaves (1). Heatwaves are sudden temperature rises relative to the expected conditions of the area at that time of the year (2). These climate changes can profoundly affect the distribution of vector-borne diseases by altering
the habitat suitability of vectors (3,4). Heatwaves can accelerate mosquito life cycles, shorten the extrinsic incubation period, and increase generational turnoverfactors that together can enhance pathogen transmission and potentially lead to more severe disease outbreaks (4)(5)(6)(7). Dengue fever is the most prevalent arboviral disease globally, with an estimated 390 million cases occurring yearly (8). The primary vector for dengue virus (DENV) is the mosquito, Aedes aegypti, an anthropophilic species highly adapted to urban environments and human settings (9). With rising global temperatures, the geographic range of these mosquitoes is expected to expand, creating a redistribu tion of the arbovirus transmission (3,4).
To curb the spread of dengue fever, an innovative vector control strategy involving Wolbachia pipientis, an insect endosymbiont, has been recently developed and tested in multiple field release programs globally (10)(11)(12)(13). In Ae. aegypti, this vertically inherited, self-spreading bacterium limits the mosquitoes' ability to transmit DENV by interfering with virus replication through a mechanism known as Wolbachia-mediated pathogenblocking (14,15). The most common use of Wolbachia for vector control involves "population replacement" or the release of Wolbachia-infected females into populations to allow the bacterium to spread and replace the wild uninfected population with a population at or near fixation with Wolbachia. This approach has shown the potential to substantially reduce DENV transmission and, hence, lower the incidence of dengue fever (16). In the most rigorous test of the method conducted in Indonesia, dengue incidence declined by 77% inside the release zones of Wolbachia (12). The reduction has ranged from 29.5% to 69% in other localities (10,11,17).
Mosquitoes, Wolbachia, and DENV all have individual thermal tolerances that may be affected by a changing climate and that may contribute to shifting transmission zones. Ae. aegypti, for example, has an operative temperature range between 15.0°C and 35.0°C, with an optimum of 28.0°C (thermal optimum) (18,19). It is predicted that some regions of the world will become too hot for this mosquito, while others that were previously not habitable, including vast sections of North America, will become ideal. DENV, given that it spends time in mammalian hosts, including those with fever, has an optimal range of 34.0-37.0 o C, with slower replication possible at more extreme temperatures (20). Wolbachia's thermal tolerance is thought to partly depend on its history of adaptation. Its persistence may be influenced by the thermal environment of its host species, as Wolbachia must thrive within the physiological limits of its host. For example, wMel and wAlbB are the leading strains transfected into Ae. aegypti for population replacement (10,11). The former appears to be less thermally tolerant, possibly given its origins in a cosmopolitan species Drosophila melanogaster, with a global temperate range (21,22). In contrast, the latter strain originated from Aedes albopictus, a sister species of Ae. aegypti, which, unlike the globally distributed Ae. aegypti, is primarily found in tropical and subtropical regions (21,23).
The effect of infection on the mosquito's thermal tolerance, however, has been less well studied. Beyond mosquitoes, there is growing evidence that invertebrates, including Daphnia and Drosophila, are more susceptible to heat stress when they have a micro bial infection (24)(25)(26). These effects may be direct, due to host immune responses or metabolic costs, or indirect, via microbiota disruption. Wolbachia and dengue infect a range of tissues throughout the mosquito's body, providing ample opportunity to affect local physiology and systemic processes (27,28). The fitness costs of harboring these microbes under optimal conditions are moderate. In general, DENV can cause some reductions in fecundity and lifespan, although this effect may vary depending on the mosquito's genetic background, age, and environmental conditions (29,30). Dengue infection can also sometimes cause behavioral changes by increasing overall activity, altering host-seeking behavior, and increasing probing and blood-feeding attempts (31). Like DENV, Wolbachia can also reduce a mosquito's lifespan, fecundity, and developmen tal time (27,32,33). DENV and Wolbachia also trigger active immune responses in the vector (34)(35)(36). Recently, our group examined how infection with DENV or Wolbachia reduced the mosquito's ability to respond to thermal stress in a static heat knockdown (KD) assay, leading to more rapid death at a critical maximum temperature (37). The magnitude of this reduction was similar for either single or co-infections of DENV or Wolbachia. This led us to hypothesize that there might be a single shared mechanism underpinning the effect, possibly due to indirect effects like energetic tradeoffs or a direct pleiotropic effect given overlapping gene sets in immune and thermal stress responses (34,35,38).
While many studies have focused on exposure to higher mean temperatures, less attention has been given to the fact that climate change includes heatwaves and that mosquitoes might experience more than one thermal stress event during their lifetime. In this study, we exposed Ae. aegypti to repeated thermal heat stresses of varying frequencies, durations, and intensities to determine the interaction between the cumulative effects of thermal stress and DENV and Wolbachia infection. To account for the fact that mosquitoes can fly and seek cooler microclimates during a heatwave, we exposed the mosquitoes to a "short" (10 min) and "long" (60 min) thermal stress event, intended to mimic transient versus extended microhabitat exposures that mosquitoes may encounter in the field. The intensities (i.e., temperatures) of repeated thermal exposure events were 35 o C (median heat stress) and 40 o C (upper thermal limit), and were selected because both temperatures are outside the mosquito's thermal optimum, which would well represent thermal stress (39). While sublethal heat exposures could bring about heat hardening (40)(41)(42), given our previous findings, we hypothesized that multiple heat stress events would have a cumulative negative effect on mosquito thermal tolerance in the presence of infection. Our findings can help us understand how DENV infection may affect mosquito survival under changing conditions in the field and whether Wolbachia-based biological control programs may cope with heatwaves.
## RESULTS
## Effect of repeated thermal exposure thermal KD
We exposed female mosquitoes (±DENV, ±Wolbachia) to heat shock of different levels of intensity (35°C or 40°C), frequency (one, two, or three heat shocks), and duration (10 or 60 min), as per Fig. 1. In this study, we refer to wild-type mosquitoes as Ae. aegypti that are free of Wolbachia and not infected with DENV. We tested all treatment groups' performance in a thermal knockdown (KD) assay at a more extreme tempera ture, the insect's CT max (42°C) (37). We aimed to test how differing levels of sublethal thermal exposures would interact with infection to affect subsequent performance. We hypothesized that worsening thermal exposures (increasing frequency, duration, and intensity) would exacerbate previously reported evidence of infection-associated increases in thermal sensitivity (37). The static heat exposure for measuring KD was previously described (37). Briefly, we submerged mosquitoes in glass vials in a 42°C water bath and recorded the time for them to immobilize (KD time) using a barcode scanner. The treatments were blocked over 2 days to manage the scale of the experimental design.
We predicted that the effects of multiple heat shocks would compound the negative effects associated with microbial infection. Instead, we found that longer heat shocks (Fig. 2: least-square analysis: "Duration": F = 1.01, df = 1, P = 0.31), increasing temperature intensity ("Intensity": F = 0.65, df = 1, P = 0.42) and a greater number of exposures ("Frequency": F = 0.49, df = 2, P = 0.61) did not influence the thermal sensitivity of the mosquitoes, either wild type or microbe infected. We did show mosquitoes infected with either Wolbachia ("Wolbachia": F = 128.56, df = 1, P < 0.0001) or DENV ("DENV": F = 99.51, df = 1, P < 0.0001) exhibited heightened sensitivity to heat across all treatments as per our previous study (37). Additionally, we showed that in 9 out of 12 thermal stress conditions tested, co-infection with both Wolbachia and DENV often resulted in an additive effect on thermal sensitivity. Specifically, at 35°C for 10 min, mosquitoes with both infections exhibited median KD time approximately twofold faster than uninfected ones across all exposure frequencies. However, under more extreme heat stress conditions (e.g., 40°C for 10 or 60 min at 2× frequency), co-infected mosqui toes did not differ significantly from singly infected groups, though all infected groups remained more sensitive than uninfected controls. Mosquitoes that were singly infected with either Wolbachia or DENV generally had similar average KD times and typically exhibited intermediate thermal sensitivity-greater than uninfected mosquitoes but less than those co-infected with both Wolbachia and DENV when exposed to prolonged and intense heat stress (e.g., 40°C for 60 min across all exposure frequencies). However, this pattern was not consistent across all thermal regimes. In 3 out of 12 thermal stress conditions, co-infected mosquitoes did not differ significantly in KD time from singly infected groups. For example, following a single exposure to 35°C for 60 min, mosquitoes co-infected with both Wolbachia and DENV did not differ significantly in KD time from singly infected groups. In summary, increasing the exposure to thermal stress by lengthening exposure time and raising the temperature over repeated exposure events did not lead to a change in subsequent thermal performance in the KD assay for wild-type mosquitoes, nor did these exposures interact with microbial infection. We did show that having both Wolbachia and DENV infections had an additive effect on KD time-co-infected mosquitoes were more thermally sensitive than singly infected ones when exposed to 40°C for 60 min across all frequencies. However, this additive effect was context-dependent and not consistent across all thermal stress conditions.
To further investigate potential heat-hardening effects, we conducted an additional experiment with a later generation of mosquitoes, comparing individuals exposed to heat shock once to those without exposure at 40°C for 60 min (selected as the most extreme treatment studied). This separate test was added to the experimental design after the fact, to specifically tease apart whether prior heat exposure provided any protective effect against subsequent thermal stress even with a single exposure. We did not include a zero pre-exposure in our original design (Fig. 2) because we expected there to be large cumulative effects from one to three exposures. The effect of a single versus no pre-exposure was significant across all treatment groups (F = 26.88, df = 1, P < 0.0001), lengthening KD times by ~1.1-to 1.2-fold (Fig. 3). The lengthening of KD time was also present (Fig. 3, beyond dashed line) when we compared each treatment group between 0× in the new experiment and 1× of the previous experiment (data from Fig. 2). We note this with the caveat that this comparison will include additional environmental variation because the mosquitoes were not tested in parallel.
## Effect of repeated thermal exposure on microbe load (DENV and Wolbachia)
We then assessed the effect of repeated heat shock on DENV viral load in mosquitoes (Fig. 4) in association with Wolbachia infection. Since the frequency of heat shock events did not significantly alter mosquito thermal sensitivity in most cases, we focused on the most extreme treatment, those exposed to three heat shock events. Mosquitoes exposed to heat shock events with a longer duration (60 min) had a lower viral load compared to those exposed for only 10 min (least square analysis: F = 22.59, df = 1, P < 0.0001). This effect was significant at 35°C (F = 8.76, P = 0.0045) but not at 40°C (F = 0.91, P = 0.35), indicating temperature-specific differences in how prolonged heat exposure influences viral load. The viral load in mosquitoes between those exposed to 35°C or 40°C did not differ significantly (F = 0.19, df = 1, P = 0.66). Wolbachia infection was also shown to reduce DENV viral load by at least an average of twofold in most cases (F = 17.24, df = 1, P < 0.0001). The same pattern can be observed for the abdomen tissue of the mosquitoes (Fig. S1). In contrast, the Wolbachia load was relatively stable, the same across all durations (F = 3.00, df = 1, P = 0.0857) and intensities (F = 1.59, df = 1, P = 0.2086) of the heat exposure (Fig. S2). In summary, longer exposures to thermal stress reduced viral loads at 35°C, but Wolbachia loads were unaffected.
## Survival
We then performed a survival assay to analyze the effect of repeated heat shock on female mosquitoes (±DENV, ±Wolbachia). We assessed their survival daily for 43 days, with data collection beginning on days post-infection (DPI) 0, and presented the results using Kaplan-Meier survival curves stratified by treatment (Fig. 5). Since there were no Different letters (a,b,c) indicate statistically significant differences (P < 0.05) among treatment groups as determined by Tukey's post hoc test. The data beyond the vertical dashed line are from the previous independent experiment (Fig. 2). We include it here to demonstrate that KD time is also greater than 1× although there will likely be some differences in environmental contributions across experiments (mosquito population generation/rearing, virus culture, etc.) that you see in the 1× vs 1×.
significant frequency effects, we pooled across these treatments. Additionally, there were multiple replicates within each treatment group, and no significant differences were detected among replicates. We used a pairwise log-rank analysis to compare the survival of these mosquitoes between treatments. The intensity and duration of treatments did not have a consistently significant effect on survival. Survival was, however, affected by infection status. For example, at 35°C and 10 min of prior thermal exposure, mosquitoes that were infected with both DENV and Wolbachia died earlier than those that had either one infection or no infection (log-rank, P < 0.05) (see Table S1 for information on individual pairwise comparison results). Mosquitoes infected with just DENV had a better survival probability than those infected with just Wolbachia. If we looked at the most extreme heat shock conditions, 40°C and 60 min, uninfected mosquitoes had the best survival probability. Double-infected mosquitoes died faster than mosquitoes infected with only Wolbachia but did not significantly differ from those infected with only DENV (log-rank, P = 0.328). Overall, the survival of these mosquitoes is highly affected by their infection status. Being infected by both Wolbachia and DENV almost always lowers the survival probability of mosquitoes relative to those that are either +DENV/-Wolbachia or uninfected. In two out of four heat shock treatments that involved longer thermal exposure, the uninfected population had a better survival probability than those single-infected with either Wolbachia or DENV.
## DISCUSSION
In this study, we varied the parameters intensity, frequency, and duration to simulate the complexity of thermal exposure events and examine their interactions with infection on mosquito thermal tolerance. Transient exposures to temperatures above the thermal optimum of most organisms, including mosquitoes, can negatively affect their overall physiology, fitness, and behavior (7,18,43,44). At higher temperatures, enzymes and essential proteins involved in regulating various systems such as respiratory, metabolism, nervous, and circulatory become disrupted, causing a cascade of stress responses that is either able to overcome the damage or succumb to it (18,45,46). Contrary to our expectations, repeated thermal exposure had no cumulative effect on Ae. aegypti's thermal sensitivity. Being exposed to greater thermal stress by either lengthening the exposure time or increasing the temperature did not make the mosquitoes more susceptible to heat than those exposed to only one thermal event. The absence of cumulative effects may also be attributed to the 4-day recovery period between exposures, which was likely sufficient for physiological recovery. This interval, chosen based on pilot data showing high mortality at shorter intervals, may have allowed mosquitoes to fully recover from each heatwave exposure. It is also possible that the thermal stress conditions in our study were not sufficiently severe to induce significant changes, that is, the mosquitoes are highly robust to thermal stress (47). Heatwaves can persist for days in the field and involve significantly elevated temperatures. While exposures to temperatures higher than 42°C for longer than an hour would have been lethal (37), longer or more frequent exposures at high but sublethal temperatures could be explored in future study designs.
While our initial design focused on repeated exposures, we later incorporated a comparison between mosquitoes with no prior heat exposure and those with at least one pre-exposure to explicitly test the possibility of heat hardening. Heat hardening is a mechanism by which an organism can develop increased thermal tolerance after exposure to a nonlethal, elevated temperature for a certain period. This mechanism has been observed in Drosophila, Apis mellifera, and other insect species, with each species having its hardening capacity based on their current geographical distribution (41,42,48). The test of heat hardening was conducted in a subsequent mosquito generation, but comparable results between the two experiments suggest both consistency of the findings and low involvement of any genetic or environmental differences between generations. We revealed evidence of a moderate hardening effect, with mosquitoes that experienced one prior heat exposure exhibiting increased thermal tolerance compared to those with no prior exposure. The moderate effect was strongest in wild-type mosquitoes, suggesting that prior heat exposure experience provides little protection against future extreme heat events, especially in the presence of infection.
Another key finding from this study was the additive effect of infection on a mosquito's thermal sensitivity. In all cases, being infected with both DENV and Wolbachia made the mosquitoes less tolerant to extreme heat, with a twofold faster median KD time compared to those that were uninfected. In 9 out of 12 cases, mosquitoes infected with just one microbe (either DENV or Wolbachia) tended to have a similar average KD time and were more sensitive than uninfected mosquitoes but less sensitive than mosquitoes with both infections. The same pattern was true for the effects on survival. The lower variability observed in uninfected mosquitoes may reflect more stable thermal responses due to unaltered metabolic processes in the absence of bacterial or viral infections. However, it is important to note that this additive effect was not consistent across all thermal stress conditions. These context-dependent outcomes suggest that additive effects of infection may only manifest under certain environmental thresholds or stress intensities. The observed increase in thermal sensitivity in mosquitoes with double infections compared to those with single infections raises several hypotheses for discussion. First, it is essential to note that a previous study by our group did not find any additive effects of Wolbachia and DENV infection on mosquitoes' thermal sensitivity (37). This could be due to two factors: one, the absence of pre-exposure to high tempera tures in the original study, which may have prevented the mosquitoes from reaching a threshold where additive effects become apparent. Second, the data from the previous study (37) did trend toward additivity; it may have lacked the statistical power to confirm it. The current study shows clear additive effects in most but not all treatment combina tions, suggesting that a single pre-exposure may push the mosquitoes into a condition that reveals these interactions more strongly. These data also show that Wolbachia does not confer protection against thermal stress induced by DENV, by reducing virus load through the action of viral blocking. One possible explanation of this additive effect relates to energetic trade-offs. The presence of Wolbachia and DENV may significantly burden the mosquito's immune system and resources, given its propensity to react to both microbes (34,(49)(50)(51). This dual burden could become overwhelming when combined with thermal stress, as activating these immune pathways could interfere with the mosquito's ability to cope with additional stressors. Wolbachia, as an obligate intracellular bacteria, lacks specific metabolic pathways, and its consumption/use of the host's resources, such as amino acids and cholesterol, could deplete essential resources needed for managing thermal stress (52,53). This competition for resources might be amplified under heat stress conditions, leading to the additive effects we observe.
Another possible hypothesis for an additive effect involves direct pleiotropic effects, where genes and pathways involved in coping with microbial infections might over lap with those required for thermal stress management. For example, infection with Wolbachia and DENV activates critical components of the mosquito's immune response, including RNA interference (RNAi) pathway, the Toll pathway, the Imd (Immune deficiency) pathway, and the JAK/STAT pathway (34,(49)(50)(51). These same immune pathways can also be triggered by thermal stress (54)(55)(56). Heat shock proteins (HSPs), such as Hsp70 and Hsp90, are involved in managing thermal stress and responding to various other stressors, including anoxia, crowding, and dehydration (57)(58)(59). It is possible that these HSPs and other generalized stress proteins like reactive oxy gen species may be produced to help manage microbial infections in mosquitoes (35,60). When these genes are activated or suppressed to address the infections, their effects on thermal stress management can be counterproductive. For instance, upregulating responses to Wolbachia or DENV might enhance pathogen defense but could simultaneously interfere with the mosquito's ability to regulate heat stress, thus explaining the additive effects seen in our study. Conversely, downregulating these responses to improve thermal tolerance might compromise the mosquito's ability to combat infections. While our previous study hypothesized these mechanisms (37), one novel finding in this study is the manifestation of additive effects. This suggests that pre-exposure to thermal stress may reduce the mosquito's physiological capacity to balance multiple stressors. Among the mechanisms discussed, resource depletion and competition between immune and stress-response pathways appear most consistent with observed additivity. Future studies should aim to quantify the extent of resource consumption by both microbes under varying thermal conditions to confirm this hypothesis.
Interestingly, while thermal stress exposure did not affect Wolbachia load in mosquitoes subjected to multiple thermal stress events, we found that extended exposure (60 min) significantly reduced viral loads compared to shorter exposure (10 min). This phenotypic reduction in viral load could reflect a decrease in viral replica tion, an increase in mosquito antiviral activity, or a combination of both. DENV has an optimal temperature range for replication. Deviations from this range, high or low, can disrupt the replication cycle, leading to decreased viral loads (20,61). Additionally, DENV replication relies on temperature-sensitive viral enzymes, such as RNA-dependent RNA polymerase (RdRp). High temperatures can affect the activity and stability of these enzymes, further contributing to reduced replication rates (62). Therefore, prolonged heatwaves in the field could potentially reduce vector competence by lowering the viral load in mosquitoes exposed to extended periods of thermal stress. This suggests that rising temperatures may inadvertently enhance the effectiveness of Wolbachia-based biocontrol strategies by selectively reducing DENV transmission in mosquitoes exposed to prolonged heat stress. If this pattern holds under field conditions, it could have important implications for vector control programs, particularly in regions experiencing more frequent and intense heatwaves due to climate change.
Our study suggests Ae. aegypti exhibit a degree of thermal tolerance, likely due to slight heat hardening or inherent robustness, which may help them withstand the effects of heatwaves. However, the magnitude of this effect appears limited under our experimental conditions. It also suggests that mosquitoes infected with Wolbachia and DENV may be especially susceptible to thermal stress, which is consistent with findings in other host-pathogen systems where co-infection or immune activation increases susceptibility to environmental stressors (63,64). Given the rarity of mosquitoes infected with DENV during an outbreak (1-2%) (65), the thermal sensitivity could potentially limit the survival of these individuals with DENV in Wolbachia release zones, enhancing Wolbachia's efficacy without significantly affecting Wolbachia persistence. Mosquitoes infected with DENV, but without Wolbachia, may also be limited by their thermal tolerance, a welcome result under rising temperatures. In the field, these effects on host fitness-singly and additively-will play their part in shifting transmission zones and Wolbachia efficacy just as much as the individual thermal tolerance ranges of the mosquito, Wolbachia, and virus. Future studies are needed to test the effects of infection and co-infection in the field, where mosquitoes can regulate temperature exposure by behavioral modification. These findings underscore Wolbachia's potential role in climate-adaptive vector control strategies, highlighting the need for integrated approaches that account for environmental variability in disease management.
## MATERIALS AND METHODS
## Mosquitoes
Ae. aegypti mosquitoes infected with wAlbB strain of Wolbachia were obtained from Zhiyong Xi (Michigan State). In 2017, the wAlbB strain was backcrossed into AFM, the wild-type Wolbachia-free line of Ae. aegypti obtained from Mérida, Mexico. As this work was carried out about 3 years later, there could be some small genetic differences between lines contributing to phenotypes. The wild-derived line served as a negative control for all experiments. Both lines were reared under standard conditions: 26°C, 60% relative humidity, and a 12-h light/dark photoperiod. During the larval phase, larvae were fed fish food (TetraMin) ad libitum. As adults, mosquitoes had access to 10% sucrose.
## Virus cultivation and mosquito blood-feeding
All experiments were performed using DENV-2 strain ET-300 (GenBank accession number EF440433.1) grown in Ae. albopictus C6/36 cells (Sigma), as previously descri bed (66).Briefly, C6/36 cells were cultured in RPMI 1640 medium (Life Technologies) supplemented with 10% heat-inactivated fetal bovine serum, 20 mM HEPES buffer (Sigma-Aldrich), and 1% penicillin-streptomycin (Life Technologies). The cells grown to 80% confluence in T75 flasks were then inoculated with DENV-2. Flasks were incubated at 27°C for 7 days, and the supernatant was harvested. The supernatant was mixed in a 1:1 ratio with human blood just prior to mosquito feeding.
Adult female mosquitoes, 7 days post-eclosion, were sugar-deprived for 24 h prior to feeding. The mosquitoes were fed using double-chamber glass feeders covered with pig intestine sausage casing that were warmed to 37°C using a water bath. The final virus concentration in the blood was 1.6e 7 DENV copies/mL for the thermal KD experiment comparing no prior heat exposure to one heat exposure, and 3.25e 7 DENV copies/mL for the multiple heat exposure thermal KD experiment. All DENV-negative mosquitoes were fed a 1:1 ratio of blood (without virus) and RPMI 1640 cell culture media to serve as a negative control. Mosquitoes were anesthetized on ice post-feeding, and only those that were fully engorged were retained for subsequent experiments. Fed mosquitoes were sorted into 32 oz paper soup cups with mesh lids and provided cotton balls soaked in 10% sucrose that were changed daily. Environmental rearing conditions were as above for the stock lines.
## Thermal disturbance
All female mosquitoes were then exposed to heat stress with varying intensities (35°C or 40°C), frequencies (one, two, or three heat shocks), and durations (10 or 60 min). With the variables "DENV infection status, " "Wolbachia infection status, " "Intensity, " "Frequency, " and "Duration, " there was a total of 48 treatment groups (Fig. 1). The first exposure to heat stress was carried out at 4 DPI, and subsequent heat shocks then occurred every 4 days (DPI 4, 8, and 12). A 4-day interval between exposures was chosen to ensure physiological recovery while enabling the evaluation of cumulative stress responses over multiple exposures. Pilot studies with 2-or 3-day intervals exhibited substantial mortality. Our repeated exposure design also means that when comparing mosquitoes after a single heat shock versus three, for example, they also differ in age. To specifically test the potential for heat hardening, an additional thermal KD experiment was later incorporated using a subsequent mosquito generation (three generations after the initial thermal KD experiment), reared under the same laboratory conditions mentioned above. This experiment followed the same thermal disturbance protocol, comparing mosqui toes with no prior heat exposure (0×) to those exposed to a single heat stress event (40°C, 60 min). Both 0× and 1× mosquitoes were tested for KD time at the same adult age to ensure age-matched comparisons.
## Thermal KD assay
After the above variable exposures to treatments, we then examined the mosquito's response to thermal stress using a static heat shock assay as described in references 37, 37. Thermal KD experiments were conducted at 42°C (a temperature representing mosquitoes' critical thermal maximum) as determined empirically in a previous study (37) for 60 min. The assay was carried out over 2 days post-blood feed (2 days after the final thermal disturbance) to manage the scale of the design. Mosquitoes from each treatment were randomly selected and placed in individual 40 mL glass vials with solid plastic lids. The vials containing each mosquito were then attached to a vertical plastic board in groups of 48 using anchored clips. The boards were then immersed in a water bath heated to 42°C, and mosquitoes were given a 60-s acclimation period. Mosquitoes were monitored visually for immobility, and time to thermal KD was scored using Brady labels and a TriColor Scanner (Worth Data Inc., Santa Cruz, CA, USA). The immobility of these mosquitoes was confirmed by tapping on the glass vials and visual inspection. These mosquitoes were then collected, and tissues were dissected for DENV and Wolbachia quantification (below).
## Survival assay
After blood-feeding, the mosquitoes that were sorted and kept under standard conditions were assessed daily for their survival until all mosquitoes were dead (43 days), with the data collection beginning on DPI 0. These mosquitoes were also subjected to the thermal disturbance regime as above (Fig. 1). Each treatment group (a total of 48) had 60 individuals divided into 4 cups of 15 subjected to heat shock events with the same regime described above. After exposing them to heat stress, these mosquitoes were returned to chambers set at 26°C, their base temperature.
## Mosquito nucleic acid extraction
Individual whole mosquitoes from the thermal KD assays were dissected for their various tissues (salivary gland, midgut, and carcass) that were placed in 300 µL of TRIzol reagent (Sigma-Aldrich). Samples were homogenized on a Bead Ruptor Elite (Omni International, USA) using a 2.8-mm ceramic bead. Total RNA was extracted with the Direct-zol RNA 96 Magbead Zymo Kit (Zymo Research) according to the manufacturer's protocol. RNA was eluted in 50 µL RNase-free water and then treated with 5 units of DNase I (Sigma-Aldrich) at room temperature for 15 min, followed by inactivation with 50 mM EDTA at 70°C for 10 min. To measure Wolbachia DNA, RNA, and DNA extractions were performed using the column-based Direct-zol DNA/RNA miniprep kit. RNA was eluted into 50 µL RNase-free water, followed by DNA elution in 50 µL Direct-zol DNA elution buffer.
## DENV quantification
DENV was quantified using TaqMan Fast Virus 1-step Master Mix (Thermo Fisher Scientific) in 10 µL reaction volumes with DENV-2 specific primers and probes for qRT-PCR as per previous (67). The following protocol was used: reverse transcription at 50°C for 5 min, followed by 95°C for 20 s, and amplification cycling at 95°C for 3 s and 60°C for 30 s. A standard reference curve of known concentration of DENV-2 genomic fragment was used for absolute qRT-PCR ( Table S2). The DENV-2 genomic fragment was inserted into a plasmid and transformed into Escherichia coli as described (67). The linearized and purified fragment was serially diluted, ranging from 10 6 to 10 2 copies, and was used to create a standard curve of DNA amplification. The standard curve was run on duplicates on each 96-well plate, along with negative controls.
## Wolbachia quantification
Wolbachia (wAlbB) quantifications by qPCR were performed on a LightCycler 480 Real-Time PCR system (Roche) using previously published primers and probes specific to wAlbB and the mosquito ribosomal subunit 17 housekeeping gene (RpS17) (68). The sequences of the primers and probes used in this study are listed in Table S2.
DNA samples were amplified using PerfeCTa Multiplex ToughMix (Quanta, 95147-250) following the manufacturer's protocol. Each qPCR reaction was prepared in a total volume of 20 µL, containing 5 µL of PerfeCTa Multiplex ToughMix, forward and reverse primers, and probes for both wAlbB and RpS17, along with 2 µL of DNA template. Cycling was performed using the LightCycler 480 Real-Time machine, with one cycle at 95°C for 30 s, followed by 40 amplification cycles of 95°C for 3 s, 60°C for 30 s, and a final melting curve analysis. Target gene to housekeeping gene ratios were calculated using the 2 -ΔCt method (69).
## Statistical analysis
All statistical analyses were performed using R (version 4.3.2) or JMP Pro (version 18.1.0, SAS Institute Inc., Cary, NC, 1989-2025), depending on the assay. For the thermal KD assay, analyses were conducted in JMP Pro. KD time was analyzed using mixed-effect models, with least squares estimation applied for model fitting via maximum likelihood. KD time was treated as the response variable, while Wolbachia infection, DENV infec tion, frequency, intensity, and duration of thermal stress were included as fixed effects. Statistical comparisons among treatment groups were performed using Tukey's post hoc tests to assess interactions. The same statistical approach was applied to analyze Wolbachia and DENV loads, with either DENV load or Wolbachia load as the response variable. For the survival assay, Kaplan-Meier survival curves were generated in R, stratified by treatment groups to visualize survival differences. Pairwise comparisons were conducted using the log-rank test to determine significant differences between treatment groups.
## References
1. Perkins-Kirkpatrick, Lewis (2020) "Increasing trends in regional heatwaves" *Nat Commun*
2. Marx, Haunschild, Bornmann (2021) "Heat waves: a hot topic in climate change research" *Theor Appl Climatol*
3. Ryan, Carlson, Mordecai et al. (2019) "Global expansion and redistribution of Aedes-borne virus transmission risk with climate change" *PLOS Negl Trop Dis*
4. Monaghan, Sampson, Steinhoff et al. (2018) "The potential impacts of 21st century climatic and population changes on human exposure to the virus vector mosquito Aedes aegypti" *Clim Change*
5. Lafferty (2009) "The ecology of climate change and infectious diseases" *Ecology*
6. Kamiya, Greischar, Wadhawan et al. (2020) "Temperature-dependent variation in the extrinsic incubation period elevates the risk of vector-borne disease emergence" *Epidemics*
7. Mordecai, Caldwell, Grossman et al. (2019) "Thermal biology of mosquito-borne disease" *Ecol Lett*
8. Bhatt, Gething, Brady et al. (2013) "The global distribution and burden of dengue" *Nature*
9. Scott, Takken (2012) "Feeding strategies of anthropophilic mosquitoes result in increased risk of pathogen transmission" *Trends Parasitol*
10. Ryan, Turley, Wilson et al. (2020) "Establishment of wMel Wolbachia in Aedes aegypti mosquitoes and reduction of local dengue transmission in Cairns and surrounding locations in northern Queensland" *Australia. Gates Open Res*
11. Nazni, Hoffmann, Noorafizah et al. (2019) "Establishment of Wolbachia strain wAlbB in Malaysian populations of Aedes aegypti for dengue control" *Curr Biol*
12. Indriani, Tantowijoyo, Rancès et al. (2020) "Reduced dengue incidence following deployments of Wolbachia-infected Aedes aegypti in Yogyakarta, Indonesia: a quasi-experimental trial using controlled interrupted time series analysis" *Gates Open Res*
13. Pinto, Riback, Sylvestre et al. (2021) "Effectiveness of Wolbachiainfected mosquito deployments in reducing the incidence of dengue and other Aedes-borne diseases in Niterói, Brazil: A quasi-experimental study" *PLOS Negl Trop Dis*
14. Walker, Johnson, Moreira et al. (2011) "The wMel Wolbachia strain blocks dengue and invades caged Aedes aegypti populations" *Nature*
15. Teixeira, Ferreira, Ashburner (2008) "The bacterial symbiont Wolbachia induces resistance to RNA viral infections in Drosophila melanogaster" *PLOS Biol*
16. Yen, Failloux (2020) "A review: Wolbachia-based population replacement for mosquito control shares common points with genetically modified control approaches" *Pathogens*
17. Andersson, Nava-Aguilera, Arosteguí et al. (2015) "Evidence based community mobilization for dengue prevention in Nicaragua and Mexico (Camino Verde, the Green Way): cluster randomized controlled trial" *BMJ*
18. Reinhold, Lazzari, Lahondère (2018) "Effects of the environmental temperature on Aedes aegypti and Aedes albopictus mosquitoes: a review"
19. Mordecai, Cohen, Evans et al. (2017) "Detecting the impact of temperature on transmission of Zika, dengue, and chikungunya using mechanistic models" *PLOS Negl Trop Dis*
20. Zhang, Sheng, Plevka et al. (2013) "Dengue structure differs at the temperatures of its human and mosquito hosts" *Proc Natl Acad Sci*
21. Ross, Ritchie, Axford et al. (2019) "Loss of cytoplasmic incompatibility in Wolbachia-infected Aedes aegypti under field conditions" *PLOS Negl Trop Dis*
22. Ulrich, Beier, Devine et al. (2016) "Heat Sensitivity of wMel Wolbachia during Aedes aegypti Development" *PLOS Negl Trop Dis*
23. Kraemer, Sinka, Duda et al. (2015) "The global distribution of the arbovirus vectors Aedes aegypti and Aedes albopictus"
24. Hector, Sgrò, Hall (2021) "Temperature and pathogen exposure act independently to drive host phenotypic trajectories" *Biol Lett*
25. Hector, Sgrò, Hall (2019) "Pathogen exposure disrupts an organism's ability to cope with thermal stress" *Glob Chang Biol*
26. Linder, Owers, Promislow (2008) "The effects of temperature on host-pathogen interactions in D. melanogaster: who benefits?" *J Insect Physiol*
27. Mcmeniman, Lane, Cass et al. (2009) "Stable introduction of a life-shortening Wolbachia infection into the mosquito Aedes aegypti" *Science*
28. Salazar, Richardson, Sánchez-Vargas et al. (2007) "Dengue virus type 2: replication and tropisms in orally infected Aedes aegypti mosquitoes" *BMC Microbiol*
29. Maciel-De-Freitas, Koella, Lourenço-De-Oliveira (2011) "Lower survival rate, longevity and fecundity of Aedes aegypti (Diptera: Culicidae) females orally challenged with dengue virus serotype 2" *Trans R Soc Trop Med Hyg*
30. Sylvestre, Gandini, Maciel-De-Freitas (2013) "Age-dependent effects of oral infection with dengue virus on Aedes aegypti (Diptera: Culicidae) feeding behavior, survival, oviposition success and fecundity" *PLoS One*
31. Xiang, Saron, Stewart et al. (2022) "Dengue virus infection modifies mosquito blood-feeding behavior to increase transmission to the host" *Proc Natl Acad Sci*
32. Maciel-De-Freitas, Sauer, Kliemke et al. (2024) "Wolbachia strains w Mel and w AlbB differentially affect Aedes aegypti traits related to fecundity" *Microbiol Spectr*
33. Duran-Ahumada, Karrer, Cheng et al. (2024) "Wolbachia pipientis (Rickettsiales: Rickettsiaceae) mediated effects on the fitness and performance of Aedes aegypti (Diptera: Culicidae) under variable temperatures and initial larval densities" *J Med Entomol*
34. Ramirez, Dimopoulos (2010) "The toll immune signaling pathway control conserved anti-dengue defenses across diverse Ae. aegypti strains and against multiple dengue virus serotypes" *Developmental & Comparative Immunology*
35. Pan, Zhou, Wu et al. (2012) "Wolbachia induces reactive oxygen species (ROS)-dependent activation of the Toll pathway to control dengue virus in the mosquito Aedes aegypti" *Proc Natl Acad Sci*
36. Xi, Ramirez, Dimopoulos (2008) "The Aedes aegypti toll pathway controls dengue virus infection" *PLOS Pathog*
37. Ware-Gilmore, Sgrò, Xi et al. (2021) "Microbes increase thermal sensitivity in the mosquito Aedes aegypti, with the potential to change disease distributions" *PLOS Negl Trop Dis*
38. Xu, James (2012) "Temperature stress affects the expression of immune response genes in the alfalfa leafcutting bee, Megachile rotundata" *Insect Mol Biol*
39. Paaijmans, Heinig, Seliga et al. (2013) "Temperature variation makes ectotherms more sensitive to climate change" *Glob Chang Biol*
40. Malmendal, Overgaard, Bundy et al. (2005) "Metabolomic profiling of heat stress: hardening and recovery of homeostasis in Drosophila" *Am J Physiol Regul Integr Comp Physiol*
41. Kellett, Hoffmann, Mckechnie (2005) "Hardening capacity in the Drosophila melanogaster species group is constrained by basal thermotolerance" *Funct Ecol*
42. Alqarni, Ali, Iqbal et al. (2019) "Expression of heat shock proteins in adult honey bee (Apis mellifera L.) workers under hotarid subtropical ecosystems" *Saudi J Biol Sci*
43. Couper, Dodge, Hemker et al. (2025) "Evolutionary adaptation under climate change: Aedes sp. demonstrates potential to adapt to warming" *Proc Natl Acad Sci*
44. Carlassara, Khorramnejad, Oker et al. (2024) "Population-specific responses to developmental temperature in the arboviral vector Aedes albopictus : implications for climate change" *Glob Chang Biol*
45. Couret, Benedict (2014) "A meta-analysis of the factors influencing development rate variation in Aedes aegypti (Diptera: Culicidae)" *BMC Ecol*
46. Neven (2000) "Physiological responses of insects to heat" *Postharvest Biol Technol*
47. Jia, Liang, Tan et al. (2019) "Potential effects of heat waves on the population dynamics of the dengue mosquito Aedes albopictus" *PLOS Negl Trop Dis*
48. Sgrò, Overgaard, Kristensen et al. (2010) "A comprehensive assessment of geographic variation in heat tolerance and hardening capacity in populations of Drosophila melanogaster from eastern Australia" *J Evol Biol*
49. Blair, Olson (2015) "The role of RNA interference (RNAi) in arbovirus-vector interactions" *Viruses*
50. Pan, Pike, Joshi et al. (2018) "The bacterium Wolbachia exploits host innate immunity to establish a symbiotic relationship with the dengue vector mosquito Aedes aegypti" *ISME J*
51. Sim, Jupatanakul, Ramirez et al. (2013) "Transcriptomic profiling of diverse Aedes aegypti strains reveals increased basal-level immune activation in dengue virus-refractory populations and identifies novel virus-vector molecular interactions" *PLOS Negl Trop Dis*
52. Caragata, Rancès, Neill et al. (2014) "Competition for amino acids between Wolbachia and the mosquito host, Aedes aegypti" *Microb Ecol*
53. Caragata, Rancès, Hedges et al. (2013) "Dietary cholesterol modulates pathogen blocking by Wolbachia" *PLOS Pathog*
54. Wrońska, Boguś (2020) "Heat shock proteins (HSP 90, 70, 60, and 27) in Galleria mellonella (Lepidoptera) hemolymph are affected by infection with Conidiobolus coronatus (Entomophthorales)" *PLOS ONE*
55. Sivan, Shriram, Muruganandam et al. (2017) "Expression of heat shock proteins (HSPs) in Aedes aegypti (L) and Aedes albopictus (Skuse) (Diptera: Culicidae) larvae in response to thermal stress" *Acta Trop*
56. Kaur (2018) "Regulation of heat shock protein responses"
57. Benoit, Lopez-Martinez, Phillips et al. (2010) "Heat shock proteins contribute to mosquito dehydration tolerance" *J Insect Physiol*
58. Michaud, Teets, Peyton et al. (2011) "Heat shock response to hypoxia and its attenuation during recovery in the flesh fly, Sarcophaga crassipalpis" *J Insect Physiol*
59. Wang, Wang, Zhou et al. (2007) "cDNA cloning of heat shock proteins and their expression in the two phases of the migratory locust" *Insect Mol Biol*
60. Rancès, Ye, Woolfit et al. (2012) "The relative importance of innate immune priming in Wolbachia-mediated dengue interference" *PLOS Pathog*
61. Lim, Chandramohan, Lim et al. (2017) "Epitope and paratope mapping reveals temperature-dependent alterations in the dengue-antibody interface" *Structure*
62. Bisht, Velthuis (2022) "Decoding the role of temperature in rna virus infections"
63. Lutterschmidt, Schaefer, Fiorillo (2007) "The ecological significance of helminth endoparasites on the physiological perform ance of two sympatric fishes" *Comp Parasitol*
64. Shokal, Yadav, Atri et al. (2016) "Effects of co-occurring Wolbachia and Spiroplasma endosymbionts on the Drosophila immune response against insect pathogenic and non-pathogenic bacteria" *BMC Microbiol*
65. Lambrechts, Reiner, Jr et al. (2023) "Direct mosquito feedings on dengue-2 virus-infected people reveal dynamics of human infectiousness" *PLOS Negl Trop Dis*
66. Terradas, Joubert, Mcgraw (2017) "The RNAi pathway plays a small part in Wolbachia-mediated blocking of dengue virus in mosquito cells" *Sci Rep*
67. Ford, Allen, Ohm et al. (2019) "Selection on Aedes aegypti alters Wolbachia-mediated dengue virus blocking and fitness" *Nat Microbiol*
68. Allman, Lin, Joubert et al. (2023) "Enhancing the scalability of Wolbachia-based vector-borne disease management: time and temperature limits for storage and transport of Wolbachia-infected Aedes aegypti eggs for field releases" *Parasit Vectors*
69. Schmittgen, Livak (2008) "Analyzing real-time PCR data by the comparative C(T) method" *Nat Protoc* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12567937&blobtype=pdf | # Advancing Arbovirus Research in the Caribbean and Latin America: 2025 Global Virus Network Regional Meeting
Tiffany Butterfield, Joshua Anzinger, John Lindo, Gene Morse, Sten Vermund, Maggie Bartlett
## Abstract
A May 2025 symposium convened leading virology experts across Latin America and the Caribbean (LAC) to advance regional research and collaborative efforts. Sessions explored cutting-edge developments in arbovirology, pressing challenges in viral surveillance, and the complexities of vector biology. Integrated networking opportunities and hands-on workshops offered mentorship and training, focused on the next generation of virologists, and strengthened scientific communication within the region. The morning session included reports from the LAC Global Virus Network (GVN) Centers of Excellence. A roundtable dialogue tackled the present challenges faced in arbovirus research. The Abbott Pandemic Defense Coalition reported on its collaborative progress. Trainees from the University at Buffalo, the State University of New York, and the University of the West Indies Global Infectious Diseases Research Training program showcased their current research projects. A session concentrated on health landscapes and the capacity for viral vaccinations within the region. A mentoring workshop focused on immune evasion methodologies and obstacles associated with arboviruses. One Health perspectives on viral zoonotic diseases addressed developments in the surveillance of vector-borne viruses in the Caribbean. Studies of mosquitoes and ticks as vectors of viruses included discussion on the neurovirulence of arboviruses and symptoms occurring after viral infections. Pediatric infectious diseases were highlighted in their environmental health context. An additional mentoring workshop centered on viruses and the microbiome. The relationship between viruses and cancer was discussed in the South American context and included recent advancements in the field of vaccinology. The Jamaica Regional GVN meeting promoted collaboration, facilitated the exchange of knowledge, and advanced research efforts throughout the region.
## 1. Overview
Dr. John Lindo (Professor at the University of the West Indies (UWI), Mona Campus) opened the meeting in Jamaica and underscored the importance of regional meetings of experts and their importance to advancing pandemic preparedness globally (Figure 1). Dr. Sten Vermund (Dean of the University of South Florida College of Public Health and GVN Chief Medical Officer) introduced the new GVN initiative to categorize expertise and capacity across Central America, South America, and the Caribbean to improve pandemic preparedness by connecting Latin America and Caribbean (LAC) experts to regional U.S. Centers for Disease Control and Prevention (CDC) offices. Dr. Vermund noted the recent substantial loss of support to global research from U.S. federal funds and called for investment from other sectors to ensure that critical LAC investments are sustained. One example is the specimen repository at the University of Texas Medical Branch (UTMB), a GVN Center of Excellence (CoE), that has lost significant funding after 60 years of continuous funding; this repository has aided in innumerable research projects that benefited the U.S. and the world [1].
## 2. GVN Regional Center of Excellence and Affiliate Updates
Dr. Calum MacPherson (Professor at St. George's University and GVN Affiliate Director) gave an overview of his and his colleagues' work, encompassing bats as viral vectors, viruses of sea turtles, mongoose rabies, and arboviruses including Zika, chikungunya, and dengue [2,3]. He underscored that the cause of approximately 80% of acute febrile illness cases in Grenada is unknown, with the other 20% being attributable to the dengue virus, and the need for better diagnostics for other causes. Dr. Eduardo Gotuzzo (Professor at the Peruvian Cayetano Heredia University Institute of Tropical Medicine "Alexander von Humboldt" and GVN CoE Director) presented the current understanding of HTLV-1 and co-infections in LAC [4][5][6][7]. Dr. Gotuzzo further highlighted the need for funding to better characterize the relationship between HTLV-1 and cancers. Dr. Joshua Anzinger (Senior Lecturer and Head of Department for UWI-Mona and GVN Affiliate Director) reviewed the utility of Jamaica as a hub for pathogen surveillance and the prevalence of dengue virus serotype 3 regionally [8,9]. Interestingly, Jamaica's rates of HTLV-1 prevalence have remained steady at 2% over the last 20+ years; however, due to a lack of routine screening, there are likely asymptomatic cases that could be detected earlier to enhance recognition of cancers and prevention of vertical transmission [10,11]. Dr. Rubens Alves (Principal Investigator at the Institut Pasteur de São Paulo) emphasized the need for One Health as a focus, using Brazil's ecosystem as a model. The keys to leveraging this include complementary expertise, capitalizing on Brazil's biodiversity, and positioning South America as a Global Health Hub [12].
## 3. Abbott Pandemic Defense Coalition Update
Dr. Lester Perez (Principal Scientist at Abbott) updated the attendees on the outcomes of the Abbott Pandemic Defense Coalition (APDC) initiative [13]. Dr. Perez underscored that time is of the essence when a pathogen emerges, and APDC has grown to include 20+ countries and 5 continents, including GVN CoE's UTMB and University at Buffalo, State University of New York. Through April 2025, the APDC has provided 690,396 on-market tests and 24,164 research tests to partner sites, developed 48 new prototype tests, collected 36,758 specimens, identified 6 significant outbreaks in 3 countries, trained 116 future virus hunters through FETP, published 108 peer-reviewed papers, and identified 24 new viruses [14][15][16][17][18][19]. The Yellow Fever Virus (YFV) causes 200,000 cases annually, with 60,000 deaths on average, with a range from asymptomatic to severe disease presentations in Colombia [20]. From 2020 to 2023, 53.3% of cases of acute febrile illness (n = 2528) were not dengue or malaria, and 52 of them were sequenced, resulting in the identification of a new strain of YFV. Analyses suggest that the emergence of the Colombia/Bolivia clade resulted from episodic positive selection in non-structural protein 2 a (NS2a). This may indicate a benefit to transmissibility, though more research is needed to further elucidate this observation. Most YFV strains circulating in Colombia originated from the country itself, suggestive of cryptic circulation [20]. Although Ecuadorian colleagues were not in attendance, there are analogous challenges in that neighboring nation [21][22][23].
## 4. Global Infectious Disease (GID) Research Training Program
Dr. John Lindo and Dr. Gene Morse (State University of New York (SUNY) Distinguished Professor at the University at Buffalo HIV and HCV Clinical Pharmacology Laboratory and GVN CoE Director) underscored the need for programs like these to train the next generation. Trainee projects encompassed diverse aspects of virology research, including studies on the clinical outcomes of viral infections, viral surveillance in human populations and vectors, as well as the application of cloud-based computing and molecular epidemiology. Samatha Mosha Miller (pediatric resident at UWI-Mona) described her work on long COVID in pediatric cases. Alton Bodley (PhD candidate at UWI-Mona) discussed his work on cloud-based computing and viral surveillance in Jamaica. Within Jamaica, sentinel sites include hospitals, health centers, and hotels; they serve as the first line of detection and investigation. Once sentinel sites detect a pathogen, they are confirmed at UWI or one of the other designated labs. Public health action then occurs via the Ministry of Health Surveillance Unit. Alton developed a data lake with analytics in a dynamic collaborative data ecosystem for West Nile Virus (WNV), as the data were siloed and lacked metadata, such as monthly cases, climate at that time, and other key variables. Dr. Gene Morse emphasized that virology requires a multidisciplinary team, and mentored training grants have been the primary way we advance the next generation, especially trainees from low-and middle-income countries (LMICs). "Brain-drain" has always been a problem, sometimes inadvertently exacerbated by training visits. New approaches are needed from academic collaborations with government and industry partnerships to support LMIC scientists in their home nations. A current challenge is that there is a lack of commercial partners for sustaining these initiatives in LMICs based on prior investments and training.
## 5. Latin America as a Hub for Vaccine Research
Dr. Arlene Calvo (Associate Professor at USF and member of the Instituto de Investigaciones Científicas y Servicios de Alta Tecnología de Panamá (INDICASAT)) highlighted the importance of the trust between the public and public health. She shared information on the implementation of the pertussis vaccine in Panamá, which had an absolute vaccine effectiveness of 99.3% after three doses [24][25][26][27][28]. This meeting also included Continuing Medical Education workshops and mentoring opportunities focused on virus discovery and surveillance, arbovirology, challenges to training the next generation, and zoonotic viruses like dengue virus (DENV) in Grenada. Dr. Sten Vermund presented a review of new frontiers in vaccinology. Vaccination against measles, pertussis, mumps, rubella, smallpox, diphtheria, human papillomavirus (HPV), hepatitis B virus (HBV), hepatitis A virus (HAV), and poliovirus decreased morbidity from each by 97-100% by 2023 [29]. While protection from infection is rare (i.e., sterilizing immunity), protection from severe disease (symptomatic infection, hospitalization, and death) is universal for so-called vaccine-preventable diseases. Rapidly emerging vaccination technologies include new whole-organism vaccines, purified macromolecules, recombinant-vector vaccines, mRNA-based vaccines, and DNA-based vaccines. These applications are being expanded from preventing severe disease from infections to preventing the progression of cancer [30]. One vaccine technology may not be ideal for all pathogens, so multiple platforms should be explored to identify the best tool for each infection.
## 6. Re-Emerging Arboviruses
Dr. Scott Weaver (John Sealy Distinguished University Chair in Human Infections and Immunity, Professor at UTMB, and GVN CoE Director) shared that this decade has seen a 96% increase in DENV cases compared to the previous decade, highlighting the growing risk this virus poses. The Butantan-DV vaccine prevented DENV-1 and DENV-2 regardless of serostatus through a 2-year follow-up, and early evidence suggests quadrivalent protection. Dr. Weaver highlighted that there is not enough surveillance for arboviruses, but that the Brazilian efforts to expand epidemiological studies have helped identify a 2022 chikungunya virus (CHIKV) outbreak [31]. Another arbovirus, western equine encephalitis virus (WEEV), disappeared for many decades after a 1988 outbreak in Argentina and re-emerged in 2024. Exploration of submergence and re-emergence of a virus is critical to public health forecasting [32]. WEEV lost fitness for mammals but not for enzootic hosts like house sparrows; this may have happened by chance or may be due to unknown factors [33]. Oropouche virus (OROV) causes 11% acute febrile illnesses in Colombia, primarily transmitted by midges and can be transmitted by mosquitoes, sloths, and through human-to-human sexual transmission [34]. OROV has been hidden below the known and frequent dengue cases; diagnosis is low due to a lack of clinical diagnostics, but it needs study as a re-emerging arboviral threat.
## 7. Field Studies of Mosquitoes
Dr. Simone Sandiford (Lecturer at UWI-Mona) highlighted the need for field studies of mosquitoes. Dr. Sandiford's team uses BG sentinel traps, aspirators, and resting shelters to collect mosquitoes from field locations to monitor arboviruses in Jamaica, in and around domiciles. Several niches exist, from urban to rural and wetlands to forested regions in Jamaica. Her team and collaborators generated the first complete Aedes vittatus mitochondrial genome, which is a critical reference for arbovirus research [35]. These data suggest that there have been multiple introductions of Ae. vittatus to Jamaica, and that this species has been present for some time. However, we often do not know where mosquitoes are located, what arboviruses are circulating, or which invasive species are present, underscoring the need to perform more field studies in Jamaica. Reintroduction of malaria to Jamaica and the struggle to eliminate it in 2006-2009 after a 44-year malaria-free period is a reminder of the vulnerability of tropical nations when vector-transmitted organisms are reintroduced [36,37].
## 8. Tick-Borne Viruses
Dr. Saravanan Thangamani (SUNY Empire Innovation Professor at SUNY Upstate Medical University) illuminated the role ticks and tick-borne viruses play in Jamaica [38].
Little is known about both the vectors and pathogens in the Caribbean, which is a significant gap in knowledge. Tick-borne viruses are accidental pathogens to humans, as 90% of ticks have preferred species specificity; however, the ones that do cause human disease cause significant mortality, as with the Crimean-Congo Hemorrhagic Fever virus. Ticks are competent vectors as they are persistent and slow feeders that can feed on a host for up to two weeks. Some tick species can live up to 20 years without a blood meal and lay as many as 18,000 eggs. Within the U.S., Heartland virus, Powassan virus (POWV), deer tick virus, Colorado tick fever virus, and Bourbon virus are concerns, but contribute to under 100 cases on average annually in humans. However, there are likely more cases, including those that are asymptomatic, but surveillance is lacking to determine the true epidemiology and disease risk. Global warming will lengthen tick seasons and expand geographic habitats, prolonging and extending the risk to humans of tick-borne diseases. POWV and deer tick virus cause distinct clinical outcomes and brain pathology [39]. There have been 340 reported human cases of POWV since 2004, with 44 deaths and 267 with neuroinvasive disease sequelae, and there has been a four-fold rise in cases from 2014 to 2023 compared to 2004 to 2013 [40]. In the U.S., POWV cases have increased from northeast to southwest following the distribution of the known vector that is across the entirety of the mid-U.S. to the east coast. To address the issue of collection for surveillance, the Upstate Tick Testing Laboratory was developed in New York State, U.S., to empower citizen science. Individuals can fill out a form if bitten by a tick and send the tick to the lab for the species to be identified, then nucleic acid is extracted and screened for 16 pathogens. Data is shared with the individual and presented in tickMAP [41]. Ticks carry more than one pathogen at one time, which is often overlooked in the one-pathogen one-vector mindset.
## 9. Dengue in Florida
Dr. Kristi Miley (Research Associate faculty at USF) described the globalization of disease risk through competent mosquito vectors with a focus on the dengue virus in Florida. From 2019 to 2024, there has been a 4× increase in dengue cases in Florida, similar to information presented by other speakers that highlighted Jamaica. These cases are primarily serotype 3, with 6:1 travel to local cases [42]. As climate shifts and salinity changes, the primary vector is expanding north, allowing for more human cases in more counties. Surveillance is critical but underperformed and requires more community involvement to address the gaps that exist. Further planning is needed to prevent outbreaks following hurricanes. Severe winds may reduce vector densities radically over the short run where wind velocities are great. However, longer-term increases in breeding sites or flooding where wind velocities are less severe can increase mosquito densities [43][44][45].
## 10. Microbiome
Dr. Christian Bréchot (Director of the USF Microbiomes Institute and Vice Chair of the GVN Board of Directors) illuminated the role viruses play in the microbiome and the need for further research, particularly across the emergence of new microbiome centers globally [46]. Integrating microbiome science into virology holds particular promise, as mounting evidence underscores the microbiome's influence across health and disease. The human phenotype is shaped not only by the human genome but also by diverse microbial populations, particularly within the gut, where mucus layers and bacterial communities play protective roles. Disruptions in this microbial balance, known as dysbiosis, are increasingly linked to various diseases. Virus-microbiome interactions have profound implications, influencing susceptibility to infection, disease severity, vaccine response, and treatment outcomes, with compelling evidence across pathogens such as HIV, HPV, HBV, influenza, and SARS-CoV-2 [47]. For example, in COVID-19, characteristics of the gut microbiome have been associated with illness severity, suggesting potential for microbiome-targeted interventions [48,49]. Yet, current microbiome research faces limitations, including methodological inconsistencies, a lack of mechanistic studies, and overreliance on cross-sectional data. In audience discussions, the importance of virome research and the accessibility of oral microbiome studies were highlighted, with calls for comparative work across body sites and organisms, with further investment in studies of microbiomes from both humans and vectors.
## 11. Proactive One Health
Dr. Jean Paul Carrera (Investigator at the Instituto Conmemorativo Gorgas de Estudios de la Salud (Gorgas Memorial Laboratory) in Panamá) endorsed the need for active One Health surveillance of viruses, vectors, and reservoirs. His team set up mosquito traps with different hosts, including chicken, frog, hamster, and mice, and collected 11,000+ mosquitoes from Darien, Panamá. Within these vectors, numerous viruses were detected circulating, including Madrid, Aruza, Aguas Calientes, and Matusgarii orthobunyaviruses. Future work includes monitoring enrolled participants living in the region for seroconversion as well as febrile cases that present to nearby healthcare facilities. This work detected three previously undescribed orthobunyaviruses. Dr. Helena Solo-Gabriele (Professor at the University of Miami) discussed the application of wastewater surveillance (WWS) and novel methods to enhance detection, monitoring, and mitigation [50][51][52][53]. Her team has deployed WWS to detect SARS-CoV-2 in the community up to two weeks ahead of a surge and has shown utility in Miami for the detection of influenza A and B, poliovirus, respiratory syncytial virus, norovirus, and monkeypox virus [50,52,53]. This work is ongoing and will be expanded to detect arthropod-borne pathogens from dengue virus to malaria.
## 12. Excellence in Regional Leadership Award
Dr. Christine Carrington (Professor and Head of the Department of Preclinical Sciences of the Faculty of Medical Sciences at UWI-St. Augustine Campus, and GVN Affiliate Director) delivered a workshop seminar on tracking viruses in real-time and was awarded the inaugural Excellence in Regional Leadership Award for her contributions to the field and in regional capacity building. Dr. Carrington spearheaded the COVID-19 IMPACT project, establishing local whole-genome sequencing capacity for SARS-CoV-2 and providing genomic surveillance for 17 Caribbean nations during a critical window before broader regional infrastructure was available [54].
## 13. Conclusions
The 2025 GVN Regional Meeting in Jamaica was a valuable convening of arbovirology experts in the region to stimulate collaborations, highlight cutting-edge findings, and nurture the next generation of virologists. This conference brought together leading experts in arbovirology from surveillance to mitigation efforts to advance pandemic preparedness. The emphasis on Jamaica's role in addressing arboviral threats was underscored by key lectures and workshops. The GVN remains steadfast in enhancing international partnerships, supporting leading virology research, and fostering pandemic preparedness solutions. Discussions will continue at the 2026 GVN Annual Meeting to be held at USF 4-6 March 2026, in Tampa, FL, USA.
## Funding:
The conference sponsors had no role in the decision to publish this report.
## References
1. Dacso, Bente, Weaver et al. (2022) "Texas professionals are employing a one health approach to protect the United States against biosecurity threats" *One Health*
2. Kiener, Cudjoe, Evans et al. (2023) "Factors Associated with Chikungunya Infection among Pregnant Women in Grenada, West Indies" *Am. J. Trop. Med. Hyg*
3. Fernandes, Evans, Cheng et al. (2023) "Does Intra-Uterine Exposure to the Zika Virus Increase Risks of Cognitive Delay at Preschool Ages? Findings from a Zika-Exposed Cohort from Grenada, West Indies. Viruses"
4. Velasco, Seas, Gotuzzo et al. "Human T-Lymphotropic Virus Type 1 and Cryptococcosis Infection, an Underdiagnosed Association: Case Series and Literature Review" *Open Forum Infect. Dis*
5. Cachay, Gil-Zacarias, Watanabe-Tejada et al. (2021) "Radiological and Functional Characteristics of Pulmonary Diseases among HTLV-1 Infected Patients without Prior Active Tuberculosis Infection"
6. Dextre, Álvarez, Castañeda et al. (2024) "Association between HTLV-1/2 infection and COVID-19 severity in a migrant Shipibo-Konibo population" *PLOS Glob. Public Health*
7. Alva, Blas, Mori et al. (2025) "Human T Cell Lymphotropic Virus Type 1 and Human T Cell Lymphotropic Virus Type 2 Mortality in the Shipibo-Konibo People" *Am. J. Trop. Med. Hyg*
8. Valenzuela, De Siqueira "Editorial: Global excellence in infectious diseases-Surveillance, prevention and treatment: Central and South America" *Front. Med*
9. Redman, Perez, Forberg et al. (2024) "Dengue Virus Serotype 3 Origins and Genetic Dynamics" *Jamaica. Emerg. Infect. Dis*
10. Figueroa, Cooper, Duncan et al. (2024) "Urgent need to prevent mother-to-child transmission of HTLV-1 in Jamaica" *Lancet Reg. Health Am*
11. Leys, Cameron, Frater et al. (2023) "Seroprevalence of Human T-Cell Lymphotropic Virus-1 in a Jamaican Antenatal Population and Assessment of Pooled Testing as a Cost Reduction Strategy for Implementation of Routine Antenatal Screening" *Am. J. Trop. Med. Hyg*
12. Alves, Amorim (2023) "Editorial: Arboviruses: Co-circulation, co-transmission, and co-infection" *Front. Microbiol*
13. Averhoff, Berg, Rodgers et al. (2022) "The Abbott Pandemic Defense Coalition: A unique multisector approach adds to global pandemic preparedness efforts" *Int. J. Infect. Dis*
14. Ciuoderis, Berg, Perez et al. (2022) "Oropouche virus as an emerging cause of acute febrile illness in Colombia" *Emerg. Microbes Infect*
15. Orf, Olivo, Harris et al. (1022) "Metagenomic Detection of Divergent Insect-and Bat-Associated Viruses in Plasma from Two African Individuals Enrolled in Blood-Borne Surveillance"
16. Perez, Cloherty, Berg (2023) "Parallel evolution of picobirnaviruses from distinct ancestral origins" *Microbiol. Spectr*
17. Perez, Forberg, Cloherty et al. (2023) "Temporal and coevolutionary analyses reveal the events driving the emergence and circulation of human mamastroviruses" *Emerg. Microbes Infect*
18. Perez-Acle, Ravello, Rosemblatt (2024) "Are we cultivating the perfect storm for a human avian influenza pandemic?" *Biol. Res*
19. Perez-Restrepo, Ciuoderis, Usuga et al. "Mayaro Virus as the cause of Acute Febrile Illness in the Colombian Amazon Basin"
20. Perez, Perez-Restrepo, Ciuoderis et al. (2025) "Emergence, persistence, and positive selection of yellow fever virus in Colombia" *Front. Microbiol*
21. Izurieta, Macaluso, Watts et al. (2009) "Assessing yellow Fever risk in the ecuadorian Amazon" *J. Glob. Infect. Dis*
22. Izurieta, Macaluso, Watts et al. (2011) "Hunting in the Rainforest and Mayaro Virus Infection: An emerging Alphavirus in Ecuador" *J. Glob. Infect. Dis*
23. Izurieta, Macaluso, Watts et al. (2009) "Anamnestic immune response to dengue and decreased severity of yellow Fever" *J. Glob. Infect. Dis*
24. Calvo, Urrutia, Vargas-Zambrano et al. "Pertussis vaccine effectiveness following country-wide implementation of a hexavalent acellular pertussis immunization schedule in infants and children in Panama" *Hum. Vaccin*
25. O'ryan, Calvo, Espinoza et al. (2020) "Parent reported outcomes to measure satisfaction, acceptability, and daily life impact after vaccination with whole-cell and acellular pertussis vaccine in Chile" *Vaccine*
26. Sáez-Llorens, Rowley, Wong et al. (2017) "Efficacy of 10-valent pneumococcal non-typeable Haemophilus influenzae protein D conjugate vaccine against acute otitis media and nasopharyngeal carriage in Panamanian children-A randomized controlled trial" *Hum. Vaccin. Immunother*
27. Azziz-Baumgartner, Veguilla, Calvo et al. (2022) "Incidence of influenza and other respiratory viruses among pregnant women: A multi-country, multiyear cohort" *Int. J. Gynecol. Obs*
28. Azziz-Baumgartner, Gonzalez, Davis et al. (2022) "Lower cognitive scores among toddlers in birth cohorts with acute respiratory illnesses, fevers, and laboratory-confirmed influenza. Influenza Other Respir" *Viruses*
29. Shattock, Johnson, Sim et al. "Contribution of vaccination to improved survival and health: Modelling 50 years of the Expanded Programme on Immunization" *Lancet*
30. Brandenburg, Heine, Brossart (2024) "Next-generation cancer vaccines and emerging immunotherapy combinations" *Trends Cancer*
31. De Souza, De Lima, Mello et al. (2023) "Spatiotemporal dynamics and recurrence of chikungunya virus in Brazil: An epidemiological study" *Lancet Microbe*
32. Bergren, Haller, Rossi et al. (2020) "Submergence" of Western equine encephalitis virus: Evidence of positive selection argues against genetic drift and fitness reductions" *PLoS Pathog*
33. Li, Plante, Lin et al. "Shifts in receptors during submergence of an encephalitic arbovirus" *Nature*
34. Scachetti, Forato, Claro et al. (2025) "Re-emergence of Oropouche virus between 2023 and 2024 in Brazil: An observational epidemiological study" *Lancet Infect. Dis*
35. Noble, Ali, Wilson-Clarke et al. "Invasive Aedes vittatus mosquitoes in Jamaica: A looming public health concern. bioRxiv 2025"
36. Webster-Kerr, Figueroa, Weir et al. (2011) "Success in controlling a major outbreak of malaria because of Plasmodium falciparum in Jamaica" *Trop. Med. Int. Health*
37. Lindo, Brown, Vickers et al. (2013) "Leptospirosis and malaria as causes of febrile illness during a dengue epidemic in Jamaica" *Pathog. Glob. Health*
38. Paine, Hermance, Thangamani (2024) "Early transcriptomic changes at the skin interface during Powassan virus transmission by Ixodes scapularis ticks" *Front. Immunol*
39. Reynolds, Hart, Nelson et al. "Comparative Pathogenesis of Two Lineages of Powassan Virus Reveals Distinct Clinical Outcome, Neuropathology, and Inflammation" *Viruses*
40. Cdc (2025) "Tickborne Disease Surveillance Data Summary"
41. Hart, Bhaskar, Reynolds et al. "Community engaged tick surveillance and tickMAP as a public health tool to track the emergence of ticks and tick-borne diseases in New York"
42. Taylor-Salmon, Hill, Paul et al. (2023) "Travel surveillance uncovers dengue virus dynamics and introductions in the Caribbean"
43. Lucas, Watkins, Phillips et al. (2019) "The Impact of Hurricane Irma on Population Density of the Black Salt-Marsh Mosquito, Aedes taeniorhynchus"
44. Weaver, Xue, Gaines (2020) "Population Outbreaks of Mosquitoes After Hurricanes Matthew and Irma and the Control Efforts in St. Johns County, Northeastern Florida" *J. Am. Mosq. Control. Assoc*
45. Moise, Huang, Mutebi et al. (2016) "Effects of Hurricane Irma on mosquito abundance and species composition in a metropolitan Gulf coastal city"
46. Martiny, Whiteson, Bohannan et al. (2020) "The emergence of microbiome centres" *Nat. Microbiol*
47. Zhang, Mei, Wang et al. (2025) "Analysis of the overall development trends and hotspots in the research field of the human gut virome" *Virol. J*
48. Smail, Albarzinji, Salih et al. (2025) "Microbiome dysbiosis in SARS-CoV-2 infection: Implication for pathophysiology and management strategies of COVID-19" *Front. Cell. Infect. Microbiol*
49. Li, Armstrong, Neff et al. (2016) "Complexities of Gut Microbiome Dysbiosis in the Context of HIV Infection and Antiretroviral Therapy" *Clin. Pharmacol. Ther*
50. Babler, Solo-Gabriele, Sharkey et al. (2025) "Novel Workflows for Separate Isolation of Pathogen RNA or DNA from Wastewater: Detection by Innovative and Conventional qPCR" *Bio-Protoc*
51. Boza, Amirali, Williams et al. (2024) "Evaluation of a field deployable, high-throughput RT-LAMP device as an early warning system for COVID-19 through SARS-CoV-2 measurements in wastewater" *Sci. Total Environ*
52. Solo-Gabriele, Guevara, Kumar et al. (2025) "Wastewater Based Measures of COVID-19 and Associations with Children's Absenteeism at Grade Schools" *Sci. Total Environ*
53. Tierney, Foox, Ryon et al. (2024) "Towards geospatially-resolved public-health surveillance via wastewater sequencing" *Nat. Commun*
54. Carrington (2025) "Preparing for the Next Pandemic"
55. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12465148&blobtype=pdf | Zhenfu Zhao, Ou Sha
## Abstract
Background Over the past four decades, the relationship between viruses and gastrointestinal (GI) mucosal immunity has gained increasing attention due to frequent viral epidemics.Objective This study explores current research trends and future directions in this field through bibliometric analysis.
Trends and hotspots in research of virus and gastrointestinal mucosal immunity: a bibliometric analysis of four decades Yu Cai 1 † , Jingqi Liu 2 † , Mingxia Zhang 3 , Nianhong Qin 4 , Mingfeng Liao 3 , Feng Gao 2 , Zhenfu Zhao 1* and Ou Sha 1*
## Background
The GI tract is a complex channel extending from the mouth to the anus, responsible not only for food digestion and nutrient absorption, but also for housing the largest mucosal immune system in the human body [1][2][3]. The GI mucosal immune system consists of a diverse array of immune cells and structures that together support the body's defense against pathogens while maintaining tolerance to non-pathogenic stimuli, such as dietary antigens and commensal microbiota [4]. Gut microbiota consisting of bacteria, archaea, fungi, and viruses, inhabit the GI tract and represent the living microbial component; by contrast, the gut microbiome encompasses not only these microorganisms but also their collective genomes, metabolites, and the ecological niches they occupy. Maintenance of microbial homeostasis is critical: under eubiosis, the gut microbiota promotes immune cell development and prevents pathogenic colonization, whereas dysbiosis can trigger immune dysregulation and increase the risk of infections and autoimmune disorders [5,6]. Mucosal epithelial cells form tight junctions, which create a physical barrier that prevents pathogens and harmful substances from entering the body [7,8]. Additionally, these epithelial cells play a crucial role in nutrient absorption and transport, and maintain stability within the GI environment [9]. The protective mucus layer, composed of mucins, provides an additional barrier to pathogens, while antimicrobial peptides such as defensins and lysozymes act to neutralize potentially harmful microorganisms [3]. Furthermore, a variety of immune cells, including B cells, T cells, natural killer cells, dendritic cells, macrophages, and neutrophils, are active in the submucosa, participating in immune responses and regulation [10]. However, dysfunction in this protective mucosal immune system can result in a variety of diseases, including acute and chronic gastroenteritis, IBD, and systemic diseases [11][12][13].
Viruses are small yet highly potent pathogens that rely entirely on host cells for survival and replication. They consist of genetic material (DNA or RNA) encased in a protein shell, lacking the complexity of bacteria or fungi as they possess no organelles or independent metabolic processes [14,15]. Viruses disrupt GI epithelial cells through various mechanisms, activating immune responses and leading to GI diseases [7,8,16]. For instance, HIV weakens the immune system by infecting immune cells, particularly CD4 + T cells, thereby increasing the risk of GI infections [17][18][19]. Persistent viruses in the gut, such as rotavirus and norovirus, have been linked to the onset and progression of IBD [20]. Coronaviruses, such as SARS-CoV-2, infect intestinal epithelial and immune cells, disrupting their structure and function, triggering the release of inflammatory mediators, and causing intestinal inflammation and tissue damage [21][22][23]. A deeper understanding of GI defense mechanisms against viral infections will aid in the development of more precise and effective prevention strategies, thereby improving patient outcomes and quality of life [24][25][26][27].
There have been two bibliometric studies related to mucosal immunity of the whole body, one of which focused on IgA nephropathy and the other explored commensal microbiota in vivo [28,29]. Hitherto no bibliometric analysis has been conducted on the critical research areas of viruses and GI mucosal immunity. This study aimed to fill this gap by creating a knowledge map of this field and to help researchers quickly observe the hot spots and prospective development directions.
## Methods
By utilizing specialized software such as CiteSpace and VOSviewer, researchers can visually map collaboration networks, knowledge frameworks, current research hotspots, and potential future trends within a specific field [30]. Web of Science (WoS) is one of the most widely accessed global academic databases, encompassing over 12,000 high-quality journals and extensive citation records [30][31][32]. Lower-quality sources could introduce noise into bibliometric analyses, WoS offers citation indexing dating back to 1900, thereby enabling comprehensive historical citation studies. Consequently, WoS serves as a vital statistical resource for bibliometric investigations and remains one of the predominant databases in this field [33,34]. To minimize errors due to daily database updates, this study conducted a literature search on January 6, 2025, exporting relevant publications from the past 40 years to the WoSCC. The screening process was shown in Fig. 1.
Bibliometrix R package (version 4.4.1) was used to clean, preprocess, and analyze the data, particularly focusing on visualizing publication volumes by country and institution [35]. The process began with package installation " install.packages("bibliometrix") ", loading " library(bibliometrix) ", and launching the user-friendly web interface using " biblioshiny() ". In addition, the coefficient of determination (R²) was employed to evaluate the goodness-of-fit of regression models used to describe publication trends over time.
VOSviewer (version 1.6.20) was employed to analyze co-authorship, institutional affiliations, and keyword cooccurrence [36]. Especially, the minimum threshold for co-authorship was set to 12, and fractional counting was used to account for author contributions in the analysis. Additional selection criteria were determined in accordance with the corresponding figure legends.
CiteSpace (version 6.3.R1) was used for citation analysis to visualize the structure by identifying highly cited literature and keywords with citation growth Fig. 1 The workflow of bibliometric analysis related to viruses and GI mucosal immunity [37]. Co-cited reference clustering analysis was used as an illustrative example. The g-index (with a scale factor k = 25) was adopted as the node selection criterion. The following parameters were applied: look-back years (LBY) = 5, Link retaining factor (LRF) = 2.5, L/N = 10, and e = 1.0. To enhance network clarity, the Pathfinder algorithm was selected for pruning. The final co-citation network comprised 2,252 nodes and 4,178 edges, with a density of 0.0016. The five largest co-citation clusters accounted for 77% of the total network. The modularity of the network was 0.8765, indicating a well-defined cluster structure. The weighted mean silhouette score reached 0.9424, and the harmonic mean of modularity and silhouette was 0.9083, reflecting high cluster homogeneity and internal consistency. Additional analytical procedures and parameter settings were detailed in the corresponding figure legends. Especially, beyond temporal stratification, research trends and thematic hotspots were further examined by categorizing publications into research and review articles.
## Results
## Annual publication volume analysis
In the past 40 years, 4,842 publications on the topic of viruses and GI mucosal immunity were retrieved. Figure 2a shows that the total number of publications in this field has steadily increased from 1985 to 2024, with an average annual growth rate of 8.65%. The overall publication trends in this field can be categorized into three distinct phases. The first 20 years (1985-2004) experienced a gradual increase in publication volume, rising from 2 articles in 1985 to 103 in 2004, accounting for 1,218 articles, which represented 25.15% of the total 4,842 publications. The middle 10 years (2005-2014) showed a consistent rise in annual publications, contributing 1,449 articles (29.93%) to the total output. The last 10 years (2015-2024) witnessed a significant surge in research output, with 2,175 publications constituting 44.92% of the total publications. Notably, the number of publications exceeded 200 annually, from 2019 to 2024. Figure 2b depicts the yearly publication trajectories of review and research articles, both of which exhibit trends that align closely with the overall growth in publication output. Trends in publication volume showed a strong upward trajectory ( Fig. 2c; slope = 6.1857, R 2 = 0.928), while the average annual citation count peaked in 2005 and gradually declined thereafter.
## Distribution of countries, regions, institutions and authors
Analysis of publication volume by the countries of the corresponding authors (Fig. 3a) revealed that there were a total of 87 countries or regions contributing to this field. Figure 3b and c list the top ten countries in terms of publication volume and their respective proportions.
The USA made the most significant contribution, with 1866 publications accounting for 38.54% of the total. Especially, its publication volume was 3.41 times greater than that of China (11.31%), the second-ranked country, and 8.48 times higher than that of Japan (4.54%), which ranked third. Figure 3d shows that the top nine institutions by publication volume were based in the USA with University of California contributing the highest number of publications (n = 605). Furthermore, Table 1 lists the top ten authors by publication output, all affiliated with institutions in the USA. The author with the highest publication output was Dandekar S, with 46 papers and an H-index of 26. Notably, his paper "Severe CD4 + T-cell depletion in gut lymphoid tissue during primary HIV type 1 infection and substantial delay in restoration following highly active antiretroviral therapy" published in JOUR-NAL OF VIROLOGY in 2003, has been cited 840 times. Veazey RS ranked second with 41 publications and an H-index of 24. In addition, Brenchley JM from National Institutes of Health (NIH) ranks third, with the highest H-index of 27.
Table 2 presents the proportion of international collaborations by publication volume among the top 20 countries. The USA had the highest number of multiinternational collaborative papers (MCP) at 365 (19.56%). Table 3 lists the top 20 international collaborations, with 15 out of 20 collaborations involving the USA. The top three countries in international collaborations led by the USA were the United Kingdom, Canada, and China. By contrast, Argentina had the highest international collaboration rate (55.56%). Among the institutional collaborations, the largest cluster (in red) was led by the University of California and comprised 12 institutions (Fig. 4a). Most research institutions in this field established their presence early on, while activity recently notably increased from the Chinese Academy of Medical Sciences, the Chinese Academy of Agricultural Sciences, and McGill University (Fig. 4b). Collaborative relationships among researchers were analyzed using cluster (Fig. 4c) and time-overlapping networks (Fig. 4d). The top three clusters (red, green, and blue) show close collaboration, with the largest cluster (red) representing 10 authors, including Brenchley JM and Silvestri G. Recently, there have also been novel scientists such as Kumar VV. As shown in Fig. 4E, the authors with the highest citation counts were Brenchley JM (n = 1,028), Veazey RS (n = 713), and Brandtzage P (n = 392).
## Distribution of journals
A total of 1143 journals were included in this study, resulting in 4,842 publications. Table 4 lists the top ten journals based on publication output and their latest impact factors (IF). Seven publishers were from the USA. The top three journals were Journal of Virology (US), Vaccine (Netherlands), and Frontiers in Immunology (Switzerland). Among these ten journals, four were from the JCR Q1 category and three had IF above 5.
## Most cited publications
Table 5 lists the top ten most cited publications. These papers were published between 2005 and 2020 and each had been cited more than 1,000 times. The most frequently cited publications focused on HIV and influenza virus. Microbial translocation was closely linked to the infection caused by these viruses. Other highly cited topics included fungal infections, oral health, oral cancer, innate immunity, and mucosal vaccines. Table 6 lists the virus types most relevant to the cited themes. "HIV" significantly affects both publication volume and total citations. The total citation of coronaviruses ranked second, with key representative articles included "Multiple organ infection and the pathogenesis of SARS" and "Pathogenesis and transmission of SARS-CoV-2 in golden hamsters".
## Co-citation analysis of references
Figure 5a describes the clustering themes and trends in co-citation of references over the past 40 years. The largest clustering theme was "#0 HIV infection, " followed by "#1 ku2-infected macaque" and "#2 cov-2 infection". Figure 5b displays the trends in research interest for the co-cited literature clustering themes, with "#0 HIV infection", "#4 t-cell response", "#12 inflammatory bowel disease" and "#14 type iii interferon" showing higher research interest between 2004 and 2019. Recent cocitation themes included "#2 Cov-2 infection" and "#6 gut microbiome". The co-citation clustering themes of research articles over the same period revealed dominant clusters such as "#0 hiv enteropathy ",followed by "#1 t-cell depletion" and "#2 fecal microbiota"(Fig S1a). As shown in Fig S1b, recent co-citation themes in review articles included #0 "Gut dysbiosis" and #2 "CoV-2 infection".
## Citation burst analysis of references
## Keywords analysis
In this study, 7, 233 keywords were collected. Figure 7a shows the top 20 keywords ranked by frequency. "HIV" was the most frequently occurring keyword, appearing 640 times, followed by "mucosal immunity" and "vaccine" (340 and 184 times, respectively). "IgA" (44 occurrences) was the only antibody-related term. "Inflammatory bowel disease" and "ulcerative colitis" pertain to relevant GI diseases, while "COVID-19" and "Epstein-Barr virus" were the most common viral terms after HIV. Keywords such as "probiotics" and "microbiome" were linked to GI microbiota, whereas "cytokine", "immune" and "immune activation" were associated with immune mechanisms. Figure 7b provides a network visualization of the keyword co-occurrence clustering. The largest cluster focused on mucosal vaccine development for different viruses. The second largest focus was the etiology, diagnosis, and prognosis of various viruses and related clinical diseases. The third cluster, in dark blue, addressed the etiology of intestinal diseases., while the fourth cluster, in yellow, highlighted studies on HIV-related immune mechanisms The fifth cluster (in purple) was centered on gut microbiota Fig. 7c illustrates recent research hotspots, including keywords such as "microbiota", "inflammation" and "COVID-19".
To investigate specific trends, research topics from the past 40 years were clustered based on publication volumes over different developmental stages. Figure 8a indicates that, from 1985 to 2004, themes such as "#0 simian immunodeficiency virus", "#1 cholera toxin" and "#2 in vitro susceptibility" exhibited relatively stable distributions over time. In the 2004-2015 period (Fig. 8b), researchers, particularly those from USA, focus not only on studying the mechanisms of viral infections at mucosal structure but also on developing edible vaccines. Figure 8c illustrates the dense clustering of themes associated with immune mechanisms and the prevention and treatment of various viruses, such as "#0 infection", "#1 human papillomavirus", "#4 HIV". Interest in diseases related to mucosal barrier disruption remained robust, particularly in areas like "inflammatory bowel disease". The emerging themes during this period included "microbial translocation" and "gut microbiome". To explore the temporal dynamics of research themes, a comparative analysis was conducted between research articles (Fig S2a-c) and review articles (Fig S2d-f ) across three developmental stages.
Figure 9 lists the top 30 keywords with the strongest citation bursts. The three longest-lasting keywords were "lamina propria" (1992-2011, lasting for 19 years), "monoclonal antibodies" (1990-2004, for 14 years), and "virus infection" (1990-2005, for 15 years). The keywords with the highest burst intensity were "AIDS", "gut microbiota" and "gastrointestinal tract" with a strength of 49.27,40.4 and 30.33, respectively. Recently, keywords such as "gut microbiota" (2017-2024), "microbiota" (2019-2024), "mechanism" (2019-2024), and "COVID-19" (2020-2024) have attracted increasing attention.
## Discussion
## Research trends in GI mucosal immunity and viruses
This study analyzed the research output related to both viruses and GI mucosal immunity over the past 40 years. There were 4655 publications published in 84 countries or regions, for which there are many reasons for the differences in publication trends. With an understanding of the structure of mucous membranes in the first 20 years, scientists have studied the pathogenic mechanisms of viruses. As major scientific discoveries emerged, publication volumes continued to increase steadily during the second decade. For instance, scientists have found that CD4 + T cells are reduced during acute simian immunodeficiency (SIV) infections [38]. From 2015 to 2024, ) pandemic [41,42]. Declining average citations per year may be caused by factors such as prolonged cycles, access issues, and shifting priorities [43,44].
## Geographic distribution of research
U.S. institutions have led this field with the highest publication volume. Among them, the University of California ranked first, which was also the affiliation of Dandekar S, who ranked the foremost in publications. Notably, Brenchley JM from the NIH and Veazey RS from Tufts University are not only the most co-cited authors, but also rank among the top 10 in publication output. These results suggest that teams led by these authors are regarded as authoritative in the field. As a super county, extensive domestic cooperation supports its publication.
Recently publications of China have surge to rank second, none of the Chinese institutions could rank in the top ten. China could enhance its research competitiveness through greater international partnerships. Notably, all top 10 journals had fewer than 250 publications, indicating a dispersed publication landscape. While published by European or American publishers, significant contributions from China, Japan, and South Korea highlight the potential of high-impact Asian journals.
## Hotspots on GI mucosal immunity and viruses
To further refine our understanding of research dynamics in this field, we performed stratified analyses based on article types distinguishing between research articles and review articles. Co-citation reference clustering, the change of keywords clustering themes, and citation burst analysis could represent the foundational research on HIV in a field, while other viruses such as EB virus, Rotavirus, and Cytomegalovirus became popular topics. Keywords and the most frequently cited publications served as valuable indicators for evaluating research hotspots and future development trends. In addition to HIV, HPV, influenza virus and COVID-19 have also aroused the interest of scientists recently in this field. In addition, "gut microbiota" and "inflammatory bowel disease" are emerging as potential research topics. This shift from single-virus focus to a more integrative understanding of mucosal immunopathology illustrates the convergence of virology, microbiome science, and systems immunology, reflecting a broader trend toward holistic mucosal health research [45,46].
HIV has been the virus most closely related to the research on GI mucosal immunity over the past 40 years. Early mechanistic studies revealed massive CD4⁺Tcell depletion in gutassociated lymphoid tissue, establishing the GI tract as a major HIV reservoir [47]. These insights have informed biomarkers for monitoring ART efficacy and mucosal immune reconstitution. Notably, Brenchley JM's group identified microbial translocation markers as predictors of systemic inflammation [38]; Veazey RS'steam elucidated early mucosal immune depletion dynamics in SIV virus models [48][49][50]. With the in-depth study of microbial pathogenic mechanisms, particularly the elucidation of the role of pattern recognition receptors such as Toll-like receptors, ART, including highly active antiretroviral therapy (HAART) and immune reconstitution therapies, have significantly prolonged the survival of HIV-infected individuals by controlling viral replication and delaying immune deterioration [47,51]. However, HIV relapse and latent infection remain major challenges in clinical management. Current translational efforts focus on broadspectrum neutralizing antibodies, nanoscale delivery of antiretrovirals, and Tcell-based immunotherapies aimed at "functional cure" of HIV [52][53][54][55][56]. Additionally, with advances in biotechnology, new vaccine platforms such as mRNA vaccines offer promising avenues for HIV mucosal vaccine development [18,19,25,57]. These translational efforts exemplify how deep molecular insights can directly inform therapeutic innovation and vaccine design, moving from bench to bedside in tackling chronic viral reservoirs.
HPV, influenza virus and COVID-19 have also been seriously considered. Investigations into GI mucosal responses to HPV have catalyzed development of mucosal vaccine platforms that enhance mucosal IgA responses that particularly valuable for mass immunization in lowresource settings. The application of the HPV vaccine has shown significant effectiveness, with the development of mucosal immune-targeting HPV vaccines playing a crucial role in this progress [24,58]. Furthermore, recognition of HPV's role in anal and oropharyngeal cancers has opened avenues for combining mucosal vaccination with early cancer screening programs [59][60][61][62]. In addition, advances in HPV detection technologies, including nextgeneration sequencing and liquid biopsy techniques, are promising for the diagnosis of HPV-related cancers [63][64][65]. With further research into the mechanisms of innate immunity in the GI tract, the association between respiratory viruses, such as influenza viruses and coronaviruses, has become increasingly close, especially in the context of several global public health outbreaks. The COVID-19 pandemic spurred research into intestinal replication of SARSCoV-2 and its GI manifestations [66]. Studies on ACE2mediated viral entry in enterocytes have directly informed the design of crossmucosal vaccines and oral antiviral therapies [67][68][69]. Research on the development of oral broad-spectrum vaccines based on gastrointestinal mucosal immunity to mitigate viral infection responses and address post-infection sequelae is ongoing [16,25]. Research indicates that nanotechnology and targeted delivery systems can enhance vaccine immunogenicity. For example, nanobodies targeting dendritic cells have shown promising results in boosting immune responses. Additionally, novel mucoadhesive films and particle-based delivery systems for oral vaccines provide more stable and controlled release, improving vaccine efficacy and patient compliance [70][71][72]. Collectively, these approaches signal a future where integrated mucosal immunization and precision diagnostics may jointly inform population-level disease control strategies, particularly in resource-constrained settings.
## Emerging research topics: gut microbiome and IBD
Metabolites of the gut microbiome are vital in viral infections, particularly short-chain fatty acids that mediate gut epithelial and immune regulation [73,74]. Disruption of the gut microbiome plays a significant role in systemic diseases [75,76]. Recent studies indicate that viruses can indirectly influence immune responses by interacting with the gut microbiota, potentially disrupting gut barrier function [23,77]. In addition, the specific mechanisms by which the interaction between gut microbiome and viruses contribute to IBD development are not yet fully understood [78,79]. This highlights the potential for integrated interventions-combining antiviral agents, microbiota modulation, and targeted immunosuppression-to improve disease management and patient quality of life. Accordingly, probiotics, fecal microbiota transplantation, and dietbased interventions are under evaluation as adjunctive therapies to restore mucosal health in virusassociated diseases [80,81]. However, the heterogeneity in microbial-host-virus interactions and the context-dependent nature of dysbiosis underscore a major research gap: elucidating causality versus correlation in IBD progression remains a key frontier [82]. Large-scale randomized controlled trials are needed to define optimal strains, dosing regimens, and patient selection for microbiotabased interventions.
Dependence on the unique algorithms of a single software package may introduce bias, whereas employing multiple tools in tandem can enhance the robustness and accuracy of the analysis. Although relying solely on the WoSCC database may have overlooked non-English publications, it is among the most commonly used databases for bibliometric research [5,32,33]. Furthermore, citation data may not fully reflect the impact of recently published studies owing to delayed accumulation. However, through the novelty of our four-decade bibliometric mapping, the use of advanced visualization software to uncover collaboration networks and thematic trends, and the identification of high-impact research hotspots.
## Conclusion
Research on viruses and GI mucosal immunity is expanding rapidly. Affiliations and authors of the USA dominate this field. Future studies should prioritize interdisciplinary collaboration. Key fields include HIV pathogenesis and ART. In addition, hotspots worth monitoring are the prevention and early diagnosis of HPV-related diseases, the role of the gut microbiome in antiviral defense, IBD pathogenesis, and prognostic prevention. These areas represent promising frontiers for translational innovation particularly in mucosal vaccine development, early detection, microbiota-targeted therapies, and integrative approaches to immune modulation.
## References
1. Hussain, Pan (2015) "Circadian regulators of intestinal lipid absorption" *J Lipid Res*
2. Wang (2007) "Regulation of intestinal cholesterol absorption" *Annu Rev Physiol*
3. Wershil, Furuta
4. (2008) "Gastrointestinal mucosal immunity"
5. Ionescu, Nagler (2024) "The role of intestinal bacteria in promoting tolerance to food" *Curr Opin Immunol*
6. Abavisani, Foroushan, Kesharwani et al. (2025) "The frontier of health: exploring therapeutic potentials of the microbiome" *PharmaNutrition*
7. Abavisani, Ebadpour, Khoshrou et al. (2024) "Boosting vaccine effectiveness: the groundbreaking role of probiotics" *J Agric Food Res*
8. Tugizov (2021) "Virus-associated disruption of mucosal epithelial tight junctions and its role in viral transmission and spread" *Tissue Barriers*
9. Paradis, Bègue, Basmaciyan et al. (2021) "Tight junctions as a key for pathogens invasion in intestinal epithelial cells" *Int J Mol Sci*
10. Kummerlowe, Mwakamui, Hughes (2022) "Single-cell profiling of environmental enteropathy reveals signatures of epithelial remodeling and immune activation" *Sci Transl Med*
11. Cerutti, Chen, Chorny (2011) "Immunoglobulin responses at the mucosal interface" *Annual Review of Immunology. Annual Reviews*
12. Velikova, Snegarova, Kukov et al. (2021) "Gastrointestinal mucosal immunity and COVID-19" *World J Gastroenterol*
13. Morse, Simister, Crowe et al. (2023) "Virus induced dysbiosis promotes type 1 diabetes onset" *Front Immunol*
15. Lu, Hitch, Zhou (2024) "A host-adapted auxotrophic gut symbiont induces mucosal immunodeficiency" *Science*
16. Rose (2014) "Mucosal immunity and acute viral gastroenteritis the example rotavirus" *Review. Human Vaccines Immunother*
17. Bányai, Estes, Martella et al. (2018) "Viral gastroenteritis" *Lancet*
18. Li, Molleston, Lovato (2024) "Sequential early-life viral infections modulate the microbiota and adaptive immune responses to systemic and mucosal vaccination" *PLoS Pathog*
20. Salinas, Gosselin, Marchand (1016) "IL-17A reprograms intestinal epithelial cells to facilitate HIV-1 replication and outgrowth in CD4 + T cells" *Article iScience*
21. Flynn, Ortiz, Vujkovic-Cvijin "Translocating bacteria in SIV infection are not stochastic and preferentially express cytosine methyltransferases"
22. (2024) *Mucosal Immunol*
23. Al-Talib, Dimonte, Humphreys (0114) "Mucosal T-cell responses to chronic viral infections: implications for vaccine design" *Cell Mol Immunol*
24. Van Den Hoven, Thaller, Bankuti (2017) "Chronic inflammatory bowel disease (CIBD) and Celiac disease, are these linked in the horse? Article" *Pferdeheilkunde*
25. Lamers, Beumer, Van Der Vaart (2020) "SARS-CoV-2 productively infects human gut enterocytes" *Science*
26. Nascimento, Aquino, Sousa (2024) "SARS-CoV-2 spike protein triggers gut impairment since mucosal barrier to innermost layers: from basic science to clinical relevance" *Mucosal Immunol*
27. Raj, Bruce, Anbalagan (2024) "COVID-19 influenced gut dysbiosis, post-acute sequelae, immune regulation, and therapeutic regimens" *Front Cell Infect Microbiol*
28. Ranasinghe (2014) "New advances in mucosal vaccination" *Immunol Lett*
29. Kocabiyik, Amlashi, Vo (2024) "Vaccine targeting to mucosal lymphoid tissues promotes humoral immunity in the gastrointestinal tract" *Sci Adv*
30. Lo, Schroeder, Roberts (2024) "CTLA-4 expressing innate lymphoid cells modulate mucosal homeostasis in a microbiota dependent manner" *Nat Commun*
31. Heidari, Vareki, Yaghobi et al. (2024) "Microbiota activation and regulation of adaptive immunity" *Front Immunol*
32. Chen, Yan, Pan (2024) "Bibliometric analysis of mucosal immunity in IgA nephropathy from 1990 to 2022. Article. Immun Inflamm Dis"
33. Wang, Wu, Ran (2023) "Study of the relationship between mucosal immunity and commensal microbiota: a bibliometric analysis" *Rev Nutrients*
34. Broadus (1987) "Toward a definition of bibliometrics. Article" *Scientometrics*
35. Jiang, Liu, Zheng (2023) "Evolutionary patterns and research frontiers in neoadjuvant immunotherapy: a bibliometric analysis" *Review. Int J Surg*
36. Brandt, Hadaya, Schuster et al. (2019) "A bibliometric analysis of top-cited journal articles in obstetrics and gynecology" *JAMA Netw Open*
37. Birkle, Pendlebury, Schnell et al. (2020) "Web of science as a data source for research on scientific and scholarly activity" *Quant Sci Stud*
38. Tian, Chen (2024) "Global research progress of gut microbiota and epigenetics: bibliometrics and visualized analysis" *Front Immunol*
39. Aria, Cuccurullo (2017) "Bibliometrix: an R-tool for comprehensive science mapping analysis" *J Informetrics*
40. Van Eck, Waltman (2007) "VOS: a new method for visualizing similarities between objects"
41. Chen, Citespace (2006) "Detecting and visualizing emerging trends and transient patterns in scientific literature" *J Am Soc Inf Sci Technol*
42. Brenchley, Schacker, Ruff (2004) "CD4 < SUP>+ T cell depletion during all stages of HIV disease occurs predominantly in the gastrointestinal tract" *Journal Experimental Medicine*
43. Shapiro, Biezuner, Linnarsson (2013) "Single-cell sequencing-based technologies will revolutionize whole-organism science" *Nat Rev Genet*
44. Soon, Hariharan, Snyder (2013) "High-throughput sequencing for biology and medicine" *Mol Syst Biol*
45. Zuo, Zhang, Lui (2020) "Alterations in gut microbiota of patients with COVID-19 during time of hospitalization" *Gastroenterology*
46. Anderson, Hollingsworth, Baggaley et al. (2020) "COVID-19 spread in the UK: the end of the beginning?" *Lancet*
47. Zeng, Shen, Zhou (2019) "Increasing trend of scientists to switch between topics" *Nat Commun*
48. De Domenico, Omodei, Arenas (2016) "Quantifying the diaspora of knowledge in the last century" *Appl Netw Sci*
49. Haneishi, Furuya, Hasegawa et al. (2023) "Inflammatory bowel diseases and gut microbiota" *Int J Mol Sci*
50. Foppa, Rizkala, Repici et al. (2024) "Microbiota and IBD: current knowledge and future perspectives" *Dig Liver Dis*
51. Li, Tubiana, Katlama et al. (1998) "Long-lasting recovery in CD4 T-cell function and viral-load reduction after highly active antiretroviral therapy in advanced HIV-1 disease" *Lancet*
53. Brenchley (2013) "Mucosal immunity in human and simian immunodeficiency lentivirus infections" *Rev Mucosal Immunology*
54. Veazey, Shattock, Klasse et al. (2012) "Animal models for microbicide studies. Article" *Curr HIV Res*
55. Veazey, Shattock, Pope "Prevention of virus transmission to macaque monkeys by a vaginally applied monoclonal antibody to HIV-1"
56. (2003) *Nature Medicine*
57. Sharland, Watkins, Dalgleish et al. (1998) "Immune reconstitution in HAART-treated children with AIDS. Highly active Anti-Retroviral therapy" *Lancet*
58. Vellas, Martres, Requena (2024) "Compartmentalized HIV-1 reservoir in intestinal monocytes/macrophages on antiretroviral therapy" *J Infect Dis*
59. Vasukutty, Jang, Han et al. (2024) "Navigating latency-inducing viral infections: therapeutic targeting and nanoparticle utilization" *Biomater Res*
60. Sperber, Raymond, Bouzidi (2023) "The hypoxia-regulated ectonucleotidase CD73 is a host determinant of HIV latency" *Cell Rep*
61. Schriek, Aldon, Van Gils et al. (2024) "Next-generation bNAbs for HIV-1 cure strategies" *Antiviral Res*
62. Garrido, Clohosey, Whitworth et al. (2018) "Soriano-Sarabia N. γδ T cells: an immunotherapeutic approach for HIV cure strategies" *JCI Insight*
63. Ahmed, Herschhorn (2024) "mRNA-based HIV-1 vaccines" *Clin Microbiol Rev*
64. Taghinezhad, Keyvani, Bermúdez-Humarán et al. (2021) "Twenty years of research on HPV vaccines based on genetically modified lactic acid bacteria: an overview on the gut-vagina axis" *Cell Mol Life Sci*
65. Ang, Harris, Wheeler (2010) "Human papillomavirus and survival of patients with oropharyngeal cancer" *N Engl J Med*
66. Palefsky, Giuliano, Goldstone (2011) "HPV vaccine against anal HPV infection and anal intraepithelial neoplasia" *N Engl J Med*
67. Templeton, Laimins (2023) "p53-dependent R-loop formation and HPV pathogenesis" *Proc Natl Acad Sci U S A*
68. Wang, Wang, Chen "Rational design of a cross-type HPV vaccine through immunodominance shift guided by a cross-neutralizing antibody"
69. (1016)
70. Ma, Gharizadeh, Cai (2022) "A comprehensive HPV-STI NGS assay for detection of 29 HPV types and 14 non-HPV sexually transmitted infections" *Infect Agent Cancer*
71. Wu, Wang, Zhang (2022) "Detection of five types of HPV genotypes causing anogenital warts (Condyloma Acuminatum) using PCR-Tm analysis technology" *Front Microbiol*
72. Ma, Li, Cai (2022) "Development of an HPV genotype detection platform based on aggregation-induced emission (AIE) and flow-through hybridization technologies" *Molecules*
73. Bao, Deng, Huang (2020) "The pathogenicity of SARS-CoV-2 in hACE2 transgenic mice" *Nature*
74. Lamers, Beumer, Van Der Vaart (2020) "SARS-CoV-2 productively infects human gut enterocytes" *Science*
75. Viana, Nunes, Reis (2020) "ACE2 imbalance as a key player for the poor outcomes in COVID-19 patients with age-related comorbidities -role of gut microbiota dysbiosis" *Ageing Res Rev*
76. Zhao, Xie, Lin et al. (2022) "The mucoadhesive nanoparticle-based delivery system in the development of mucosal vaccines" *Int J Nanomed*
77. Jia, Sun, Ge (2023) "Chicken dendritic cell-targeting nanobodies mediated improved protective effects against H9N2 influenza virus challenge in a homologous sequential immunization study" *Article. Vet Microbiol*
78. Arafat, El Rahman, Naguib et al. (2020) "Coinfection of < i > salmonella enteritidis with H9N2 avian influenza virus in chickens" *Article. Avian Pathol*
79. Jones, Swayne (2004) "Comparative pathobiology of low and high pathogenicity H7N3 Chilean avian influenza viruses in chickens" *Article. Avian Dis*
80. Qing, Xie, Su (2019) "Gut microbiome, short-chain fatty acids, and mucosa injury in young adults with human immunodeficiency virus infection. Article" *Dig Dis Sci*
81. Wu, Zhang, Wang (2023) "Short-chain fatty acids alleviate vancomycin-caused humoral immunity attenuation in rabies-vaccinated mice by promoting the generation of plasma cells via Akt-mTOR pathway" *Article. J Virol*
82. Zhang, Ma, Luo (2024) "Low-medium polarity ginsenosides from wild ginseng improves immunity by activating the ahr/mapk pathway through tryptophan metabolism driven by gut microbiota" *J Agric Food Chem*
83. Yan, Sun, Fan et al. (2024) "Unveiling the role of gut microbiota and metabolites in autoimmune thyroid diseases: emerging perspectives" *Int J Mol Sci*
84. Alvarez, Stork, Sayoc-Becerra et al. (2019) "A simulated microgravity environment causes a sustained defect in epithelial barrier function" *Article. Sci Rep*
85. Beheshti-Maal, Shahrokh, Ansari (2021) "Gut mycobiome: the probable determinative role of fungi in IBD patients" *Rev Mycoses*
86. Alzahrani, Hussain, Simar (2019) "Inflammatory and immunometabolic consequences of gut dysfunction in HIV: parallels with IBD and implications for reservoir persistence and non-AIDS comorbidities" *Rev EBioMedicine*
87. Wang, Moon, Huang et al. (2022) "Antiviral effects and underlying mechanisms of probiotics as promising antivirals" *Review. Front Cell Infect Microbiol*
88. Brunse, Deng, Pan (2022) "Fecal filtrate transplantation protects against necrotizing enterocolitis. Article" *Isme J*
89. (1107)
90. Tian, Li, Wang (2024) "Gut virome-wide association analysis identifies cross-population viral signatures for inflammatory bowel disease. Microbiome" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12602096&blobtype=pdf | # A case of secondary lymphangiectasia due to ancylostoma intestinal infection The authors
Agostino Cosenza, Luca Elli, Anna Maraschini, Elia Fracas, Lucia Scaramella
## Abstract
reporting on interesting cases and new techniques in gastroenterological endoscopy. All papers include a high-quality video and are published with a Creative Commons CC-BY license. Endoscopy E-Videos qualify for HINARI discounts and waivers and eligibility is automatically checked during the submission process.
A 33-year-old patient presented to an outpatient clinic with a diagnosis of intestinal lymphangiectasia following the onset of dependent edema, hypoalbuminemia, and altered liver transaminases in January/2024. Further evaluation revealed hypogammaglobulinemia, IgA/IgG/IgM deficiency, negative anti-TTG antibodies, normal fecal calprotectin/elastase, and persistent hypoalbuminemia.
The patient was on biweekly human albumin infusions, medium-chain triglyceride (MCT) oil, and vitamin D supplementation.
To further investigate the suspected gastrointestinal pathology, the patient underwent the following diagnostic procedures: Esophagogastroduodenoscopy (EGD): Duodenal bulb/D2 with epithelial disruption. Histology showed normotrophic villi with lymphoplasmacytic/eosinophilic infiltration. Helicobacter pylori and Tropheryma whipplei were negative, and intra-epithelial T-lymphocyte count (CD3+) was not increased. Video capsule endoscopy (VCE): Diffuse lymphangiectasia in the duodenum/proximal jejunum, pseudopolypoid appearance in some frames, and mild distal villous edema. Due to the persistence of lymphangiectasia with minimal improvement following albumin and MCT oil supplementation, further evaluation of the small intestine was pursued: Capsule endoscopy (PillCam SB3, Given Medtronic): demonstrated mild lymphangiectasia in the proximal and mid-small intestine with mild villous edema (▶ Fig. 1 and ▶ Fig. 2). Anterograde double-balloon enteroscopy (DBE) (Fujifilm EN-580 T): Diffusely edematous mucosa with shortened, whitish villi, consistent with lymphangiectasia in the explored segments. Presence of two pseudopolyps in the proximal jejunum, which were biopsied, and a vermiform parasite, which was removed (▶ Video 1). Histopathological analyses showed no dysplasia in the two pseudopolyps, which exhibited only mild inflammation of the lamina propria. Parasitological analysis identified the specimen as two hook-▶ Fig. 1 Diffuse edema of the villi with lymphangiectasia in the duodenum. Article published online: 2025-11-10 worms of Ancylostoma spp. (male and female in reproductive phase) (▶ Fig. 3 and ▶ Fig. 4). Despite negative stool cultures for hookworm (repeated twice), only Blastocystis hominis was detected. Infectious disease specialists advised treating Ancylostoma with Albendazole (400 mg TID) with a rapid improvement of Immunoglobulin and albumin blood levels. Endoscopy_UCTN_Code_CCL_1AC_2AG |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12821075&blobtype=pdf | # Geographical and Ecological Drivers of Zoonotic Viral Spillover: A Review of Emerging and Reemerging Outbreaks
Mahendra Verma, Harjeet Maan, Shravya Konatam, Yogendra Verma, Rohit Kumar, Deepti Chaurasia, Lokendra Dave, Shweta Sharma
## Abstract
Over the past two decades, outbreaks of zoonotic viruses have become increasingly frequent and severe, posing substantial threats to public health systems and the global economy. The viruses responsible for these outbreaks, such as SARS-CoV, MERS-CoV, Zika, Ebola, Nipah, avian influenza, and, most recently, SARS-CoV-2, typically originate in wildlife, highlighting the complex relationship between ecological systems and human activities. Human-wildlife interactions have markedly increased due to disruptions in environmental and geographic boundaries, primarily driven by urbanization, deforestation, intensified agricultural practices, and climate change. These factors contribute to an environment that facilitates zoonotic transmission spillover. This narrative review summarizes current research on the ecological, geographic, and human factors influencing zoonotic virus transmissions. It emphasizes how these viruses adapt to human hosts and cross species barriers via direct contact, vector-borne transmission, intermediate carriers, and environmental contamination. Moreover, the review discusses how the genomic plasticity of viruses enhances their transmissibility and facilitates adaptation to new hosts, thereby increasing the risk of epidemics and pandemics.The review further underscores the importance of ecological boundaries in mitigating spillover events and advocates for a One Health approach that integrates human, animal, and environmental health. This approach is essential for predicting, detecting, and preventing future outbreaks. In conclusion, the review emphasizes the importance of interdisciplinary research, proactive surveillance, habitat preservation, and policy interventions that address the underlying ecological factors contributing to zoonotic outbreaks. Restoring ecological barriers and implementing sustainable practices to minimize the interaction between wildlife and humans, while bolstering global biosecurity, are essential measures to mitigate the risk of future pandemics.
## Introduction And Background
The twenty-first century has witnessed several zoonotic disease outbreaks, which are infectious illnesses transmitted from animals to humans. Many of these outbreaks have had significant global impacts, highlighting the complex connections between human and animal populations [1]. Research from 1970 onward has clearly shown that, out of approximately 1,500 recorded diseases, nearly 70% originated from animals (wildlife) [2]. Furthermore, the World Health Organization identified 15 viruses of zoonotic origin that pose a worldwide threat to human health. Among these pathogens, respiratory viruses are the most common, with the potential for causing both epidemic and endemic outbreaks [3]. Moreover, zoonotic pathogens are not limited to causing respiratory infections; they also cause other deadly diseases, such as HIV and the Zika virus. Over the past twenty years, zoonotic viruses, including SARS-CoV, MERS-CoV, Ebola, Influenza, and COVID-19, have affected the global population, leading to significant global mortality. COVID-19 alone has resulted in over 7 million fatalities worldwide [4]. The primary zoonotic pathogens impacting the global population include SARS-CoV, MERS-CoV, Zika, Ebola, Nipah [5], and Influenza, along with the more recent SARS-CoV-2 and Monkeypox [6], as discussed here (Table 1). It is important to remember that most zoonotic viruses originate from a common ancestor: wild animals. Given the recent and sharp rise in zoonotic viral infections in humans, understanding how pathogens spread from wildlife to humans across different regions is crucial [7]. Over the past few decades, human-wildlife interactions have advanced beyond points where ecological and geographic barriers once contained these viruses, now permitting them to cross species boundaries [8,9]. Does restoring barriers between wildlife and domestic animals help prevent future viral outbreaks?
## Review
## Zoonotic virus
Over the last two decades, the world has witnessed outbreaks of several viruses with the potential to become endemic, epidemic, or pandemic, with most of these viruses originating from zoonotic sources [10]. Zoonotic viruses are viruses that can be transmitted between animals and humans. These viruses could cross species barriers, causing infections in both animals and humans. Zoonotic transmissions can occur through direct contact with infected animals, the consumption of contaminated food or water, or exposure to vectors such as mosquitoes or ticks that carry the virus between species [11]. While some zoonotic viruses may only cause minor infections, others can lead to severe conditions that have a significant impact on public health. It is essential to investigate the variables and causes that contribute to the transmission of viruses from wild types to humans, as evidenced by the re-emergence and emergence of zoonotic viral infections and their ability to infect humans despite ecological limits [12]. It is noteworthy to mention that the viral genome is highly unstable, and changes influence genomic plasticity in the environment. Zoonotic viruses pose challenges for public health and require a coordinated One Health approach, involving collaboration between human and animal health sectors, as well as environmental scientists [13]. Monitoring, surveillance, and research are crucial for understanding the dynamics of zoonotic diseases and preventing their spread. The ongoing COVID-19 pandemic, caused by the SARS-CoV-2 virus, is another example of a zoonotic virus that likely originated in bats with an unknown intermediate host [14].
Animals, especially birds and mammals, naturally carry many viruses. Typically, these viruses don't cause symptoms or only minor ones in their animal hosts, as they have coevolved with them [15]. Many viruses found in wild animals have the potential to infect humans. These viruses are pretty diverse. A study by Pearce-Duvet highlights that bats harbor a wide range of viruses and have been linked to several zoonotic outbreaks [16]. Wild animals act as natural reservoirs for various infections, making them a key part of the ecology of these viruses. Typically, zoonotic viruses that transfer from animals to humans originate in wildlife. To prevent and control the spread of infectious diseases, understanding the relationships between animals and these viruses is essential [11]. Human interaction with natural environments, often beyond ecological limits, is a significant factor driving the emergence, reappearance, and re-emergence of zoonotic viruses as they cross species boundaries [17]. These zoonotic events occur when viruses that naturally infect animals, mostly wildlife, spill over to humans. Such events can lead to the emergence of new human diseases and occasionally result in large outbreaks [18]. To effectively prevent and mitigate the impact of zoonotic events, it is essential to understand these factors and closely monitor high-risk areas and species.
Addressing the complex dynamics of viral spread and zoonotic transmission requires a One Health approach,
## Crossing species boundaries and zoonotic transmission
According to Zumla and Hui, close contact with wildlife constitutes a critical determinant in the occurrence of viral spillover events. Regardless of whether the interaction is direct or indirect, it significantly elevates the risk of zoonotic viral transmission [20]. Human encroachment into natural habitats through activities such as hunting, farming, and other anthropogenic actions further increases human proximity to wildlife [21]. The presence of wildlife itself also heightens the risk of zoonotic spillover, as numerous studies have demonstrated that wildlife functions as a natural reservoir for zoonotic viruses and other pathogenic microorganisms [22]. It is noteworthy that these zoonotic viruses are typically nonpathogenic to the wildlife hosts; however, alterations in host species can pose severe health threats, potentially leading to the emergence of life-threatening diseases [23]. For instance, bats are recognized as a reservoir for zoonotic viruses with pandemic potential. The transmission of zoonotic viruses largely depends on intermediate hosts, which can carry the viruses with epidemic and pandemic potential [24]. Notably, intermediate hosts can amplify the virus, thereby facilitating its transmission to humans. An example of this is the 2002-2003 SARS-CoV outbreak, where civet cats were identified as intermediate hosts. The consequences of globalization and urbanization, which result in the loss of wildlife habitats, further exacerbate the risk of zoonotic virus spillover [25].
Urbanization and globalization have gradually led wildlife to venture outside their usual habitats, which can sometimes result in zoonotic viral spillovers. Another challenge in our modern era is food scarcity, which has prompted a rise in the popularity and expansion of animal husbandry over recent decades [26]. It has been reported that intensive farming practices, especially when animals are kept in close quarters, can create conditions conducive to the spread of zoonotic infections from animals to humans, for example, swine flu and avian flu. Additionally, changes to natural ecosystems, including deforestation and climate change, impact the distribution and behavior of wildlife, often increasing interactions with domestic animals [27]. Genetic modifications can also enhance viruses' ability to infect new hosts, including humans, as they evolve.
This evolution is driven by factors such as genetic variations, selection pressures, and the fitness landscape, which collectively heighten the risks of pandemics, drug resistance, and expanding host ranges. As viruses and their variants evolve, the likelihood of zoonotic transmission increases-consider examples such as SARS-CoV-2, avian influenza, and Ebola, where viral evolution has demonstrated a broad host range [28]. The use of antimicrobial drugs in agriculture can also lead to the development of drug-resistant viral strains, posing additional challenges.
## Zoonotic transmission pathways
Robert and Baylis examined how human activities, such as deforestation, urbanization, and agricultural development, constitute significant factors in habitat loss, which in turn facilitate viral spillover and frequent outbreaks [11]. Habitat destruction is a critical element that not only precipitates viral spillovers but also enables species jumps and transmission to humans. In a study, Zumla and Hui demonstrated that disease outbreaks within wildlife populations can often originate from other animal species [20]. To mitigate the occurrence of recurrent viral spills, it is essential to preserve wildlife ecosystems. It is reported that within these ecosystems, particular species serve as reservoirs for viruses that possess pandemic potential, such as bats. Coronaviruses are naturally found in bat habitats, and previous reports have indicated that outbreaks of SARS-CoV and SARS-CoV-2, as well as filoviruses such as the Ebola virus, are associated with these reservoirs. Notably, despite the presence of viruses in bats, their immune systems confer protection against disease. According to a study by Spencer et al., viral spillover is a widespread phenomenon whereby viruses transmit from one host species to another, including humans [19]. In cases of viral spillover, the route of transmission may vary, with direct interaction between hosts and humans being particularly significant. Other mechanisms of transmission include fomites, aerosols, and water. Although viral spillover has posed a considerable threat to human health over the past two decades, there has been a notable increase in instances where viruses have crossed species barriers to infect humans, resulting in various diseases [9]. Understanding the ecology of wildlife hosts and the factors influencing zoonotic transmission is vital for early detection of outbreaks and the implementation of appropriate measures [29].
The ongoing research focuses on the dissemination and dynamics of viruses, which affect communities.
Although wildlife naturally hosts viruses with pandemic potential, transmission to humans requires either an intermediate host or occurs through aerosolized particles and contaminated surfaces, known as fomites.
The transmission of zoonotic viruses with pandemic potential depends on the mode of transmission. Kampf et al. demonstrated that intermediate hosts or vectors facilitate not only the rapid propagation of viruses but also subsequent infections [30]. When inanimate objects contaminated by an infected person come into contact with an animal or person who is susceptible to the disease, this is known as fomite transmission. The surfaces in COVID-19 were identified as a high-risk zone for viral dissemination. Exam tables, cages, kennels, medical equipment, environmental surfaces, clothing, and other items may all be categorized as fomites [33]. The term aerosol transmission pertains to the dissemination of illnesses via minute particles or droplet nuclei. Aerosol particles may settle on mucosal membranes or environmental surfaces, or they may be inhaled by a susceptible host [34]. Such transmission can occur during specialized medical procedures, including suctioning, bronchoscopy, dentistry, and inhalation anesthesia, or when an infected individual breathes, coughs, sneezes, or vocalizes. Air currents within a room or facility can disperse small particles, which can remain suspended in the environment for prolonged durations [35]. Nonetheless, most pathogens relevant to veterinary care in companion animals are diminutive and are not capable of surviving for extended periods in the environment; consequently, they are unlikely to spread disease through close or prolonged contact.
## Ecological boundaries and virus spillover
Ecological boundaries and changes within ecosystems can significantly influence the dynamics of viral outbreaks, particularly in the context of zoonotic diseases transmitted from animals to humans [11]. Various ecological factors contribute to the emergence and dissemination of infectious diseases, and a comprehensive understanding of these dynamics is crucial for effective prevention and mitigation of outbreaks. A recent study by Holmes elucidates the ecology of viral emergence and the role of ecological boundaries in the rapid appearance and reemergence of viral pathogens [12]. When an animal's migration results in species crossing, it is often straightforward to identify cases involving clear pathology. Zoonosis refers to the transmission of viruses from non-human hosts to humans [36]. Unpredictable events, involving complex interactions between the virus and the newly adopted host, occur during species crossings (Figure 1). Notable examples of viruses that have crossed the species barrier and become established within the human population include HIV and the contemporary human influenza virus, which do not necessarily require the initial animal reservoir to persist [37]. Fortunately, it is uncommon for a new virus to adapt naturally and spread extensively among humans. Frequently, the virus encounters difficulties in successfully infecting and transmitting between individuals. Human activities that alter the environment, technology, and ecological systems often precipitate the emergence of novel viral infections [38].
## FIGURE 1: An overview of the interconnections between human activities and zoonotic spillover, highlighting the various mechanisms involved in the zoonotic process.
An Illustration of the zoonotic spillover process and mechanism, and engagement of various drivers. Further, the figure decipher the role of geographical and ecological boundaries in the emergence and reemergence of zoonotic infections to the human. In the zoonotic process, the spillover of the pathogen from the reservoir to the host has been illustrated. The figure also enumerates the drivers, such as human engagements and climate changes in progressing zoonotic spillover.
Image created by the authors, Dr. Mahendra Kumar Verma and Dr. Harjeet Singh Maan, using Microsoft Artwork tools.
These activities may lead to an increase in human-animal contact with animal hosts that act as reservoirs for zoonotic viruses [39]. Agricultural development, increased exploitation of environmental resources, population growth and mobility, and the trade and transportation of food and livestock have all been linked to the emergence and spread of several new viruses throughout the human population [40]. It seems plausible that the COVID-19 pandemic has led to increased awareness of the growth in zoonotic illnesses. Therefore, this awareness may also skew our knowledge of the ecology and evolution of viruses, particularly concerning the direction and timing of host-jumping events [41]. Understanding the processes that lead to disease onset and averting future zoonotic catastrophes requires acknowledging that viruses have been an integral part of global ecosystems for a long time, long before they became clinically or agriculturally significant [42]. This understanding also underscores the close relationship between ecological disturbance and the onset of disease. As humans become a larger part of the global ecosystem that contains viruses, the key issue is not merely that zoonotic viruses infect humans; rather, it is the increasing frequency and how modern human society influences it. When viruses are viewed through the lens of ecosystems, the widely accepted idea of One Health expands [43]. The ecosystems perspective is more expansive than One Health, which focuses primarily on viruses as disease agents, particularly those that connect illnesses in humans and animals [28].
Considering all major viruses, including those that have adapted to coexist harmoniously with their hosts, as well as the variables that disrupt ecosystems and increase the likelihood of disease outbreaks, along with the complex repercussions that follow ecological disturbances [24]. An essential principle of the ecosystem approach is that, although zoonotic virus transmission to humans is a common and regular occurrence, disease outbreaks are infrequently caused by this process. While the primary causes remain unclear, metagenomic research consistently demonstrates that healthy animal species can harbor a diverse array of viruses without exhibiting apparent health effects [44]. Traditionally, virologists have predominantly focused on studying viruses that infect humans, domesticated animals, or plants of significant importance to human society. Although this anthropocentric perspective is understandable, it may have contributed to a limited understanding of the scope and interconnectedness of the global virosphere [45]. The term zoonosis inherently implies directionality: viruses possess animal reservoirs, representing the source populations, and then transmit to humans as novel hosts. This view naturally considers humans as the endpoint of an evolutionary process. Although all respiratory viruses infecting humans ultimately originate from those found in other animals, humans are not the sole recipients of these viruses [46].
The endemic barrier associated with cross-species migration, which relates to the emergence of infectious diseases, has been extensively examined [47]. From micro and macro perspectives, these barriers represent two aspects of the ecological barrier that influence the probability of emerging viruses spreading within human populations. The cross-species barrier signifies the infrequency with which viruses effectively transmit among novel hosts that have not previously been exposed to them or are not susceptible [44]. Viral mutation or evolution is principally responsible for breaching the cross-species barrier, facilitating spillover infections into other hosts. Such events enable viruses to gradually adapt to new host cells and ultimately disseminate into additional populations. Conversely, overcoming the endemic barrier depends on the likelihood and frequency of viral transmission, which is closely linked to interactions between natural hosts of viruses and humans or potential hosts [28]. Despite numerous investigations into early outbreaks and epidemics of infectious diseases and their association with cross-species or endemic barriers [39], a comprehensive and systematic analysis of viral migration and transmission within ecosystems from a macro perspective remains lacking. These investigations mainly emphasize epidemiology and immunology.
The precise zoonotic origin of viruses, such as those with pandemic potential (coronavirus and influenza), possesses a higher degree of genomic plasticity that facilitates molecular evolution. During the COVID-19 pandemic, rapid surges of several variants were observed in a short period. Several factors, including host, environment, and therapeutics, affect viral genomic plasticity and its molecular evolution. Additionally, there are natural mechanisms by which microbes regularly change their genome for improved survival. He et al. note that genomic plasticity is primarily influenced by a changing environment, where human interference plays a critical role [38]. Human interference with wildlife causes habitat loss, where constant climate change shows the emergence and reemergence of viral outbreaks. Therefore, by altering interactions between humans and the natural environment, increased human activity may reduce ecological barriers and accelerate the spread of viruses within human civilizations [48]. Zhang et al. demonstrated how human interference breached the ecological barrier that serves as a protective boundary between wildlife and humans, thereby increasing the possibility of zoonotic viral spillover [36]. More precisely, the four main ecological barrier elements essential for viral transmission, which act as a molecular or endemic barrier, are transmission pathways, contact likelihood, contact frequency, and viral characteristics. The environmental barrier integrates all possible obstacles to viral transmission from the virus's natural or intermediate hosts to human society, making it a vital hub for recently developing infectious illnesses [37].
## One Health approach to tackle viral spillover
The relationships among animal, human, and environmental health are particularly evident in the context of zoonoses [38]. These refer to illnesses or infections that vertebrate animals can naturally transmit to humans. The emergence or re-emergence of pathogens can occur through various mechanisms. For example, the migration of known pathogens to new regions (e.g., the Ebola virus spreading to West Africa) and the adaptation of pathogens to new hosts (e.g., the H5N1 avian influenza virus adapting to humans) exemplify these mechanisms [25]. At the microscopic level, emergence and re-emergence involve genetic modifications that may lead to increased resistance to antimicrobial agents. Therefore, drivers of emergence can impact our environment at multiple scales, potentially resulting in pathogenic alterations that give rise to novel disease entities [28]. A critical challenge in recognizing and detecting newly emerging zoonoses is inherently linked to this broad scope. Additionally, biological factors are not the sole contributors. The initial human case marks the point at which a new disease from an animal reservoir host transmits to humans. This juncture is likely to influence the subsequent dissemination of disease within human populations, considering factors such as ecology, sociology, and human and animal behaviors [49]. The One Health approach is an interdisciplinary framework that acknowledges the interconnectedness of human, animal, and environmental health in addressing complex health issues, such as the prevention and control of viral spill-over events. It emphasizes collaboration across various sectors, including public health, veterinary medicine, environmental science, and human medicine [50].
In circumstances where surveillance and early detection systems for both human and animal health, in conjunction with environmental surveillance, require an integrated approach, a One Health strategy may serve as a pivotal preventive instrument to mitigate viral spillover. It is also imperative to undertake collaborative studies with specialists from various disciplines to understand the ecology of viruses, their natural reservoirs, and the factors influencing spillover events. To enhance collective knowledge and early warning mechanisms, fostering the exchange of data and information across environmental, animal, and human health sectors is revolutionary [25]. Additional measures are essential, such as identifying high-risk regions with sensitive ecological interfaces and assessing areas or ecosystems with significant potential for viral spillovers, considering variables including wildlife diversity, human-wildlife interactions, and land-use changes. The implementation of risk-reduction strategies, such as sustainable land management, wildlife management, and habitat conservation, is vital. Cross-species diagnostics constitute another health strategy that addresses ongoing viral spillover events [51]. To facilitate early detection of potential spillover incidents, it is necessary to develop and adopt diagnostic techniques capable of identifying and detecting viruses in both humans and animals. This approach helps pinpoint high-risk populations and supports surveillance efforts in communities where the likelihood of viral spillover is heightened, particularly those residing in regions undergoing ecological changes or close to wildlife habitats [52].
## Conclusions
Wildlife serves as a natural reservoir of viruses and remains confined within an ecological niche. Zoonosis is a phenomenon that permits viruses to transfer from native species to humans, often via an intermediate host. Zoonosis does not occur spontaneously; instead, it is driven by various factors and drivers that promote viral spillover, with human interference in wildlife ecosystems being a key factor. In wildlife populations, viruses generally do not cause significant harm to their reservoir hosts. The evolution of wildlife and viruses is a mutually beneficial process that often occurs without detrimental effects on the host (reservoir). To establish a more effective system, it is essential to consider risk factors at the interface between humans, animals, and the environment, as well as the sources of zoonotic disease outbreaks. Moreover, legal coherence between environmental treaties and animal health regulations must be enhanced. By reducing the likelihood of pathogen spillover, a One Health approach-promoting and safeguarding the health of animals and the environment-also enhances biosecurity in food production, ultimately benefiting both human and animal health.
## References
1. Bhadoria, Gupta, Agarwal (2021) "Viral pandemics in the past two decades: an overview" *J Family Med Prim Care*
2. Patel, Goel, Bhardwaj (2021) "Emerging and re-emerging viral infections in India" *J Prev Med Hyg*
3. Piret, Boivin (2020) *Pandemics throughout history. Front Microbiol*
4. (2023) "WHO coronavirus (COVID-19) dashboard"
5. Luby, Gurley, Hossain (2009) "Transmission of human infection with Nipah virus" *Clin Infect Dis*
6. Sklenovská, Van Ranst (2018) "Emergence of monkeypox as the most important orthopoxvirus infection in humans"
7. Baker, Mahmud, Miller (2022) "Infectious disease in an era of global change" *Nat Rev Microbiol*
8. Li, Liu, Yu et al. (2020) "Tang CK: Coronavirus disease 2019 (COVID-19): current status and future perspectives" *Int J Antimicrob Agents*
9. Wu, Chen, Chan (2020) "The outbreak of COVID-19: an overview" *J Chin Med Assoc*
10. Hui (2006) "Reasons for the increase in emerging and re-emerging viral infectious diseases" *Microbes Infect*
11. Robert, Baylis (2003) "Crossing species boundaries" *Am J Bioeth*
12. Holmes (2022) "The ecology of viral emergence" *Annu Rev Virol*
13. Zhang, Yang, Li (2022) "Ecological barrier deterioration driven by human activities poses fatal threats to public health due to emerging infectious diseases. Engineering (Beijing)"
14. Mahy, Brown (2000) "Emerging zoonoses: crossing the species barrier" *Rev Sci Tech*
15. Bengis, Leighton, Fischer et al. (2004) "The role of wildlife in emerging and reemerging zoonoses" *Rev Sci Tech*
16. (2006) "The origin of human pathogens: evaluating the role of agriculture and domestic animals in the evolution of human disease" *Biol Rev Camb Philos Soc*
17. Parrish, Holmes, Morens (2008) "Cross-species virus transmission and the emergence of new epidemic diseases" *Microbiol Mol Biol Rev*
18. Shuman (2011) "Global climate change and infectious diseases" *Int J Occup Environ Med*
19. Spencer, Finucane, Fox et al. (2020) "Emerging infectious disease, the household built environment characteristics, and urban planning: Evidence on avian influenza in Vietnam. Landsc Urban Plan"
20. Zumla, Hui (2019) "Emerging and re-emerging infectious diseases: global overview" *Infect Dis Clin North Am*
21. Escudero-Pérez, Lalande, Mathieu et al. (2023) "Host-pathogen interactions influencing zoonotic spillover potential and transmission in humans" *Viruses*
22. Lytras, Xia, Hughes et al. (2021) "The animal origin of SARS-CoV-2" *Science*
23. Plowright, Reaser, Locke (2021) "Land use-induced spillover: a call to action to safeguard environmental, animal, and human health" *Lancet Planet Health*
24. Keesing, Ostfeld (2021) "Impacts of biodiversity and biodiversity loss on zoonotic diseases" *Proc Natl Acad Sci U S A*
25. Vora, Hannah, Walzer (2023) "Interventions to reduce risk for pathogen spillover and early disease spread to prevent outbreaks, epidemics, and pandemics. Emerg Infect Dis"
26. Williams, Spruill-Harrell, Taylor (2021) "Common themes in zoonotic spillover and disease emergence: lessons learned from bat-and rodent-borne RNA viruses" *Viruses*
27. Milbank, Vira (2022) "Wildmeat consumption and zoonotic spillover: contextualising disease emergence and policy responses"
28. Ellwanger, Chies (2021) "Zoonotic spillover: understanding basic aspects for better prevention" *Genet Mol Biol*
29. French, Holmes (2020) "An ecosystems perspective on virus evolution and emergence" *Trends Microbiol*
30. Kampf, Brüggemann, Kaba et al. (2020) "Potential sources, modes of transmission and effectiveness of prevention measures against SARS-CoV-2" *J Hosp Infect*
31. Leung (2021) "Transmissibility and transmission of respiratory viruses" *Nat Rev Microbiol*
32. Rahman, Aziz, Hussein (2020) "The transmission modes and sources of COVID-19: A systematic review" *Int J Surg Open*
33. Karia, Gupta, Khandait et al. (2020) "COVID-19 and its Modes of Transmission" *SN Compr Clin Med*
34. Bushman, Mccormick, Sherrill-Mix (2019) "Virus structures constrain transmission modes" *Nat Microbiol*
35. Fermin (2018) "Host range, host-virus interactions, and virus transmission" *Viruses*
36. Zhang, Shi, Holmes (2018) "Using metagenomics to characterize an expanding virosphere" *Cell*
37. Carlson, Zipfel, Garnier et al. (2019) "Global estimates of mammalian viral diversity accounting for host sharing" *Nat Ecol Evol*
38. He, Hou, Zhao (2022) "Virome characterization of game animals in China reveals a spectrum of emerging pathogens" *Cell*
39. Cohen (2000) "Changing patterns of infectious disease" *Nature*
40. Daszak, Cunningham, Hyatt (2000) "Emerging infectious diseases of wildlife--threats to biodiversity and human health" *Science*
41. Kruse, Handeland (2004) "Wildlife as source of zoonotic infections" *Emerg Infect Dis*
42. Osterhaus (2001) "Catastrophes after crossing species barriers" *Philos Trans R Soc Lond B Biol Sci*
43. Ludwig, Kraus, Allwinn et al. (2003) "Viral zoonoses -a threat under control?" *Intervirology*
44. Domingo (1998) "Quasispecies and the implications for virus persistence and escape" *Clin Diagn Virol*
45. Lemieux, Colby, Poulain (2022) "Aris-Brosou S: Viral spillover risk increases with climate change in High Arctic lake sediments" *Proc Biol Sci*
46. Grange, Goldstein, Johnson (2021) "Ranking the risk of animal-to-human spillover for newly discovered viruses" *Proc Natl Acad Sci U S A*
47. Johnson, Hitchens, Evans (2015) "Spillover and pandemic properties of zoonotic viruses with high host plasticity" *Sci Rep*
48. Gavotte, Gaucherel, Frutos (2023) "Environmental spillover of emerging viruses: is it true?" *Environ Res*
49. Horefti "The importance of the One Health concept in combating zoonoses" *Pathogens. 2023*
50. Peterson, Bakuza, Standley (2020) "One Health approach: a data-driven priority for mitigating outbreaks of emerging and re-emerging zoonotic infectious diseases" *Trop Med Infect Dis*
51. Kelly, Machalaba, Karesh (2020) "Implementing One Health approaches to confront emerging and re-emerging zoonotic disease threats: lessons from PREDICT. One Health Outlook"
52. Reddy, Mathur (2021) "Pandemics & One Health: India's evolving response" *Indian J Med Res*
53. Verma (2025) *Cureus* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12472737&blobtype=pdf | # LMP7 as a Target for Coronavirus Therapy: Inhibition by Ixazomib and Interaction with SARS-CoV-2 Proteins Nsp13 and Nsp16
Yi Ru, Yue Ma-Lauer, Chengyu Xiang, Pengyuan Li, Brigitte Von Brunn, Anja Richter, Christian Drosten, Andreas Pichlmair, Susanne Pfefferle, Markus Klein, Robert Damoiseaux, Ulrich Betz, Albrecht Von Brunn
## Abstract
The emergence of human coronaviruses has led to three epidemics or pandemics in the last two decades, collectively causing millions of deaths and thus highlighting a long-term need to identify new antiviral drug targets and develop antiviral therapeutics. In this study, a compound library was screened to uncover novel potential inhibitors of coronavirus replication. Three lead compounds, designated #16-14, #16-23, and #16-24, which were Ixazomib and its analogs, were identified based on their potent antiviral activity and minimal cytotoxicity. These compounds were found to inhibit the immunoproteasome subunit LMP7, a target whose subcellular localization and expression are altered in Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2)-infected Huh7 cells. Yeast two-hybrid assays and co-immunoprecipitation further revealed that LMP7 interacts with the viral proteins Nsp13 and Nsp16. In addition, Nsp13 and Nsp16 disrupted the expression of LMP7 in response to pathogen attacks. Functional studies showed that LMP7 knockout in BEAS-2B-ACE2 cells resulted in enhanced replication of attenuated SARS-CoV-2, highlighting the role of this subunit in restricting viral replication. Taken together, these findings position LMP7 as a novel therapeutic target and highlight Ixazomib and its analogs as potential antiviral agents against current and future coronavirus threats.
## 1. Introduction
The COVID-19 pandemic has lasted five years and resulted in more than seven million deaths worldwide [1]. Prior to the SARS-CoV outbreak in China in 2003, human coronaviruses were always considered to be mild viruses that cause cold-like symptoms. Over the past two decades, highly virulent coronaviruses have emerged every few years. Following the SARS-CoV outbreak, MERS-CoV was reported in 2012 and has so far caused 935 deaths, with a crude case fatality rate of 36% [2]. Just seven years later, SARS-CoV-2 was transmitted between humans and quickly became a pandemic. Therefore, the future emergence of novel human coronaviruses causing disease is a reasonable speculation, especially given the fact that several SARS-CoV-and SARS-CoV-2-related viruses are already known to exist in bats [3,4]. Although COVID-19 has moved from a pandemic emergency to an endemic state, the threat to public health from future coronaviruses remains. Further efforts are needed to discover new drug targets and develop new antiviral drugs.
Low-molecular-mass protein-7 (LMP7) is a catalytic subunit specifically present in the immunoproteasome induced under inflammatory conditions. The immunoproteasome is a special type of proteasome with a cylindrical structure composed of a 19S regulatory unit and a 20S proteolytic core complex [5,6]. The proteolytic core complex consists of two α outer rings and two β inner rings with seven subunits per ring. The two β inner rings contain three catalytically active subunits, PSMB6 (β1c), PSMB7 (β2c), and PSMB5 (β5c), in the classical constitutive proteasome, and these subunits are replaced by LMP2 (β1i), LMP10 (β2i), and LMP7 (β5i) in the immunoproteasome induced by proinflammatory cytokines.
The main function of the immunoproteasome is to process major histocompatibility complex (MHC) class I peptides for antigen presentation [7]. Previous research has demonstrated the anti-cancer role of immunoproteasome inhibitors. For example, the immunoproteasome-targeting compound PR-924 increases apoptosis in human hematological malignant cells and exhibits significant anti-leukemic properties [8]. Another inhibitor of the immunoproteasome that targets subunit LMP2, UK-101, was observed to induce apoptosis in PC-3 prostate cancer cells and inhibit tumor growth in vivo [9]. The antiviral activity of immunoproteasome inhibitors, including those that act against coronaviruses, is not yet understood. Ixazomib was the first oral proteasome inhibitor to be approved by the Food and Drug Administration (FDA) [10]. It has been shown to prolong progression-free survival in patients with multiple myeloma [11], but little is known about its potential role in antiviral therapy.
In this study, we identify Ixazomib and its analogs as novel LMP7 inhibitors that can significantly suppress coronavirus replication. LMP7 is targeted by SARS-CoV-2 viral proteins, and SARS-CoV-2 infection alters the subcellular distribution and protein levels of LMP7. Knockout of LMP7 significantly enhances the replication of the attenuated SARS-CoV-2 virus sCPD9. Our work suggests LMP7 as a potential drug target for treatments preventing human coronavirus replication.
## 2. Materials and Methods
## 2.1. Cell Culture and Transfection
HEK293, Huh7, and Human Angiotensin-converting enzyme 2 (ACE2)-RFP-transduced A549 cells (A549_ACE2_RFP) [12] were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco, ThermoFisher, Item Number: 11965092, Location: Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. BEAS-2B-ACE2 [13] cells were maintained in DMEM/F-12 (Gibco, ThermoFisher, Item Number: 11320033, Location: Waltham, MA, USA) containing 10% FBS, 1% penicillin-streptomycin, and 1% N-2-Hydroxyethylpiperazine-N-2-Ethane Sulfonic Acid (HEPES). Transfection of HEK293 cells was performed using Lipofectamine3000 (ThermoFisher, Item Number: L3000015, Location: Waltham, MA, USA) or 25 KD polyethyleneimine (PEI) according to the manufacturer's protocols.
## 2.2. Plasmid and Constructions
GFP-and HA-fused LMP7 and RFP-fused HLA-A were generated by gateway cloning as described previously [14]. SARS-CoV-2 Nsp13 and Nsp16 were subsequently cloned into expression plasmids, following the method described recently in the literature [14].
For LMP7-knockout constructs, LMP7 sgRNAs were designed using the Synthego design tool: https://design.synthego.com/#/ (accessed on 2 October 2023)with the enzyme restriction sites of BsmbI. The oligo-dimers of LMP7 sgRNAs were cloned into a lentiCRISPR v2 vector, which was a gift from Feng Zhang (Addgene plasmid # 52961; http://n2t.net/addgene:52961, accessed on 22 July 2019; RRID: Addgene_52961) [15]. The packaging vector, psPAX2, was a gift from Didier Trono (Addgene plasmid # 12260; http: //n2t.net/addgene:12260, accessed on 22 July 2019; RRID: Addgene_12260). pCMV-VSV-G was a gift from Bob Weinberg (Addgene plasmid # 8454; http://n2t.net/addgene:8454, accessed on 22 July 2019; RRID: Addgene_8454) [16].
## 2.3. Generation of LMP7-Knockout Cells
The successfully cloned pLentiCrispr v2 LMP7 was co-transfected with psPAX2 and pCMV-VSV-G plasmids into HEK293T cells. Three days after transfection, the lentivirus supernatant was harvested and BEAS-2B-ACE2 cells cultured in a 12-well plate were infected with this supernatant. Three days after infection, cells were selected using a growth medium containing 2 µg/mL puromycin for a period of between five and seven days. Surviving cells were harvested for Western blot analysis to determine the expression levels of LMP7.
## 2.4. Virus Infection
For a recombinant HCoV-229E virus expressing Renilla luciferase (HCoV-229E RLuc) infection, Huh7 or HEK293 cells transfected with aminopeptidase N (APN) receptors were infected with HCoV-229E RLuc (MOI = 1) [17]. At 24 h post-infection (24 h.p.i.), cells were harvested and replication of HCoV-229E RLuc was quantified using a Promega Renilla luciferase assay kit.
SARS-CoV-2 virus inhibition assays were performed using the IncuCyte S3 Live-Cell Analysis System (Essen Bioscience), as described in detail in [18]. A549_ACE2_RFP cells were infected with SARS-CoV-2-GFP at MOI = 1. The GFP levels were quantified every 4 h until 72 h.p.i. Cell confluency was measured in parallel as a measure of cell viability.
For infection with the attenuated SARS-CoV-2 virus sCPD9 [19,20], BEAS-2B-ACE2 cells were challenged with sCPD9 (MOI = 0.0001). Two hours after infection, the cells were washed with DPBS and supplied with fresh growth medium. Five days after infection, the cells were harvested for RNA isolation. Viral RNAs were quantified by qPCR using primers and probes specific to the SARS-CoV-2 nucleocapsid protein.
For vaccinia virus vTF-7 infection, HEK293 cells were infected with a recombinant vaccinia virus, vTF-7, expressing the T7 RNA polymerase as described previously [21]. At 24 h.p.i., cells were harvested for immunofluorescence staining or Western blot analysis.
## 2.5. Cell Viability Assay
The viability of the cells was determined using a CellTiter-Glo ® 2.0 Cell Viability Assay (Promega, Item Number: #G9242, Location: Madison, WI, USA). The cell lines were plated in 96-well plates and incubated with inhibitor concentrations corresponding to each inhibition experiment.
## 2.6. Immunofluorescence Staining and Fluorescence Microscopy
Immunofluorescence staining and fluorescence microscopy were performed as previously described [14]. Briefly, mock and infected cells were seeded on coverslips in a 24-well plate, fixed with 4% paraformaldehyde (PFA) in PBS for 15 min at room temperature, and washed with PBS. Cell membranes were permeabilized by 0.1% Triton X-100 in PBS for 15 min. The cells were washed again with PBS and blocked in PBS containing 5% bovine serum albumin (BSA) and 0.2% Tween-20 for one hour at room temperature. The cells were then incubated with primary antibodies diluted in PBS containing 5% BSA and 0.2% Tween-20 at 4 • C overnight. After washing with PBS and incubation with secondary antibodies for one hour in the dark, the cells were washed and stained with DAPI diluted to 1:1000 in PBS for 10 min in the dark and mounted on glass slides. Images were captured using a Leica DM4000 B fluorescence microscope with a 40× objective.
## 2.7. Co-Immunoprecipitation and Western Blot Analysis
HEK293 cells were transfected with plasmids in a 6-well plate using Lipofectamine 3000 (Thermofisher). Twenty-four or forty-eight hours after transfection, cells were lysed and their protein was purified with GFP-Trap_A bead-based co-immunoprecipitation (Co-IP) (ChromoTek, Planegg-Martinsried, Germany) according to the manufacturer's protocol. The Western blot protocol has been described elsewhere [22]. The details of antibodies used in Western blot are described in Table S2.
## 2.8. Statistics
Statistical analysis was performed using one-way analysis of variance (one-way ANOVA) with Dunnett's test for multiple comparisons or Student's t-test for two-group analysis using GraphPad Prism 10.5.0 software with the significance level set at α = 0.05. Symbols in the figures represent p-values: ns indicates not significant and p ≥ 0.05, * indicates 0.01 < p < 0.05, ** indicates 0.001 < p < 0.01, and *** indicates p < 0.001.
## 3. Results
## 3.1. Coronavirus Replication Inhibitors Selected from Screening Are LMP7 Inhibitors
To identify new drug candidates that suppress coronavirus replication, the recombinant human coronavirus HCoV-229E RLuc, carrying the Renilla luciferase reporter [23], was used to rapidly screen a compound library. As a result, three compounds (#16-14, # 16-23, and #16-24) were selected due to their potent antiviral activity and low cytotoxicity in both Huh7 cells and HCoV-229E receptor APN-transfected HEK293 cells (Figure 1A). Detailed HCoV-229E RLuc inhibition assays in Huh7 cells showed that the IC50 values of compounds #16-14, #16-23, and #16-24 are 26.88 nM, 28.40 nM, and 5.729 nM, respectively (Figure 1B). Interestingly, all three selected compounds are LMP7 inhibitors with remarkable inhibition efficiencies for LMP7 in cell-free biochemical assays [24] (Table 1). Compound #16-14 is the well-known drug Ixazomib. Compounds #16-23 and #16-24 are analogs of Ixazomib. Their structural formulas are shown in Figure 1C. C for 1 h. After inoculation, the cells were washed with PBS and cultured in a growth medium containing each compound for 24 h before harvesting for luciferase activity measurements. For the cell viability assays, Huh7 and HEK293 cells in 96-well plates were treated with the indicated individual compound for 24 h before cell viability was measured using a CellTiter-Glo ® 2.0 Cell Viability Assay Kit (Promega). Means and standard deviations were calculated based on three biological replicates (n = 3) for both the inhibition screening and cell viability assays. Candidate compounds were selected according to their cell viability (more than 80%) and inhibition rates (below 0.1% for Huh7 cells and below 1% for HEK293 cells). (B) For the inhibition assays, Huh7 cells in 96-well plates were infected with HCoV-229E RLuc (MOI = 1) and treated with the indicated inhibitors for luciferase activity measurement as described above. Three biological replicates were performed for each compound concentration to calculate the means and standard deviations for the graphs (n = 3). Inhibition curves and corresponding logIC50 and IC50 values were generated using GraphPad Prism. Cell viability assays were performed as Table 1. IC50 values of immunoproteasome subunit LMP7 inhibitors #16-14 (Ixazomib), #16-23, and #16-24 in a cell-free biochemical assay [24] and in HCoV-229E-infected Huh7 cells. The selected compounds were then tested for their ability to inhibit the replication of highly virulent human coronaviruses. A SARS-CoV replicon carrying the Renilla luciferase reporter pBAC-REP-RLuc [25,26], which mimics SARS-CoV replication, was transfected into HEK293 cells with or without post-transfection drug treatment. As shown in Figure 2A, all selected compounds significantly inhibited the replication of this SARS-CoV replicon. The compounds were also tested for their ability to suppress SARS-CoV-2 replication. A549_ACE2_RFP cells were challenged with a recombinant SARS-CoV-2 virus carrying a GFP reporter (SARS-CoV-2-GFP) [18] in the presence of the compounds. The ratio of GFP and RFP intensity, representing viral spread and replication, was measured every 4 h for 72 h. The RFP intensity of uninfected cells treated with the drugs was measured to assess the cytotoxicity of the drugs. As shown in the lower graphs in Figure 2B, the RFP intensity remained constant with a high-concentration drug treatment and even increased with a low-concentration drug treatment. This indicates the drugs have low cytotoxicity in A549_ACE2_RFP cells such that no cells died after 72 h of drug treatment. The upper graphs in Figure 2B reveal that all compounds inhibited viral growth of SARS-CoV-2-GFP, although less efficiently than HCoV-229E-RLuc in Huh7 cells when compared to Figure 1B.
## Inhibitor ID IC50 (nM) of LMP7 in
## 3.3. SARS-CoV-2 Infection Reduces LMP7 Protein Levels, and SARS-CoV-2 Viral Proteins Target LMP7
We then investigated whether SARS-CoV-2 infection alters the subcellular localization or protein level of LMP7. Huh7 cells infected with SARS-CoV-2 at an MOI of 0.1 for 24 h were fixed and examined by immunofluorescence microscopy. As shown in Figure 2C and Figure S1, in uninfected Huh7 cells, LMP7 was mainly located in the cytosol, although it was occasionally found in the nucleus. However, in SARS-CoV-2-infected Huh7 cells, as indicated by dsRNA staining, LMP7 was mainly located in the nucleus and normally showed low expression. To confirm that SARS-CoV-2 infection reduced LMP7 expression, Western blot analysis was performed on Huh7.5 cells. Huh7.5 is more susceptible to SARS-CoV-2 than Huh7 and therefore always displays a stronger and more convincing phenotype upon infection. As shown in Figure 2D, SARS-CoV-2 infection caused a significant reduction in the expression of LMP7.
To further investigate how LMP7 is regulated by SARS-CoV-2, we used LMP7 as a bait to screen SARS-CoV-2 viral proteins in yeast two-hybrid (Y2H) assays [14]. As a result, LMP7 was found to interact with SARS-CoV-2 Nsp13 a.a.1-259 and Nsp16 (Figure S2). The interactions were then verified in mammalian cells by Co-IP. As Nsp13 a.a.1-259 showed relatively weak expression under the control of the CMV promoter, pDEST-GADT7 Nsp13 a.a.1-259, in which Nsp13 a.a.1-259 is under the control of the T7 promoter, was used. After induction with the vaccinia vTF-7 virus, optimal expression of Nsp13 a.a.1-259 was achieved. As shown in Figure 2E, SARS-CoV-2 Nsp13 a.a.1-259 clearly interacted with LMP7 upon vaccinia vTF-7 stimulation. In addition, the interaction between SARS-CoV-2 Nsp16 and LMP7 was also confirmed in HEK293 cells by Co-IP (Figure 2F).
## 3.4. SARS-CoV-2 Nsp13 and Nsp16 Inhibit LMP7 Expression, and Nsp16 Disrupts LMP7-Mediated Antigen Presentation
The vaccinia virus has been shown to infect HEK293 cells and induce an antiviral host response [27]. Therefore, we infected HEK293 cells with the vaccinia virus vTF-7 (VACV) and observed an increase in endogenous LMP7 expression compared to the uninfected control group (Figure 3A, compare lane 1 and lane 4). This stimulation was inhibited when Nsp13 a.a.1-259 or Nsp16 was transfected. LMP7 is a catalytic subunit of the immunoproteasome, whose primary function is to process peptides for antigen presentation via MHC class I. During the process of antigen presentation, MHC class I migrates from the endoplasmic reticulum (ER) to the cell membrane to expose the bound peptide to T-cell receptors [28]. To further investigate the effect of these two SARS-CoV-2 proteins on LMP7 function in relation to antigen presentation, we co-transfected HEK293 cells with GFP fused to Nsp13 or Nsp16 and RFP fused to HLA-A, HLA-B, and HLA-C, which are major members of the MHC class I heavy chain paralogs, and then infected the cells with the vaccinia virus vTF-7 for 24 h. Figures 3B andS3 demonstrate that GFP-Nsp16 altered the subcellular localization of HLA-A, HLA-B, and HLA-C compared to an empty-vector group. We speculate that Nsp16 interferes with the process of antigen presentation mediated by LMP7.
## 3.5. LMP7 Is an Antiviral Factor That Acts Against SARS-CoV-2
To further understand the role of LMP7 in human coronavirus replication, LMP7 was knocked out in human ACE2-transgenic bronchial epithelial cells (BEAS-2B-ACE2), as this cell line shows abundant endogenous LMP7 expression. In addition, this cell line is susceptible to infection with SARS-CoV-2 and mild coronaviruses. The endogenous protein level of LMP7 was greatly reduced in the LMP7-knockout (BEAS-2B-ACE2 LMP7-KO) cells (Figure 4A) compared to the empty-vector control (BEAS-2B-ACE2 control). After infection with the attenuated SARS-CoV-2 virus sCPD9, cells that were deficient in LMP7 produced more virus (Figure 4B). LMP7 is an antiviral factor that acts against SARS-CoV-2. (A) Western blot analysis of the endogenous LMP7 levels in BEAS-2B-ACE2-LMP7-knockout (LMP7-KO) and control cells. (B) BEAS-2B-ACE2-LMP7-KO and control cell lines were infected with SARS-CoV-2 SCPD9 (MOI = 0.0001) in a 24-well plate. At 2 h post-infection, the cells were washed twice with DPBS and replenished with fresh growth medium. Then, 5 days post-infection, the cells were harvested for RNA isolation, followed by qPCR using virus-specific primers and probes. The mRNA levels were normalized using β-actin. Means and standard deviations were calculated based on three biological replicates (n = 3). Statistical analysis was performed using Student's t-test. *** indicates p < 0.001.
## 4. Discussion
LMP7 is a catalytic subunit of an immunoproteasome that degrades viral protein into peptides during infection for downstream antigen presentation at the cell surface. In this way, infected cells can be recognized by CD8+ T cells. In our study, we demonstrated that LMP7 negatively regulates SARS-CoV-2 replication, as LMP7 knockout promoted viral growth (Figure 4B). This indicates that LMP7 is vital for proteasome activity and subsequent antigen presentation. To antagonize LMP7 or degradation by immunoproteasomes, SARS-CoV-2 Nsp13 and Nsp16 bind to (Figure 2E,F) and decrease the expression of (Figure 3A) LMP7. As a consequence, expression of Nsp16 disturbs antigen presentation by altering the subcellular localization of MHC class I molecules (Figure 3B and Figure S3A,B). But it is not clear why expression of Nsp13 does not lead to an altered subcellular distribution upon viral infection.
LMP7 has been shown to be an antiviral factor in several research studies. During rhinovirus infection, airway epithelial immunoproteasome subunit LMP7 was observed to exert anti-inflammatory and antiviral effects, thereby facilitating the resolution of inflammation and a reduction in the viral load [29]. In addition, LMP7 can interact with and positively regulate the expression of A20, which can fight respiratory viral infections such as the influenza A virus [30]. These findings also provide potential explanations for why LMP7 negatively regulates coronavirus replication.
Our study demonstrated that LMP7 deficiency promotes SARS-CoV-2 replication (Figure 4B). However, the compound Ixazomib and its analogs act as LMP7 inhibitors (Table 1) and clearly suppress coronavirus replication (Figures 1B and2A,B). The reason for this pseudo-contradiction is that the pharmaceutical mechanism of these compounds is very likely independent from LMP7, although they can also inhibit LMP7 activity. It has been shown that Ixazomib is not only an LMP7-specific inhibitor; it can also inhibit other subunits of the immunoproteasome such as LMP2 and some subunits of the constitutive classic proteasome [24]. The proteasome inhibitors MG132, epoxomicin, and Velcade (bortezomib) have shown clear coronavirus-inhibitory activity in cell culture, mainly by affecting the early steps of coronavirus infection [31]. It has been demonstrated that functional proteasomeubiquitin systems facilitate the release of coronavirus from endosomes into the cytosol after viral entry, as pretreatment with MG132 can retain murine coronavirus (murine hepatitis virus, MHV) in endosomes after virus internalization into cells [32,33]. Besides immunoproteasomes, Ixazomib also exhibits inhibitory activity against constitutive classic proteasomes [24], similar to MG132. It can be speculated that Ixazomib and its analogs act similarly to MG132 to effectively restrict coronavirus release from endosomes into the cytosol after viral entry through endocytosis. This mechanism is independent of LMP7 and immunoproteasome inhibition and should be the main pharmaceutical mechanism for Ixazomib and its analogs to inhibit coronavirus replication, as this is a very early step after viral entry, before the virus unpacks to release structural proteins and starts to express nonstructural proteins in the cytosol for degradation by immunoproteasomes. In addition to endocytosis, the coronavirus can also enter host cells via direct membrane fusion [34], preventing it from being retained in endosomes by proteasome inhibitors. But if this is the major case, inhibition of LMP7 by our compounds should not lead to a reduction in viral replication.
Antigen presentation mediated by MHC class I molecules is downstream of immunoproteasomes and is a key process in the immune response to SARS-CoV-2 infection, influencing both viral replication and disease progression. The SARS-CoV-2 virus shows a high capacity to suppress the induction of MHC-I in infected cells in vivo [35]. The SARS-CoV-2 accessory protein ORF7a has been observed to interact specifically with the MHC-I heavy chain, thereby competing with the binding of the small protein β2-microglobulin. This interaction has been shown to reduce antigen presentation by the human MHC-I allele HLA-A*02:01 [36]. It is not surprising that Nsp16, which decreases LMP7, also disturbed the distribution of HLA proteins upon viral infection in our study. But whether Nsp16 affects the subsequent process, such as activation of antigen-specific CD8+ T cells, requires further investigation. Furthermore, we observed that Nsp13 did not alter the subcellular localization of HLA-A, HLA-B, and HLA-C. Nevertheless, we cannot rule out the possibility that Nsp13 affects other stages of the MHC-I antigen presentation pathway, such as peptide loading or transport to the cell surface [37]. This needs further exploration.
## Limitations of This Study
One of the limitations of this study is the unresolved paradox between the antiviral effects of pharmacological inhibition of LMP7 and the proviral phenotype observed upon LMP7 knockout. Our data suggest that Ixazomib and its analogs likely inhibit coronavirus replication through a mechanism independent of LMP7. However, direct experimental evidence supporting this proposed off-target, immunoproteasome-independent mechanism is currently limited. Future studies are needed to precisely define the molecular basis of this early-stage antiviral activity and to further distinguish between LMP7-specific and LMP7independent pathways. Additionally, while we used multiple human cell lines to evaluate the consistency of the observed phenotypes, we acknowledge that the use of different cell types with varying levels of susceptibility introduces some degree of variability. Future studies using more physiologically relevant models, such as primary cells or organoids, will be valuable to further validate these findings. Another limitation of this study is the lack of in vivo validation of the antiviral efficacy of Ixazomib and its analogs. While our in vitro results demonstrate potent inhibition of multiple human coronaviruses, including SARS-CoV-2, further studies in appropriate animal models, such as SARS-CoV-2-infected hamsters or transgenic mice, are necessary to confirm the therapeutic potential and safety profile of these compounds in vivo.
## 5. Conclusions
In summary, our study uncovered the role of LMP7 during SARS-CoV-2 infection. LMP7 deficiency promotes viral replication, indicating that LMP7 is a restriction factor for SARS-CoV-2. To counteract this restriction factor, SARS-CoV-2 Nsp13 and Nsp16 interact with LMP7, and expression of Nsp13 or Nsp16 reduces LMP7 at the protein level upon infection. Additionally, Nsp16 alters the subcellular localization of MHC class I molecules responsible for antigen presentation during infection. In parallel, three compounds, #16-14 (Ixazomib), #16-23, and #16-24, are LMP7 inhibitors and can effectively suppress coronavirus replication. But their pharmaceutical mechanism should be LMP7-independent.
## References
1. (2025) "Number of COVID-19 Deaths Reported to WHO (Cumulative Total). Available online"
2. (2022) "Middle East Respiratory Syndrome: Global Summary and Assessment of Risk"
3. Li, Shi, Yu et al. (2005) "Bats are natural reservoirs of SARS-like coronaviruses" *Science*
4. Temmam, Vongphayloth, Baquero et al. (2022) "Bat coronaviruses related to SARS-CoV-2 and infectious for human cells" *Nature*
5. Basler, Groettrup (2021) "On the Role of the Immunoproteasome in Protein Homeostasis" *Cells*
6. Murata, Takahama, Kasahara et al. (2018) "The imm unoproteasome and thymoproteasome: Functions, evolution and human disease" *Nat. Immunol*
7. Groettrup, Kirk, Basler (2010) "Proteasomes in immune cells: More than peptide producers?" *Nat. Rev. Immunol*
8. Niewerth, Van Meerloo, Jansen et al. (2014) "Anti-leukemic activity and mechanisms underlying resistance to the novel immunoproteasome inhibitor PR-924" *Biochem. Pharmacol*
9. Wehenkel, Ban, Ho et al. (2012) "A selective inhibitor of the immunoproteasome subunit LMP2 induces apoptosis in PC-3 cells and suppresses tumour growth in nude mice" *Br. J. Cancer*
10. Raedler, Ninlaro (2016) "First Oral Proteasome Inhibitor Approved for the Treatment of Patients with Relapsed or Refractory Multiple Myeloma" *Am. Health Drug Benefits*
11. Moreau, Masszi, Grzasko et al. (2016) "Oral Ixazomib, Lenalidomide, and Dexamethasone for Multiple Myeloma" *N. Engl. J. Med*
12. Bergant, Yamada, Grass et al. (2022) "Attenuation of SARS-CoV-2 replication and associated inflammation by concomitant targeting of viral and host cap 2 ′ -O-ribose methyltransferases" *EMBO J*
13. Mathieu, Touret, Jacquemin et al. (2021) "A Bioluminescent 3CL(Pro) Activity Assay to Monitor SARS-CoV-2 Replication and Identify Inhibitors"
14. Ma-Lauer, Li, Niemeyer et al. (2024) "Oxysterole-binding protein targeted by SARS-CoV-2 viral proteins regulates coronavirus replication" *Front. Cell Infect. Microbiol*
15. Sanjana, Shalem, Zhang (2014) "Improved vectors and genome-wide libraries for CRISPR screening" *Nat. Methods*
16. Stewart, Dykxhoorn, Palliser et al. (2003) "Lentivirus-delivered stable gene silencing by RNAi in primary cells" *Rna*
17. Van Den Worm, Eriksson, Zevenhoven et al. (2012) "Reverse genetics of SARS-related coronavirus using vaccinia virus-based recombination" *PLoS ONE*
18. Stukalov, Girault, Grass et al. (2021) "Multilevel proteomics reveals host perturbations by SARS-CoV-2 and SARS-CoV" *Nature*
19. Kunec, Osterrieder, Trimpert (2022) "Synthetically recoded virus sCPD9-A tool to accelerate SARS-CoV-2 research under biosafety level 2 conditions" *Comput. Struct. Biotechnol. J*
20. Nouailles, Adler, Pennitz et al. (2023) "Live-attenuated vaccine sCPD9 elicits superior mucosal and systemic immunity to SARS-CoV-2 variants in hamsters" *Nat. Microbiol*
21. Von Brunn, Teepe, Simpson et al. (2007) "Analysis of intraviral protein-protein interactions of the SARS coronavirus ORFeome" *PLoS ONE*
22. Carbajo-Lozoya, Ma-Lauer, Malešević et al. (2014) "Human coronavirus NL63 replication is cyclophilin A-dependent and inhibited by non-immunosuppressive cyclosporine A-derivatives including Alisporivir" *Virus Res*
23. Ma-Lauer, Zheng, Malesevic et al. (2020) "Influences of cyclosporin A and nonimmunosuppressive derivatives on cellular cyclophilins and viral nucleocapsid protein during human coronavirus 229E replication" *Antivir. Res*
24. Sanderson, Friese-Hamim, Walter-Bausch et al. (2021) "Delivering Efficacy in Multiple Myeloma Models"
25. Kusov, Tan, Alvarez et al. (2015) "A G-quadruplex-binding macrodomain within the "SARS-unique domain" is essential for the activity of the SARS-coronavirus replication-transcription complex" *Virology*
26. Lei, Ma-Lauer, Han et al. (2021) "The SARS-unique domain (SUD) of SARS-CoV and SARS-CoV-2 interacts with human Paip1 to enhance viral RNA translation" *EMBO J*
27. Bartel, Doellinger, Darsow et al. (2011) "Proteome analysis of vaccinia virus IHD-W-infected HEK 293 cells with 2-dimensional gel electrophoresis and MALDI-PSD-TOF MS of on solid phase support N-terminally sulfonated peptides" *Virol. J*
28. Leone, Shin, Perosa et al. (2013) "MHC class I antigen processing and presenting machinery: Organization, function, and defects in tumor cells" *J. Natl. Cancer Inst*
29. Dimasuay, Schaunaman, Berg et al. (2022) "Airway epithelial immunoproteasome subunit LMP7 protects against rhinovirus infection" *Sci. Rep*
30. Maelfait, Roose, Vereecke et al. (2016) "A20 Deficiency in Lung Epithelial Cells Protects against Influenza A Virus Infection" *PLoS Pathog*
31. Raaben, Posthuma, Verheije et al. (2010) "The ubiquitin-proteasome system plays an important role during various stages of the coronavirus infection cycle" *J. Virol*
32. Longhitano, Tibullo, Giallongo et al. (2020) "Proteasome Inhibitors as a Possible Therapy for SARS-CoV-2" *Int. J. Mol. Sci*
33. Yu, Lai (2005) "The ubiquitin-proteasome system facilitates the transfer of murine coronavirus from endosome to cytoplasm during virus entry" *J. Virol*
34. Jackson, Farzan, Chen et al. (2022) "Mechanisms of SARS-CoV-2 entry into cells" *Nat. Rev. Mol. Cell Biol*
35. Moriyama, Lucas, Monteiro et al. "Enhanced inhibition of MHC-I expression by SARS-CoV-2 Omicron subvariants"
36. Arshad, Laurent-Rolle, Ahmed et al. "SARS-CoV-2 accessory proteins ORF7a and ORF3a use distinct mechanisms to down-regulate MHC-I surface expression"
37. Blum, Wearsch, Cresswell (2013) "Pathways of antigen processing" *Annu. Rev. Immunol*
38. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548440&blobtype=pdf | # The E3 ubiquitin ligase STUB1 inhibits Senecavirus A replication by mediating VP1 ubiquitination and proteasomal degradation
Penghui Zeng, Jingyu Mao, Jinshuo Guo, Xiaoyu Yang, Yongyan Shi, Xiaoyu Wang, Jiangwei Song, Jianwei Zhou, Lei Hou, Jue Liu
## Abstract
Senecavirus A (SVA), an emerging vesicular pathogen, poses a significant threat to the global pig industry. STIP1 homology and U-box-containing protein 1 (STUB1), a chaperone-dependent E3 ubiquitin ligase, plays a pivotal role in protein quality control by mediating target protein degradation. However, its precise role of STUB1 in regulating SVA replication remains undefined. In this study, we combined liquid chromatography-mass spectrometry, confocal imaging, and Western blotting to demonstrate that STUB1 interacts with the SVA VP1 protein and negatively regulates SVA replication. Mechanistically, STUB1 promotes the ubiquitination-dependent degradation of VP1 by specifically targeting lysine residues at positions 177 and 260 (K177 and K260). This degradation process is significantly enhanced by heat shock protein 70 (HSP70) and heat shock cognate protein 70 (HSC70), which strengthen the STUB1-VP1 interaction. Notably, the SVA 3C protease (3Cpro) counteracts this antiviral defense by enzymatically reducing STUB1 expression. In vivo studies using a mouse model showed that a VP1 mutant virus lacking STUB1-targeted ubiquitination sites replicates more efficiently than the wild-type strain, resulting in significantly higher viral loads across multiple tissues and more severe pulmonary pathology. Together, these findings reveal that STUB1 inhibits SVA replication through ubiquitination-dependent degradation of VP1, a process that is antagonized by viral 3C protease via suppression of STUB1 expression.
IMPORTANCEViruses have evolved diverse strategies to enhance their replication efficiency. Senecavirus A (SVA), an emerging porcine pathogen associated with vesicular disease outbreaks, has become increasingly prevalent in swine populations worldwide. As a chaperone-dependent E3 ubiquitin ligase, STUB1 plays a crucial role in maintain ing cellular protein homeostasis. In this study, we elucidated the functional interplay between STUB1 and SVA replication. Our results demonstrate that STUB1 directly interacts with the viral protein VP1 and mediates its ubiquitination-dependent degra dation through specific targeting of lysine residues at positions 177 and 260 (K177 and K260), thereby significantly inhibiting viral replication. However, SVA has evolved a countermeasure, whereby its 3C protease (3Cpro) downregulates STUB1 expression, effectively blocking VP1 degradation and subverting this host antiviral defense to promote viral propagation. These findings not only reveal novel host-virus interaction mechanisms but also provide valuable molecular targets for developing innovative strategies to control SVA infection.
contaminant in 2002, its complete genome shows a close relationship to that of Cardiovirus (1). The viral genome, approximately 7.2 kb in length, contains a single open reading frame (ORF) encoding a polyprotein. This polyprotein is proteolytically processed into structural proteins (VP1, VP2, VP3, and VP4) and nonstructural proteins (Lpro, 2A, 2B, 2C, 3A, 3B, 3C, and 3D) (1). SVA infections can cause vesicular diseases, leading to substantial economic losses in the global swine industry (2)(3)(4). Since the first detection of Chinese strains of SVA in 2015, increasing numbers of infections have been reported across various provinces in China, indicating widespread transmission (5,6). Additionally, SVA has attracted attention as a potential oncolytic virus for cancer therapy (7,8). However, no specific or effective treatment for SVA is currently available in clinical practice. A deeper understanding of the molecular mechanisms underlying viral replication is essential for developing effective strategies to prevent and control SVA infections.
The structural proteins of SVA are derived from the initially synthesized P1 polymeric protein. Once mature P1 is produced, it is cleaved by SVA 3C protease into VP1, VP3, and VP0; VP0 is subsequently cleaved into VP2 and VP4 (1,(9)(10)(11). Similar to other members of the Picornaviridae family, the antigenic regions of SVA are primarily composed of VP1 and VP3, with VP1 exhibiting the strongest antigenicity (12). Furthermore, the SVA VP1 protein collaborates with VP3 and 3Cpro to activate the AKT-AMPK-MAPK-mTOR signaling pathway (13), facilitating viral replication. It also plays a critical role in viral attachment and entry into host cells, making it a potential target for SVA intervention (14). The 3Cpro is a multifunctional protein involved in viral pathogenesis: it promotes viral protein maturation and suppresses host antiviral responses by cleaving histone deacetylase 4 (HDAC4) to inhibit type I interferon signaling and by degrading interferon regulatory factors 3 and 7 (IRF3 and IRF7) to disrupt the innate immune response (15,16).
The ubiquitin-proteasome system is a highly efficient protein degradation pathway in eukaryotic cells, responsible for the selective degradation of proteins and crucial for maintaining cellular homeostasis (17). It contributes to antiviral defense by degrading viral proteins; for example, the host E3 ligase Hrd1 ubiquitinates and degrades the H protein of the canine distemper virus to inhibit viral replication (18). RNF5 restricts SARS-CoV-2 replication by targeting its envelope protein for degradation (19); and TRIM7 inhibits enterovirus replication by mediating the degradation of the enteroviral 2C protein (20).
STIP1 homology and U-box containing protein 1 (STUB1), formerly known as the C-terminus of HSC70-interacting protein (CHIP), is a chaperone-dependent E3 ubiquitin ligase (21). It is highly expressed in metabolically active tissues and regions prone to protein misfolding, such as the skeletal muscle, brain, and heart (22). The N-terminal domain of STUB1 interacts with heat shock proteins to enhance the resilience to stress (23), while its C-terminal U-box domain mediates the ubiquitination and subsequent degradation of specific substrate proteins (24). STUB1 has been implicated in various immune responses, including the regulation of immune cell differentiation, maturation, and inflammation (25)(26)(27)(28). It is also involved in viral replication: for instance, STUB1 modulates antiviral RNA interference by inducing the ubiquitination and degradation of Dicer and AGO2 in mammals (29); mediates the degradation of the porcine deltacorona virus nucleocapsid protein (30); and is targeted by the SUMO-interacting motif of EBNA1 to maintain the latency of Epstein-Barr virus (31). However, the role of STUB1-mediated ubiquitination in SVA infection remains unclear.
In this study, we demonstrate that STUB1 exerts antiviral activity by mediating ubiquitination-dependent degradation of VP1 through specific targeting of lysine residues at positions 177 and 260 (K177/K260). However, SVA infection counteracts this host defense mechanism through 3Cpro-mediated downregulation of STUB1 expression, thereby facilitating viral infection. Collectively, our results reveal a novel host-virus interaction mechanism that provides important insights for the prevention and control of SVA infection.
## RESULTS
## STUB1 interacts with SVA VP1
During viral infection, the host leverages the ubiquitin-proteasome system to degrade viral proteins, thereby effectively inhibiting viral replication (18)(19)(20). A previous report demonstrated that the SVA VP1 protein is ubiquitinated (32), which consequently raises the question of whether the host can inhibit SVA replication by ubiquitinating VP1 protein. To identify host proteins that interact with VP1, HEK-293T cells were transfected with either a GFP-tagged VP1 plasmid or a control plasmid (GFP-C1). After lysis, the cells were subjected to immunoprecipitation using anti-GFP agarose, followed by liquid chromatography-tandem mass spectrometry (LC-MS) analysis (Table S1). Meanwhile, immunoprecipitates from cells transfected with the control plasmid served as nega tive controls to account for nonspecific interactions. Additionally, silver staining was performed on the immunoprecipitated proteins to visualize their interactions with GFP-VP1 (Fig. 1A). This screen identified two candidate E3 ubiquitin ligases (Fig. 1B), and STUB1 was selected for further analysis primarily due to its enrichment. Comparative analysis of STUB1 protein sequences revealed significant conservation among porcine, mice, and human orthologs (Fig. 1C). Accordingly, we constructed a plasmid express ing porcine STUB1 for subsequent experiments. To confirm the interaction between the SVA VP1 and STUB1, HEK-293T cells were coexpressed with GFP-C1 or GFP-VP1 and HA-STUB1, followed by forward and reverse co-immunoprecipitation (co-IP) assays using anti-GFP agarose or anti-HA agarose, respectively. Specific bands corresponding to GFP-VP1 and HA-STUB1 were detected (Fig. 1D andE), thus indicating an interaction between these two proteins. Furthermore, confocal imaging revealed colocalization of viral VP1 and HA-STUB1 in SVA-infected BHK-21 and ST cells (Fig. 1F). To further validate the interaction of STUB1 with SVA VP1 during SVA infection, ST cells infected with SVA were harvested at 12 hours post-infection and subjected to immunoprecipitation with an anti-VP1 antibody. The result showed that endogenous STUB1 interacted with VP1 during SVA infection (Fig. 1G). Moreover, confocal microscopy demonstrated colocaliza tion of endogenous STUB1 with VP1 in BHK-21 and ST cells during SVA infection (Fig. 1H).
Collectively, these results confirm an interaction between STUB1 and SVA VP1.
## STUB1 negatively regulates SVA replication
To investigate the effect of STUB1 on SVA replication, BHK-21 and ST cells transfected with HA-STUB1 plasmids were infected with SVA for 6 or 12 hours. As shown in Fig. 2A through D, overexpression of STUB1 inhibited SVA replication, as evidenced by reduced viral VP1 expression and lower viral titers at both time points compared to the control group. Additionally, enhanced green fluorescent protein (eGFP)-tagged recombinant SVA (rSVA-eGFP) was used to assess the impact of STUB1 on viral infectivity in BHK-21 and ST cells. The number of rSVA-eGFP-positive cells gradually decreased with increas ing HA-STUB1 expression (Fig. 2E). Next, BHK-21 and ST cells were transfected with three distinct siRNAs targeting STUB1 (siSTUB1-1, siSTUB1-2, and siSTUB1-3). Among these, siSTUB1-1 was selected for subsequent experiments due to its optimal silencing efficiency in both cell types (Fig. 2F andI). Western blotting and viral titer assays showed that STUB1 silencing increased VP1 expression and viral titers in BHK-21 and ST cells compared to the control group (Fig. 2G, H, J and K). Taken together, these results demonstrate that STUB1 functions as a negative regulator of SVA replication.
## K177 and K260 in VP1 are necessary for STUB1-mediated, ubiquitinationdependent degradation of VP1
We hypothesized that STUB1 restricts SVA replication by promoting VP1 degradation.
To test this, HEK-293T cells were co-transfected with GFP-VP1 or GFP-C1 and varying concentrations of HA-STUB1, and protein extracts were analyzed by Western blotting 36 hours later. The results showed that elevated HA-STUB1 expression caused a dosedependent decrease in GFP-VP1 levels, with no effect on GFP-C1 (Fig. 3A andB). To further examine the effect of STUB1 on VP1 stability, cycloheximide (CHX) chase assays were performed in HEK-293T cells expressing GFP-VP1, either transfected with HA-STUB1 or STUB1-targeting siRNA (siSTUB1). Western blot analysis revealed that overexpression of HA-STUB1 significantly accelerated GFP-VP1 degradation, whereas STUB1 knockdown markedly stabilized the viral protein (Fig. 3C andD).
STUB1-mediated protein degradation primarily occurs through two major path ways in eukaryotic cells: the ubiquitin-proteasome system and autophagy-lysosome machinery (33). To clarify the mechanism underlying STUB1-mediated VP1 degradation, we evaluated the contribution of these pathways using pharmacological inhibitors. MG132 (a specific proteasome inhibitor) substantially rescued VP1 from degradation, whereas chloroquine (CQ, an autophagy inhibitor) had no protective effect (Fig. 3E), indicating that STUB1 promotes VP1 degradation primarily through the ubiquitin-protea some system.
Subsequently, ubiquitination of SVA VP1 was confirmed via co-IP and Western blotting, verifying direct ubiquitin-protein conjugation (Fig. 3F). To determine whether STUB1 is involved in VP1 ubiquitination, HEK-293T cells were co-expressed with GFP-VP1, HA-tagged ubiquitin (Ub), pCMV-Myc, or varying concentrations of Myc-STUB1. Co-IP results showed that VP1 ubiquitination increased in a STUB1 concentration-dependent manner (Fig. 3G).
To further elucidate the molecular basis of STUB1-mediated VP1 ubiquitination, HEK-293T cells were co-transfected with GFP-VP1, Myc-STUB1, and either wild-type HA-ubiquitin (HA-Ub) or HA-Ub mutants containing only K48 (HA-Ub-K48) or K63 (HA-Ub-K63) lysine residues. Immunoprecipitation with anti-GFP agarose revealed that VP1 ubiquitination specifically required K48-linked ubiquitin chains, as evidenced by robust ubiquitination signals in the presence of HA-Ub-K48, but not HA-Ub-K63 (Fig. 3H). These results demonstrate that STUB1 specifically promotes K48-linked polyubiquitina tion of VP1, a modification that typically targets substrates for proteasomal degradation.
To systematically identify potential ubiquitination sites on VP1, we combined bioinformatics prediction with experimental validation. Since ubiquitination predomi nantly targets lysine residues, we first utilized an online ubiquitination site prediction tool (http://www.ubpred.org) to identify the top five candidate lysine residues in VP1 based on their scores (Fig. 3I). To determine which lysine residues are targeted by STUB1-mediated ubiquitination, we generated a series of GFP-tagged VP1 mutants: a 12R mutant (all lysines mutated to arginines) and five single-lysine mutants retaining only K34, K174, K177, K233, or K260. When co-expressed with HA-STUB1 in HEK-293T cells, only the K177-and K260-retaining mutants exhibited significant degradation (Fig. 3J andK).
For definitive verification, ubiquitination assays were performed by transfecting cells with GFP-VP1 (wild-type or mutants), HA-Ub, and Myc-STUB1, followed by MG132 treatment to accumulate ubiquitinated species. As shown in Fig. 3L, STUB1 specifi cally mediated ubiquitination of VP1 at K177 and K260, as indicated by robust ubiqui tin conjugation to the K177-and K260-retaining mutants. Furthermore, simultaneous mutation of K177 and K260 in VP1 abolished STUB1-mediated ubiquitination, confirming that these sites are specifically targeted by STUB1 (Fig. 3M).
To further demonstrate that STUB1 suppresses SVA replication primarily by degrad ing VP1, BHK-21 cells were co-transfected with GFP-VP1 (or GFP-C1) and HA-STUB1 and then infected with SVA for 12 hours. Western blotting and TCID₅₀ assays showed that exogenous VP1 expression reversed the STUB1-mediated suppression of SVA yield (Fig. 3N andO). These results indicate that K177 and K260 in VP1 are essential for the STUB1-mediated, ubiquitination-dependent degradation of VP1.
## Heat shock protein 70 (HSP70) or heat shock cognate protein 70 (HSC70) is critical for STUB1-mediated VP1 degradation
STUB1-mediated ubiquitination and degradation of substrates often depends on interactions with molecular chaperones (34), typically HSP70 and HSC70 (35). The STUB1-K30A mutant, which harbors a mutation in the tetratricopeptide repeat domain (a chaperone-binding domain), fails to bind chaperones and exhibits reduced substrate interaction (36). Conversely, the STUB1-H260Q mutation in the U-box domain impairs E3 ubiquitin ligase activity, diminishing its ability to ubiquitinate substrates for degradation (36). These two sites are highly conserved in porcine, human, and murine STUB1 (Fig. 4A). To investigate whether these residues are involved in VP1 degradation, HEK-293T cells were co-transfected with GFP-VP1 and either wild-type STUB1 or STUB1 mutants. Compared to wild-type STUB1, both mutants reduced the ability to degrade viral VP1 (Fig. 4B), suggesting that chaperone binding and ubiquitin ligase activity of STUB1 are critical for VP1 degradation.
To explore whether the K30 or H260 mutation affects the STUB1-VP1 interaction, co-IP assays were performed. STUB1-K30A failed to interact with VP1, HSP70, or HSC70, whereas STUB1-H260Q retained interactions with all three (Fig. 4C). This indicates that the STUB1-K30A mutation abrogates chaperone binding, which may underlie its inability to interact with VP1, highlighting the essential role of STUB1's chaperone-binding activity in mediating VP1 interaction.
Reciprocal co-IP assays using anti-GFP agarose further revealed that while the STUB1-K30A mutation disrupted STUB1-VP1 binding, it did not affect interactions between VP1 and HSP70 or HSC70 (Fig. 4D). This supports a model in which heat shock proteins act as molecular adapters, bridging STUB1 and SVA VP1.
To examine the role of HSP70 and HSC70 in the STUB1-VP1 interaction, ST cells cotransfected with GFP-VP1 and Flag-HSP70 or Flag-HSC70 were lysed 36 hours and subjected to immunoprecipitation with an anti-GFP antibody. VP1 exhibited a stronger interaction with STUB1 when HSP70 or HSC70 was overexpressed (Fig. 4E), whereas this interaction was weakened when HSP70 or HSC70 was knocked down using siRNA (Fig. 4F).
To further define the role of these chaperones in STUB1-mediated VP1 degradation, HEK-293T cells were co-transfected with Flag-HSP70, Flag-HSC70, HA-STUB1, HA-STUB1-K30A, and GFP-VP1. Both HSP70 and HSC70 significantly enhanced STUB1-mediated VP1 degradation, whereas they had no effect when combined with STUB1-K30A (Fig. 4G). Importantly, disruption of either chaperone-binding (K30) or catalytic (H260) sites abolished STUB1's antiviral activity, as evidenced by increased VP1 expression and viral titers (Fig. 4H andI). Collectively, these results indicate that HSP70 and HSC70 play critical roles in STUB1-mediated VP1 degradation.
## SVA infection downregulates STUB1 expression.
To investigate the regulatory role of HSP70 and HSC70 in SVA replication, cells overex pressing these chaperones were infected with SVA for 6 or 12 hours. Despite their role in enhancing STUB1-mediated VP1 degradation (Fig. 4), both HSP70 and HSC70 promoted viral replication (Fig. 5A through D), consistent with our previous report (37). This paradox prompted us to investigate the underlying mechanism.
STUB1 expression was downregulated in SVA-infected BHK-21 and ST cells, whereas HSP70 and HSC70 levels remained unchanged (Fig. 5E andF). This reduction in STUB1 was dependent on active SVA replication (Fig. 5G andH), with no change in STUB1 mRNA levels (Fig. 5I andJ), suggesting post-translational regulation of STUB1.
To determine whether STUB1 is required for HSP70-/HSC70-mediated VP1 degrada tion, STUB1 was silenced in cells expressing HSP70 or HSC70. STUB1 silencing attenu ated VP1 degradation induced by HSP70 or HSC70 (Fig. 5K andL), indicating that STUB1 is necessary for chaperone-mediated VP1 degradation. Taken together, these results suggested that SVA infection downregulates STUB1 expression to disrupt HSP70-/ HSC70-mediated VP1 degradation.
## SVA 3C protein antagonizes STUB1-mediated VP1 degradation by reducing STUB1 expression.
To identify the SVA protein responsible for STUB1 downregulation, BHK-21 cells were transfected with various viral proteins (13). Only SVA 3C protease (3Cpro) induced STUB1 degradation, mimicking the pattern observed during SVA infection (Fig. 6A andB). Moreover, increased GFP-3C expression caused a dose-dependent decrease in both endogenous STUB1 and HA-STUB1 levels (Fig. 6C andD).
We also examined the effect of 3Cpro from other picornaviruses, including coxsack ievirus B3 (CVB3), encephalomyocarditis virus (EMCV), foot and mouth disease virus (FMDV), enterovirus 71 (EV71), and human rhinovirus (HRV). SVV, FMDV, and HRV 3Cpro decreased STUB1 expression, whereas CVB3, EMCV, and EV71 3Cpro had no effect (Fig. 6E).
SVA 3Cpro typically degrades host proteins directly via its protease activity (38). To investigate this, we co-expressed HA-STUB1 with GFP-3C or GFP-3C mutants harboring mutations in critical protease residues (GFP-3C-H48A, GFP-3C-C160A, and GFP-3C-DM [H48A-C160A]) (39). Mutational analysis showed that catalytically inactive 3Cpro (H48A/ C160A) failed to promote STUB1 degradation (Fig. 6F), demonstrating that protease activity is essential.
To assess the functional consequences, we examined 3Cpro's impact on STUB1mediated antiviral activity. Co-expression experiments in HEK-293T cells showed that wild-type 3Cpro, but not the proteasedeficient mutant (3Cpro-DM), effectively counteracted STUB1-mediated VP1 degradation via specific reducing STUB1 protein levels (Fig. 6G).
## Replication capacity of SVA mutants is enhanced in vitro and in vivo.
Given the high sequence homology between murine and porcine STUB1 proteins, we investigated the effect of murine STUB1 on SVA replication. Murine STUB1 also inhibited SVA replication (Fig. 7A andB) and degraded VP1 containing only 177 or 260 (Fig. 7C). To evaluate the role of VP1 K177 and K260 in SVA replication, we engineered viruses harboring K177R, K260R, or double mutant (Fig. 7D). The number of plaques formed by rSVA (VP1 K260R) exceeded that of wild-type rSVA, indicating higher viral titers (Fig. 7E). In contrast, no plaques were detected for rSVA (VP1 K177R), suggesting that K177 mutation is lethal to SVA. Additionally, rSVA (VP1 K260R) exhibited significantly higher titers than the wild-type rSVA (Fig. 7F).
To further characterize the functional role of we used a mouse model of SVA infection established via oral administration (11). At 5 days post-infection (dpi), rSVA (VP1 K260R) showed higher viral loads in the heart, liver, spleen, kidney, and lung tissues compared to wild-type rSVA (Fig. 7G through K). Notably, lung tissue from mice infected with rSVA (VP1 K260R) exhibited more severe damage (thickened alveolar walls with focal granulocyte infiltration and compensatory alveolar dilation with widened interalveolar septa) than from wild-type rSVA-infected mice, with no significant pathology in the liver or kidney (Fig. 7L). Taken together, these findings indicate that rSVA (VP1 K260R) has enhanced replication capacity and pathogenicity in vitro and in vivo.
## DISCUSSION
Senecavirus A (SVA), an emerging virus associated with porcine idiopathic vesicular disease, causes significant global economic losses (2)(3)(4)40). The SVA VP1 protein is critical for viral attachment and entry, making it a promising target for anti-SVA strategies (14). For example, the selective autophagy receptor SQSTM1/p62 inhibits SVA replication by targeting VP1 (32), and OPTN suppresses SVA replication by promoting VP1 degrada tion (41). Components of the ubiquitin-proteasome system, particularly E3 ligases, often act as intrinsic antiviral factors (42). Here, we demonstrate that the E3 ubiquitin ligase STUB1 binds to and degrades the SVA VP1, with VP1 K177 and K260 being essential for STUB1-mediated, ubiquitination-dependent degradation. Previous studies on STUB1 in viral replication have focused on regulating immune signaling pathways. For instance, mixed-lineage leukemia 5 inhibits the antiviral innate immune response by facilitating STUB1-mediated degradation of retinoic acid-induci ble gene I (43); receptor for activated C kinase 1 enhances bovine ephemeral fever virus replication by upregulating STUB1 expression, which degrades mitochondrial antiviral signaling proteins (MAVS) (44); and cytoplasmic signal transducer and activator of transcription 4 promotes antiviral type I interferon (IFN) production by inhibiting STUB1-mediated degradation of retinoic acid-inducible gene I (45). These findings highlight STUB1's complex roles in viral infections via immune modulation. In contrast, here we identify STUB1 as a key host restriction factor that negatively regulates SVA replication by targeting VP1 for ubiquitination and degradation, consistent with reports that STUB1 mediates degradation of the HIV-1 Tat protein (46). STUB1 is an E3 ubiquitin ligase that ubiquitinates and degrades substrates ( 21), and our results confirm that STUB1-mediated VP1 degradation occurs primarily via the ubiquitin-proteasome degradation pathway. Ubiquitin contains seven lysine residues (K6, K11, K27, K29, K33, K48, and K63) that form chains on target proteins (47,48). K48-linked ubiquitin chains predominantly target proteins for proteasomal degradation, whereas K63-linked chains are involved in processes such as DNA damage repair, protein stabilization, and kinase activation (49,50). We show that STUB1 overexpres sion increases VP1 ubiquitination in a dose-dependent manner, specifically inducing K48-linked (but not K63-linked) ubiquitination, consistent with proteasomal targeting.
Identifying specific ubiquitination sites on target proteins helps elucidate the relationship between the ubiquitin-proteasome system and protein function (51). For example, dengue virus nonstructural protein 3 (NS3) is ubiquitinated at K104 by TRIM69 (52), and STING is modified with K63-linked chains at K150 by TRIM56 (53). SVA VP1 contains 12 lysine residues, with K177 and K260 identified as STUB1 targets.
Molecular chaperones, particularly HSP70 and HSC70, are known to participate in STUB1-mediated protein degradation (54,55). STUB1 residues K30 and H260 are crucial for chaperone binding and ubiquitination activity, respectively (36). We explored the effects of HSP70, HSC70, and these residues on VP1 degradation and the STUB1-VP1 interaction. Consistent with previous reports (56,57), STUB1 mutants at these resi dues failed to degrade VP1, and the K30A mutant disrupted interactions with VP1 or chaperones but not between VP1 and HSP70/HSC70. This suggests that STUB1 does not directly interact with VP1 but instead requires HSP70 or HSC70 as intermediaries. Furthermore, we found that HSP70 and HSC70 promote STUB1-mediated VP1 degrada tion by enhancing their interaction.
During evolution, viruses have evolved to exploit host proteins that facilitate replication while degrading those that hinder survival. Here, HSP70 and HSC70 enhance the STUB1-VP1 interaction to promote VP1 degradation, suggesting that they inhibit SVA replication. This contradicts our previous finding that HSP70 positively regulates SVA replication by stabilizing viral L and 3D proteins (37). This paradox implies a mecha nism that inhibits STUB1-mediated VP1 degradation in the presence of HSP70/HSC70 during SVA infection. STUB1 expression is downregulated in SVA-infected cells, whereas HSP70/HSC70 levels remain unchanged. Thus, SVA-mediated STUB1 downregulation may shift HSP70/HSC70 function toward stabilizing viral nonstructural proteins rather than degrading VP1, thereby promoting replication.
SVA 3Cpro typically cleaves or degrades host proteins via its protease activity to evade antiviral mechanisms. For example, it mediates nucleolin cleavage and redistrib ution (39), disrupts mitochondrial DNA-mediated immune sensing (58), and cleaves hnRNPK (59) to facilitate replication. Our results show that catalytically inactive 3Cpro loses its ability to degrade STUB1 and antagonizes STUB1-mediated VP1 degradation, indicating that protease activity is critical for counteracting host antiviral responses. However, like previous studies (16,38), we did not detect cleaved fragments of STUB1, possibly due to rapid degradation of small fragments; the mechanism of 3Cpro-mediated STUB1 degradation requires further investigation. Reverse genetics is pivotal for validating the functional roles of specific amino acid residues in viral proteins. For example, recombinant FMDV with the VP1-K200R mutation abrogates RNF5-mediated antiviral suppression (60), and recombinant Tembusu virus with NS1-K141R shows enhanced replication in avian hosts (61). We engineered VP1 mutants to explore the roles of K177 and K260 in SVA replication. The K177R mutation was lethal, whereas K260R enhanced replication. Mice infected with rSVA (VP1 K260R) showed higher viral loads in multiple tissues and more severe lung damage than those infected with wild-type virus. Consistent with the findings of Li et al. (62), the highest viral loads were observed in the heart, likely due to the close genomic relationship between SVA and Cardioviruses.
In conclusion, our results demonstrate that STUB1 inhibits SVA replication by promoting VP1 degradation, while SVA 3Cpro antagonizes this process by reducing STUB1 expression (Fig. 8). Understanding STUB1 regulatory mechanisms provides insights into strategies for preventing and controlling SVA infections.
## MATERIALS AND METHODS
## Cells, viruses, and reagents
BHK-21, ST, and HEK-293T cells were obtained from the American Type Culture Collection (ATCC) and cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, Waltham, MA, USA) supplemented with 5%-10% fetal bovine serum (FBS; Gibco), streptomycin, and penicillin at 37°C in a 5% CO 2 incubator. SVA CHhb17 strain and an anti-SVA VP1 monoclonal antibody were preserved in our laboratory. The eGFP-tagged recombinant SVA (rSVA-eGFP) was kindly provided by Dr. Fuxiao Liu (Qingdao Agricultural University). Commercially sourced antibodies included the following: rabbit anti-GFP (D110008-0200; Sangon, Shanghai, China), mouse anti-β-actin (D191047; Sangon), rabbit anti-STUB1 (A11751; ABclonal), rabbit anti-HSP70 (A20819; ABclonal, Woburn, MA, USA), rabbit anti-HSC70 (A0415; ABclonal), mouse anti-Flag (F4049; Sigma, St. Louis, MO, USA), mouse anti-Myc (05-419;Merck, St. Louis, MO, USA), mouse anti-HA (H3663; Sigma), horserad ish peroxidase (HRP)-conjugated anti-rabbit and -mouse secondary (A0545 or A9044; Sigma), and tetramethylrhodamine isothiocyanate (TRITC)-conjugated rabbit anti-mouse antibodies (ab6799; Abcam, Cambridge, UK).
## Plasmid construction and transfection
Plasmids encoding the SVA VP1 gene were preserved in our laboratory (13). STUB1, HSP70, and HSC70 genes derived from PK-15 cells or RAW cells were cloned into pCMV-HA, pCMV-Flag, or pCMV-Myc vectors. All primers used for plasmid construction are listed in Table 1. For transfection, BHK-21, ST, or HEK-293T cells were cultured to 70%-80% confluence and transfected with the indicated plasmids using Lipofectamine 2000 (11668019; Invitrogen, Carlsbad, CA, USA), following the manufacturer's protocols.
## Silver staining and mass spectrometric identification of proteins
HEK-293T cells were transfected with plasmids expressing GFP-VP1 or GFP (control) and lysed 36 h post-transfection in NP 40 buffer (10 mM Tris pH 7.5, 150 mM NaCl, 0.5% NP-40, 0.5 mM EDTA). Lysates were incubated with anti-GFP mAb-conjugated beads (GAN-50-1000; Lablead, Beijing, China) for immunoprecipitation. Immunoprecipitated proteins were washed three times with washing buffers (10 mM Tris pH 7.5, 150 mM NaCl, 0.05% NP-40, 0.5 mM EDTA), separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and visualized using a silver staining kit (#24612; Thermo Fisher Scientific) according to the manufacturer's instructions. Differential protein bands GFP-VP1 and GFP control lanes were manually excised for mass spectrometric analysis.
## Viral infection and 50% tissue culture infectious dose (TCID 50 ) assay
BHK-21 or ST cells (transfected or untransfected with plasmid) were washed with phosphatebuffered saline (PBS) and infected with SVA at a multiplicity of infection (MOI) of 1 for 1 h at 37°C. After removing unbound virus, cell culture supernatants were collected at specified time points, and viral titers were determined. Monolayer BHK-21 or ST cells in 96-well plates were inoculated with 100 μL of 10-fold serial dilutions of samples (eight replicates per dilution) and cultured until cytopathic effects (CPE) were observed. The 50% tissue culture infectious dose (TCID 50 ) was determined using the Spearman and Karber's method.
## siRNA transfection (RNAi)
siRNAs targeting STUB1 and HSP70 genes were designed and synthesized by Gene Pharma. The siRNA targeting HSC70 (siHSC70, sc-29349) was purchased from Santa Cruz. Cells were transfected with siRNA using Lipofectamine RNAiMAX (13778; Invitrogen) for 36 h, followed by SVA infection or plasmid transfection. Samples were then analyzed by Western blotting and TCID 50 assay. Sequences of siRNAs are listed in Table 1.
## RNA extraction and reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
Total RNA was extracted from cells using TRIzol reagent (15596018; Invitrogen) and reverse-transcribed into cDNA using HiScript III RT SuperMix (R323-01; Vazyme, China).
Quantitative PCR was performed using Taq Pro Universal SYBR qPCR Master Mix (Q712-02; Vazyme) on a LightCycler 96 system (Roche, Basel, Switzerland), with three technical replicates per sample. Relative mRNA levels of the STUB1 gene were calculated using the compare active cycle threshold (2 -ΔΔCT ) method and normalized to glyceral dehyde-3-phosphate dehydrogenase (GAPDH) mRNA. SVA RNA copy numbers were quantified using a standard curve (Y = -3.3814X + 36.59) generated from conserved sequences of the SVA 3D gene. Primer sequences are listed below: porcine GAPDH: F (TCG GAGTGAACGGATTTGGC) and R (TGACAAGCTTCCC-GTTCTCC), porcine STUB1: F (GGAGAACGAGCTGCACTCTT) and R (TGGTTCCGCTGACACTCTTC), hamster GAPDH: F GTCATC ATCTCCGCCCCTTC and R CCGTGGTCATGAGTCCTTCC, and hamster STUB1: F (TGC CCTTCGCATTGCTAAGA) and R TCCTCCAGTTCCCTCTCTCG, and SVA-3D: F (CCAACAAGGG TTCCGTCTTC) and R (TTGGACGAATTTGCGTTTTAGA).
## Cycloheximide (CHX) chase assay
Cells co-transfected with GFP-VP1, HA-STUB1, or siSTUB1-1 were treated with CHX (100 μg/mL) for various durations to block de novo protein synthesis. Proteins were extracted and analyzed by Western blotting.
## Immunofluorescence assays and confocal microscopy
BHK-21 or ST cells (80%-90% confluence in 24-well plates) were transfected with the indicated plasmids for 12 h (with or without rSVA-eGFP infection). Cells were fixed with 4% paraformaldehyde, washed, and incubated with primary antibodies, followed by TRITC-conjugated secondary antibodies and DAPI (nuclear stain). Fluorescent images were captured using an immunofluorescence microscope (IX73; Olympus, Tokyo, Japan) or confocal microscope (TCS SP8 STED; Leica, Wetzlar, Germany).
## Co-immunoprecipitation and western blotting
HEK-293T or ST cells were co-transfected with the indicated plasmids for 36 h, lysed with NP40 buffer containing phenylmethanesulfonyl fluoride (PMSF) (ST506; Beyotime), and immunoprecipitated with anti-GFP agarose (PGA025; Lablead) or anti-HA agarose (HNA-50-1000; Lablead) at 4°C for 1 h with rotation. For endogenous co-IP, the mouse anti-VP1 monoclonal antibody was conjugated to Protein A/G Plus Agarose overnight at 4°C and then incubated with lysates of SVA-infected ST cells for 12 h at 4°C. Immunopre cipitates or total cell lysates were separated by SDS-PAGE, transferred to nitrocellulose membranes (66485; Pall, Port Washington, NY, USA), blocked with 5% non-fat milk, and probed with primary and secondary antibodies. Signals were detected using an AMERSHAM ImageQuant800 chemiluminescence imaging system (GE, Chicago, IL, USA).
## Plaque assays
BHK-21 cells were seeded in 6-well plates and grown to 100% confluence. Cells were inoculated with 100 TCID 50 of virus and incubated for 1 h at 37°C and then overlaid with DMEM containing 2% heat-inactivated FBS and 1% low-melting-point agarose. After 24 h, plaques were visualized by staining with 1% crystal violet in methanol for 5 h at 37°C.
## Construction of infectious SVA cDNA clones
Based on the SVA CHhb17 genome (MG983756), two fragments were synthesized: SVA Frag I (positions 1-3,514, flanked by PacI and NheI) and SVA Frag II (positions 3,515-7,280, flanked by NheI and NotI). These were cloned into pUC-GW-Kan and then ligated subsequently into pcDNA-rSVAuni (a modified pcDNA3.1(+) vector containing a CMV promoter, T7 promoter, SVA hammerhead ribozyme [HamRz], restriction sites, hepati tis delta virus ribozyme [HdvRz], and T7 terminator) using double enzyme digestion. Plasmids pUC-GW-Kan and pcDNA-rSVAuni were kindly provided by Dr. Shichong Han (International Joint Research Center of National Animal Immunology, College of Veterinary Medicine, Henan Agricultural University). To rescue viruses, full-length cDNA plasmids were transfected into BHK-21 cells using Lipofectamine 2000. When ~90% CPE was observed, supernatants were collected and serially passaged five times in BHK-21 cells. Virus stocks were stored at -80°C.
## Animal experiments
Fifteen 3-week-old female BALB/c mice (Yangzhou University Laboratory Animal Center) were housed in pathogen-free facilities and randomly divided into three groups (n = 5 each): negative control, rSVA-infected, and rSVA (VP1K260R)-infected. Mice were infected with 2 × 10 7 TCID 50 of virus via oral administration, as described previously (11). At 5 days post-infection (dpi), mice were euthanized, and tissues (heart, liver, spleen, kidneys, and lungs) were collected. Tissues were either processed for RT-qPCR (viral load quantification) or fixed in 4% paraformaldehyde, embedded in paraffin, sectioned (4 μm), and stained with hematoxylin and eosin (HE) for histopathological analysis.
## Statistical analysis
Statistical differences were evaluated using one-way analysis of variance (ANOVA) or Student's t-test with GraphPad Prism 9.0 software (GraphPad Software, Boston, USA). P<0.05 was considered statistically significant.
## References
1. Hales, Knowles, Reddy et al. (2008) "Complete genome sequence analysis of Seneca Valley virus-001, a novel oncolytic picornavirus" *J Gen Virol*
2. Leme, Zotti, Alcântara et al. (2015) "Senecavirus A: an emerging vesicular infection in Brazilian Pig Herds"
3. Saeng-Chuto, Rodtian, Temeeyasen et al. (2016) "The first detection of Senecavirus A in pigs in Thailand"
4. Sun, Vannucci, Knutson et al. (2017) "Emergence and whole-genome sequence of Senecavirus A in Colombia" *Trans bound Emerg Dis*
5. Qian, Qian, Chen et al. (2016) "Isolation and full-genome sequencing of Seneca Valley virus in piglets from China" *Virol J*
6. Zhu, Yang, Chen et al. (2017) "Emergence of novel Seneca Valley virus strains in China"
7. *Transbound Emerg Dis*
8. Miles, Burga, Gardner et al. (2017) "Anthrax toxin receptor 1 is the cellular receptor for Seneca Valley virus" *J Clin Invest*
9. Kennedy, Denslow, Hewett et al. (2022) "Development of intravenously administered synthetic RNA virus immunotherapy for the treatment of cancer" *Nat Commun*
10. Liu, Huang, Wang et al. (2020) "Construction of eGFP-tagged Senecavirus A for facilitating virus neutralization test and antiviral assay" *Viruses*
11. Joshi, Mohr, Clement et al. (2016) "Detection of the emerging picornavirus Senecavirus A in Pigs, mice, and houseflies" *J Clin Microbiol*
12. (2025) *Full-Length Text Journal of Virology*
13. Li, Qiao, Guo et al. (2021) "Evaluation of immunogenicity and protective efficacy of a novel Senecavirus A strain-based inactivated vaccine in mice" *Res Vet Sci*
14. Segalés, Barcellos, Alfieri et al. (2017) *Senecavirus A. Vet Pathol*
15. Song, Hou, Quan et al. (2022) "Synergetic contributions of viral VP1, VP3, and 3C to activation of the AKT-AMPK-MAPK-MTOR signaling pathway for seneca valley virus-induced autophagy" *J Virol*
16. Cao, Zhang, Liu et al. (2018) "Seneca Valley virus attachment and uncoating mediated by its receptor anthrax toxin receptor 1" *Proc Natl Acad Sci*
17. Li, Yang, Ma et al. (2025) "Seneca Valley virus 3C protease cleaves HDAC4 to antagonize type I interferon signaling" *J Virol*
18. Xue, Liu, Zhu et al. (2018) "Seneca Valley Virus 3C pro abrogates the IRF3-and IRF7-mediated innate immune response by degrading IRF3 and IRF7" *Virology (Auckl)*
19. Nakamura (2018) "Ubiquitin system" *Int J Mol Sci*
20. Wang, Bi, Song (2023) "Host E3 ligase Hrd1 ubiquitinates and degrades H protein of canine distemper virus to inhibit viral replication" *Vet Res*
21. Li, Hao, Zhao et al. (2023) "The E3 ligase RNF5 restricts SARS-CoV-2 replication by targeting its envelope protein for degradation" *Sig Transduct Target Ther*
22. Fan, Mar, Sari et al. (2021) "TRIM7 inhibits enterovirus replication and promotes emergence of a viral variant with increased pathogenicity" *Cell*
23. Meacham, Patterson, Zhang et al. (2001) "The Hsc70 co-chaperone CHIP targets immature CFTR for proteasomal degradation" *Nat Cell Biol*
24. Narayan, Landré, Ning et al. (2015) "Protein-protein interactions modulate the docking-dependent E3-ubiquitin ligase activity of carboxy-terminus of Hsc70-interacting protein (CHIP)" *Mol Cell Proteomics*
25. Fan, Park, Nephew (2005) "CHIP (carboxyl terminus of Hsc70interacting protein) promotes basal and geldanamycin-induced degradation of estrogen receptor-alpha" *Mol Endocrinol*
26. Xiong, Liu, Cai et al. (2017) "The carboxyl terminus of heat shock protein 70-interacting protein (CHIP) participates in high glucose-induced cardiac injury" *Free Radic Biol Med*
27. Chen, Barbi, Bu et al. (2013) "The ubiquitin ligase Stub1 negatively modulates regulatory T cell suppressive activity by promoting degradation of the transcription factor Foxp3" *Immunity*
28. Yang, Wang, Zhu et al. (2011) "E3 ubiquitin ligase CHIP facilitates Toll-like receptor signaling by recruiting and polyubiquitinating Src and atypical PKC{zeta}" *J Exp Med*
29. Wang, Li, Hu et al. (2013) "STUB1 is essential for T-cell activation by ubiquitinating CARMA1" *Eur J Immunol*
30. Shao, Li, Song et al. (2016) "E3 ubiquitin ligase CHIP interacts with C-type lectin-like receptor CLEC-2 and promotes its ubiquitin-proteasome degradation" *Cell Signal*
31. Zhang, Zhang, Bie et al. (2022) "STUB1 regulates antiviral RNAi through inducing ubiquitination and degrada tion of Dicer and AGO2 in mammals" *Virol Sin*
32. Yang, Kong, Qin et al. (2023) "PGAM5 degrades PDCoV N protein and activates type I interferon to antagonize viral replication" *J Virol*
33. Wang, Du, Zhu et al. (2020) "STUB1 is targeted by the SUMOinteracting motif of EBNA1 to maintain Epstein-Barr Virus latency" *PLoS Pathog*
34. Wen, Li, Yin et al. (2021) "Selective autophagy receptor SQSTM1/ p62 inhibits Seneca Valley virus replication by targeting viral VP1 and VP3" *Autophagy*
35. Kwon, Ciechanover (2017) "The ubiquitin code in the ubiquitinproteasome system and autophagy" *Trends Biochem Sci*
36. Liu, Zhou, Tang (2023) "STUB1/CHIP: new insights in cancer and immunity" *Biomedicine & Pharmacotherapy*
37. Soss, Rose, Hill et al. (2015) "Biochemical and proteomic analysis of ubiquitination of Hsc70 and Hsp70 by the E3 ligase CHIP" *PLoS One*
38. Xu, Marcu, Yuan et al. (2002) "Chaperone-dependent E3 ubiquitin ligase CHIP mediates a degradative pathway for c-ErbB2/Neu" *Proc Natl Acad Sci*
39. Hou, Zeng, Wu et al. (2024) "Heat shock protein 70 enhances viral replication by stabilizing Senecavirus A nonstructural proteins L and 3D" *Vet Res*
40. Song, Wang, Quan et al. (2021) "Seneca Valley virus 3C pro degrades heterogeneous nuclear ribonucleoprotein A1 to facilitate viral replication" *Virulence*
41. Song, Quan, Wang et al. (2022) "Seneca Valley Virus 3C pro mediates cleavage and redistribution of nucleolin to facilitate viral replication" *Microbiol Spectr*
42. Wang, Prarat, Hayes et al. (2016) "Detection and genomic characterization of Senecavirus A" *Emerg Infect Dis*
43. Song, Guo, Wang et al. (2024) "Seneca Valley virus 3C protease cleaves OPTN (optineurin) to Impair selective autophagy and type I interferon signaling" *Autophagy*
44. Patil, Li (2019) "Tripartite motif proteins: an emerging antiviral protein family" *Future Virol*
45. Zhou, Ding, Wan et al. (2018) "MLL5 suppresses antiviral innate immune response by facilitating STUB1-mediated RIG-I degradation" *Nat Commun*
46. Zhang, Hou, He et al. (2021) "RACK1 degrades MAVS to promote bovine ephemeral fever virus replication via upregulating E3 ubiquitin ligase STUB1" *Vet Microbiol*
47. Zhao, Zhang, Li et al. (2016) "Cytoplasmic STAT4 promotes antiviral type I IFN production by blocking CHIP-mediated degradation of RIG-I" *J Immunol*
48. Ali, Farooqui, Banerjea (2019) "The host cell ubiquitin ligase protein CHIP is a potent suppressor of HIV-1 replication" *Journal of Biological Chemistry*
49. Joazeiro, Weissman (2000) "RING finger proteins: mediators of ubiquitin ligase activity" *Cell*
50. (2025) *Full-Length Text Journal of Virology*
51. Clague, Urbé (2010) "Ubiquitin: same molecule, different degrada tion pathways" *Cell*
52. Chen, Sun (2009) "Nonproteolytic functions of ubiquitin in cell signaling" *Mol Cell*
53. Raiborg, Stenmark (2009) "The ESCRT machinery in endosomal sorting of ubiquitylated membrane proteins" *Nature*
54. Wang, Zhang (2019) "Towards computational models of identifying protein ubiquitination sites" *Curr Drug Targets*
55. Wang, Zou, Wang et al. (2018) "Interferon-stimulated TRIM69 interrupts dengue virus replication by ubiquitinating viral nonstructural protein 3" *PLoS Pathog*
56. Tsuchida, Zou, Saitoh et al. (2010) "The ubiquitin ligase TRIM56 regulates innate immune responses to intracellular double-stranded DNA" *Immunity*
57. Li, Kurata, Maharani et al. (2015) "E3 ligase CHIP and Hsc70 regulate Kv1.5 protein expression and function in mammalian cells" *J Mol Cell Cardiol*
58. Luo, Zhong, Chang et al. (2010) "Hsp70 and CHIP selectively mediate ubiquitination and degradation of hypoxiainducible factor (HIF)-1alpha but Not HIF-2alpha" *J Biol Chem*
59. Wei, Sha, Bhattacharya et al. (2014) "Regulation of IL-4 receptor signaling by STUB1 in lung inflammation" *Am J Respir Crit Care Med*
60. Sha, Rao, Settembre et al. (2017) "STUB1 regulates TFEB-induced autophagy-lysosome pathway" *EMBO J*
61. Yan, Wu, Yuan et al. (2023) "Speciesspecific cleavage of cGAS by picornavirus protease 3C disrupts mitochondria DNA-mediated immune sensing" *PLoS Pathog*
62. Song, Quan, Wang et al. (2022) "Seneca Valley virus 3C (pro) cleaves heterogeneous nuclear ribonucleoprotein K to facilitate viral replication" *Front Microbiol*
63. Zhang, Li, Yang et al. (2025) "RING finger protein 5 is a key anti-FMDV host factor through inhibition of virion assembly" *PLoS Pathog*
64. Zhou, Zhang, Yang et al. (2025) "TRIM14 restricts tembusu virus infection through degrading viral NS1 protein and activating type I interferon signaling" *PLoS Pathog*
65. Li, Lin, Qi et al. (2023) "Senecavirus A-induced glycolysis facilitates virus replication by promoting lactate production that attenuates the interaction between MAVS and RIG-I" *PLoS Pathog* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724268&blobtype=pdf | # A newly isolated giant virus, ushikuvirus, is closely related to clandestinovirus and shows a unique capsid surface structure and host cell interactions
Jiwan Bae, Narumi Hatori, Raymond Burton-Smith, Kazuyoshi Murata, Masaharu Takemura
## Abstract
The family Mamonoviridae, assigned in 2023, consists of three strains of medusavirus that infect acanthamoeba. A closely related species, clandestinovirus, which infects vermamoeba, was reported in 2021. Here, we report a newly identified clandes tinovirus-like virus, named ushikuvirus, isolated from a freshwater pond in Japan. The ushikuvirus genome was at least 666,605 bp and contained 784 genes. Annotation revealed that a substantial proportion (58%) of open reading frames (ORFs) are ORFans, and 25% of ORFs are similar to those of other viruses in the phylum Nucleocytoviricota. Among ORFs sharing sequence similarity with other viruses, a large proportion (80%) were similar to the clandestinovirus sequences. However, ushikuvirus shows several remarkable features. (i) The capsid surface has multiple spike-like "cap" structures, some of which exhibit a fibrous structure. (ii) The infection cycle is longer than that of medusavirus and clandestinovirus, and the virus exhibits a unique cytopathic effect (CPE) that causes enlargement of host vermamoeba cells. (iii) The virus forms a viral factory for duplication and destroys the nuclear membrane of vermamoeba cells, a phenomenon not observed with medusavirus and clandestinovirus. These characteristics indicate that this newly isolated giant virus related to the family Mamonoviridae and Clandestinovirus may represent a key taxon for elucidating virus-host interactions and the evolution of this virus group.
IMPORTANCEThe family Mamonoviridae consists of only one genus, including three species: Medusavirus medusae, Medusavirus sthenus, and the recently described medusavirus euryale. These three medusaviruses have been reported to infect Acantha moeba castellanii. Meanwhile, clandestinovirus, a closely related species in the family Mamonoviridae, infects vermamoeba. In these viruses, genome replication takes place in the nucleus of the host cell, and like eukaryotes, the genome encodes a full set of histones and has numerous spikes on the capsid surface. Here, we report a new member of this unique virus group, ushikuvirus, which displays distinct features including cytopathic effects in vermamoeba cells. These findings improve our understanding of the biological significance of the family Mamonoviridae and closely related taxa and provide a basis for elucidating the evolutionary relationships of giant viruses with their eukaryotic hosts. KEYWORDS ushikuvirus, giant virus, virus isolation, vermamoeba, family Mamonoviri dae, clandestinovirus, cytopathic effect, spike proteins, virus infection and proliferation, host-virus interaction S ince the discovery of Acanthamoeba polyphaga mimivirus in 2003 (1), a wide variety of complex dsDNA viruses in the phylum Nucleocytoviricota have been isolated from water and soil environments worldwide (2-4). Several families whose hosts are
mainly eukaryotic unicellular organisms (e.g., Acanthamoeba spp.) have been assigned to this large virus group, including the families Mimiviridae, Allomimiviridae, Shizomimi viridae, Mesomimiviridae, Marseilleviridae, Pithoviridae, and Orpheoviridae, in addition to the well-known virus families (documented since the 20th century) Poxviridae, Ascoviri dae, Asfarviridae, Phycodnaviridae, and Iridoviridae, which have a wide range of hosts, from unicellular to multicellular eukaryotes (3).
In the virus taxonomy released in 2023 by the International Committee on Taxonomy of Viruses (ICTV), the family Mamonoviridae has been newly added to this large virus group (5). This family consists of a single genus Medusavirus, including Medusavirus medusae, Medusavirus sthenus, and a newly isolated putative third species, medusavirus euryale (6)(7)(8). The first strain in the genus Medusavirus was originally isolated from hot spring water in Hokkaido, Japan, in 2019 (Acanthamoeba castellanii medusavirus, species Medusavirus medusae), followed by the discovery of a strain isolated from a freshwa ter river in Kyoto, Japan, in 2021 (medusavirus stheno, species Medusavirus sthenus), respectively (6,7). The genus Medusavirus has several characteristic features (6)(7)(8). The genomes encode a full set of histones (H2A, H2B, H3, H4, and linker histone H1), except for medusavirus euryale. Genomic DNA is replicated in the host cell nucleus, without constructing a virion factory in the cytoplasm, as observed in mimivirus and marseillevi rus. The family B DNA polymerase gene is more similar to eukaryotic DNA polymerase δ than to B DNA polymerases of other viral families. Based on these properties and the ancestral features of the medusavirus genome, it has been proposed that the ancestor of medusavirus may have contributed to the emergence of eukaryotes (9). However, the genome of the recently isolated medusavirus euryale does not encode linker histone H1 (8). This report and a recent cryo-electron microscopy (cryo-EM)-based nucleosome reconstruction have revealed that the linker histone H1 of medusavirus does not function as a linker as in eukaryotes, indicating that the linker histone is not essential to the infection cycle of medusavirus (10).
In 2021, a Mamonoviridae-related giant virus named clandestinovirus was discovered from a wastewater sample in Saint-Pierre-de-Mézoargues, France (11). Clandestinovirus infects the unicellular eukaryote Vermamoeba vermiformis, not Acanthamoeba spp. The linear dsDNA genome of 581,987 bp contains 617 genes, exceeding the estimated number of genes for other members in the family Mamonoviridae. Although it is very closely related to the family Mamonoviridae, phylogenetically and its genomic DNA replicates in the host nucleus, similar to medusaviruses (6,9), clandestinovirus is not included in the family Mamonoviridae because it shows substantial divergence at the intra-and inter-family levels from the other two Mamonoviridae species, M. medusae and M. sthenus (5). As mentioned previously, the infectivity of clandestinovirus to vermamoeba clearly distinguishes this taxon from the family Mamonoviridae, which infects acanthamoeba. To date, several giant viruses have been isolated by co-culture with vermamoeba, including faustovirus, kaumoebavirus, yasminevirus, orpheovirus, tupanvirus, and fadolivirus (11)(12)(13)(14)(15)(16)(17)(18). It is possible that the attachment and invasion mechanisms of these giant viruses into vermamoeba differ from those of giant viruses infecting acanthamoeba, despite the fact that tupanvirus can infect both amoebae (16).
Here, we present a giant virus newly isolated from a freshwater pond in Ibaraki Prefecture near Tokyo, Japan, named ushikuvirus. The genomic, phylogenetic, and structural features of the virus are characterized as well as its cytopathic effects in vermamoeba cells. The newly identified features of ushikuvirus provide insights into the evolution of vermamoeba-infecting giant viruses and the mechanism underlying host-virus interactions.
## RESULTS
## Isolation of new giant virus
We isolated a new giant virus infecting Vermamoeba vermiformis from the freshwater pond "Ushiku-numa" in Ibaraki Prefecture near the Tokyo metropolitan area of Japan. This study reports, for the first time, a giant virus isolated from aquatic environments in Japan that infects Vermamoeba. Vermamoeba vermiformis, formerly known as Hartman nella vermiformis, belongs to the class Tubulinea (whereas Acanthamoeba spp. belongs to the class Discosea) and forms two shapes in the healthy state: globular and fusiform (Fig. 1a andb) (19). Under microbial infection, vermamoebae often change from globular to fusiform and finally form a rounded shape, as observed in the endosymbiotic bacte rial Candidatus phylum Dependentiae (20). Similarly, after infection with ushikuvirus, vermamoebae showed a shape change from globular to partly fusiform and finally rounded (Fig. 1a andb).
## Morphological features of ushikuvirus
Cryo-transmission electron microscopy (cryo-TEM) and conventional transmission electron microscopy (c-TEM) revealed that ushikuvirus particles are morphologically very similar to medusavirus infecting acanthamoeba (6), with an icosahedral shape and numerous short spikes on the capsid surface. Excluding the surface spikes, the particle diameter was approximately 250 nm (Fig. 1c andd). After ushikuvirus infection, the vermamoebae exhibited a CPE, with cell rounding (Fig. 1b). Icosahedral-shaped virus particles were observed in the cytoplasm of the virus-infected round cells (Fig. 1e). In the cytoplasm of some vermamoeba infected with ushikuvirus at 5 hpi, virus particles with a core containing genomic DNA were observed in smaller numbers than those without a core (Fig. 2a), suggesting the early-stage cells of virus infection. In contrast, in the cytoplasm of some vermamoeba-infected ushikuvirus, virus particles with a core were observed in larger numbers than those without a core. Cryo-EM observations of released ushikuvirus particles revealed no empty particles outside the cells (Fig. 1c andd), suggesting that these clustered-filled viral particles represent the late stage of viral infection (Fig. 2b). In addition, at these late-stage cells, newly assembled viruses were accumulated and surrounded with a membrane (Fig. 2b), which have been observed in clandestinovirus and marseilleviruses (11,(21)(22)(23).
## Morphological change of vermamoeba under ushikuvirus infection
The morphological changes in vermamoeba cells after ushikuvirus infection were observed by comparison with the endosymbiotic bacterium Candidatus phylum Dependentiae strain Noda2021, isolated from a Japanese freshwater pond in 2021 (20). As mentioned above, vermamoeba cells have two morphological forms: globular and fusiform, even in healthy conditions. However, vermamoebae cultured long term in our laboratory using PYG medium initially exhibit a typical globular shape, and the number of globular cells increased throughout the culture period (Fig. 3a). In the case of Noda2021 infection, the number of vermamoeba cells exhibiting a fusiform shape gradually increased from 3 days post-infection (dpi), and eventually, most of the cells underwent lysis (Fig. 3a). In contrast, ushikuvirus infection showed a distinct CPE on vermamoeba, with cells becoming round and some fusiform by 3 dpi, which was maintained up to 6 dpi (Fig. 3a). Cell lysis was not detected in ushikuvirus infection, suggesting that ushikuviruses are released from infected cells without cell lysis, probably by exocytosis. Cell numbers (measured using a cell counting chamber) did not change under ushikuvirus infection throughout the experiment (Fig. 3b), suggesting that ushikuvirus infection immediately inhibits cell proliferation. Vermamoeba cells increased moderately in size and gradually changed to a smooth and round shape during the early stage of ushikuvirus infection, i.e., up to 36 hours post-infection (hpi) (Fig. 4). The cells were progressively enlarged, a trend that persisted until 60 hpi, when the average cell dimensions were approximately two times larger than those of uninfected cells (0 hpi). Thereafter, the cells decreased in size or lost their typical morphology. Although the cell dimension varied, there was a general decreasing trend across the population (Fig. 4b).
## Proliferation of ushikuvirus in vermamoeba
A c-TEM observations of the ushikuvirus-infected vermamoeba cells provided clues into the infection and proliferation processes. (i) Ushikuviruses were taken up into the vermamoeba cells by endocytosis or phagocytosis (Fig. 5a). (ii) Virion uncoating occurred, followed by the formation of virion factories (VFs) within the cytoplasm of vermamoeba (Fig. 5b andc). In parallel, the nuclear membrane of virus-infected vermamoebae started to disappear (Fig. 6a). (iii) Progeny virions were produced from VFs and mature virions accumulated in the cytoplasm (Fig. 5d ande). (iv) Virion particles were finally released from the vermamoeba cells by exocytosis (Fig. 5f and6b). As shown in Fig. 5c through e, the VFs of ushikuvirus were detected as electron-dense globular regions in the vermamoeba cytoplasm, such as the VFs of the family Mimiviridae (2,(24)(25)(26). Progeny virus particles are thought to form from the surface of the VFs, with initial capsid assembly followed by genomic packaging, as observed in mimivirus (Fig. 5c through e). As ushikuvirus infection progressed, the nuclear membrane of host vermamoeba cells disappeared, although vestiges of putative heterochromatin remained (Fig. 6a). This feature has also been observed in pandoravirus infection (27), but not in clandestinovirus as the host nuclear membrane does not disappear in the infection cycle (11). Furthermore, we observed some cysts containing ushikuvirus particles in the cytoplasm (Fig. 6a). The number of rounded, viable cells (cells possessing pseudopods) infected with ushikuvirus was moderately decreased at 96 hpi (Fig. 6b). The TCID 50 (tissue culture infectious dose) value of ushikuvirus in the culture supernatant of virus-infected vermamoeba cells was maintained up to 96 hpi, despite a decrease in viable cells (Fig. 6c). These findings suggest that ushikuvirus is released by exocytosis from infected cells over time and not rapidly released from lysed cells. To confirm this, we visualized cells from a culture infected at a multiplicity of infection (MOI) of 10 using electron microscopy (Fig. 6d) and monitored their movement via time-lapse imaging (Movies S1 and S2).
## Genomic and phylogenetic characterizations of ushikuvirus
Analysis of the whole-genome sequence of ushikuvirus resulted in the reconstruction of two contigs of 652,555 and 14,050 bp, indicating that the ushikuvirus genome is at least 666,605 bp in length, has a GC content of 47.90%, and contains 784 genes (and two tRNA genes) (Fig. 7). Annotation of each open reading frame (ORF) revealed that the majority were classified as ORFans (58%), and 25% of the ORFs shared sequences similar to other viruses in the phylum Nucleocytoviricota (Fig. 8a). Among the ORFs with similarity to other viruses, the majority (80%) shared similarity with sequences from clandestinovi rus (11) (Fig. 8b). Clandestinoviruses are closely related to the family Mamonoviridae, which includes the genus Medusavirus, but their host is different from that of medusavi rus, Acanthamoeba castellanii. From our observations, ushikuvirus, like clandestinovirus, infects only vermamoeba and not acanthamoeba. A functional enrichment analysis of the ORF profile showed a pattern consistent with other giant viruses, and thus no specific functional genes were reported for ushikuvirus (Fig. 8c). Molecular phylogenetic analyses based on major capsid protein (MCP), mRNA capping enzyme, and family B DNA polymerase genes supported the close relationship between ushikuvirus and clandesti novirus (Fig. 9). The genomes of the family Mamonoviridae, except for newly discovered medusavirus euryale (8), encode a full set of histones (H1, H2A, H2B, H3, and H4), as observed in eukaryote genomes and clandestinovirus (6,7,11). The ushikuvirus genome was also found to encode a full set of histones, although the genes encoding H2A and H2B were fused together, similar to the H2A-H2B fusions of clandestinovirus and marseillevirus (Table 1). A recent study suggests that the linker histones of medusavirus do not function as nucleosome linkers (10). Thus, the linker histones of ushikuvirus and clandestinovirus may not function as "linkers. " These genomic features strongly suggest that ushikuvirus is closely related to clandestinovirus and the family Mamonoviridae, and these findings are further supported by proteomic trees constructed using ViPTree (Fig. 10).
## Unique structure of the ushikuvirus capsid
As mentioned above, the morphological features of the ushikuvirus viron, determined using cryo-TEM and c-TEM analyses, were considerably similar to those of medusavirus (6), including the capsid diameter and the presence of numerous spikes on the capsid surface (Fig. 11). To elucidate the detailed structural characteristics of the viral capsid, a cryo-EM single-particle analysis (SPA) of ushikuvirus was performed. As a result, the capsid array on the viral surface was reconstructed at 9.3 Å resolution using a capsid-spe cific mask (Fig. 11a through c). Notably, ushikuvirus capsid with a diameter of 250 nm, except for the surface spikes, showed T = 309 icosahedron consisting of h = 7 and k = 13 (Fig. 11a), which is similar to those of Marseilleviridae viruses, not Mamonoviridae viruses with a diameter of ~260 nm showing T = 277 icosahedron consisted of h = 7 and k = 12 (6). The fact suggested that MCPs of ushikuvirus are more closely packed in the capsid than that of medusaviruses. To elucidate why the MCP of ushikuvirus is closely packed, we modeled its structure using AlphaFold2 (28) and compared it to the MCPs of tokyovirus (Marseilleviridae) and medusavirus (Mamonoviridae) (Fig. S1). The results revealed that these structures are highly similar, consisting primarily of two jelly-roll motifs. This suggests that the difference in capsid packing is attributable to the formation of a minor capsid protein (mCP) that maintains the MCP array, with ushikuvirus possessing a unique mCP structure, warranting further structural analysis in future studies. The viral DNA was surrounded by an inner membrane, similar to other giant viruses (Fig. 11b). The surface of the ushikuvirus capsid was covered with multiple spikes (Fig. 11d), with several longer spikes distributed around the fivefold vertices (arrow in Fig. 11d), similar to those of the medusavirus capsid (6,29,30), and the diameter of the capsid including the spikes reached approximately 270 nm (Fig. 11a andb). However, the other spikes were relatively shorter and diverse compared to those of medusavirus. The shortest spikes formed a unique "cap" structure (white circle in Fig. 11d) that was not observed in medusavirus. Among these, sixteen of the capso meres arranged in a straight line of six each around the threefold axis were particularly decorated with the small amounts of fibrous structures on their surfaces (asterisks in Fig. 11d). To examine whether the capsid surface, including the flexible fibrous structures, is composed or arranged with glycans, periodic acid-Schiff (PAS) staining was performed using ushikuvirus proteins. The result suggested that the capsid of ushikuvirus may be decorated with some glycans attached to the capsid proteins with a molecular weight slightly larger than that of MCP (Fig. 12). Data shown in Fig. 12a andb indicate that Mimivirus shirakomae and tokyovirus have PAS signals, consistent with previous reports showing that mimivirus has surface fibrils containing glycoprotein and tokyovirus has carbohydrate chains on its surface (21,24,(31)(32)(33).
## Putative role of GMC-oxidoreductase
In the family Mimiviridae, GMC-oxidoreductase is responsible for the structure and function of surface fibrils (24,32,34). Ushikuvirus harbored two GMC-oxidoreductase genes that are suggested to contribute to the formation of surface fibrils in mimi viruses and are also encoded by giant viruses possessing fibrous structures on the capsid surface, such as pandoravirus (27) and vermamoeba-infecting orpheovirus (15). A molecular phylogenetic analysis revealed that ushikuvirus GMC-oxidoreductases are closely related to orpheovirus GMC-oxidoreductases (Fig. 13). Among the three GMC-oxidoreductases encoded by the orpheovirus genome, ORPV 177 was similar to ushikuvirus GMC-oxidoreductase H5 167, and ORPV 129 was similar to ushikuvirus H5 445 (Fig. 13). Protein structure prediction using AlphaFold3 also revealed that these two "sets" of ushikuvirus-orpheovirus GMC-oxidoreductases are structurally homologous to them (Fig. 14). These proteins involved in the capsid surface fibers may be a key component to infect vermamoeba.
## DISCUSSION
Since the discovery of the first medusavirus strain, Acanthamoeba castellanii medusavi rus (Medusavirus medusae), in 2019, there has been a gradual increase in the reports of members of the family Mamonoviridae and closely related taxa, including medusavi rus stheno (Medusavirus sthenus) discovered from a freshwater river in Japan in 2021, medusavirus euryale (without an assigned species name) discovered from a freshwater river in South Korea in 2025, and clandestinovirus discovered in wastewater in France in 2021 (6)(7)(8)11). Clandestinovirus is phylogenetically closely related to medusaviruses but was not classified in the family Mamonoviridae because its host is vermamoeba rather than acanthamoeba (5,11). These results prompted us to evaluate the differences between Mamonoviridae viruses that infect acanthamoeba and clandestinovirus that infects vermamoeba. The discovery of ushikuvirus, which was closely related to clandestinovirus, in this study further encouraged us to evaluate the characteristics of these viruses that contribute to differences in host species. Besides the host, medusavirus and ushikuvirus showed important differences in many aspects, including genome sizes, number of ORFs, and capsid surface structure. The genome of ushikuvirus is 666 kbp (Fig. 7), which is similar to the genome size of clandestinovirus (i.e., 582 kbp) (11), but different from viruses in the family Mamo noviridae (medusae: 381 kbp, sthenus: 362 kbp, euryale: 369 kbp) (6)(7)(8). Similarly, the number of ORFs of ushikuvirus (Fig. 8) is slightly higher than that of clandestinovirus (11). The relationship between genome size and number of ORFs and the differences in host amoeba has not been established. Alternatively, it is possible that the capsid structure of ushikuvirus (Fig. 11) is involved in the host difference between ushikuvirus and medusavirus.
Ushikuvirus exhibited a unique structure on the capsid surface (Fig. 11). Both ushikuvirus and medusavirus have numerous spike structures on the capsid surface, including regular and long spikes (22,33). In addition, wide spikes have been identified in medusavirus (29, 30) but were not observed in ushikuvirus particles (Fig. 11). In both ushikuvirus and medusavirus, long spikes were observed around the fivefold vertices of the icosahedral capsid (Fig. 11) (29,30). Interestingly, the MCPs of ushikuvirus have unique "cap" structures (Fig. 11) that are not observed in medusavirus particles (29,30). Cryo-EM SPA of ushikuvirus suggested that all MCPs of ushikuvirus have "cap" structures, while some of them possess additional fibrous structures on the top (Fig. 11). The most studied fibrous structure on the giant virus capsid is the surface fibrils of mimivirus, which is thought to be involved in viral attachment to the surface of acanthamoeba cells. A study of a mutated mimivirus (M4 strain) lacking surface fibrils revealed several genes involved in surface fibril structure and function, including the GMC-oxidoreductase gene (32,35). It was suggested that the GMC-oxidoreductase functions in the assembly of the surface fibrils in mimiviruses through an unknown mechanism (24). In addition, orpheovirus has thin fibrous structures on the capsid, and its genome is known to encode GMC-oxidoreductase (15). The ushikuvirus genome also contains a GMC-oxidor eductase gene, which is similar to the corresponding gene in orpheovirus, suggesting that, as in mimiviruses, there is a relationship between the fibrous structure of ushikuvi rus particles and ushikuvirus GMC-oxidoreductase (Fig. 13 and14). During cell imaging, we were unable to obtain data on the genome packaging mechanism of ushikuvirus; therefore, the localization of this protein to the genome fiber-a phenomenon observed during mimivirus genome packaging (34, 36)-was not detected. It has been reported that CPE in orpheovirus-infected vermamoeba cells shows a fusiform shape (37) that is different from CPE in ushikuvirus-infected vermamoeba (Fig. 3a). Furthermore, as shown by the PAS staining results (Fig. 12), if the fibrous structure of ushikuvirus contains glycans, this may be involved in the unique infection characteristics, leading to the slightly enlarged cells (Fig. 4). In conclusion, the fibrous structure of ushikuvirus might significantly affect its infection cycle. Compared to medusavirus and clandestinovirus, a distinctive characteristic of ushikuvirus is its long infection cycle, exhibiting a notably slower rate of CPE induction than its close relatives. We hypothesize that this surface structure causes a deceleration in the infection rate; this scenario is supported by observations within the family Mimiviridae, where the absence of viral surface fibers in infection kinetics results in a faster overall rate of viral replication and induction of CPE (38). Notably, giant viruses infecting acanthamoeba (e.g., mimivirus, marseillevirus, medusavirus, and pandoravirus) do not induce cell enlargement when they infect healthy cells. Instead, virus infection induces cell compaction, and this change occurs within 24 hpi (39). In contrast, cells infected with mimivirus and moumouvirus belonging to the family Mimiviridae shrink to less than 70% of their original size at 6 hpi, whereas cells infected with megaviruses show no change in dimensions (40). On the other hand, cells infected with ushikuvirus enlarged to approximately twice the dimensions of uninfected cells on average, with some individual cells increasing to more than seven times their original size. The curve of cell dimensions (Fig. 4b) suggested that the duration of the replication cycle of ushikuvirus infection was approximately 60 hpi or more. This is consistent with the finding that vermamoeba-infected giant viruses have a longer multiplication cycle than giant viruses infected with acanthamoeba (37). However, we assessed the replication cycle under CPE without viral titer measurements at each infection time point (Fig. 6c). Although we did not measure the volume of infected cells, ushikuvirus-infected cells may have become flatter than non-infected cells, resulting in enlargement of cell dimensions when viewed from above. In that case, it can be hypothesized that the cell flattening is caused by increasing attachment of cells to flasks via the release of ushikuvirus particles with fibrous structures containing glycans (Fig. 11 and12). It is believed that ushikuvirus particles are gradually released by exocytosis from infected vermamoeba cells without cell lysis, which promote cell flattening. Another possible scenario for cell enlargement is that ushikuvirus particles have not yet been released from the cells, leading to their accumulation in the cytoplasm of infected cells, causing cell swelling. A hypothetical model for the mechanism of ushikuvirus prolifera tion and release is schematically shown in Fig. 15. It is also possible that in an aqueous environment, swollen, floating cells may excrete virions over a wide area for a long period of time.
While the replication and proliferation strategy of ushikuviruses is similar to that of members of the family Mamonoviridae, notable differences are also observed. However, the possible evolutionary scenario is unclear because the molecular phylogenies of ushikuvirus' core proteins have not been elucidated. For example, sequence similarity in MCP and B family DNA polymerases was detected between ushikuvirus and clandes tinovirus and their sister clade, medusaviruses (Fig. 9a andc), but both ushikuvirus and clandestinovirus genomes encode mRNA capping enzymes that are not encoded by medusaviruses of the family Mamonoviridae (Fig. 9b). The proteomic tree of DNA viruses suggests that the genomic evolution of ushikuvirus and clandestinovirus is earlier than the diversification of the family Mamonoviridae (Fig. 10), prompting us to hypothesize that the divergence of the ancestors of ushikuvirus and clandestinovirus from the ancestor of the family Mamonoviridae coincides with the diversification of the class Discosea, which includes acanthamoeba, and the class Tubulinea, which includes vermamoeba. Regarding the effects of these viruses on the host nucleus, interesting evolutionary scenarios are possible since medusaviruses of the family Mamonoviridae and clandestinovirus do not disrupt the nuclear membranes and replicate within the host cell nucleus, while ushikuvirus disrupts the host nuclear membrane (Fig. 6). In pandoravirus, whose particle size is significantly larger than that of other giant viruses, loss of nuclear membranes has been observed following virus infection (27). Although it is not clear why the disassembly of the nucleus occurs upon infection with these viruses, the genomes of medusavirus do not encode RNA polymerase, mRNA capping enzyme, or DNA topoisomerase II, resulting in a strong dependency on the host cell nucleus, which they then use as a "virion factory" (6,9). Ushikuvirus does not need to use host enzymes because the virus itself encodes these proteins and therefore may not need to persist in the host cell nucleus. Moreover, we observed cyst formation in the ushikuvirusinfected culture environment. Encystment of infected cells has been uniquely reported for faustovirus mariensis (41), a virus infecting vermamoeba, and is a common feature among genus Medusavirus species infecting acanthamoeba (6). However, a unique characteristic of these viruses is their ability to induce encystment in almost all cells in the culture environment. Specifically, F. mariensis induced encystment in most cells in the culture environment at an MOI exceeding 1. However, similar observations have not yet been reported in the ushikuvirus-infected environment, necessitating further studies.
Several giant viruses infecting vermamoeba have been isolated, including ushikuvi rus, clandestinovirus, faustovirus, fadolivirus, orpheovirus, kaumoebavirus, yasminevius, and tupanvirus (11)(12)(13)(14)(15)(16)(17)(18). Tupanvirus, in particular, behaves as "dual-acting" viruses, capable of infecting both acanthamoeba and vermamoeba (16). Tupanvirus exploits both classes of amoebae via shared features or via a set of genes for infection targeting both classes. However, ushikuvirus and most vermamoeba-infecting viruses appear to have only one set of genes for infection targeting the class Tubulinea. Nevertheless, the molecular similarities between ushikuvirus and the family Mamonoviridae indicate that the basic infection mechanism is shared between these taxa, and differences in infection strategies, including the capping of capsid proteins, addition of fibrous structures, mechanisms of nuclear membrane disassembly, and moderate release of particles using exocytosis, likely arose during ushikuvirus evolution. Taken together, this newly isolated relative of the family Mamonoviridae and clandestinovirus may provide key insights into virus-host interactions and the evolution of this virus group.
## MATERIALS AND METHODS
## Culture of vermamoeba, virus isolation, and virus cloning
Vermamoeba vermiformis was purchased from American Type Culture Collection (ATCC) as Hartmannella vermiformis Page strain CDC-19 (ATCC 50237). Vermamoeba cells were cultured in proteose peptone-yeast extract-glucose (PYG) medium at 26°C, similar to the conditions for acanthamoeba described previously (6,21,42). Water samples (50 mL) were collected from a freshwater pond, Ushiku-numa, in Ibaraki Prefecture of Japan and then stored at 4°C until inoculation to vermamoeba cells. A portion of the sample (4.5 mL) was filtered using Whatman filter paper 43, followed by filtration using a 1.2 µm syringe filter. Samples were then mixed with 2 × PYG medium (4.5 mL) and an antibiotic solution (360 µL), as described previously (6,21,42). The mixed solution containing the vermamoeba suspension (50 µL) was inoculated into a 96-well microplate and incubated at 26°C. After 4 days, CPE in vermamoeba cells was observed in only one well in a 96-well microplate. The supernatant of this well was added to fresh vermamoeba cells, and putative viruses were cloned by serial dilution, as described previously (6,21,42). The finally isolated virus was named "ushikuvirus, " reflecting the name of the freshwater pond "Ushiku-numa" where it was collected.
## c-TEM analyses
Vermamoeba cells were infected with ushikuvirus at an unknown multiplicity of infection (MOI) and incubated at 26°C. After 5 dpi, cells were collected by centrifugation at 500 × g for 5 min, washed with PBS, and fixed with 2% glutaraldehyde (GA) as described previously (6,21,42). The fixed cells were washed again with PBS, stained with 2% osmium tetroxide, and dehydrated with increasing ethanol concentrations and propylene oxide, as described previously (6,21,42). Dehydrated cells were embedded in EPON-812 resin (TAAB Laboratory Equipment, Aldermaston, UK), sectioned, and then visualized using a transmission electron microscope (H-7600; Hitachi, Tokyo, Japan). The c-TEM analysis was performed at Hanaichi UltraStructural Research Institute (Okazaki, Aichi, Japan).
## Cryo-EM and single-particle analysis
After 5 dpi, ushikuvirus particles released from infected vermamoeba cells were collected from the supernatant by centrifugation at 8,000 × g for 35 min, 4°C. After washing with PBS three times, virions were resuspended with PBS. Then, 2 µL of the virus suspension was applied to QuantiFoil R1.2/1.3 copper grids (QuantiFoil GmbH, Thuringia, Germany) and plunge-frozen using a Vitrobot Mk. IV (Thermo Fisher Scientific: TFS). Cryo-TEM imaging and grid screening for SPA were performed on a 200kV JEM-2200FS electron microscope (JEOL) using a 626 cryo-specimen holder (Gatan). For SPA, a 300 kV Titan Krios G4 electron microscope (TFS) equipped with a C-FEG electron source, Falcon 4i direct detector, and Selectris-X energy filter. Micrographs were acquired using Tomogra phy 5 software rather than EPU software both implemented on the microscope as the high contrast of the viruses caused the hole finding and centering algorithm of EPU to misidentify most holes and skip acquisitions as a result. Micrographs were acquired at 64,000× (effective magnification 1.912 Å/pixel) with a total dose of 8 e -/Å 2 . Micrograph movies were imported into RELION 5 and motion-corrected using the RELION (43)(44)(45)(46) implementation of the MotionCor2 algorithm (47), before contrast transfer function (CTF) estimation with CTFFIND 4 (version 4.1.14) (48). Micrographs with a poor CTF fit were removed manually, and 79 particles were picked across 40 micrographs. These particles were extracted into 1,600 pixel boxes (downsampled to 160 pixels) with an effective pixel size of 19.12 Å/pixel. 2D classification into five classes was performed, and 54 virus particles were selected in one class for template-based autopicking. This resulted in 8,598 automatically picked particles, which were extracted into 1,536 pixel boxes (downsampled to 192 pixels) and 2D-classified into 40 classes with a mask diameter of 2,650 Å using the expectation/maximization algorithm. Two rounds of 2D classification were performed, with a total of 1,600 particles finally selected. An initial model was generated with no symmetry imposition and aligned to I1 symmetry, which the selected particles were classified and refined against in 3D. Particles were extracted into 1,536 pixel boxes (downsampled to 384 pixel boxes, effective pixel size 7.648 Å/pixel), 3D-classified, and re-extracted into 1,536 pixel boxes (downsampled to 768 pixel boxes, effective pixel size 3.824 Å/pixel), resulting in a 10 Å resolution reconstruction (Fig. S2). CTF refinement was carried out (defocus and astigmatism refinement, followed by beam tilt estimation), followed by further 3D refinement. Finally, particles were re-extracted into 1,536 pixel boxes (downsampled to 1,024 pixel boxes, effective pixel size 2.868 Å/ pixel), and a last round of 3D refinement was carried out. Ewald sphere correction (45) was performed on the half-sets, and post-processing resulted in a final resolution of 9.3 Å with a capsid mask (Fig. 11c; Fig. S2).
## Infection cycle analysis
Vermamoeba cells were cultured in PYG medium in a 25 cm 2 culture flask and exposed to ushikuvirus particles. Images of infected and non-infected cells at 0, 1, 2, 3, 4, 5, and 6 dpi were captured using an all-in-one fluorescence microscope (BZ-X800/X810, Keyence Co., Tokyo, Japan) with a 20× objective lens. For comparison, the vermamoeba-infecting Dependentiae strain Noda2021 (20) was independently used to evaluate the culture cells. Images of infected vermamoeba cells were captured as described above. Furthermore, c-TEM images of ushikuvirus-infected vermamoeba cells at 2, 4, 8, and 120 h post-infec tion (hpi) were obtained as described above. Independently, ushikuvirus was inoculated to vermamoeba cells at an MOI of 10. The supernatants of infected cells were collected at different time points and centrifuged at 400 × g for 5 min. The supernatants were titrated in 96-well plates to 5 × 10 5 vermamoeba cells using the TCID 50 calculator (49).
## Cell counts and dimensions
Vermamoeba cells were infected by ushikuvirus at an MOI of 5 in a 25 cm 2 culture flask. Ushikuvirus-infected cells at each time point were pipetted and quantified using a disposable hemocytometer (WAKENBTECH Co. Ltd., Tokyo, Japan). Simultaneously,
## References
1. Scola, Audic, Robert et al. (2003) "A giant virus in amoebae" *Science*
2. Claverie, Abergel (2018) "Mimiviridae: an expanding family of highly diverse large dsDNA viruses infecting a wide phylogenetic range of aquatic eukaryotes" *Viruses*
3. Bosmon, Abergel, Claverie (2025) "20 years of research on giant viruses" *Npj Viruses*
4. Gaïa, Forterre (2023) "From Mimivirus to mirusvirus: the quest for hidden giants" *Viruses*
5. Zhang, Takemura, Murata et al. (2023) "Mamonoviridae", a proposed new family of the phylum Nucleocytoviricota" *Arch Virol*
6. Yoshikawa, Blanc-Mathieu, Song et al. (2019) "Medusavirus, a novel large DNA virus discovered from hot spring water" *J Virol*
8. Yoshida, Zhang, Garcia et al. (2021) "Draft genome sequence of medusavirus Stheno, isolated from the Tatakai River of Uji" *Japan. Microbiol Resour Announc*
9. Bae, Takemura (2025) "Near-complete genome of medusavirus euryale, putative third species of the genus Medusavirus, isolated from the TaeHwaGang river, Ulsan, South Korea" *Microbiol Resour Announc*
10. Takemura (2020) "Medusavirus ancestor in a proto-eukaryotic cell: updating the hypothesis for the viral origin of the nucleus" *Front Microbiol*
11. Toner, Hoitsma, Weerawarana et al. (2024) "Characterization of Medusavirus encoded histones reveals nucleosome-like structures and a unique linker histone" *Nat Commun*
12. Rolland, Andreani, Sahmi-Bounsiar et al. (2021) "Clandestinovirus: a giant virus with chromatin proteins and a potential to manipulate the cell cycle of its Host Vermamoeba vermiformis" *Front Microbiol*
13. Klose, Reteno, Benamar et al. (2016) "Structure of faustovirus, a large dsDNA virus" *Proc Natl Acad Sci*
14. Bajrai, Benamar, Azhar et al. (2016) "Kaumoebavirus, a new virus that clusters with faustoviruses and Asfarviridae" *Viruses*
15. Bajrai, Mougari, Andreani et al. (2019) "Isolation of yasminevirus, the first member of Klosneuvirinae isolated in coculture with Vermamoeba vermiformis, demonstrates an extended arsenal of translational apparatus compo nents" *J Virol*
16. Andreani, Khalil, Baptiste et al. (2017) "Orpheovirus IHUMI-LCC2: a new virus among the giant viruses" *Front Microbiol*
17. Abrahão, Silva, Silva et al. (2018) "Tailed giant Tupanvirus possesses the most complete translational apparatus of the known virosphere" *Nat Commun*
18. Andreani, Schulz, Pinto et al. "La Scola B. 2021. Morphological and genomic features of the new Klosneuvirinae isolate fadolivirus IHUMI-VV54" *Front Microbiol*
19. Geballa-Koukoulas, Scola, Blanc et al. (2022) "Diversity of giant viruses infecting Vermamoeba vermiformis" *Front Microbiol*
20. Delafont, Rodier, Maisonneuve et al. (2018) "Vermamoeba vermiformis: a free-living amoeba of interest" *Microb Ecol*
21. Takemura (2022) "Genome sequence of a new "Candidatus" phylum "Dependentiae" isolate from Chiba" *Japan. Microbiol Resour Announc*
22. Takemura (2016) "Morphological and taxonomic properties of Tokyovirus, the first Marseilleviridae member isolated from Japan" *Microbes Environ*
23. Aoki, Hagiwara, Akashi et al. (2019) "Fifteen marseilleviruses newly isolated from three water samples in Japan reveal local diversity of Marseilleviridae" *Front Microbiol*
24. Sahmi-Bounsiar, Rolland, Aherfi et al. (2021) "Marseilleviruses: an update in 2021" *Front Microbiol*
25. Klose, Herbst, Zhu et al. (2015) "A Mimivirus enzyme that participates in viral entry" *Structure*
26. Suzan-Monti, Scola, Barrassi et al. (2007) "Ultrastructural characterization of the giant volcano-like virus factory of Acanthamoeba polyphaga Mimivirus" *PLoS One*
27. Mutsafi, Shimoni, Minsky (2013) "Membrane assembly during the infection cycle of the giant Mimivirus" *PLoS Pathog*
28. Philippe, Legendre, Doutre et al. (2013) "Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes" *Science*
29. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
30. Watanabe, Song, Kayama et al. (2022) "Particle morphology of medusavirus inside and outside the cells reveals a new maturation process of giant viruses" *J Virol*
31. (2025) *Full-Length Text Journal of Virology*
32. Watanabe, Song, Takemura et al. (2024) "Subnanometer structure of medusavirus capsid during maturation using cryo-electron microscopy" *J Virol*
33. Takemura, Mikami, Murono (2016) "Nearly complete genome sequences of two Mimivirus strains isolated from a Japanese freshwater pond and river mouth" *Genome Announc*
34. Boyer, Azza, Barrassi et al. (2011) "Mimivirus shows dramatic genome reduction after intraamoebal culture" *Proc Natl Acad Sci*
35. Xiao, Kuznetsov, Sun et al. (2009) "Structural studies of the giant mimivirus"
36. Alempic, Bisio, Villalta et al. (2024) "Functional redundancy revealed by the deletion of the mimivirus GMC-oxidoreductase genes"
37. Rodrigues, Silva, Dornas et al. (2015) "Mimivirus fibrils are important for viral attachment to the microbial world by a diverse glycoside interaction repertoire" *J Virol*
38. Villalta, Schmitt, Estrozi et al. (2022) "The giant mimivirus 1.2 Mb genome is elegantly organized into a 30-nm diameter helical protein shield"
39. Souza, Rodrigues, Reis et al. (2019) "Indepth analysis of the replication cycle of Orpheovirus" *Virol J*
40. Nunes, Oliveira, Essus et al. (2024) "Cytopathic effects in Mimivirus infection: understanding the kinetics of virus-cell interaction" *Mem Inst Oswaldo Cruz*
41. Fukaya, Takemura (2021) "Kinetic analysis of Acanthamoeba castellanii infected with giant viruses quantitatively revealed process of morpho logical and behavioral changes in host cells" *Microbiol Spectr*
42. Aquino, Reis, Moreira et al. (2024) "Giant viruses inhibit superinfection by downregulating phagocytosis in Acanthamoeba" *J Virol*
43. Borges, Rodrigues, Dornas et al. (2019) "Trapping the enemy: Verma moeba vermiformis circumvents Faustovirus mariensis dissemination by enclosing viral progeny inside cysts" *J Virol*
44. Takahashi, Fukaya, Song et al. (2021) "Morpholog ical and taxonomic properties of the newly isolated Cotonvirus japonicus, a new lineage of the subfamily Megavirinae" *J Virol*
45. Scheres (2012) "RELION: implementation of a Bayesian approach to cryo-EM structure determination" *J Struct Biol*
46. Kimanius, Forsberg, Lindahl (2017) "Accelerated cryo-EM structure determination with parallelisation using GPUs in Relion-2" *Biophys J*
47. Zivanov, Nakane, Forsberg et al. (2018) "New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7:e42166"
48. Zivanov, Nakane, Scheres (2020) "Estimation of high-order aberrations and anisotropic magnification from cryo-EM data sets in RELION-3.1" *IUCrJ*
49. Zheng, Palovcak, Armache et al. (2017) "MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy" *Nat Methods*
50. Rohou, Grigorieff (2015) "CTFFIND4: fast and accurate defocus estimation from electron micrographs" *J Struct Biol*
51. Reed, Muench (1938) "A simple method of estimating fifty per cent endpoints" *Am J Epidemiol*
52. Bankevich, Nurk, Antipov et al. (2012) "SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing" *J Comput Biol*
53. Camacho, Coulouris, Avagyan et al. (2009) "BLAST+: architecture and applications" *BMC Bioinformatics*
54. Tamura, Stecher, Kumar (2021) "MEGA11: molecular evolutionary genetics analysis version 11" *Mol Biol Evol*
55. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
56. Nishimura, Yoshida, Kuronishi et al. (2017) "ViPTree: the viral proteomic tree server" *Bioinformatics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645930&blobtype=pdf | # PRRSV GP2a blocks the RLR signaling pathway by targeting RIG-I
Yingjie Xiang, Chunxiao Mou, Xing Zhao, Chen Zhuo, Kaichuang Shi, Yun Go, Zhenhai Chen
## Abstract
The strategy of viruses targeting RIG-I to disrupt the interferon (IFN) system represents an effective mechanism for evading innate immune responses. In this study, we observed that porcine reproductive and respiratory syndrome virus (PRRSV) GP2a inhibited IFN production by targeting RIG-I. Further studies revealed that GP2a blocks the RIG-I-like receptor signaling pathway through two mechanisms: (i) GP2a induces K48-linked ubiquitination of RIG-I by promoting the interaction between RIG-I and RING finger protein 125 (RNF125), resulting in RIG-I degradation, and (ii) GP2a hijacks zinc finger CCHC-type containing 3 (ZCCHC3) to disrupt the formation of the tripar tite motif-containing 25 (TRIM25)-RIG-I complex, thereby inhibiting RIG-I K63-linked ubiquitination. This inhibition effectively prevents the activation and expression of RIG-I. In conclusion, our findings demonstrate novel mechanisms by which PRRSV GP2a inhibits IFN production, thereby improving our understanding of PRRSV immune evasion strategies.IMPORTANCE Porcine reproductive and respiratory syndrome is an important viral disease that affects the swine industry worldwide. PRRSV glycoproteins (GPs) play a crucial role in the viral infection process. However, it remains largely unknown about what roles PRRSV GPs play in antagonizing the innate immune response. In this study, we found that GP2a targets RIG-I to inhibit IFN production through a dual-faceted mechanism. GP2a promotes the RNF125-mediated degradation of RIG-I and competi tively interacts with ZCCHC3 to impede TRIM25-induced RIG-I activation. This research contributes to a deeper understanding of the immune escape mechanisms employed by PRRSV.
recognize the viral RNA (10). Subsequently, RIG-I interacts with mitochondrial antiviral signaling protein (MAVS) to recruit downstream signaling molecules, such as TANK-bind ing kinase 1 (TBK1) and inhibitor of NF-κB kinase-ε (IKKε), thereby inducing phosphory lation of IFN regulatory factor 3 (IRF3) and IRF7 (11,12). Once phosphorylated, IRF3 and IRF7 translocate to the nucleus and associate with CREB-binding protein (CBP) to form transcriptional enhancers that stimulate IFN transcription (13,14). Subsequently, IFN activates the Janus kinase-signal transducer and activator of the transcription (JAK-STAT) signaling pathway, leading to the production of interferon-stimulated genes (ISGs) (15). Notably, RIG-I is an ISG that establishes a positive feedback loop, further increasing RIG-I production (16). It is worth noting that the ubiquitin (Ub) system plays an important role in regulating RIG-I activation (17). E3 Ub ligases, such as tripartite motif-containing 25 (TRIM25), positively regulate RIG-I by inducing K63-linked ubiquitination (18). This modification can promote RIG-I activation, thereby promot ing IFN production (18). Conversely, some E3 Ub ligases function as negative regu lators of RIG-I. RING finger protein 125 (RNF125) and tripartite motif-containing 40 (TRIM40) promote RIG-I degradation by inducing its K48-linked ubiquitination, ultimately inhibiting IFN production (19,20). Furthermore, RIG-I ubiquitination is modulated by various host factors. For instance, Src homology 3 domain-containing kinase-binding protein 1 (SH3KBP1) and zinc finger CCHC-type containing 3 (ZCCHC3) induce K63-linked polyubiquitination of RIG-I by improving its interaction with TRIM25 (21,22). Moreover, Ras-GTPase-activating protein (GAP)-binding protein 1 (G3BP1) interacts with RNF125 to promote its degradation, thereby inhibiting RNF125-mediated RIG-I degradation (23).
Previous studies have demonstrated that PRRSV develops various strategies to evade host innate immune defenses by disrupting the IFN system (24). For instance, PRRSV nsp1α inhibits IFN production by interacting with TRIM25, thereby preventing RIG-I activation (25). Furthermore, nsp4 blocks the NF-κB signaling pathway by cleaving the NF-κB essential modulator (NEMO). Nsp11 inhibits the formation and nuclear transloca tion of interferon-stimulated gene factor 3 (ISGF3) by targeting interferon regulatory factor 9 (IRF9) (26). Moreover, nsp11 degrades ISG15 through endoribonuclease activity (27). The N protein interferes with RIG-I activation by targeting TRIM25 and inhibits IRF3 activation by preventing its phosphorylation and nuclear translocation (28). Recent studies have identified that PRRSV GPs also play a role in immune evasion. Specifically, GP3 inhibits IFN production by blocking phosphorylation of TBK1 and IRF3 (29). GP5 inhibits chaperone-mediated autophagy by blocking K63-linked polyubiquitination of lysosome-associated membrane GP 2 (LAMP2A), thereby suppressing IFN production (30). However, it remains uncertain whether GP2a is involved in PRRSV immune evasion.
This study demonstrated that GP2a induces RIG-I degradation. However, the specific mechanism by which GP2a induces this degradation remains unknown. Therefore, we attempted to address the following questions. (i) Which pathway does GP2a induce RIG-I degradation through? (ii) What mechanism does GP2a induce RIG-I degradation by? (iii) Does GP2a affect RIG-I ubiquitination? Addressing these issues can help us gain a deeper understanding of the immune evasion mechanisms of PRRSV.
## RESULTS
## PRRSV-induced RIG-I degradation
To determine the effect of PRRSV on IFN production and response, we initially assessed whether PRRSV infection influences the expression of IFN-β and ISG15. Marc-145 cells were infected with PRRSV at varying multiplicities of infection (MOIs), and cell samples were collected to measure the mRNA levels of IFN-β and ISG15. PRRSV did not signifi cantly increase the mRNA levels of IFN-β and ISG15 compared with the Poly(I:C) group (Fig. 1A). PRRSV reduced the expression of IFN-β and ISG15 mRNA in cells treated with Poly(I:C) (Fig. 1A). Further analysis was performed to determine the mRNA levels of IFN-β and ISG15 at different time points post-infection (hpi). PRRSV significantly inhibited the expression of IFN-β and ISG15 mRNA at 36 and 48 hpi (Fig. 1B). To further validate the results, we measured IFN-β levels in the culture supernatant using an enzyme-linked (Fig. 1D). To validate this finding, Marc-145 cells were infected with PRRSV at various MOIs. The degradation of RIG-I induced by PRRSV became more pronounced with an increase in the infection dose (Fig. 1E). These findings suggested that PRRSV inhibits RIG-I expression to block IFN production.
Previous studies have reported that PRRSV nsp1α, nsp2, nsp5, nsp11, and N inhibit IFN production by targeting RIG-I. However, it remains unclear whether PRRSV GPs are involved in this antagonistic function. HEK-293T cells were transfected with plasmids encoding different GPs. Cells transfected with plasmids encoding N proteins served as controls. Western blot analysis revealed that GP2a significantly inhibited RIG-I expression (Fig. 1F). The effect of PRRSV structural proteins on IFN-β expression was detected using ELISA. The results showed that GP2a treatment significantly inhibited IFN-β expression (Fig. 1G). To verify this result, we evaluated the effects of PRRSV GPs on IFN-β and ISG15 mRNA expression levels. GP2a inhibited IFN-β and ISG15 mRNA expression stimulated by Poly(I:C) (Fig. 1H andI). These findings suggest that PRRSV GP2a plays a crucial role in inhibiting IFN production.
## PRRSV GP2a induced RIG-I degradation through the Ub proteasome pathway
To elucidate the mechanism by which PRRSV GP2a inhibits IFN production, we investi gated the effect of GP2a on the activation of the IFN-β promoter. GP2a inhibited the activation of the IFN-β promoter induced by Poly(I:C) or Sendai virus (SeV) (Fig. 2A andB). Notably, GP2a inhibited the activation of the IFN-β promoter induced by RIG-I, suggesting that GP2a may impede the IFN production through targeting RIG-I (Fig. 2C andD). Subsequent analyses revealed that GP2a induced RIG-I degradation and inhibited phosphorylation of IRF3, IRF7, and NF-κB (Fig. 2E). Similarly, GP2a also induced the degradation of RIG-I in Marc-145 cells (Fig. 2F). To verify these findings, HEK-293T cells were transfected with varying doses of a plasmid encoding GP2a. The results demonstra ted that GP2a significantly induced RIG-I degradation in a dose-dependent manner (Fig. 2G). Furthermore, we detected the IFN-β production in the supernatants by analyzing the infection efficiency of vesicular stomatitis virus green fluorescent protein (VSV-GFP). In the empty plasmid + Poly(I:C) group, Poly(I:C) effectively stimulated the cells to produce substantial amounts of IFN-β, resulting in a marked inhibition of VSV-GFP infection. In the Poly(I:C) + GP2a group, GP2a induced the degradation of RIG-I, thereby impairing the ability of Poly(I:C) to induce IFN-β production. The results showed that the infection efficiency of VSV-GFP was enhanced (Fig. 2H).
## PRRSV GP2a promoted K48-linked ubiquitination of RIG-I and inhibited K63linked ubiquitination of RIG-I
To investigate the impact of GP2a on the downregulation of RIG-I expression, HEK-293T cells were treated with cycloheximide (CHX) to inhibit RIG-I production. The RIG-I expression was decreased in GP2a-transfected cells, suggesting that GP2a plays a role in inducing the RIG-I degradation process (Fig. 3A). We then detected the GP2a-mediated RIG-I degradation pathway. Cells were treated with MG132 (a proteasome pathway inhibitor), NH 4 Cl (an autophagy pathway inhibitor), or CQ (an apoptosis pathway inhibitor). GP2a consistently induced RIG-I degradation in cells treated with NH 4 Cl and CQ (Fig. 3B andC). However, this degradation was attenuated in cells treated with MG132, implying that MG132 effectively antagonized GP2a-induced RIG-I degradation (Fig. 3D). These findings indicate that GP2a induces RIG-I degradation via the proteaso mal pathway.
K63-linked ubiquitination is essential for the activation of RIG-I, whereas K48-linked ubiquitination leads to the proteasomal degradation of RIG-I (31,32). We performed coimmunoprecipitation (Co-IP) assays to investigate whether PRRSV GP2a promoted RIG-I ubiquitination. As expected, GP2a promoted the interaction between Ub and RIG-I, suggesting that GP2a increased the ubiquitination of RIG-I (Fig. 3E andF). Subsequently, we identified the type of ubiquitination of RIG-I induced by GP2a. GP2a promoted the interaction between Ub-K48O and RIG-I, indicating that GP2a facilitates the K48-linked ubiquitination of RIG-I (Fig. 3G). Furthermore, GP2a inhibited the interaction between Ub-K63O and RIG-I, indicating that GP2a suppresses K63-linked ubiquitination of RIG-I (Fig. 3H). To validate these findings, we assessed the interaction between RIG-I and the Ub mutants (Ub-K48R and Ub-K63R). GP2a induced the interaction between Ub-K63R and RIG-I and inhibited the interaction between Ub-K48R and RIG-I (Fig. 3I andJ). These results demonstrated that GP2a induces RIG-I degradation by promoting K48-linked ubiquitination and prevents RIG-I activation by inhibiting K63-linked ubiquitination.
## PRRSV GP2a affected the interaction between RIG-I and E3 Ub ligases
Previous studies have demonstrated that TRIM25 plays a crucial role in the K63-linked ubiquitination of RIG-I, whereas TRIM40 and RNF125 act as negative regulatory factors that promote the K48-linked ubiquitination of RIG-I (Fig. 4A) (18)(19)(20). To further elucidate the mechanism by which GP2a interferes with RIG-I ubiquitination, we examined whether GP2a blocked the interaction between RIG-I and these E3 Ub ligases. Immuno precipitation experiments revealed that GP2a blocked the interaction between RIG-I and TRIM25, implying that GP2a inhibited TRIM25-mediated K63-linked ubiquitination of RIG-I (Fig. 4B). Furthermore, GP2a facilitated RNF125-mediated K48-linked ubiquitination of RIG-I by promoting the interaction between RIG-I and RNF125 (Fig. 4B). Subsequent studies demonstrated that GP2a does not interact with E3 Ub ligases or RIG-I (Fig. 4C).
Western blot analysis revealed that GP2a did not influence the endogenous expression of TRIM25, TRIM40, or RNF125 (Fig. 4D). In summary, these findings suggest that GP2a disrupts TRIM25-mediated ubiquitination and promotes RNF125-mediated RIG-I ubiquitination.
To further substantiate the hypothesis that GP2a hijacks RNF125 to promote RIG-I degradation, we synthesized small interfering RNAs (siRNAs) to downregulate RNF125 expression. The results showed that GP2a reduced RIG-I expression by 60.8% in the Dulbecco's modified Eagle medium (DMEM) group and 62.5% in the mock siRNA group, whereas GP2a only decreased RIG-I expression by 38.7% and 52.3% in the RNF125-siRNA1 and RNF125-siRNA2 groups, respectively (Fig. 4E). These findings indicated that GP2a induced the degradation of RIG-I through hijacking RNF125.
## PRRSV GP2a interacted with ZCCHC3
RIG-I ubiquitination is modulated by various host factors. We speculate that GP2a hijacks unidentified regulatory factors that influence E3 Ub ligases. To identify the potential host factors involved in this degradation process, we performed liquid chromatographytandem mass spectrometry analysis (Fig. 5A). This analysis identified 228 potential factors that could bind to GP2a. Gene ontology (GO) enrichment analysis indicated a significant enrichment of genes within the RIG-I-like receptor (RLR) signaling pathway, suggesting that PRRSV GP2a blocks the RIG-I signaling pathway to suppress IFN production (Fig. 5B). Subsequently, we compared the mass spectrometry (MS) results with various databases to identify the potential host factors implicated in GP2a-mediated RIG-I degradation. The BioGRID database was used to screen proteins that interact with RIG-I through biological analysis. The Ubibrowser database was used to screen Ub ligases involved in RIG-I ubiquitination and deubiquitination. Additionally, the proteins reported in the literature to interact with RIG-I were systematically screened utilizing the UniProt database. A A-D) The cells were collected to detect IFN-β promoter activity using a dual-luciferase reporter assay. The experiment was conducted three times, and the data are shown as mean ± SD from triplicate wells in the same experiment (one-way ANOVA; *, P < 0.05). (E) HEK-293T cells were transfected with pCAGGS-3×Flag-GP2a. The cells were collected to detect the expression of RIG-I, IRF3, phosphorylated IRF3, NF-κB, phosphorylated NF-κB, IRF7, phosphorylated IRF7, GAPDH, and N proteins using Western blot assay. (F) Marc-145 cells were transfected with pCAGGS-3×Flag-GP2a. (G) HEK-293T cells were transfected with pCAGGS-3×Flag-GP2a at 250, 500, or 1,000 ng. (F and G) The cells were collected to detect RIG-I, GAPDH, and Flag using Western blot assay. (H).
HEK-293T cells were transfected with pCAGGS-3×Flag-GP2a and treated with Poly(I:C). Cell supernatants were collected and added to new plates. Fluorescence was observed after infecting the cells with VSV-GFP. (E-G) Relative band intensities were quantified using ImageJ software. Data are shown as mean ± SD from three independent experiments (one-way ANOVA; *, P < 0.05; **, P < 0.01; ***, P < 0.001). ns, not significant. comparative analysis between the mass spectrometry results and the BioGRID database reveals an overlap of 45 proteins (Fig. 5C). These 45 proteins were further cross-refer enced with the UniProt database, resulting in the identification of five proteins (Fig. 5C). Co-IP assays demonstrated that GP2a interacts with ZCCHC3 (Fig. 5D). In addition, none of the results of the mass spectrometry analysis overlapped with the Ubibrowser database, suggesting that GP2a does not directly interact with the Ub ligases (Fig. 5C). Next, we investigated the effect of ZCCHC3 on RIG-I. ZCCHC3 promotes the interac tion between RIG-I and Ub(K63O), suggesting that ZCCHC3 induces the activation of RIG-I by promoting K63-linked ubiquitination of RIG-I (Fig. 5E). Notably, ZCCHC3 upregulated RIG-I mRNA and induced RIG-I expression (Fig. 5F andG). In conclusion, these results indicate that GP2a targeted RIG-I by hijacking ZCCHC3.
## PRRSV GP2a hijacked ZCCHC3 to inhibit RIG-I activation
First, we investigated the specific functions of ZCCHC3 in RIG-I activation. Our results indicated that ZCCHC3 did not affect the expression of TRIM25, TRIM40, or RNF125 (Fig. 6A). Subsequent analyses revealed that ZCCHC3 improved the interaction between TRIM25 and RIG-I, implying that ZCCHC3 facilitated TRIM25-mediated ubiquitination of RIG-I (Fig. 6B). Furthermore, ZCCHC3 did not alter the interactions between TRIM40, RNF125, and RIG-I, suggesting that it does not affect TRIM40-and RNF125-mediated RIG-I degradation (Fig. 6B). Notably, GP2a-mediated RIG-I degradation was reduced in the ZCCHC3 overexpressing cells, underscoring the critical role of ZCCHC3 in this degrada tion pathway (Fig. 6A).
Subsequently, we examined the inhibitory effect of GP2a on ZCCHC3-mediated RIG-I activation. Our results demonstrate that GP2a did not suppress ZCCHC3 expression (Fig. 6C). Immunoprecipitation assays revealed that ZCCHC3 interacted with TRIM25 and RIG-I, whereas GP2a disrupted this interaction (Fig. 6D). These findings indicated that GP2a is associated with ZCCHC3, which impedes the formation of the ZCCHC3-TRIM25-RIG-I complex. Further analysis revealed that GP2a blocks the ZCCHC3-promoted interaction between TRIM25 and RIG-I. (Fig. 6E). In summary, these results indicate that GP2a inhibits TRIM25-mediated RIG-I activation by targeting ZCCHC3.
## PRRSV GP2a hijacked ZCCHC3 to inhibit RIG-I production
Subsequently, we investigated the potential inhibitory effects of GP2a on ZCCHC3mediated RIG-I production. To exclude the possibility of RIG-I degradation due to ZCCHC3 knockdown, CHX was administered to the cells, and samples were collected at various time points. The rate of RIG-I degradation in ZCCHC3 knockdown cells was comparable to that of the negative control, indicating that ZCCHC3 knockdown does not contribute to RIG-I degradation (Fig. 7A). Notably, ZCCHC3 knockdown leads to a decrease in RIG-I expression, implying that ZCCHC3 is crucial for RIG-I production (Fig. 7A).
The inhibition of K63-linked ubiquitination of RIG-I effectively impedes the RLR signaling pathway, thereby suppressing the production of ISGs and ultimately diminish ing RIG-I production. We subsequently explored whether GP2a suppressed RIG-I production by interacting with ZCCHC3. Notably, in the DMEM and mock siRNA groups, GP2a reduced RIG-I expression by 64.1% and 62.7%, respectively. Conversely, in ZCCHC3-siRNA1 and ZCCHC3-siRNA2 groups, GP2a decreased RIG-I expression by only 45.8% and 44.7%, respectively (Fig. 7B). These findings suggest that GP2a inhibits RIG-I production, whereas ZCCHC3 knockdown impairs the ability of GP2a to suppress RIG-I production.
We hypothesized that ZCCHC3 induces RIG-I production by facilitating TRIM25mediated RIG-I activation, whereas GP2a hijacks ZCCHC3 to inhibit RIG-I production. To verify this hypothesis, we synthesized siRNAs to downregulate TRIM25 expression. GP2a reduced RIG-I expression by 45.5% and 49.2% in the DMEM and mock siRNA groups, respectively, whereas GP2a only decreased RIG-I expression by 38.1% and 38.5% in the TRIM25-siRNA1 and TRIM25-siRNA2 groups, respectively (Fig. 7C). These findings indicated that GP2a hijacked ZCCHC3 to inhibit TRIM25-mediated RIG-I activation, thereby preventing RIG-I production. Furthermore, immunoprecipitation experiments were performed to exclude the possibility that the knockdown of TRIM25 and ZCCHC3 led to RNF125-mediated RIG-I degradation. Knockdown of TRIM25 and ZCCHC3 did not enhance the interaction between RNF125 and RIG-I. This result suggests that the function of GP2a hijacking of ZCCHC3 to block RIG-I activation and production is not associated with its promotion of RNF125-mediated RIG-I degradation (Fig. 7D andE).
## Identifying the key amino acid sites of GP2a that induce RIG-I degradation
The predicted structures of GP2a and ZCCHC3 were generated using AlphaFold to identify the critical amino acids in GP2a that facilitate its interaction with ZCCHC3. Protein-protein docking was performed using the Docking Web Server (GRAMM), and the interaction between GP2a and ZCCHC3 was predicted and visualized using PyMOL. The results indicated that Tyr59, Arg121, Val229, and Arg239 were essential residues for the binding of GP2a to ZCCHC3 (Fig. 8A). Sequence alignment analysis revealed that these amino acids were conserved across different strains, indicating that the function of GP2a in interacting with ZCCHC3 is conserved (Fig. 8B). To confirm this result, we analyzed whether the GP2a proteins of different PRRSV strains had the same effect and found that they all inhibit RIG-I expression (Fig. 8C). To evaluate the functional signifi cance of these conserved amino acid sites, they were mutated to alanine to generate a plasmid encoding GP2a mut . As expected, GP2a mut did not inhibit IFN-β and ISG15 mRNA expression stimulated by Poly(I:C) (Fig. 8D). The ability of GP2a mut to inhibit RIG-I production was weakened (Fig. 8E). GP2a mut did not promote RIG-I ubiquitination (Fig. 8F). Immunoprecipitation experiments revealed that GP2a mut did not interact with ZCCHC3 (Fig. 8G). Collectively, these findings underscore the significance of Tyr59, Arg121, Val229, and Arg239 as crucial residues for the hijacking of ZCCHC3 by GP2a.
## GP2a inhibits the function of ZCCHC3 in suppressing PRRSV replication
We investigated the effect of ZCCHC3 on IFN activity. The results demonstrate that ZCCHC3 overexpression significantly promoted the production of IFN-β, whereas the levels of IFN-β were reduced in ZCCHC3 knockdown cells (Fig. 9A). We further investiga ted the role of ZCCHC3 in PRRSV replication. DDX3X, a known facilitator of PRRSV replication, was used as the positive control (33). DDX3X increased the PRRSV titer during the early stages of infection, whereas ZCCHC3 significantly decreased the PRRSV titer (Fig. 9B). Notably, GP2a inhibited the ability of ZCCHC3 to suppress PRRSV replication (Fig. 9B). Our findings were further validated using the immunofluorescence assay (IFA) results (Fig. 9C). Moreover, ZCCHC3 inhibited PRRSV-induced suppression of the IFN response. This inhibition was diminished in the cells overexpressing GP2a (Fig. 9D). In summary, our findings suggest that ZCCHC3 acts as an antiviral factor against PRRSV, whereas GP2a counteracts this antiviral function.
## DISCUSSION
Viruses have developed sophisticated strategies to evade innate immune responses in the host. One of the most efficient strategies is the inhibition of IFN production by targeting RIG-I. Human papillomavirus E6 protein promotes K48-linked ubiquitination of TRIM25, leading to its degradation and inhibition of TRIM25-mediated activation of RIG-I (34). Nonstructural protein 5 of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) cleaves RIG-I to disrupt the formation of the RIG-I-MAVS complex (35).
Besides, the SARS-CoV-2 N protein interacts with G3BP1 to block the G3BP1-mediated degradation of RNF125, thereby inducing the degradation of RIG-I (35). The foot-andmouth disease virus 2B protein recruits RNF125 to promote the degradation of RIG-I (36).
Similarly, PRRSV has evolved strategies to disrupt the IFN system by targeting RIG-I. Nsp1α and N protein hijack TRIM25 to inhibit TRIM25-mediated activation of RIG-I (25,28). Nsp2 interacts with SH3KBP1 to induce autophagic degradation, ultimately inhibit ing the K63-linked ubiquitination of RIG-I ( 21). Nsp11 inhibits the expression of RIG-I (37).
In this study, we observed that PRRSV significantly induced RIG-I degradation, with GP2a playing a critical role in this process. Further studies have revealed that GP2a mediates RIG-I degradation via the Ub-proteasome pathway. Specifically, GP2a enhances the K48linked ubiquitination of RIG-I by promoting the interaction between RIG-I and RNF125. Notably, GP2a did not directly bind to RNF125, implying that GP2a modulates the interaction between RNF125 and RIG-I through indirect mechanisms. It is imperative to elucidate the specific mechanisms involved in the GP2a-induced RIG-I degradation. RIG-I activation is dependent on TRIM25-mediated K63-linked ubiquitination, which is crucial for initiating the innate immune response (20). In this study, we observed that ZCCHC3 recruits TRIM25 to facilitate the formation of the TRIM25-RIG-I complex, thereby increasing the K63-linked ubiquitination of RIG-I. Furthermore, ZCCHC3 did not affect the interaction between RIG-I and either RNF125 or TRIM40, indicating that ZCCHC3 did not interfere with RIG-I degradation. PRRSV GP2a did not affect ZCCHC3 expression. Subse quent studies have revealed that GP2a hijacked ZCCHC3 to inhibit RIG-I activation. Specifically, GP2a interacted with ZCCHC3 to disrupt the interaction between ZCCHC3 and TRIM25. This disruption inhibits TRIM25-mediated K63-linked ubiquitination of RIG-I, thereby preventing RIG-I activation.
IFN-β is believed to facilitate RIG-I production through a positive feedback mecha nism (18,38). Numerous studies have highlighted the importance of this feedback loop. The upregulation of IFN-induced RIG-I expression subsequently induces the activation of antiviral proteins and the production of IFN (38,39). This regulatory mechanism ensures that upon detection of viral infection using RIG-I, the IFN system is rapidly activated to inhibit viral propagation (38,39). In this study, we observed that the knockdown of TRIM25 and ZCCHC3 resulted in a reduction in RIG-I production, indicating that K63linked ubiquitination is crucial for RIG-I production. ZCCHC3 induces IFN production by enhancing TRIM25-mediated K63-linked ubiquitination of RIG-I. Subsequently, IFN activates the JAK-STAT signaling pathway to induce the expression of ISGs, ultimately promoting RIG-I production. PRRSV GP2a hijacked ZCCHC3 to inhibit K63-linked ubiquitination of RIG-I, disrupting the RIG-I positive feedback loop and consequently inhibiting RIG-I production. Collectively, these findings suggested that GP2a hijacked ZCCHC3 to reduce RIG-I production.
PRRSV is characterized by its high genetic variability, and the GP2a sequence exhibits divergence among various PRRSV strains (40). In this study, we used GRAMM software to conduct docking analyses between GP2a and ZCCHC3, identifying Tyr59, Arg121, Val229, and Arg239 as critical residues for the binding interaction. Notably, these key amino acid residues were conserved across different PRRSV strains, indicating that NADC30-like, classical, and highly pathogenic strains of PRRSV GP2a play similar roles in RIG-I degrada tion. Additionally, ZCCHC3 significantly enhanced IFN production, thereby inhibiting PRRSV replication. ZCCHC3 overexpression counteracted PRRSV-induced reduction in IFN-β and ISG15 expression, whereas GP2a diminished this function. These results indicate that ZCCHC3 functions as an antiviral factor against PRRSV, while GP2a counter acts this antiviral activity.
In conclusion, this study demonstrates that PRRSV GP2a inhibits RIG-I expression through two mechanisms: (i) PRRSV GP2a facilitates the K48-linked ubiquitination of RIG-I by improving the interaction between RIG-I and RNF125, thereby promoting the degradation of RIG-I, and (ii) GP2a hijacks ZCCHC3 to inhibit TRIM25-mediated activation of RIG-I, thereby suppressing IFN production and ultimately inhibiting RIG-I production (Fig. 10). This study fills a gap in our understanding of the mechanisms by which PRRSV GPs block the IFN system. It is imperative to understand these mechanisms at the molecular level to develop effective strategies to combat PRRSV infections.
## MATERIALS AND METHODS
## Virus, cells, and plasmids
The PRRSV-JX02 strain, PRRSV-JS strain, PRRSV-SD strain, PRRSV-YN strain, VSV-GFP, and SeV were stored in our laboratory. HEK-293T, pulmonary alveolar macrophage (PAM), and African green monkey kidney epithelial cell line (Marc-145) cells were cultured in DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin and incubated at 37°C with 5% CO 2 . Full-length cDNAs encoding PRRSV structural proteins were amplified from the PRRSV cDNA and cloned into the pCAGGS-3×Flag vector. Fulllength cDNAs encoding ZCCHC3 and DDX3X were amplified from HEK-293T cDNA and cloned into the pCAGGS-2×HA vector. The pIFN-β-luc, pRL-TK, and plasmids expressing Ub were stored in our laboratory.
## Antibodies and reagents
The following antibodies were used in this study: RIG-I/DDX58 rabbit mAb (A22478), MDA5 rabbit mAb (A2419), NF-κB p65/RelA rabbit mAb (A19653), phospho-NF-κB p65/ RelA-S536 rabbit mAb (AP1294), IRF7 rabbit pAb (A0159), IRF3 rabbit mAb (A19717), phospho-IRF3-S396 rabbit mAb (AP1412), DDDDK-tag rabbit mAb (AE092), GAPDH mouse mAb (AC002), ZCCHC3 rabbit pAb (A17235), TRIM25 rabbit mAb (A25846), HRPconjugated rabbit antigoat IgG (H + L) (AS029), and HRP-conjugated goat anti-mouse IgG (H + L) (AS003) from ABclonal Biotechnology; phospho-IRF-7 (Ser471/472) antibody (5184) from CST; and TRIM40 monoclonal antibody (67073) and RNF125 polyclonal antibody (13290) from Proteintech. Dylight-conjugated 488 goat anti-mouse IgG (A23210) was obtained from Abbkine. The PRRSV-N mouse polyclonal antibodies were kept in our laboratory. RIPA lysis buffer (P0013D), MG132 (Y210207), NH4Cl (ST2030), and Z-VAD-FMK (Y062551) were obtained from Beyotime. CHX710 (HY-112951) was procured from MedChemExpress. Poly(I:C) LMW (31852-29-6) was obtained from InvivoGen. Lipofectamine 3000 (L3000015) was obtained from Thermo Fisher Scientific. Protein A/G magnetic beads (B23202) were obtained from SelleckChem. Anti-Flag M2 magnetic beads (M8823) and anti-HA magnetic beads (SAE0197) were obtained from Sigma.
## ELISA assay
For the viral infection experiment, Marc-145 cells were seeded in 24-well plates over night. The cells were infected with PRRSV at an MOI of 0.02. The medium was replaced at 2 hpi, and the supernatants were collected at 0, 12, 24, 36, 48, 60, and 72 hpi. For the plasmid transfection experiment, HEK-293T cells were cultured overnight in 12-well plates. The cells were transfected with plasmids expressing PRRSV structural proteins at a concentration of 1,000 ng/well. After 24 h, the cells were incubated with Poly(I:C) (1,000 ng/mL) for 12 h. Supernatants were then collected, and ELISA assays were performed using an ELISA kit (BOSTER, EK2286) according to the manufacturer's instructions.
## Virus growth kinetics
The Marc-145 cells were cultured overnight in 12-well plates. The cells were then transfected with pCAGGS-3×Flag (Mock group), pCAGGS-2×HA-DDX3X + pCAGGS-3×Flag (DDX3X group), pCAGGS-2×HA-ZCCHC3 + pCAGGS-3×Flag (ZCCHC3 group), and pCAGGS-2×HA-ZCCHC3 + pCAGGS-3×Flag-GP2a (ZCCHC3 + GP2 a group), respectively. At 24 h post-transfection, the cells were infected with PRRSV at an MOI of 0.02. The medium was replaced at 2 hpi, and the supernatants were collected at 0, 12, 24, 32, 48, and 60 hpi. Virus titers were calculated as log10 50% tissue culture infectious dose using the Reed-Muench method.
## Western blot assay
Marc-145 cells were seeded in 12-well plates and incubated overnight for viral infec tion experiments. The cells were infected with PRRSV at an MOI of 0.1. After 48 h, cell samples were lysed, sonicated, and centrifuged. Subsequently, the samples were combined with the loading buffer and heated for 10 min. The cell samples were then subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes (Millipore). Membranes were blocked with 7.5% skim milk in Trisbuffered saline with 0.1% Tween 20 detergent and incubated with primary antibodies overnight at 4°C. Following this, the membranes were incubated with secondary antibodies for 2 h at room temperature and treated with enhanced chemiluminescence (Thermo Fisher Scientific). The protein bands were visualized using the Tanon imaging system (Biotanon) and analyzed using the ImageJ software. For the plasmid transfection experiment, HEK-293T cells were seeded in six-well plates overnight. The cells were then transfected with plasmids. After 24 h, the cells were stimulated with SeV or Poly(I:C) for 12 h. Cell samples were then collected, and subse quent procedures were conducted as described previously.
## VSV-GFP interferon bioassay
HEK-293T cells were cultured overnight in 12-well plates. The cells were then transfec ted with pCAGGS-3×Flag-GP2a. At 24 h, the cells were stimulated with Poly(I:C) for 12 h. Supernatants from the transfected cells were diluted twofold and added to fresh HEK-293T cells. The cells were then infected with VSV-GFP (MOI = 0.1), and expression was examined using inverted fluorescence microscopy.
## Co-IP assay
HEK-293T cells were seeded in six-well plates overnight and transfected with plasmids. After 24 h, the cells were stimulated with Poly(I:C) for 12 h. Cell lysates were collected using RIPA lysis buffer and centrifuged. Subsequently, the samples were incubated with anti-Flag M2 or anti-HA magnetic beads. The beads were washed thrice with RIPA lysis buffer and then treated with SDS loading buffer. The samples were then ana lyzed using Western blot assay. For the endogenous Co-IP experiments, HEK-293T cells
## Luciferase reporter gene assay
HEK-293T cells were seeded into 24-well plates and incubated overnight. The cells were then co-transfected with plasmids expressing RIG-I/MAVS (500 ng), pIFN-β-luc (100 ng), pRL-TK (25 ng), or pCAGGS-3×Flag-GP2a (200 ng). At 24 h post-transfection, the cells were collected to detect promoter activity using a dual-luciferase reporter assay. Data are expressed as the mean ± standard deviation from three independent experiments.
## Relative quantitative real-time PCR
For the viral infection experiment, Marc-145 cells were cultured overnight in 12-well plates. The cells were infected with PRRSV at an MOI of 0.1 or 0.5 for 36 h and at an MOI of 0.1 for 0, 36, or 48 h. The cells were then collected. RNA was extracted from the samples using an RNA Easy Fast Tissue/Cell Kit, treated with DNase I to remove genomic DNA, and converted to cDNA via reverse transcription PCR. IFN-β and ISG15 mRNA levels were analyzed using relative quantitative real-time PCR (RT-qPCR) with the LineGene9600 RT-PCR system. To determine whether ZCCHC3 inhibits PRRSVmediated suppression of the IFN response, Marc-145 cells were cultured in 12-well plates overnight. The cells were then transfected with pCAGGS-3×Flag (Mock group), pCAGGS-2×HA-DDX3X + pCAGGS-3×Flag (DDX3X group), pCAGGS-2×HA-ZCCHC3 + pCAGGS-3×Flag (ZCCHC3 group), and pCAGGS-2×HA-ZCCHC3 + pCAGGS-3×Flag-GP2a (ZCCHC3 + GP2 a group), respectively. At 24 h post-transfection, the cells were infec ted with PRRSV at an MOI of 0.1. The cell samples were then collected, and subse quent procedures were conducted as described previously. In the plasmid transfection experiment, HEK-293T cells were cultured in 12-well plates overnight. The cells were transfected with plasmids expressing PRRSV structural proteins at a concentration of 1,000 ng/well. After 24 h, the cells were incubated with Poly(I:C) (1,000 ng/mL) for 12 h. Cell samples were then collected, and subsequent procedures were performed as described previously.
## RNA interference experiments
To exclude the possibility of RIG-I degradation due to ZCCHC3 knockdown, HEK-293T cells were cultured overnight in 12-well plates. Cells were transfected with siRNA at a concentration of 30 pm/well. After 48 h, the cells were treated with CHX. Cells were collected for Western blot analysis. For ZCCHC3 and TRIM25 knockdown experi ments, HEK-293T cells were cultured in 12-well plates overnight. Cells were transfected with siRNA at a concentration of 30 pm/well. After 24 h, cells were transfected with pCAGGS-3×Flag-GP2a at 2,500 ng/well. After 24 h, the cells were stimulated with Poly(I:C) for 12 h and collected for Western blot analysis.
## Molecular docking experiment
GP2a and ZCCHC3 structures were predicted using Alphafold (version 3.0, https:// alphafold.com/) and prepared using AutoDockTools (version 1.5.7, https://autodock suite.scripps.edu/adt/) by removing water and adding polar hydrogen. GRAMM (https:// gramm.compbio.ku.edu/) was used for protein-protein docking, and the resulting complex was optimized similarly. PyMOL (https://pymol.org/) was used to predict interactions and create a visualization diagram: GP2a is presented as a slate cartoon, whereas ZCCHC3 is presented as a cyan cartoon, and the binding sites are depicted in matching stick structures.
## MS analysis
Protein digestion was performed according to the filteraided sample preparation procedure. Experiments were performed using a Q Exactive HF-X mass spectrometer coupled to an Easy nLC (Thermo Fisher Scientific). MS data were analyzed using MaxQuant software (version 1.3.0.5). The data were searched against the UniProtKB Homo sapiens database. The initial search was performed using a precursor mass window of 6 ppm. This search followed the enzymatic cleavage rule for trypsin. Carbamidomethyl (C) was defined as a fixed modification, whereas oxidation (M) and phosphorylation (S/T/Y) were defined as variable modifications for database searching. The cutoff global false discovery rate for peptide and protein identification was 0.01.
## IFA
Marc-145 and PAM cells were infected with PRRSV. The cells were washed thrice using phosphatebuffered saline, fixed in 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 2% bovine serum albumin. The cells were incubated overnight with the primary antibody at 4°C and then incubated for 2 h with the secondary antibody at 37°C. All statistical experiments were performed in triplicate.
## Statistical analysis
Statistical analyses were conducted using Student's t-test to analyze the two groups of data. Furthermore, statistical analyses were conducted using one-way analysis of variance (ANOVA) to analyze the multiple groups of data. Differences were considered statistically significant at *P < 0.05, **P < 0.01, and ***P < 0.001.
## References
1. "After 24 h, the cells were stimulated with Poly(I:C) for 12 h. Cell samples were collected using RIPA lysis buffer, centrifuged, and incubated with anti-RIG-I mAb or anti-ZCCHC3 mAb"
2. "HEK-293T cells were co-transfected with siRNA or pCAGGS-2×HA-ZCCHC3, pIFN-β-luc, and pRL-TK. The cells were collected to detect IFN-β promoter activity using a dual-luciferase reporter assay. (B) Marc-145 cells were co-transfected with pCAGGS-3×Flag-GP2a, pCAGGS-2×HA-ZCCHC3, and pCAGGS-2×HA-DDX3X. The cell supernatants were collected to detect the TCID 50 . The experiment was conducted three times. (C) Marc-145 and PAM cells were infected with PRRSV at an MOI of 0.1 for 36 h. The cells were stained with PRRSV N mouse-pAb, followed by Alexa Fluor 488-conjugated goat antimouse staining (green). The nuclei were stained with 4′,6-diamidino-2-phenylindole (blue). (D) Marc-145 cells were co-transfected with pCAGGS-3×Flag-GP2a and pCAGGS-2×HA-ZCCHC3. The cells were infected with PRRSV and collected to detect the IFN-β mRNA level using RT-qPCR. (A and D) The experiment was conducted three times, and the data are shown as mean ± SD from triplicate wells in the same experiment"
3. (2025) *Full-Length Text Journal of Virology*
4. Lunney, Fang, Ladinig et al. (2016) "Porcine reproductive and respiratory syndrome virus (PRRSV): pathogenesis and interaction with the immune system" *Annu Rev Anim Biosci*
5. Zhao, Xu, Xu et al. (2022) "Emergence and spread of NADC34-like PRRSV in southwest China" *Transbounding Emerging Dis*
6. Chen, Ye, Li et al. (2017) "Emergence of a novel highly pathogenic recombinant virus from three lineages of porcine reproductive and respiratory syndrome virus 2 in China" *Transbound Emerg Dis*
7. Kappes, Faaberg (2015) "PRRSV structure, replication and recombination: Origin of phenotype and genotype diversity" *Virology (Auckl)*
8. Stoian, Rowland, Brandariz-Nuñez (2022) "Identification of CD163 regions that are required for porcine reproductive and respiratory syndrome virus (PRRSV) infection but not for binding to viral envelope glycoproteins" *Virology (Auckl)*
9. Jiang, Xiao, Fang et al. (2006) "DNA vaccines co-expressing GP5 and M proteins of porcine reproductive and respiratory syndrome virus (PRRSV) display enhanced immunogenicity" *Vaccine (Auckl)*
10. Zheng, Li, Luo et al. (2024) "Research progress on the N protein of porcine reproductive and respiratory syndrome virus" *Front Microbiol*
11. Xiang, Mou, Shi et al. (2023) "SADS-CoV nsp1 inhibits the IFN-β production by preventing TBK1 phosphorylation and inducing CBP degradation" *J Med Virol*
12. Xiang, Mou, Zhu et al. (2024) "SADS-CoV nsp1 inhibits the STAT1 phosphorylation by promoting K11/ K48-linked polyubiquitination of JAK1 and blocks the STAT1 acetylation by degrading CBP" *J Biol Chem*
13. Rehwinkel, Gack (2020) "RIG-I-like receptors: their regulation and roles in RNA sensing" *Nat Rev Immunol*
14. Cao, Zhou, Zhu et al. (2016) "Pyruvate carboxylase activates the RIG-I-like receptor-mediated antiviral immune response by targeting the MAVS signalosome" *Sci Rep*
15. Wu, Peisley, Tetrault et al. (2014) "Molecular imprinting as a signal-activation mechanism"
16. (2025) *Full-Length Text Journal of Virology*
17. "of the viral RNA sensor RIG-I" *Mol Cell*
19. Jing, Zhao, Xu et al. (2020) "The structural basis of IRF-3 activation upon phosphorylation" *J Immunol*
20. Ma, Huang, Wang et al. (2023) "IRF7: role and regulation in immunity and autoimmunity" *Front Immunol*
21. Hu, Li, Fu et al. (2021) "The JAK/STAT signaling pathway: from bench to clinic" *Sig Transduct Target Ther*
22. Lee, Chiang, Gack (2019) "Endogenous nucleic acid recognition by RIG-I-like receptors and cGAS" *J Interferon Cytokine Res*
23. Sun, Xian, Tian et al. (2016) "A hierarchical mechanism of RIG-I ubiquitination provides sensitivity, robustness and synergy in antiviral immune responses" *Sci Rep*
24. Gack, Shin, Joo et al. (2007) "TRIM25 RINGfinger E3 ubiquitin ligase is essential for RIG-I-mediated antiviral activity" *Nature*
25. Chang, Hou, Wang et al. (2023) "CD97 negatively regulates the innate immune response against RNA viruses by promoting RNF125-mediated RIG-I degradation" *Cell Mol Immunol*
26. Shen, Tang, Chen et al. (2021) "Riok3 inhibits the antiviral immune response by facilitating TRIM40-mediated RIG-I and MDA5 degradation" *Cell Rep*
27. Li, Zhang, Sun et al. (2024) "Porcine reproductive and respiratory syndrome virus nonstructural protein 2 promotes the autophagic degradation of adaptor protein SH3KBP1 to antagonize host innate immune responses by enhancing K63-linked polyubiquitination of RIG-I" *PLoS Pathog*
28. Lian, Zang, Wei et al. (2018) "The zincfinger protein ZCCHC3 binds RNA and facilitates viral RNA sensing and activation of the RIG-Ilike receptors" *Immunity*
29. Yang, Ru, Ren et al. (2019) "G3BP1 inhibits RNA virus replication by positively regulating RIG-I-mediated cellular antiviral response" *Cell Death Dis*
30. Yoo, Song, Sun et al. (2010) "Modulation of host cell responses and evasion strategies for porcine reproductive and respiratory syndrome virus" *Virus Res*
31. Zheng, Jiang, Sui et al. (2024) "PRRSV NSP1α degrades TRIM25 through proteasome system to inhibit host antiviral immune response" *Vet Microbiol*
32. Wang, Chen, Yu et al. (2019) "Porcine reproductive and respiratory syndrome virus nsp11 antagonizes type I interferon signaling by targeting IRF9" *J Virol*
33. Jiang, He, Sui et al. (2023) "PRRSV nonstructural protein 11 degrades swine ISG15 by its endoribonuclease activity to antagonize antiviral immune response" *Vet Microbiol*
34. Zhao, Li, Jiang et al. (2019) "Nucleocapsid protein of porcine reproductive and respiratory syndrome virus antagonizes the antiviral activity of TRIM25 by interfering with TRIM25-mediated RIG-I ubiquitination" *Vet Microbiol*
35. Li, Cui, Li et al. (2025) "IFN-β production induced by PRRSV is affected by GP3 quantity control and CLND4 interaction" *Vet Res*
36. Li, Zhang, Wang et al. (2024) "PRRSV GP5 inhibits the antivirus effects of chaperone-mediated autophagy by targeting LAMP2A"
37. Okamoto, Kouwaki, Fukushima et al. (1942) "Regulation of RIG-I activation by K63-linked polyubiquitination" *Front Immunol*
38. Wang, Jia, Zhao et al. (2021) "RNF39 mediates K48-linked ubiquitination of DDX3X and inhibits RLRdependent antiviral immunity" *Sci Adv*
39. Mao, Ma, Li et al. (2024) "PRRSV hijacks DDX3X protein and induces ferroptosis to facilitate viral replication" *Vet Res*
40. Chiang, Pauli, Biryukov et al. (2018) "The human papillomavirus E6 oncoprotein targets USP15 and TRIM25 to suppress RIG-I-mediated innate immune signaling" *J Virol*
41. Zheng, Deng, Han et al. (2022) "SARS-CoV-2 NSP5 and N protein counteract the RIG-I signaling pathway by suppressing the formation of stress granules" *Signal Transduct Target Ther*
42. Weerawardhana, Uddin, Choi et al. (2022) "Foot-and-mouth disease virus nonstructural protein 2B downregulates the RLR signaling pathway via degradation of RIG-I and MDA5" *Front Immunol*
43. Sun, Ke, Han et al. (2016) "Nonstructural protein 11 of porcine reproductive and respiratory syndrome virus suppresses both MAVS and RIG-I expression as one of the mechanisms to antagonize type I interferon production" *PLoS One*
44. Chen, Fang, Wang et al. (2019) "Porcine deltacoronavirus nucleocapsid protein antagoni zes IFN-β production by impairing dsRNA and PACT binding to RIG-I" *Virus Genes*
45. Zou, Wu, Wang et al. (2021) "5-hydroxymethylfurfural enhances the antiviral immune response in macrophages through the modulation of RIG-I-mediated interferon production and the JAK/STAT signaling pathway" *ACS Omega*
46. Zhou, Han, Yang (2024) "The evolution and diversity of porcine reproductive and respiratory syndrome virus in China" *Vet Microbiol*
47. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12649918&blobtype=pdf | # Calprotectin, Azurocidin, and Interleukin-8: Neutrophil Signatures with Diagnostic and Prognostic Value in Sepsis
Simona Gigliotti, Michele Manno, Francesca Divenuto, Grazia Pavia, Cinzia Peronace, Francesca Trimboli, Concetta Zangari, Valentina Tancrè, Francesca Greco, Manuela Colosimo, Pasquale Minchella, Luigi Principe, Nadia Marascio, Francesca Licata, Aida Bianco, Alessandro Russo, Federico Longhini, Angela Quirino, Giovanni Matera
## Abstract
Background: Sepsis remains a major cause of morbidity and mortality in both developed and limited-resource countries. Despite over a century of research, accurate biomarkers for reliable diagnosis and prognosis in critically ill patients have yet to be established. Methods: This multicenter retrospective observational study aims to evaluate serum levels of Calprotectin, Azurocidin, cytokines, chemokines, procalcitonin (PCT) and C-Reactive Protein (CRP) in 15 healthy volunteers (controls), 15 non-infectious SIRS patients, 92 alive septic patients (Sepsis_A) and 29 dead septic patients (Sepsis_D). Results: Most biomarkers showed significantly higher serum concentrations in septic patients compared with controls, with IL-4 being increased only in the Sepsis_D group. In addition, several markers, including Calprotectin, Azurocidin, IL-6, IL-8, IL-10, TNF-α, and IL-35, were progressively elevated from SIRS to Sepsis_A and Sepsis_D cohorts, reflecting disease severity. All biomarkers showed good diagnostic performance for predicting Gram-negative bacteremia, although their accuracy in discriminating survivors from non-survivors was relatively low. Conclusions: In conclusion, calprotectin, azurocidin, IL-8, TNF-α, and IL-35 may assist clinicians in identifying Gram-negative bacteremia in septic patients; however, their prognostic value appears to be limited.
## 1. Introduction
Sepsis is one of the leading causes of death worldwide. It is still difficult to diagnose the infection quickly and accurately, and the diagnosis of sepsis is often not timely. Although more than 250 biomarkers have been studied over the last years, no biomarkers accurately differentiate between sepsis and sepsis-like syndrome. It is necessary to find one or more biomarkers useful to assist clinicians in the diagnosis and prognosis of sepsis. Herein, we provide current data on the clinical utility of host-response biomarkers, offer guidance on optimizing their use, and propose the need for future research [1].
The lack of a biomarker that serves as a generally accepted gold standard for sepsis is a common limitation of all sepsis studies. Biomarkers can play a key role in the timely diagnosis and management of sepsis. C-reactive protein (CRP) and procalcitonin (PCT) are widely used biomarkers, but their diagnostic accuracy has been questioned [2]. Indeed, they cannot easily distinguish infection from inflammation. Therefore, biomarkers with high diagnostic sensitivity and specificity are indispensable [3]. Azurocidin appears to play an important role in the pathophysiology of severe bacterial infections, thus representing a potential diagnostic marker and target for the treatment of sepsis [4]. Calprotectin is an acute-phase protein released into the circulation by monocytes and neutrophils that is thought to be more sensitive than CRP in detecting minimal residual inflammation [5]. Such polypeptide can stimulate the TLR4 receptor on several types of host cells, and such a feature would initiate an inflammatory cascade, relevant to sepsis pathogenesis [6,7].
Following the pioneering work of Bone et al. [8], which primarily identified pathogenic mediators and sepsis biomarkers of macrophage/monocyte origin, Hotchkiss et al. later shifted the focus to alterations in lymphocyte counts and to the diverse mediators derived from distinct lymphocyte subsets (Th1, Th2, Treg, and Breg) [9].
Neutrophils have been reported to play a major role in the pathogenesis of sepsis, both through the direct release of mediators and by modulating the activity of other immune cells [10]. Therefore, we evaluated mediators released from neutrophil granules (Calprotectin and azurocidin) and neutrophil chemokine (IL-8) to address the critical role of neutrophil mediators as potential sepsis biomarkers. Moreover, several groups of cytokines (Th1, Th2, Treg, Breg) were also evaluated as potential sepsis biomarkers. Also, IL-35, a novel Breg cytokine, has been rarely assessed as a sepsis biomarker. Those few publications, which include IL-35, never focused specifically on the diagnostic or prognostic role of such cytokine in sepsis and on the possibility of being used as a helpful tool to assist clinicians' decisions [11,12].
We therefore designed this multicenter, retrospective observational study to assess serum levels of calprotectin, azurocidin, PCT, CRP, Interleukins and Tumor Necrosis Factor α (TNF-α), in healthy controls, patients with non-infectious Systemic Inflammatory Response Syndrome (SIRS), or sepsis/septic shock with blood cultures positive for Gramnegative bacteria. In addition, we aimed to determine whether these biomarkers possess diagnostic value (for identifying the presence or absence of sepsis) and prognostic value (for predicting survival or not).
## 2. Materials and Methods
The Ethics Committee of the Calabria Region approved the study protocol (approval number 128/2023, on 22 December 2023); given the retrospective study design, written informed consent was waived. All procedures were conducted in accordance with the principles of the Declaration of Helsinki. The anonymized dataset generated and analyzed during the current study is available from the corresponding author upon reasonable request.
This multicenter, retrospective observational study was conducted from February to June 2024. We enrolled only adult participants (i.e., >18 years old), classified into the following groups: (1) 15 healthy volunteers; (2) 15 patients with non-infectious SIRS;
(3) 92 alive septic patients (Sepsis_A), as defined by the Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) [13]; and (4) 29 non-survivor (i.e., died) septic patients (Sepsis_D). Both Sepsis_A and Sepsis_D patients were required to have blood cultures positive for Gram-negative bacteria. Exclusion criteria included malignancies, alcohol or drug abuse, pregnancy, and lactating women.
Non-infectious SIRS, defined as an exaggerated defense response of the body to a noxious stressor (including trauma, surgery, acute inflammation or ischemia/reperfusion), was diagnosed by the presence of two or more of the following clinical criteria in the absence of infection: body temperature > 38 • C or <36 • C; heart rate > 90 beats/min; respiratory rate > 20 breaths/min or arterial PaCO 2 < 32 mmHg; and white blood cell count > 12 × 10 9 /L, < 4 × 10 9 /L, or > 10% immature (band) forms [14].
For all participants, blood samples for biomarker assessment were obtained at hospital admission, corresponding to the time of SIRS or sepsis diagnosis, before the initiation of antibiotic treatment. Mortality was determined at 28 days following SIRS diagnosis or, in the case of sepsis, 28 days after the first positive blood culture result.
## 2.1. Serum Biomarker Assessment
Serum Calprotectin levels were measured using the Calprest NG-S (ECL) ELISA kit (Eurospital Diagnostic, Trieste, Italy). This enzyme immunoassay employs colorimetric detection based on polyclonal and monoclonal antibodies directed against Calprotectin. The immobilized antibody captures Calprotectin from the diluted serum sample within the wells. Subsequently, peroxidase-conjugated (HRP) antibodies bind to the captured Calprotectin, and the enzyme catalyzes the conversion of the substrate into a colored product. The color intensity is proportional to the amount of conjugate bound, and therefore to the Calprotectin concentration. Serum Calprotectin levels were calculated by interpolation from a calibration curve. After centrifugation of the primary collection tube (3000× g for 10 min at room temperature), serum samples were obtained within 4-6 h of collection. A minimum volume of 500 µL was required for analysis. Samples not tested immediately were stored at -20 • C in 0.5-2 mL screw-capped tubes and were not subjected to more than three freeze-thaw cycles. The assay measuring range was 0-150 ng/mL. Intra-and inter-assay coefficients of variation on serum samples were <7% and <12.2%, as per the manufacturer's documentation.
Standard and serum samples were added to microplate wells coated with a biotinconjugated antibody specific for Azurocidin (AZU1) (Enzyme-linked Immunosorbent Assay Kit for AZU1, Cloud-Clone Corp., Katy, TX, USA). Avidin conjugated to horseradish peroxidase (HRP) was then added, and the plate was incubated. After the addition of the TMB substrate solution, only wells containing AZU1 bound to the biotin-conjugated antibody and enzyme-conjugated Avidin developed a color change. The enzyme-substrate reaction was stopped by adding sulfuric acid solution, and the resulting color intensity was measured spectrophotometrically at 450 ± 10 nm. The concentration of AZU1 in each sample was determined by comparing the optical density (O.D.) values to the standard calibration curve. The assay measuring range was 0.312-20 ng/mL. Intra-and inter-assay coefficients of variation were <10% and <12%, as per the manufacturer's documentation.
Clinical biochemistry data regarding soluble serum levels factors were detected by chemiluminescence assay (CLIA) for PCT and CRP [15].
To quantify serum concentrations of IL-4, IL-6, IL-8, IL-10, and TNF-α, samples were analyzed using a biochip array incorporating specific primary antibodies (Cytokine and Growth Factors Array, Randox Laboratories, Crumlin, UK), following the manufacturer's instructions. Briefly, serum samples were incubated on the biochip for 1 h at 37 • C, followed by washing and a subsequent 1 h incubation with a horseradish peroxidase (HRP)-conjugated secondary antibody. Chemiluminescent signals were then detected using the Evidence Investigator biochip analyzer (Randox Laboratories, Crumlin, UK) with the Cytokines Array I and High Sensitivity kit and quantified with the dedicated Randox software [16]. The assay measuring range was 0-450 pg/mL for IL-4, 0-400 pg/mL for IL-6, 0-1450 pg/mL for IL-8, 0-450 pg/mL for IL-10, and 0-600 pg/mL for tumor necrosis factor (TNF)-α. Intra-and inter-assay coefficients of variation were9.5% and 11.8% for IL-4,11.9% and 8.4% for IL-6, 9.4% and 9.2% for IL-8, 5.6% and 6.5% for IL-10, and 7.1% and 6.7% for TNF-α, respectively [16].
Human IL-35 levels were measured using a sandwich ELISA kit (Human IL-35 ELISA Kit, Assay Genie, Dublin, Ireland). The ELISA plate was pre-coated with an antibody specific for human IL-35. Standards and serum samples were added to the wells and allowed to bind to the immobilized antibodies. Subsequently, a biotinylated detection antibody specific for human IL-35 and an avidin-horseradish peroxidase (HRP) conjugate were added sequentially and incubated. After washing to remove unbound components, a substrate solution was added; wells containing the biotinylated detection antibody and avidin-HRP conjugate developed a blue color. The reaction was terminated by adding stop solution, resulting in a color change from blue to yellow. Optical density (OD) was measured spectrophotometrically at 450 nm, and IL-35 concentrations were determined by interpolating the sample OD values from the standard calibration curve. The detection range of the assay was 15.63-1000 pg/mL. Intra-and inter-assay coefficients of variation were <8% and <10%, as per manufacturer's documentation.
All laboratory personnel performing the biomarker analyses were blinded to the patients' clinical status and outcomes. Samples were processed and analyzed according to standardized procedures, and no clinical information was available to the operators at any stage of the laboratory workflow. This approach was adopted to minimize bias and ensure the objectivity of the measurements.
## 2.2. Statistical Analysis
Continuous variables are expressed as mean ± standard deviation (SD) or as median [25th-75th percentile], while categorical variables are presented as percentages. The normality of data distribution was assessed using the Kolmogorov-Smirnov test. Given the non-parametric distribution of the data, the Kruskal-Wallis test was used to assess differences between groups. Multiple pairwise comparisons were performed using the corrected Dunn's test.
Sensitivity, specificity, positive (PLR) and negative (NLR) likelihood ratios, and positive (PPV) and negative (NPV) predictive values were computed for all biomarkers for both diagnostic and prognostic evaluations. Receiver Operating Characteristic (ROC) curves and the corresponding areas under the curve (AUCs) with 95% confidence intervals (95% CI) were calculated using the nonparametric approach described by DeLong et al. [17]. The Youden index (J = max [sensitivity + specificity -1]) was applied to determine the optimal cut-off values. p-values of less than 0.05 were considered statistically significant. Statistical analysis was done using STATA software program, version 18.
## 3. Results
The demographic characteristics of the study population are summarized in Table 1. The cohort comprised 15 healthy volunteers, 15 patients with non-infectious SIRS, and 92 and 29 patients assigned to the Sepsis_A and Sepsis_D groups, respectively. Gender and age were comparable among groups, except for the healthy control group, which consisted of younger individuals and a higher proportion of females (80%) (Table 1). Some baseline imbalances were observed between the SIRS and Sepsis groups, particularly regarding the admission diagnosis and the need for organ support, including mechanical ventilation, vasoactive agents, renal replacement therapy, and/or extracorporeal membrane oxygenation (ECMO). The source of infection and the isolated pathogens were similar between the Sepsis_A and Sepsis_D cohorts; however, SOFA and APACHE II scores were higher in the latter group compared to the former, as expected.
## 3.1. Serum Biomarkers
Table 2 reports the median [IQR] values of the measured serum biomarkers together with the corrected p-values obtained from multiple group comparisons, whereas Figure 1 displays the corresponding boxplots with exact p-values. All biomarkers, except IL-4, were significantly increased in the Sepsis_A and Sepsis_D cohorts compared with Controls. Specifically, IL-4 levels were higher only in Sepsis_D patients (p = 0.0153) but not in Sepsis_A patients (p = 0.2590). CRP (p = 0.0003) and IL-6 (p = 0.0176) were also significantly elevated in the SIRS cohort compared with Controls. When comparing SIRS and Sepsis_A patients, higher concentrations were observed in the latter group for Azurocidin (p = 0.0137), PCT (p = 0.0114), IL-8 (p = 0.0153), TNF-α (p = 0.0192), and IL-35 (p < 0.0001).
Compared with the SIRS cohort, Sepsis_D patients exhibited higher levels of Calprotectin (p = 0.0072), Azurocidin (p < 0.0001), IL-4 (p = 0.0112), IL-8 (p = 0.0015), IL-10 (p = 0.0007), TNF-α (p = 0.0198), and IL-35 (p < 0.0001). Finally, the Sepsis_D cohort showed significantly higher concentrations of Azurocidin (p = 0.0009) and IL-6 (p = 0.0452) compared with Sepsis_A patients.
## 3.2. Diagnostic and Prognostic Values
The predictive performance of all biomarkers for Gram-negative bacteremia was evaluated, and the results of the ROC analyses are reported in Table 3, with the corresponding curves shown in 3 for further explanation and data.
The prognostic performance of the same biomarkers for mortality prediction was also assessed, with results presented in Table 3 and the corresponding ROC curves illustrated in Figure 3. All biomarkers except PCT were able to predict mortality, although the AUC values were lower than those observed in the diagnostic analyses. 3 for further explanation and data.
## 4. Discussion
In our study, serum concentrations of calprotectin, azurocidin, PCT, CRP, interleukins, and TNF-α were generally increased in septic patients, regardless of outcome, compared with healthy controls. However, their patterns in SIRS patients and between the Sepsis_A and Sepsis_D groups varied for each biomarker analyzed. Moreover, all biomarkers demonstrated good diagnostic performance for Gram-negative bloodstream infection, whereas their ability to predict mortality was relatively limited.
In this study, we sought to investigate distinct classes of sepsis biomarkers, stratified according to their functional characteristics. We first focused on Th1-associated mediators, including TNF-α and IL-6. Th1 markers are typically associated with the proinflammatory response and have been reported to correlate with a more severe course of sepsis and poorer clinical outcomes [18].
TNF-α levels among septic patients were significantly higher compared with both controls and SIRS cohorts, supporting the diagnostic value of this cytokine, as also confirmed by the ROC analysis. Conversely, TNF-α concentrations did not differ significantly between the Sepsis_A and Sepsis_D cohorts, suggesting a limited prognostic role, consistent with our ROC findings and previous reports in the literature [19].
Based on our results, IL-6 was able to discriminate both SIRS and septic patients from controls, as well as to differentiate between the Sepsis_A and Sepsis_D cohorts. These findings suggest that IL-6 may serve as both a diagnostic and prognostic biomarker in sepsis [20]. The ROC analysis further confirmed the strong predictive value of IL-6 for the diagnosis of Gram-negative bloodstream infection and for outcome prediction. Therefore, IL-6 appears to represent a promising diagnostic and prognostic tool that could assist clinicians in clinical decision-making [21,22].
Our data showed that IL-4 levels were significantly higher in Sepsis_D patients compared with healthy controls and SIRS, whereas differences among controls, SIRS, and Sepsis_A patients were not statistically significant. The absence of significant differences between SIRS and Sepsis_A groups suggests that IL-4 alone may have limited diagnostic value in distinguishing sepsis from non-infectious SIRS. Recent literature has highlighted the often-contradictory role of IL-4 in sepsis [23]. Moreover, IL-4 has been associated with immunoparalysis and partial stimulation of certain proinflammatory cytokines, such as IL-6 and TNF-α, suggesting a potentially modulatory and even protective role in the septic response [23]. The diagnostic and prognostic roles of IL-4 in human sepsis have been poorly investigated to date. However, available evidence suggests that IL-4 is more closely associated with Gram-positive bacteremia than with sepsis caused by Gram-negative bacteria or fungi. As a Th2 cytokine, IL-4 may contribute diagnostically to distinguishing between SIRS and bacteremic patients. Nevertheless, our data indicate that IL-4 is not a reliable biomarker for assessing disease progression or predicting clinical outcomes in septic patients.
Among inhibitory cytokines, IL-10 was evaluated in our study, and the most relevant finding was the significantly higher levels observed in Sepsis_D patients compared with Sepsis_A. This suggests that IL-10 may serve as a potential biomarker for estimating mortality risk among septic patients [24]. Conversely, the absence of significant differences in IL-10 levels between Sepsis_A and SIRS groups indicates that this cytokine may have limited diagnostic value in distinguishing early sepsis from non-infectious SIRS. This finding could be explained by the predominant role of IL-10 during the later stages of sepsis, whereas our study evaluated patients primarily during the early phase of the disease.
Based on the ROC analysis for both diagnostic and prognostic evaluations, IL-10 appears to be a useful biomarker for predicting Gram-negative bloodstream infection, as previously reported [25], and for estimating mortality risk.
Calprotectin has been reported as a pivotal mediator during the immunosuppressive stage of late sepsis [26], an effect that may also be mediated by . A novel finding of our study is the potential involvement of both calprotectin and IL-10 in this late immunosuppressive phase of sepsis. However, direct evidence for this mechanism remains limited, and further dedicated studies are warranted to specifically investigate this aspect.
From a mechanistic perspective, the dual and seemingly opposite roles of calprotectin may result from the involvement of different innate immune cells active during distinct stages of sepsis. In the acute phase, calprotectin is predominantly associated with the activity of dendritic cells and neutrophils, whereas in the immunosuppressive phase, its effects are mainly mediated by myeloid-derived suppressor cells (MDSCs), which play a key role in sustaining immune dysfunction during late sepsis [26,28,29]. In addition, calprotectin has been reported to act as a damage-associated molecular pattern (DAMP) during sepsis [30], exerting its effects through the Toll-like receptor 4 (TLR4) signaling complex [31].
Data on IL-35 showed that its serum concentrations were significantly higher in both groups of septic patients compared with healthy controls and SIRS subjects. However, no significant differences were observed between the Sepsis_A and Sepsis_D cohorts. Consistent with previous reports [11], IL-35 demonstrated good predictive value for the diagnosis of Gram-negative bacteremia, whereas its ability to predict mortality was limited. This pattern, in which IL-35 serves as a prominent diagnostic marker but lacks prognostic utility, suggests its involvement in the early proinflammatory phase of sepsis. Moreover, IL-35 may act during the initial immune response, potentially in coordination with neutrophil activation at the site of infection, as suggested by recent studies linking IL-35 release to neutrophil activity [32].
Regarding IL-8, a noteworthy finding was the significant difference in IL-8 concentrations between septic and non-septic cohorts. Although the role of IL-8 in sepsis has been investigated for more than three decades [33], conclusive and consistent evidence remains limited [34]. Holub et al. reported elevated IL-8 levels in septic patients, and more recently, increasing evidence has highlighted both the pathogenetic and diagnostic roles of neutrophil-derived mediators, for which IL-8 serves as a key chemotactic factor. Consistent with these observations, our ROC analysis demonstrated that IL-8 was a reliable predictor for the diagnosis of Gram-negative bloodstream infection, whereas it did not show prognostic significance.
We also found that azurocidin levels were significantly higher in septic patients compared with both SIRS patients and healthy controls. Furthermore, the Sepsis_D cohort exhibited significantly higher azurocidin concentrations than the Sepsis_A cohort. ROC analysis demonstrated good diagnostic and prognostic performance of azurocidin in our population, consistent with previous reports [35,36].
We also investigated two reference biomarkers, PCT and CRP, used for many years to assist clinicians in the diagnosis of sepsis. Regarding PCT, our data suggest a diagnostic role of PCT in septic patients. On the other hand, CRP has a minor role in both diagnosis and prognosis of this population, as previously reported by a similar study [3].
Finally, it should be noted that calprotectin, azurocidin, IL-4, and IL-8, although acting through different mechanisms, may all contribute significantly to the innate defense against invading pathogens. The binding of azurocidin to pathogen-associated molecules such as lipopolysaccharide (LPS) may facilitate pathogen clearance by professional phagocytes. Conversely, calprotectin exerts antimicrobial activity through the chelation of essential nutrient metals (Zn 2+ , Cu 2+ ), a process known as nutritional immunity [37,38].
Of note, the elevation of several biomarkers observed in our study could also be influenced by patients' comorbidities and by the overall severity of illness. Chronic conditions such as diabetes, cardiovascular disease, or renal dysfunction are known to sustain a proinflammatory state, which may contribute to higher baseline levels of cytokines and acute-phase proteins [39][40][41]. Likewise, disease severity, reflected by higher SOFA and APACHE II scores, is typically associated with a more pronounced systemic inflammatory response and organ dysfunction [14,42], which could amplify biomarker concentrations independently of infection status. Therefore, these factors should be considered when interpreting biomarker elevations in septic patients.
## Strengths and Limitations
Before drawing our conclusions, several strengths and limitations of this study should be discussed.
One notable strength is its multicenter design, which enhances the robustness of the data and supports the generalizability of our findings to a broader population of patients [43][44][45]. Furthermore, few studies to date have simultaneously investigated novel diagnostic and prognostic biomarkers alongside established and widely recognized sepsis biomarkers within the same patient cohort. This comparative approach provides valuable insight into the relative clinical performance of emerging biomarkers and strengthens the translational relevance of our results.
However, this study also has several limitations that should be acknowledged. First, its retrospective observational design may be affected by selection bias, and the findings may not be fully representative of the general population. In addition, the presence of potential confounding variables limits the ability to establish causal relationships [46]. Second, biomarker measurements were obtained at a single time point, which prevented the assessment of dynamic changes during the course of illness that might provide additional prognostic information. Furthermore, as this was a retrospective study, the exact timing of sample collection in relation to symptom onset, hospital admission, and blood culture positivity could not be systematically retrieved for all patients. Nevertheless, all samples were obtained at the time of hospital admission and before the initiation of antibiotic therapy, as verified in the clinical records. This standardized sampling procedure helps to limit variability associated with timing and ensures consistency across the study cohort. We acknowledge, however, that the absence of precise temporal data may limit the detailed interpretation of biomarker dynamics. Future prospective investigations with predefined and recorded time points are warranted to better assess the impact of sampling timing on biomarker performance and reproducibility. To address these limitations, we have already planned a further study aimed at evaluating biomarker kinetics in this patient population. Third, this study exclusively enrolled septic patients with an initial Gram-negative bloodstream infection, as detailed in the Materials and Methods section. Consequently, the present findings primarily characterize the host response and biomarker dynamics in Gram-negative sepsis and may not be fully generalizable to sepsis of Gram-positive or fungal etiology. This limitation should be taken into account when interpreting the results. Finally, although the overall sample size was adequate for the primary analyses [3,47,48], the number of patients in some subgroups was limited, potentially reducing the statistical power to detect more subtle associations. These limitations should be taken into account when interpreting our findings and highlight the need for future prospective, longitudinal studies, for further validation, with larger and more balanced cohorts to confirm and expand upon our results.
## References
1. "All ROC analyses were performed on the entire study population. In the diagnostic ROC analysis, 121 patients were classified as having a positive outcome (bloodstream infection) and 30 as having a negative outcome (no bloodstream infection)"
2. Tnf-Α "tumor necrosis factor-α; PLR, positive likelihood ratio; NLR, negative likelihood ratio; PPV, positive predictive value; NPV, negative predictive value"
3. Povoa, Coelho, Dal-Pizzol et al. (2023) "How to use biomarkers of infection or sepsis at the bedside: Guide to clinicians" *Intensive Care Med*
4. Hung, Lan, Han et al. (2020) "Current evidence and limitation of biomarkers for detecting sepsis and systemic infection" *Biomedicines*
5. Yang, Xie, Guo et al. (2016) "Combination of c-reactive protein, procalcitonin and sepsisrelated organ failure score for the diagnosis of sepsis in critical patients" *Ann. Intensive Care*
6. Taha, Najah, Omar et al. "Abd-ElGawad, M. Diagnostic and prognostic value of heparin-binding protein in sepsis: A systematic review and meta-analysis" *Medicine*
7. Gao, Jia, Han et al. (2022) "Calprotectin as a diagnostic marker for sepsis: A meta-analysis" *Front. Cell. Infect. Microbiol*
8. Eggers, Sikora, Lorenz et al. (2011) "Rage-dependent regulation of calcium-binding proteins s100a8 and s100a9 in human thp-1" *Exp. Clin. Endocrinol. Diabetes*
9. Edgeworth, Gorman, Bennett et al. (1991) "Identification of p8,14 as a highly abundant heterodimeric calcium binding protein complex of myeloid cells" *J. Biol. Chem*
10. Bone, Balk, Cerra et al. (1992) "Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. The accp/sccm consensus conference committee. American college of chest physicians/society of critical care medicine" *Chest*
11. Hotchkiss, Tinsley, Swanson et al. (2001) "Sepsis-induced apoptosis causes progressive profound depletion of b and cd4+ t lymphocytes in humans" *J. Immunol*
12. Shen, Cao, Zhao et al. (2021) "Targeting neutrophils in sepsis: From mechanism to translation" *Front. Pharmacol*
13. Hu, Tang, Zhang "The diagnostic value of interleukin 35 as a septic biomarker: A meta-analysis" *Front. Med*
14. Li, Liu, Fu et al. (2023) "Alterations in levels of cytokine following treatment to predict outcome of sepsis: A meta-analysis" *Cytokine*
15. Singer, Deutschman, Seymour et al. (2016) "The third international consensus definitions for sepsis and septic shock (sepsis-3)" *JAMA*
16. Baddam, Burns (2025) "Systemic Inflammatory Response Syndrome"
17. Cione, Siniscalchi, Gangemi et al. "Neuron-specific enolase serum levels in COVID-19 are related to the severity of lung injury" *PLoS ONE*
18. Perticone, Zito, Miceli et al. (2019) "Immunity, inflammation and heart failure: Their role on cardiac function and iron status" *Front. Immunol*
19. Delong, Delong, Clarke-Pearson (1988) "Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach" *Biometrics*
20. Marin, Van Wijk, Chambliss (2024) "Advances in sepsis biomarkers" *Adv. Clin. Chem*
21. Li, Xiao, Zheng (2024) "Prognostic value of inflammatory cytokine detection for sepsis patients in icu: A meta-analysis" *Am. J. Transl. Res*
22. Song, Moon, Park et al. (2020) "Biomarker combination and sofa score for the prediction of mortality in sepsis and septic shock: A prospective observational study according to the sepsis-3 definitions" *Medicine*
23. Liu, Wang, She et al. "Effects of neutrophil-to-lymphocyte ratio combined with interleukin-6 in predicting 28-day mortality in patients with sepsis" *Front. Immunol. 2021*
24. Behnes, Bertsch, Lepiorz et al. (2014) "Diagnostic and prognostic utility of soluble cd 14 subtype (presepsin) for severe sepsis and septic shock during the first week of intensive care treatment" *Crit. Care*
25. Schrijver, Roring, Deckers et al. (2023) "Resolving sepsis-induced immunoparalysis via trained immunity by targeting interleukin-4 to myeloid cells" *Nat. Biomed. Eng*
26. Kellum, Kong, Fink et al. (2007) "Understanding the inflammatory cytokine response in pneumonia and sepsis: Results of the genetic and inflammatory markers of sepsis (genims) study" *Arch. Intern. Med*
27. Matera, Puccio, Giancotti et al. (2013) "Impact of interleukin-10, soluble cd25 and interferon-gamma on the prognosis and early diagnosis of bacteremic systemic inflammatory response syndrome: A prospective observational study" *Crit. Care*
28. Wang, Long, Luo et al. (2023) "An emerging player in sepsis and sepsis-induced organ injury" *Biomed. Pharmacother*
29. Liu, Huang, Sun et al. "Sepsis-induced immunosuppression: Mechanisms, diagnosis and current treatment options" *Mil. Med. Res*
30. Nedeva, Menassa, Duan et al. (2020) "Treml4 receptor regulates inflammation and innate immune cell death during polymicrobial sepsis" *Nat. Immunol*
31. Venet, Monneret (2018) "Advances in the understanding and treatment of sepsis-induced immunosuppression" *Nat. Rev. Nephrol*
32. Jakobsson, Papareddy, Andersson et al. (2023) "Therapeutic s100a8/a9 blockade inhibits myocardial and systemic inflammation and mitigates sepsis-induced myocardial dysfunction" *Crit. Care*
33. Riva, Kallberg, Bjork et al. (2012) "Induction of nuclear factor-κb responses by the s100a9 protein is toll-like receptor-4-dependent" *Immunology*
34. Cao, Xu, Lin et al. (2015) "Il-35 is elevated in clinical and experimental sepsis and mediates inflammation" *Clin. Immunol*
35. Hack, Hart, Van Schijndel et al. (1992) "Interleukin-8 in sepsis: Relation to shock and inflammatory mediators" *Infect. Immun*
36. Holub, Dzupova, Ruzkova et al. (2018) "Selected biomarkers correlate with the origin and severity of sepsis" *Mediat. Inflamm*
37. Zuo, Li, Wang et al. (2024) "Heparin-binding protein as a biomarker for the diagnosis of sepsis in the intensive care unit: A retrospective cross-sectional study in china" *BMJ Open*
38. Wu, Yo, Hsu et al. (2021) "Accuracy of heparin-binding protein in diagnosing sepsis: A systematic review and meta-analysis" *Crit. Care Med*
39. Xia, Ji, Yan et al. (2024) "Roles of s100a8, s100a9 and s100a12 in infection, inflammation and immunity" *Immunology*
40. Murdoch, Skaar (2022) "Nutritional immunity: The battle for nutrient metals at the host-pathogen interface" *Nat. Rev. Microbiol*
41. Domingo, Marques, Francisco et al. (2024) "Targeting systemic inflammation in metabolic disorders. A therapeutic candidate for the prevention of cardiovascular diseases?" *Pharmacol. Res*
42. Mihai, Codrici, Popescu et al. (2018) "Inflammationrelated mechanisms in chronic kidney disease prediction, progression, and outcome" *J. Immunol. Res*
43. Lopez-Candales, Hernandez Burgos, Hernandez-Suarez et al. (2017) "Linking chronic inflammation with cardiovascular disease: From normal aging to the metabolic syndrome" *J. Nat. Sci*
44. Wang, He, Jin et al. (2024) "Immune dysfunction-associated elevated rdw, apache-ii, and sofa scores were a possible cause of 28-day mortality in sepsis patients" *Infect. Drug Resist*
45. Sa-Couto, Ericsson, Lazarovici "Conducting multicenter simulation-based experimental research: Lessons drawn from the quality cpr european project"
46. Das (2022) "Multicenter studies: Relevance, design and implementation" *Indian. Pediatr*
47. Cheng, Kessler, Mackinnon et al. "Conducting multicenter research in healthcare simulation: Lessons learned from the inspire network"
48. Hammer, Du Prel, Blettner (2009) "Avoiding bias in observational studies: Part 8 in a series of articles on evaluation of scientific publications" *Dtsch. Arztebl. Int*
49. Keller, Mester, Rath et al. (2024) "Calprotectin, a promising serological biomarker for the early diagnosis of superinfections with multidrug-resistant bacteria in patients with COVID-19" *Int. J. Mol. Sci*
50. De Guadiana-Romualdo, Botella, Rodriguez Rojas et al. (2024) "Mortality prediction model from combined serial lactate, procalcitonin and calprotectin levels in critically ill patients with sepsis: A retrospective study according to sepsis-3 definition" *Med. Intensiv*
51. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548392&blobtype=pdf | # Rescue of tomato yellow leaf curl virus mutants harboring heterologous iterons through in planta evolution
Khwannarin Khemsom, Ruifan Ren, Junping Han, Camila Carvalho, Eric Snider, Deyong Zhang, Feng Qu, Feng Qu
## Abstract
The single-stranded, circular DNA genomes of geminiviruses contain iterated motifs of five to six nucleotides, known as iterons, upstream of the replication protein (Rep) coding region. Iterons were previously found to interact with cognate Rep in a sequence-specific manner, and the iteron-Rep interaction was needed for viral DNA replication. Nonetheless, iterons of closely related viruses often have different sequences, suggesting diversifying selection. To identify selection pressures driving iteron diversification, we constructed tomato yellow leaf curl virus (isolate SH2) mutants in which the iteron motifs were replaced with those of closely related tobacco curly shoot virus (isolate Y35). All mutants replicated in inoculated leaves of Nicotiana benthamiana, but some failed to spread systemically. However, the systemic movement defects were mostly rescued by de novo mutations. Intriguingly, these de novo mutations did not restore the iterons to SH2 sequences. Rather, they likely enabled viral escape from repression exerted by the heterogeneous Y35 iterons in the absence of a matching Rep. These results suggest that iterons probably act as sites of competitive binding by host-encoded transcription factors (TFs) and the cognate Rep. We further speculate that iteron-TF binding commences as soon as viral genomes enter cell nuclei, committing genome copies to Rep mRNA transcription and protein translation, but also blocking them from replication. Conversely, iteron-Rep binding would be possible only after Rep is produced and likely repels TFs from some genome copies, permitting replica tion initiation. Testing this model through future research should clarify the intricate evolutionary interplays between geminiviruses and their crop hosts and inform novel management strategies. IMPORTANCE Geminiviruses are important crop pathogens worldwide for which effective control measures are lacking due to an incomplete understanding of their evolutionary dynamics in infected plants. The current study focuses on a class of short sequence repeats in geminiviral genomic DNA, known as iterons, located immediately upstream of the viral gene encoding replication protein (Rep). Iterons are interesting because, although their positions and repeat patterns are conserved across all gemini viruses, their sequence identities are highly diverse. Our investigations revealed that, contrary to previous reports, the sequence identity of iterons is non-essential for tomato yellow leaf curl virus replication. Rather, they are repressors of replication, and this repression is overcome by their binding with cognate Rep. Future investigations will likely unveil novel targets for more effective management of crop diseases caused by geminiviruses.
V iruses of the family Geminiviridae are among the most common pathogens of crop plants (1), causing devastating losses in staple crops such as cassava, cotton, maize, soybean, tomato, and wheat. These viruses have relatively small, single-stranded (ss), circular DNA genomes, encoding up to a dozen viral proteins (2). Geminiviruses replicate their genomes in host cell nuclei by recruiting a host-encoded DNA-dependent DNA polymerase (DNA Pol). Nevertheless, they do encode an accessory replication protein, designated variously as Rep, C1, AC1, or AL1 depending on viruses under investigation, which recruits DNA Pol to geminiviral DNA. The Rep protein is also responsible for specifying the rolling circle replication mode of these viruses. Specifically, upon entering host cells, the single-stranded, circular DNA genome of a geminivirus, designated as (+) strand, enters the nucleus to prime the synthesis of a complementary (or [-]) strand, thereby creating a double-stranded, circular DNA molecule referred to as replicative form (RF). During rolling-circle replication, the (+) strand of RF is cleaved by Rep, and the (-) strand is used as the template for the synthesis of a linear, single-stranded, multimeric (+) strand, which is subsequently cleaved by Rep into unit genomes. The resulting unitlength linear genomes are subsequently re-circularized by Rep into progeny genomes (2,3).
In addition to encoding proteins, the circular ssDNA genomes of geminiviruses also harbor various cis-acting elements. The most important among them is a DNA stem-loop consisting of a double-stranded stem of 11 base pairs (bp) and a single-stranded loop featuring a highly conserved TAATATTAC motif. Inside this motif lies the site of Rep-medi ated nicking of circular genomes (TAATATT/AC), a key step required for the initiation of rolling circle replication (4,5). This DNA stem-loop is indispensable for geminivirus replication and indeed must be duplicated at both ends of a linear, unit-length genome to ensure the infectivity of a geminiviral infectious clone (6) (also see Fig. 1A).
A second class of cis-acting motifs, referred to as iterons, are short (five to six nucleotides or nt) sequence repeats located upstream of the Rep protein coding region (8,9). Iterons typically repeat for three to four times, with at least one of the repeats being antisense to others (9)(10)(11)(12). Iterons have been shown to mediate sequence-specific binding with the Rep protein of the same virus (13)(14)(15)(16), and this iteron-Rep interaction was found to be needed for successful viral replication. Intriguingly, the iteron-Rep binding required the iteron-containing DNA to be double-stranded (12,15), suggesting that the double-stranded iteron motifs in RFs are preferred binding sites for Reps (17).
Although iterons are usually conserved among isolates of the same geminivirus species, they are highly divergent among different, though still closely related, virus species. Listed in Fig. S1A are four examples of closely related geminiviruses/isolates, namely tomato yellow leaf curl virus (TYLCV) isolate SH2, tobacco curly shoot virus (TbCSV) isolates Y41 and Y35, and tomato yellow leaf curl Sardina virus (TYLCSV). These viruses share whole genome nt level identities of more than 75% (Fig. S1B, percentages in black fonts) and Rep protein amino acid (aa) level similarities of more than 87% (Fig. S1B, percentages in green fonts). In particular, TbCSV Y41 and Y35 are more than 96% identical at both the whole genome nt level and the Rep aa level. Nevertheless, Y41 and SH2, but not Y41 and Y35, share the same, though differently spaced, iteron sequences (10, 12) (Fig. S1C, the iterons are painted light blue and yellow, respectively). Moreover, iterons of TYLCSV have entirely different sequences (Fig. S1C, iterons are painted purple).
Thus, it appears that iterons, despite their involvement in sequence-specific interactions with cognate Reps, are under rapid diversifying selection. Consistent with this view, Reps, or more specifically the iteron-interacting domains of Reps, also appear to evolve rapidly to maintain the sequence-specific interactions (8,9,11). Indeed, comprehensive bioinformatic analyses of large numbers of geminiviruses and nanovi ruses identified two separate Rep domains whose aa sequences co-varied with iteron sequences (9, 11) (Fig. S1D). However, it has not been thoroughly examined as to exactly what drives the rapid diversification of iteron motifs and the co-varying iteron-interact ing Rep domains.
Here, we report an attempt to identify such selection pressures, using TYLCV SH2 as the model virus and Nicotiana benthamiana as the model host (18,19). TYLCV is a monopartite member of the genus Begomovirus, family Geminiviridae. Its circular ssDNA genome of approximately 2,800 nt encodes at least six proteins (2,20,21). The four proteins (Rep, C2-C4) encoded on the (-) strand of the genome are early expressing, participating in various aspects of viral genome replication, transcriptional activation, and host defense mitigation (Fig. 1A) (2,22). In particular, Rep is absolutely required for the rolling circle replication of the TYLCV genome (23)(24)(25). The two proteins encoded on the (+) strand of the TYLCV genome, known as V1 and V2, are late expressing and function as capsid protein (CP) and suppressor of RNA silencing (26), respectively (Fig. 1A). V2 has also been implicated in viral cell-to-cell movement (27). V1 (CP) is not essential for the replication of TYLCV (28), but is needed for the intra-and intercellular trafficking of the virus (29).
In the current investigation, we modified the TYLCV SH2 genome by replacing its iterons with those of TbCSV Y35 (Fig. 1). We then infected N. benthamiana plants with the modified viruses and subjected infected plants to systematic analyses for up to 9 weeks. Identification of de novo mutations, and subsequent examinations of these new mutations in plants, led us to conclude that, in the absence of a cognate Rep, iterons act to repress virus replication. Such replicational repression was probably caused by recruitment of host-encoded transcription factors by iterons or other tightly linked sequence motifs. The recruited TFs not only enhanced Rep mRNA transcription to enable rapid accumulation of Rep proteins but also sequestered viral genome copies from replication. While wild-type TYLCV could ease this sequestration through iteron-Rep binding, some of the viral mutants we examined probably escaped the same sequestration through spontaneous mutations that weakened iteron-TF binding. Such mutation-driven escape may prime the diversification of iteron motifs in closely related geminiviruses.
## RESULTS
## Mutants m1 and m3, but not m2, elicit systemic symptoms in N. benthamiana
Earlier studies have shown that iterons and Rep of the same geminivirus interact with each other in a highly sequence-specific manner, and this iteron-Rep interaction is necessary for the replication of viral genomes (8,(10)(11)(12)(13)(14)(15)(16). However, these studies mostly used viruses with bipartite genomes, or those hosting satellite DNAs, with iteron-disrupt ing mutations engineered in the non-Rep-encoding genome segments (e.g., DNA-B of bipartite geminiviruses, or satellite DNAs). To determine whether iteron perturbation in cis compromises replication of a monopartite geminivirus, we manipulated the genome of TYLCV SH2 to generate three mutants-m1, m2, and m3 (Fig. 1C). In m1, the 97-nt genome section of SH2 encompassing all three iterons, a 13-nt spacer, and the 24 nt encoding the N-terminal 8 aa of SH2 Rep was replaced with the 105-nt counterpart in TbCSV Y35 (Fig. 1B andC). The reason for also exchanging the Rep N-terminus was because this domain, along with another Rep domain approximately 60 aa downstream, was found to co-vary with iteron sequences (9, 11) (Fig. S1D). By contrast, in m2 only the 60-nt iteron region was exchanged, whereas in m3 only the 8-aa (24 nt) Rep N-termi nus and 13-nt spacer were replaced (Fig. 1B andC, sections originated from Y35 are highlighted with blue fonts).
We incorporated three sets of mutations into the wild-type SH2 infectious clone (LM5, Fig. 1A) and tested the resulting mutants in N. benthamiana. The m3-infected plants, similar to SH2-infected plants, began to exhibit systemic symptoms at 2 weeks post-inoculation (wpi) (Fig. 2A). The m1-infected plants started to show symptoms by 3 wpi, representing a modest delay of 1 week (Fig. 2A). By contrast, m2 failed to cause visible diseases in any of the six infected plants. The symptom severity differences were consistent with real-time, quantitative PCR (qPCR) detection of viral genomic DNA. As shown in Fig. 2B, m2 genomic DNA levels in 1 wpi inoculated leaves (ILs) were approximately 30% of that of wild-type SH2 control. On the other hand, m1 and m3 genome levels were not statistically different from SH2 (Fig. 2B). Sequence analysis of the PCR products verified that all mutations remained stable in ILs at 1 wpi. When the systemically infected leaves (SLs) were examined at 6 and 9 wpi, both m1 and m3 genomes accumulated to levels indistinguishable from SH2 (Fig. 2C). However, m2 was undetectable at 6 wpi and was present in just one plant by 9 wpi at a very low titer. Thus, among the three mutants, m2 was most debilitated at the systemic infection level.
## Both m1 and m2 mutants incur de novo mutations in systemically infected leaves
To determine whether the mutations of m1 and m2 were stable in SLs, we next subjected the PCR-amplified genome fragments to sequence analyses. At 6 wpi, viral DNA obtained from three m1-infected plants all contained a new mutation at the fourth position of the middle iteron, changing the Y35-borne GGTCCT motif to GGTTCT (Fig. 3A, de novo mutated nt in red font). Furthermore, one of the three DNA samples also contained a second mutation at the sixth position of distal iteron (AGGACC to AGGACA). Remarkably, neither of the de novo mutations returned the iteron motifs to the sequences of wildtype SH2 (GGTTCT vs GGTGTC; AGGACA vs GACACC; Fig. 1C). Even more interestingly, at 9 wpi, the first mutation was detected in all six plants analyzed, and the second mutation was also stable in the plant in which it first emerged. Also worth noting was that the original nt (a C) at the fourth position of the middle iteron was undetectable in any of the samples. It should be noted that direct sequencing of the PCR products could mask the presence of original mutations if they are present at very low concentrations. Nonethe less, overall, these results suggest that robust systemic infection does not require the specific sequences of middle and distal iteron motifs to conform to those of either SH2 or Y35.
None of the three m2-inoculated plants examined at 6 wpi yielded detectable levels of viral DNA. However, at 9 wpi, one of the six plants examined contained low levels of viral DNA (Fig. 2C). When subjected to sequence analysis, the PCR-amplified viral DNA contained two mutations at third and fourth positions of the middle iteron, changing GGTCCT to GGGTCT, deviating the motif sequence even further from that of wild-type SH2 (GGTGTC) or Y35 (GGTCCT). To corroborate this result, we then collected leaf samples from 20 different branches of this plant for DNA extraction and PCR detection of viral DNA. As summarized in Fig. 3B, among the 20 branches, 12 contained descendants that harbored the single C-to-T change also found in m1 descendants, whereas the remaining 8 branches contained variants that had the TC-to-GT change. Together, these results showed that when present in the SH2 backbone, the Y35-borne middle iteron incurred de novo mutations that rendered it dissimilar from both SH2 and Y35.
Finally, progeny of m3 maintained the original m3 mutations, suggesting that the 10-aa N-terminus originating from Y35 was stable in the SH2 background and had minimal impact on viral systemic infections. Together, these results demonstrated that when the three iteron motifs were exchanged in concert, the middle motif probably hindered SH2 replication, and its adverse impact must be mitigated through de novo mutations in order to restore systemic infection to affected viruses. To reiterate, such new mutations did not convert the motif sequence to that of SH2 or Y35, suggesting a relief from repressive activity exerted by Y35 iterons in the absence of the cognate Y35 Rep.
## De novo mutations bolster the infectivity of the m1 mutant
It is worth emphasizing that the original middle iteron of the m1 mutant, of Y35 origin, was not detected in the systemic leaves of plants we examined. We thus speculated that the single C-to-T de novo mutation was needed for m1 descendants to spread systemically. To test this, we introduced the C-to-T mutation back into m1 to create m1a (Fig. 4A). For comparison, we also introduced the TC-to-GT mutations recovered from m2 descendants into the m1 backbone, creating m1b (Fig. 4A). Additionally, we incorporated the C-to-A mutation found in the distal iteron into the m1a backbone to create m1f (Fig. 4A). As shown in Fig. 4B, m1a and m1b accumulated genomic DNAs to levels rivaling that of SH2 and m1 in 1 wpi ILs, whereas the m1f accumulation level was significantly higher than even the SH2 wild type. Both m1a and m1f caused symptoms that emerged at about the same time as SH2, thus 1 week earlier than m1 (Fig. 4C). Curiously, the m1b mutant harboring the TC-to-GT double mutations developed symptoms 1 week later than m1a and m1f, causing the infected plants to be taller (Fig. 4C).
When subjected to sequence analyses at 6 wpi, the three new m1-infected plants once again yielded viral progeny that incurred the C-to-T de novo mutation (Fig. 4A). Combined with the earlier onset of m1a disease symptoms, the highly reproducible emergence of C-to-T mutation strongly suggested that this mutation was responsible for the systemic infections of m1 descendants. Conversely, the original m1 mutant must have been easily overtaken by the m1a variant containing the C-to-T mutation. Further more, this C-to-T mutation, upon emergence in the m1 backbone, was apparently very stable, as it did not evolve further into TC-to-GT. This was unlike the same C-to-T mutation incorporated in the m2 backbone, where it did evolve further into TC-to-GT (see later). Finally, plants infected with the m1f mutant were as severely infected as m1ainfected ones, with both of its mutations remaining stable in plants (Fig. 4B andC). Together, these results indicate that the de novo mutations detected in m1 descendants correlated with superior infectivity.
## De novo mutations restore infectivity to the m2 mutant without converting the middle iteron to the SH2 motif
Unlike the m1-infected plants, those infected with m2 did not have any systemic symptoms, and most of them also failed to accumulate viral genomic DNA (Fig. 2C and5C). Indeed, even the single plant in which m2 acquired the C-to-T and TC-to-GT mutations did not have visible systemic symptoms. We thus set out to resolve whether these de novo mutations also improved the infectivity of the m2 mutant. To this end, the m2a and m2b mutants, carrying the C-to-T and TC-to-GT mutations, respectively, were constructed (Fig. 5A) and used to infect N. benthamiana plants. As shown in Fig. 5B, in ILs, both m2a and m2b were able to replicate. However, their accumulation levels were only modestly higher than m2 (not statistically significant). We then examined the systemic leaves through 6 wpi. Unlike m2-infected plants, the m2a-and m2b-infected plants developed clearly visible systemic symptoms (Fig. 5C). Consistently, the TYLCV-specific PCR products obtained from 6 wpi SLs of m2a-and m2b-infected plants reached high levels (Fig. 5D, lanes 14-24). Since the m1 mutant was also included in the current experiment as one of the controls, we analyzed the sequences of PCR products obtained from m1-infected samples once again. As shown in Fig. 5A, four of four analyzed sequences contained the C-to-T mutation that was identified repeatedly (Fig. 3A, 4A, and5A). Interestingly, while the C-to-T mutation was very stable in the m1 background (e.g., m1a in Fig. 4), it was less so in the m2 background (m2a). By 6 wpi, the m2a mutant acquired an additional T-to-G change at the adjacent position in three of the five plants, rendering the descendants identical to m2b (Fig. 5A). Consistent with the increased infectivity attributable to the TCto-GT mutations in m2 (but not m1) descendants, the m2b-infected plants were more stunted than the m2a-infected ones. Furthermore, the m2b mutations remained stable for at least 9 weeks. Collectively, these data indicate that the de novo mutations occurring in the m2 genome were necessary and sufficient for the new mutants to gain robust systemic infections.
## The Y35-borne middle motif represses TYLCV systemic infections even without the proximal and distal motifs, and this repression is always relieved by the TC-to-GT mutations
Given that all three iteron motifs in m2 were of Y35 origin, we next examined if the proximal and distal motifs also contributed to the systemic movement failure of the m2 mutant. To this end, we generated a series of mutants that altered the proximal or distal motif alone or in combination, with or without the TC-to-GT change in the middle motif. As summarized in Fig. 6A, the failure to spread systemically persisted even when one or both flanking motifs were mutated to entirely unrelated sequences (m2g, m2h, m2gh, and m2g2h). While m2g replication in 1 wpi ILs was significantly lower than m2, m2bg containing the TC-to-GT mutations restored replication to m2 levels (Fig. 6B). More strikingly, m2bg, but not m2g, caused robust systemic infections in N. benthamiana plants (Fig. 6C). Similar outcomes were also observed with m2bh vs m2h, in which the proximal iteron motif was abolished, as well as m2bgh vs m2gh and m2bg2h vs m2g2h, in which both proximal and distal iteron motifs were mutated (Fig. 6A). Thus, regardless of the sequence identities of proximal and distal motifs, the TC-to-GT mutations in the middle motif alone were enough to restore systemic infections. These results suggest that the flanking motifs were unnecessary for either the middle-iteron-mediated repression or its relief by the TC-to-GT mutations.
## Systemic infection of the m2 mutant is variably bolstered by other mutations at the third and fourth positions of the middle iteron
So far, we showed that the TC-to-GT mutations at the third and fourth positions of the m2 middle iteron caused the resulting m2b mutant to infect N. benthamiana plants systemically, leading to clear symptoms indistinguishable from wild-type SH2.
The intriguing puzzle was that the exact sequence of m2b middle iteron differed from that of SH2 or Y35 (GGGTCT in m2b as opposed to GGTGTC in SH2, and GGTCCT in Y35). We thus wondered whether substituting the TC doublet within GGTCCT-the Y35-borne m2 middle motif-with two random nts would be enough to restore systemic infection to m2. To resolve this question, we generated three new mutants-m2c, m2d, and m2e. As shown in Fig. 7A andB, m2c and m2d changed the TC doublet to AG and CA, respectively. The m2e mutant altered the last three nucleotides from CCT to GTC so that this motif was now identical to that of SH2 (GGTGTC), though the rest of the iteron-encompassing section was still of Y35 origin. Furthermore, since the proximal and distal iterons were deemed inconsequential in m2b variants, we also tested three additional mutants in which m2c, m2d, and m2e changes were combined with those of m2g2 and m2h, yielding mutants m2cg2h, m2dg2h, and m2eg2h (Fig. 7A andB). The m2c mutant replicated in local agro-infiltrated leaves (data not shown). When the upper young leaves were assessed at 6 wpi with PCR, viral DNA was detected in three of five, two of three, and two of three plants in three independent repeat experiments (Fig. 7A). Notably, a de novo G-to-T mutation at the fourth position was recovered from one plant in the first trial and another in the third trial, changing the middle iteron motif from GGAGCT to GGATCT (Fig. 7A). More interestingly, in both first and third trials, the plants from which the de novo mutation was recovered showed delayed symptoms that emerged at 4-5 wpi, whereas all plants containing the original m2c mutants were asymptomatic.
To further assess the impact of this de novo mutation, we created and tested the m2c2 mutant harboring this mutation (Fig. 7A). Indeed, the m2c2 mutant caused systemic symptoms in one of three infected plants, though the symptoms were still delayed, only visible after 4-5 wpi (Fig. 7A). Nonetheless, the m2c2 mutations were stable in this plant. Thus, the fourth position of the middle iteron motif converged to a T residue in plants receiving m1, m2, and m2c mutants, even though the respective starting residues were different (C in m1 and m2, G in m2c). By contrast, the third position residue could be a T (m1 progeny), G (m2 progeny), or A (m2c2 progeny). These results indicate that for systemic infections to occur, it was not necessary for the middle iteron to acquire a definitive sequence identity. Finally, the m2cg2h mutant, in which both the proximal and distal iteron motifs (of Y35 origin) were perturbed with mutations, could still be detected in unaltered form in one of three plants in three independent repeats, though none of the mutant-containing plants showed any symptoms (Fig. 7A). By contrast, the TC-to-CA mutations of the m2d mutant, with or without the flanking Y35 iterons, led to complete failure of systemic infections (Fig. 7B). Curiously, the m2e mutant, in which the SH2 middle iteron motif had been restored, albeit in the middle of Y35 context, accumulated very low levels of viral DNA in just two of the five inoculated plants by 6 wpi (Fig. 7B, the sixth plant died of injury before reaching 6 wpi), with none of the plants showing any symptoms. Furthermore, eliminating the two flanking Y35 iterons led the resulting mutant (m2eg2h, Fig. 7B) to be completely absent from systemic leaves. The m2e and m2eg2h results indicated that restoring the middle iteron alone to the SH2 sequence was insufficient for the virus to regain symptomatic infections. These results contrasted with those of m2b and m2c2, where the departure of the middle motif from the SH2 and Y35 consensuses with merely two nt changes was enough to restore symptomatic infections. Overall, they argue against a critical role of unique sequence identity for iterons. Rather, they support the argument that these de novo mutations primarily enabled viral escape from repression conferred by Y35 iterons in the absence of Y35 Rep.
## GG doublet at first and second positions of m2b middle iteron is not essential for viral systemic spread
The TC-to-GT change at third and fourth positions led to an altered motif (GGGTCT) that differed from Y35 (GGTCCT) and SH2 (GGTGTC). Nevertheless, all three motifs still shared the GG doublet at the first and second positions. We next tested whether this GG doublet was needed for systemic infections by mutating them to AC. More specifically, the m2hi mutant changed the proximal, middle, and distal motifs to AGA, ACGTCT, and AGACGT, respectively (Fig. 1C, m2h-i). As a result, none of the three iterons retained the sequences of SH2 (GGTGTC/GACACC) or Y35 (GGTCCT/AGGACC), although the flanking non-iteron sequences were of Y35 origin (Fig. 7C, m2h-i). None of the six plants inoculated in two separate attempts showed any systemic symptoms. However, viral DNA was detected in one plant in the first trial and two in the second trial (Fig. 7C). Interestingly, all three viral DNA samples incurred de novo mutations within the middle motif and/or Rep N-terminus (Fig. 7C). In one plant, the ACGTCT was changed to AAGTCT. In another plant, it was changed to ATGTCT, plus an aa change in TYLCV Rep-position 46 arginine (R) was changed to isoleucine (I). Interestingly, the R46I change was also independently detected in another plant (Fig. 7C). Thus, the m2hi mutant appeared to enrich compensatory changes in the Rep protein. Together, these findings suggested that the GG doublet was not always required for viral systemic spread, as none of the de novo mutations restored the GG doublet or even just one G.
To test whether some of the de novo mutations incurred in m2hi descendants enhanced viral infectivity, we introduced one of the mutations, ACGTCT to AAGTCT, into m2hi to obtain m2hi2. The m2hi2 mutant was still weak, detectable in one of the three plants in each of two independent experiments. Interestingly, another de novo mutation was detected in the 6-nt motif in one plant, further changing AAGTCT to GAGTCT, thus restoring one G at the first position. These results suggest a stepwise evolutionary trajectory toward the restoration of at least one of the two Gs.
## Putative iteron-borne, intramolecular DNA secondary structures do not predict the systemic infection differences of the mutants
Results described above demonstrated that while the Y35-borne middle iteron must incur de novo mutations to regain systemic infections, the exact nature of the new nt exhibited a certain preference that defies easy interpretation. We hence consid ered the possibility that the three iteron motifs might fold into intramolecular DNA secondary structures, which could then be differentially perturbed by the mutations introduced (Fig. S2). We thus used the mFold algorithm (http://www.unafold.org/mfold/ applications/dna-folding-form.php) to predict potential DNA secondary structures in the iteron-encompassing sequences of SH2 (m3), Y35 (m1, m2), m1a/m2a, m1f, m1b/m2b, m2c, m2d, and m2e. As shown in Fig. S2, the m1 and m2 iterons (of Y35 origin) were indeed predicted to fold into a relatively stable structure (ΔG = -5.96 kcal/mol). However, this structure was weakened by m1a, m2a, m2b, m2b, and m2e mutations to similar extents; hence, it could not explain their differences in symptom severities. Furthermore, despite folding into a much weaker structure, the m1f mutant symptoms were indistin guishable from those of m1a. Finally, the m2bg2h mutant, with both flanking motifs deleted, was not expected to fold into stable secondary structures, yet still elicited visible symptoms (Fig. 6A). Thus, the potential secondary structures the iterons could assume in single-stranded genomes do not provide satisfactory explanations for the observed infectivity differences.
## DISCUSSION
## Sequence identity of iteron motifs is not essential for TYLCV replication
The critical importance of iterons with specific nucleotide sequences in geminivirus replication was first reported years ago (13,15,16). These earlier studies found that iterons with specific sequence motifs were required for replication of the bipartite tomato golden mosaic virus (TGMV) DNA-B segment, which depended on DNA-A for the replication protein AL1 (13,15). Similar sequence specificity requirements for monopar tite geminiviruses such as TYLCV and TbCSV were inferred from the fact that these viruses likewise harbor reiterated short motifs upstream of the Rep coding region (8,9,11). However, whether iteron motifs of monopartite geminiviruses are essential for replication has not been carefully examined. To the best of our knowledge, Xu and colleagues (12) were the first to address this question. They found that mutants of the TbCSV isolate Y35 with any two of the three iterons deleted still infected plants systemically, but a mutant with all three iterons deleted was non-infectious (12).
The current study built on these earlier findings and established that none of the iteron motifs were absolutely required for TYLCV (isolate SH2) replication in initially infected cells. Compared with a mutant of SH2 containing a loss-of-function mutation in Rep (C1fs1), all of our iteron mutants, ranging from m2, in which all three of the SH2 iterons were replaced by their Y35 counterparts, to m2hi, in which all three of them were mutated to even more distinct sequences, were able to replicate to varying levels in agro-inoculated primary leaves. Moreover, some of the mutants, such as m1, m2, m2c, m2cg2h, and m2hi, were detectable in systemic leaves and incurred de novo mutations. Yet others, such as m1a/b, m2a/b, m2bg, m2bh, m2bg2h, and m2c2, elicited systemic infections despite extensive perturbation of some or all iteron motifs. Together, our results indicate that specific iteron sequences are not essential for TYLCV replication in infected cells. It will be interesting to find out whether these findings can be repro duced in tomato, the natural host of TYLCV. As kindly reminded by one reviewer of the previous submission, N. benthamiana has been found to be highly permissive to many geminiviruses, including TYLCV. Thus, some of our mutants that infected N. benthamiana systemically might have a harder time doing so in tomato.
## Heterologous iterons without a matching Rep repress TYLCV replication
An important revelation of our study is that iteron motifs by themselves most likely repress replication of the cognate geminivirus. To explain our reasoning, let us first note that Rep proteins are thought to contain two separate domains mediating iteron binding (11). The first domain corresponds to the N-terminal 8-10 aa, whereas the second maps to aa 66-75 of SH2 Rep or 64-73 of Y35 Rep (Fig. S1). The m1 mutant acquired from Y35, along with the iteron-containing non-coding region, contains the N-terminal iteron-binding domain of Y35 Rep. Thus, a low level of iteron-Rep binding may still occur between Y35 iterons and the matching Rep N-terminus. While the m1 mutant reproducibly elicited systemic symptoms, the descendant viruses always contained the C-to-T de novo mutation at the fourth position of middle iteron, converting the motif from GGTCCT to GGTTCT. Put differently, a single mutation within a 105-nt heterologous sequence was enough to rescue robust infections. Therefore, the imported Y35 iterons must have repressed viral replication, necessitating the C-to-T change to escape the repression.
This interpretation is further supported by the fact that the m2 mutant, in which the imported Y35 iterons had to co-exist with the non-matching SH2 Rep, failed systemic infections in most plants. Instead, only the descendants that acquired one or two de novo mutations within the middle iteron (GGTCCT to GGTTCT or GGGTCT, m2a and m2b) regained symptomatic infections. Therefore, compared to m1, the further absence of the Y35 Rep N-terminus in m2 caused m2 to be more debilitated, making it neces sary to acquire two mutations to relieve the repression imposed by Y35 iterons. We hasten to note that the original m1 and m2 mutants were both able to replicate in cells they first entered, and we surmise that such low-level replication was necessary for de novo mutations to emerge and proliferate. Moreover, the possibility of their low-level co-existence in SLs with the dominant, de novo mutation-containing derivatives cannot be ruled out. Nevertheless, the original mutants by themselves are unlikely to be competent for efficient systemic spread.
To reiterate, even though the m1 and m2 mutants differed from the wild-type SH2 by sequence stretches of 105 and 60 nt, respectively, acquiring one and two de novo mutations within the imported sequences was enough to restore symptomatic infections. Especially, given that both the proximal and distal motifs still retained Y35 sequences in the infection-generated m1 and m2 descendants, it is unlikely that a stimulative role could be regained by altering just one of two nt. It is much more conceivable to foresee such minor changes relieving a certain repressive activity conferred by iterons and/or nearby sequences.
## The de novo mutations likely overcome the iteron-mediated repression by evading certain replication-blocking features of iterons and/or nearby DNA sequence(s)
If iterons by themselves repress replication, it can be inferred that in wild-type TYLCV infections, the cognate Rep acted to defeat this repression through iteron-Rep binding, thereby facilitating genome replication (30,31). How could iteron-mediated repression be defeated when a matching Rep was unavailable? Results with our mutants illustrated that in the absence of a matching Rep, iteron-mediated repression could be overcome by incurring one or two de novo mutations within the middle iteron motif. It is worth repeating that the new middle motifs (GGTTCT, GGGTCT, and GGATCT) created through de novo mutations did not match that of wild-type SH2 (GGTCTC), thus unlikely to have rescued the original iteron-Rep binding. Further disputing the involvement of a specific sequence was that one of the new middle motifs, GGGTCT, potentiated symptomatic infections even when both proximal and distal iterons (of Y35 origin) were eliminated (the m2bg2h mutant). Conversely, restoring the middle motif to that of SH2 in the Y35 context was insufficient to rescue symptomatic infections (the m2e mutant), even after the Y35 proximal and distal iterons were both eliminated (the m2eg2h mutant). Thus, instead of recreating a new sequence motif to suit SH2 Rep, the de novo mutations must have bolstered systemic infections by overcoming repression conferred by the Y35 iterons. As an aside, we consider it unlikely that the mutations we introduced could have disrupted viral movement without compromising viral replication because they were all introduced in the Rep-proximal noncoding region, which is quite distant from the coding sequences of V1 and V2 responsible for intra-and intercellular spread of TYLCV.
How do Y35 iteron motifs repress TYLCV replication? To resolve this puzzle, we considered the possibility that iteron motifs, or other sequence motifs that either overlap with iterons or are tightly linked to them, serve as the binding sites for host-encoded transcription factors. The rationale for this idea is that recruitment of TFs to iterons, which are closely linked to the TATA box of the promoter driving Rep mRNA transcrip tion, likely maximizes Rep production immediately after viral genomes enter host cell nuclei (2). Nevertheless, heavy TF attachment to this iteron-containing genome section likely blocks the same section from being accessed by Rep and other replication-related proteins. Consistent with this idea, iteron motifs are invariably found upstream of the Rep coding sequence (8). Furthermore, their repetition for three to four times within a relatively short stretch closely resembled the arrangement of TF binding sites in the promoters or enhancers of many cellular genes (32)(33)(34). For example, the 6-nt auxinresponsive promoter element (AuxRE, TGTCTC) was frequently found multiple times in promoters of auxin-responsive genes, in the form of tandem and/or inverted repeats. Indeed, synthetic promoters with repeated AuxRE motifs were shown to be much more potent than native ones and were frequently used to screen TFs that bind to the motif (32,35).
Further supporting this idea was the identification of diverse TF binding sites within genomes of various geminiviruses, many of which were shown to mediate transcriptional enhancement of downstream genes (31,36,37). Particularly relevant to our discussion is the G-box motif (CACGTG) found in the TGMV DNA-A segment, within the 5′ non-cod ing region of the TGMV-encoded Rep (AL1). G-box motifs are binding sites for TFs of two large families-the basic helix-loop-helix (bHLH) family and basic leucine zipper (bZIP) family (13,30,38). The G-box motif in TGMV DNA-A is separated from iterons by a 28-nt region that contains the TATA box (13). This G-box-plus-TATA-box-plus-iterons region was verified as a functional promoter driving strong transcription of a reporter gene (13,30,31). Strikingly, this promoter activity was all but abolished when the G-box motif was mutated, suggesting the involvement of G-box-binding TFs in AL1 mRNA transcription (31). Even more interestingly, this promoter activity was potently repressed by the TGMV AL1 protein through sequence-specific iteron-AL1 binding (30). These observations strongly suggest that TF binding to G-box motif activates AL1 mRNA transcription, whereas Rep binding to iteron motifs represses the same transcription activity, likely by peeling TFs off the promoter through competitive binding (31).
The non-coding region of the SH2 genome encompassing iterons does not contain a G-box motif. However, it does contain two consecutive repeats of a different motif, AATTCAAA, 36 nt upstream of iterons. AATTCAAA is a near-perfect match for the binding sites of three Arabidopsis TFs: TSO1, TCX2, and TCX3 (https://jaspar.elixir.no/search? q=&collection=CORE&tax_group=plants). TSO1, TCX2, and TCX3 are closely related TFs of the CPP (cysteine-rich polycomb-like) family found to play crucial roles in Arabi dopsis reproduction by coordinating the cell fate determination in both male and female reproductive organs and controlling stem cell division (39,40). Separately, the middle iteron motif of SH2, if extended by 2 nt upstream, has the sequence of TCGGTGTC, or GACACCGA on the complementary strand. Within these 8 nt, TCGGTG/ CACCGA is highly enriched in the binding sites of multiple ethylene response factors (ERFs), including ERF011, 019, 037, 038, and 043. Conversely, the Y35 middle iteron, along with two upstream nt, would have the sequence of TGGGTCCT/AGGACCCA, containing TGGGTCC/GGACCCA, a motif enriched in the binding sites of multiple TCP TFs, including Arabidopsis TCP7, 9, 21, and 22 (https://jaspar.elixir.no/search?q=&collec tion=CORE&tax_group=plants). TCPs are TFs implicated in diverse processes such as floral symmetry control, branching, lateral organ development, as well as defense responses (41). Notably, the TGGGTCC motif would have been changed by m2b mutations into TGGGGTC, possibly compromising the binding by the same set of TCPs.
## The antagonistic binding of iterons by TFs and Reps hypothesis
Together, these observations prompted us to propose the antagonistic binding of iterons by TFs and Reps (ABITR) hypothesis (Fig. 8). ABITR postulates that iterons are tightly linked to, or overlap with, binding sites of certain host-encoded TFs. Geminiviruses evolve TF-binding sites to lure TFs to the Rep promoter, ensuring rapid and efficient production of Rep protein. Yet, occupation of the Rep promoter by TFs also makes the occupied genome copies recalcitrant to replication proteins. This, in turn, selects for iteron motifs nearby, facilitating iteron-Rep interaction that peels off TFs, availing the TF-free genome copies for replication (Fig. 8).
This hypothesis predicts that in the absence of a matching Rep, geminiviral genomes are trapped in a state that is actively transcribed but difficult to transit to replication. Consistent with this prediction, we found that the m2 mutant, harboring Y35 iterons but SH2 Rep, transcribed C1 mRNA to levels exceeding wild-type SH2, yet replicated to just 30% of SH2 (Fig. S3). Similarly, the m1 mutant, though transcribing C1 mRNA to levels three times higher than SH2, replicated to levels barely matching SH2 (Fig. S3). This model further predicts that de novo mutations within the TF-binding sites/iterons, such as those in m2b, probably relieved the replicational repression by lessening TF binding to Rep promoters, permitting replication initiation despite the absence of specific iteron-Rep binding.
We hasten to note that the ABITR hypothesis must be rigorously tested with additional investigations to (i) identify the host TFs interacting with the iteron-contain ing noncoding regions of various geminiviruses; and (ii) demonstrate the competitive binding with iteron-containing DNA by the identified TFs and corresponding Reps in vitro and in vivo. Finally, it remains to be determined whether this model also applies to bipartite geminiviruses (e.g., TGMV), where Reps must also function in trans to replicate the B components of the virus genomes.
## Antagonistic iteron binding by TFs and Reps probably bottlenecks geminivi rus replication and drives iteron motif diversification
We previously reported that TYLCV replication was intracellularly bottlenecked, so that among dozens of genome copies that entered a cell, no more than three could initiate replication (19). Such stringent intracellular reproductive bottlenecking is critical for sustaining viral viability as it enables swift purging of defective genome copies harboring lethal and deleterious errors (42)(43)(44). However, it is not yet known how geminiviruses like TYLCV establish such intracellular reproductive bottlenecks. The ABITR arrangement, if verified through additional future research, could constitute an early stage of intra cellular bottlenecking. This is because, upon cellular entry, the iteron-encompassing TF-binding sites on the Rep promoter should be immediately available for binding with host-encoded TFs. Such promoter-TF binding should be sufficiently stable, hence preventing the subsequently synthesized Rep proteins from de-repressing more than a few genome copies. For the bottlenecks to sustain, there is abundant evidence showing that Rep over-accumulation later during geminivirus cellular infection blocks, rather than facilitates, viral replication. Indeed, Rep overexpression has been widely used to engineer resistance to geminiviruses (45)(46)(47)(48)(49). In short, intracellular reproductive bottlenecking of geminiviruses is likely first established with the help of host-encoded TFs and later refortified by over-accumulated Rep proteins. Conversely, novel TF-binding specificity could emerge if a viral genome copy incurs mutations allowing it to escape TF-imposed reproductive bottlenecking, hence replicating to dominance in cells containing Rep-supplying sister copies. However, such bottleneck-evading cheater copies are bound to accumulate excessive numbers of lethal errors through unconstrained replication. The descendant lineages consisting of such cheaters, upon entering new cells, are expected to either cease replication or acquire additional mutations that together recreate binding sites for different TFs. This then set in motion the evolution of new iteron motifs and new binding specificities in Rep. Such a "Red-Queen" race would explain the rapid diversification of geminivirus iterons.
To summarize, our extensive investigations of TYLCV iterons yielded the surprising revelation that specific iteron sequence motifs are not required for viral replication. Rather, they act as repressors of TYLCV replication, and their interaction with cognate Rep protein serves to overcome this repressive activity. Moreover, in the absence of a cognate Rep, the repressive activity of iterons can also be overcome by incurring relatively few (one or two) de novo mutations, thus unveiling a different sequence specificity constraint in addition to iteron-Rep binding. Based on these results, we propose the ABITR model, postulating that sequence motifs in the close vicinity of iterons likely evolve sequence specificity, matching certain host-encoded TFs, thereby recruiting TFs to enhance Rep mRNA transcription and Rep protein production. Once Rep accumulates to a certain concentration threshold, it probably expels TFs from a limited number of viral genome copies through competitive iteron binding, availing these genome copies to bottlenecked replication. Testing predictions of this new ABITR model through future research will likely unveil novel targets for more effective management of crop diseases caused by geminiviruses.
## MATERIALS AND METHODS
## Constructs
The original TYLCV infectious clone (isolate SH2, GenBank accession number: AM282874.1) was kindly provided by Dr. Xueping Zhou of China Institute of Plant Protection (18). The full-length, double-stranded form of the TYLCV genome, with a 50 bp duplication at the 5′ end and another 50 bp duplication at the 3′ end, was subcloned into pAI101, an Escherichia coli-Agrobacterium tumefaciens shuttle vector modified from pCambia1300 in our lab (7,50), resulting in a new TYLCV infectious clone named LM5. The sequence of LM5 encompassing the entire TYLCV DNA (and its modified forms) was verified with Sanger sequencing (51)(52)(53).
Most of the mutants were generated by producing the mutation-containing PCR fragments through overlapping PCR (sequences of the primers used are available upon request). The resulting PCR fragments were cloned into the LM5 infectious clone digested with SgsI and Bsp119I. A few mutants were generated with custom-synthesized, mutation-containing DNA fragments (TWIST Biosciences).
## Agrobacterium infiltration
All DNA constructs destined for testing in N. benthamiana plants were transformed into the electrocompetent A. tumefaciens strain C58C1 via electroporation using the AGR setting on the Bio-Rad Micropulser Electroporator. Briefly, 1 µL of the plasmid DNA (100-250 ng) was mixed with 40 µL of agro cells and maintained on ice until electro poration. After electroporation, 900 µL of SOB media was added, and the suspension was incubated at 28°C for 1 hour. Selection was carried out on solid Terrific Broth (TB) media containing rifampicin, gentamycin, and kanamycin. Successful introduction of the plasmid was confirmed using colony PCR. A single colony confirmed to have the desired plasmid was used to inoculate 3 mL of TB liquid media with the same antibiotics and incubated overnight at 28°C. The culture was diluted 1:100 with fresh TB liquid media and incubated under the same conditions for another night. The second culture was centrifuged at 4,000 rpm for 20 min and resuspended in agroinfiltration buffer (10 mM MgCl 2 , 10 mM MES, and 100 µM acetosyringone). All suspensions were diluted to OD 600 = 0.5 and incubated at room temperature for 3 hours. Agrobacterium suspensions were then mixed and introduced into leaves of young N. benthamiana plants via a small wound, using a needleless syringe.
## Assessing viral replication in ILs with real-time quantitative PCR
To obtain DNA samples for qPCR analysis, three different plants per treatment group were selected, and from each of them one agro-infiltrated leaf (1 wpi) was chosen to serve as a biological replicate. More specifically, 0.1 g of tissues was cut out of each of the chosen leaves and subjected to DNA extraction using the Quick-DNA Plant Kit manu factured by Zymo Research. The concentrations of the DNA samples were adjusted to 10 ng/µL prior to qPCR. Primers for detecting TYLCV (SH2) genomic DNA were SH2-84R (5′-GAGGATGCAATTTGATTGGTTGACA) and SH2-2693F (5′-GGCTATTTGGTAATTTTGTAAA AGTACATTGC), which together should generate a circular-genome-specific fragment of 173 bp. Control primers for calibrating qPCR conditions were NbITS2-qF (5′-CATT CTCATGGTTGCGGTGC) and NbITS2-qR (5′-CATGACGTTTGTCGCTGTGG), which together amplified an N. benthamiana ribosomal DNA fragment of 102 bp. The primers were designed using the NCBI Primer-BLAST tool. The qPCR experiments were carried out using the Luna Universal qPCR Master Mix (New England Biolabs) on a Bio-Rad CFX96 Real-Time System, following the manufacturer's instructions. Prior to the experiments, the efficiencies of both primer pairs were tested using a protocol recommended by the kit: a five-tube dilution series, each a 10-fold dilution of the prior one, was prepared from a sample known to contain SH2 genomes and subjected to PCR to generate a standard curve through linear regression analysis. The efficiencies were determined to be 87.14% for SH2-84R/2693F and 91.51% for NbITS2-qF/qR. The data obtained were analyzed with R Studio. The threshold cycle (Ct) values of TYLCV variants were normal ized against SH2 (wt). Fold change data were calculated with the 2 -ΔΔCt method. The fold change data were then subjected to normality test and constant variance test prior to ANOVA and Tukey's honestly significant difference treatments to determine the statistical significance of the differences among the treatments.
## Detecting de novo mutations via Sanger sequencing
The systemic leaf DNA samples were obtained with the Quick-DNA Plant Kit (Zymo Research). For the purpose of Sanger sequencing, a pair of PCR primers that amplified the entire Rep coding region plus the intergenic region between Rep and V2 were used (SH2-1475F and SH2-196R; Fig. 1A; sequences of primers are available upon request). PCR was usually repeated for 35 cycles, and the amplified fragments were sequenced by the OSU Genomics Core Facility.
## Supplemental Material
## References
1. Zerbini, Briddon, Idris et al. (2017) "ICTV virus Taxonomy profile: Geminiviridae" *J Gen Virol*
2. Hanley-Bowdoin, Bejarano, Robertson et al. (2013) "Geminiviruses: masters at redirecting and reprogramming plant processes" *Nat Rev Microbiol*
3. Stenger, Revington, Stevenson et al. (1991) "Replica tional release of geminivirus genomes from tandemly repeated copies: evidence for rolling-circle replication of a plant viral DNA" *Proc Natl Acad Sci*
4. Lazarowitz, Shepherd (1992) "Geminiviruses: genome structure and gene function" *CRC Crit Rev Plant Sci*
5. Orozco, Hanley-Bowdoin (1996) "A DNA structure is required for geminivirus replication origin function" *J Virol*
6. Urbino, Thébaud, Granier et al. (2008) "A novel cloning strategy for isolating, genotyping and phenotyping genetic variants of geminiviruses" *Virol J*
7. Lin, Guo, Finer et al. (2014) "The bean pod mottle virus RNA2-encoded 58-kilodalton protein P58 is required in cis for RNA2 accumulation" *J Virol*
8. Argüello-Astorga, Guevara-González, Herrera-Estrella et al. (1994) "Geminivirus replication origins have a groupspecific organization of iterative elements: a model for replication" *Virology (Auckl)*
9. Argüello-Astorga, Ruiz-Medrano (2001) "An iteron-related domain is associated to Motif 1 in the replication proteins of geminiviruses: identification of potential interacting amino acid-base pairs by a comparative approach" *Arch Virol*
10. Zhang, Xu, Huang et al. (2016) "A novel DNA motif contributes to selective replication of a geminivirus-associated betasatellite by a helper virus-encoded replication-related protein" *J Virol*
11. Londoño, Riego-Ruiz, Gr (2010) "DNA-binding specificity determinants of replication proteins encoded by eukaryotic ssDNA viruses are adjacent to widely separated RCR conserved motifs" *Arch Virol*
12. Xu, Qian, Wang et al. (2019) "Iterons homologous to helper geminiviruses are essential for efficient replication of betasatellites" *J Virol*
13. Fontes, Luckow, Hanley-Bowdoin (1992) "A geminivirus replication protein is a sequence-specific DNA binding protein" *Plant Cell*
14. Fontes, Gladfelter, Schaffer et al. (1994) "Geminivirus replication origins have a modular organization" *Plant Cell*
15. Fontes, Eagle, Sipe et al. (1994) "Interaction between a geminivirus replication protein and origin DNA is essential for viral replication" *J Biol Chem*
16. Gladfelter, Eagle, Fontes et al. (1997) "Two domains of the AL1 protein mediate geminivirus origin recognition" *Virology (Auckl)*
17. Preiss, Jeske (2003) "Multitasking in replication is common among geminiviruses" *J Virol*
18. Zhang, Gong, Zhou (2009) "Molecular characterization and pathogenicity of tomato yellow leaf curl virus in China" *Virus Genes*
19. Ren, Zheng, Han et al. (2023) "Intracellular bottlenecking permits no more than three tomato yellow leaf curl virus genomes to initiate replication in a single cell" *PLOS Pathog*
20. Prasad, Sharma, Hari-Gowthem et al. (2020) "Tomato yellow leaf curl virus: impact, challenges, and management" *Trends Plant Sci*
21. Zhou (2013) "Advances in understanding begomovirus satellites" *Annu Rev Phytopathol*
22. Fondong (2013) "Geminivirus protein structure and function: geminivirus proteins" *Mol Plant Pathol*
23. Orozco, Hanley-Bowdoin (1998) "Conserved sequence and structural motifs contribute to the DNA binding and cleavage activities of a geminivirus replication protein" *J Biol Chem*
24. Orozco, Miller, Settlage et al. (1997) "Functional domains of a geminivirus replication protein" *J Biol Chem*
25. Orozco, Kong, Batts et al. (2000) "The multifunctional character of a geminivirus replication protein is reflected by its complex oligomerization properties" *J Biol Chem*
26. Zrachya, Glick, Levy et al. (2007) "Suppressor of RNA silencing encoded by Tomato yellow leaf curl virus-Israel" *Virology (Auckl)*
27. Rojas, Jiang, Salati et al. (2001) "Functional analysis of proteins involved in movement of the monopartite begomovirus, Tomato Yellow Leaf Curl Virus" *Virology (Auckl)*
28. Hanley-Bowdoin, Elmer, Rogers (1990) "Expression of functional replication protein from tomato golden mosaic virus in transgenic tobacco plants" *Proc Natl Acad Sci*
29. Zhao, Wu, Barton et al. (2020) "Tomato Yellow Leaf Curl Virus V2 protein plays a critical role in the nuclear export of V1 protein and viral systemic infection" *Front Microbiol*
30. Eagle, Orozco, Hanley-Bowdoin (1994) "A DNA sequence required for geminivirus replication also mediates transcriptional regulation" *Plant Cell*
31. Eagle, Hanley-Bowdoin (1997) "cis elements that contribute to geminivirus transcriptional regulation and the efficiency of DNA replication" *J Virol*
32. Chapman (2009) "Mechanism of auxin-regulated gene expression in plants" *Annu Rev Genet*
33. Ezer, Zabet, Adryan (2014) "Homotypic clusters of transcription factor binding sites: a model system for understanding the physical mechanics of gene expression" *Comput Struct Biotechnol J*
34. (2025) *Full-Length Text Journal of Virology*
35. Espley, Brendolise, Chagné et al. (2009) "Multiple repeats of a promoter segment causes transcription factor autoregula tion in red apples" *Plant Cell*
36. Ulmasov, Hagen, Guilfoyle (1997) "ARF1, a transcription factor that binds to auxin response elements" *Science*
37. Tu, Sunter (2007) "A conserved binding site within the Tomato golden mosaic virus AL-1629 promoter is necessary for expression of viral genes important for pathogenesis" *Virology (Auckl)*
38. Sun, Han, Zheng et al. (2020) "The AC2 protein of a bipartite geminivirus stimulates the transcription of the BV1 gene through abscisic acid responsive promoter elements" *Viruses*
39. Ezer, Shepherd, Brestovitsky et al. (2017) "The G-box transcriptional regulatory code in arabidopsis" *Plant Physiol*
40. Clark, Buckner, Fisher et al. (2019) "Stem-cell-ubiquitous genes spatiotemporally coordinate division through regulation of stem-cell-specific gene networks" *Nat Commun*
41. Andersen, Algreen-Petersen, Hoedl et al. (2007) "The conserved cysteine-rich domain of a tesmin/TSO1-like protein binds zinc in vitro and TSO1 is required for both male and female fertility in Arabidopsis thaliana" *J Exp Bot*
42. Li (2015) "The Arabidopsis thaliana TCP transcription factors: a broadening horizon beyond development" *Plant Signal Behav*
43. Carvalho, Ren, Han et al. (2022) "Natural selection, intracellular bottlenecks of virus populations, and viral superinfection exclusion" *Annu Rev Virol*
44. Qu, Zheng, Zhang et al. (2020) "Bottleneck, Isolate, Amplify, Select (BIAS) as a mechanistic framework for intracellu lar population dynamics of positive-sense RNA viruses" *Virus Evol*
45. Qu, Khemsom, Carvalho et al. (2024) "Quasispecies are constantly selected through virus-encoded intracellular reproductive population bottlenecking" *J Virol*
46. Yang, Sherwood, Patte et al. (2004) "Use of Tomato yellow leaf curl virus (TYLCV) Rep gene sequences to engineer TYLCV resistance in tomato" *Phytopathology*
47. Shepherd, Dugdale, Martin et al. (2014) "Inducible resistance to maize streak virus" *PLoS One*
48. Brunetti, Tavazza, Noris et al. (2001) "Transgenically expressed T-rep of Tomato yellow leaf curl Sardinia virus acts as a trans-dominant-negative mutant, inhibiting viral transcription and replication" *J Virol*
49. Sardo, Lucioli, Tavazza et al. (2011) "An RGG sequence in the replication-associated protein (Rep) of Tomato yellow leaf curl Sardinia virus is involved in transcriptional repression and severely impacts resistance in Rep-expressing plants" *J Gen Virol*
50. Hanson, Maxwell (1999) "trans-dominant inhibition of geminiviral DNA replication by bean golden mosaic geminivirus rep gene mutants" *Phytopathology*
51. Zhang, Sun, Guo et al. (2019) "Repression of turnip crinkle virus replication by its replication protein p88" *Virology (Auckl)*
52. Guo, Zhang, Sun et al. (2020) "Superinfection exclusion by p28 of turnip crinkle virus is separable from its replication function" *Mol Plant Microbe Interact*
53. Qu, Ren, Morris (2003) "The coat protein of turnip crinkle virus suppresses posttranscriptional gene silencing at an early initiation step" *J Virol*
54. Sun, Zhang, Zheng et al. (2020) "Translation-independent roles of RNA secondary structures within the replication protein coding region of turnip crinkle virus" *Viruses*
55. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12509547&blobtype=pdf | # Closed genome sequence of an Arthrobacter globiformis phage, SilentRX
Ashna Siddiqui, Raunak Vijay, Ethan Edwards, Meliha Ulker, Fardeen Siddiqui, Thomas Gerton, Rachel Anastasi, Dylan Conroy, Isabelle Laizure, Joshua Reynolds, Kelsie Duggan, Kristen Johnson, Kyle Maclea
## Abstract
We sequenced the complete genome of SilentRX, an actinobacteriophage with siphovirus morphology in cluster AP that infects Arthrobacter globiformis NRRL B-2979. The phage possessed a capsid width of 55 nm and a tail length of 215 nm. With a length of 65,232 bp, the genome contained 102 predicted protein-coding genes.
lytic protein endolysin (15). Furthermore, there is no evidence of a temperate lifestyle since it lacks obvious genes for lysogeny (16).
Bacteriophages typically exhibit an equal or reduced GC content percentage compared to their hosts (17)(18)(19). A higher GC content of a phage genome or gene could be evidence of horizontal gene transfer (16,17). SilentRX has an average GC content of 67.12% and A. globiformis NRRL B-2979 has an average GC content of 66.2%. Gene 13 of SilentRX, a putative DNA-binding protein, has a higher GC content (73.54%). This could be the result of foreign, horizontally transferred DNA. SilentRX shows genetic diversity, including orphan homologs and GC shifts, suggesting gene acquisition from other sources.
## References
1. Hatfull (2015) "Innovations in undergraduate science education: going viral" *J Virol*
2. Poxleitner, Pope, Jacobs-Sera et al. (2018) "HHMI SEA-PHAGES phage discovery guide"
3. Ackermann (2009) "Basic Phage Electron Microscopy"
4. Jordan, Burnett, Carson et al. (2014) "A broadly implementable research course in phage discovery and genomics for first-year undergraduate students" *mBio*
5. Russell (2018) "Sequencing, assembling, and finishing complete bacteriophage genomes" *Methods Mol Biol*
6. Pope, Jacobs-Sera (2018) "Annotation of bacteriophage genome sequences using DNA Master: an overview" *Methods Mol Biol*
7. Delcher, Harmon, Kasif et al. (1999) "Improved microbial gene identification with GLIMMER" *Nucleic Acids Res*
8. Besemer, Borodovsky (2005) "GeneMark: web software for gene finding in prokaryotes, eukaryotes and viruses" *Nucleic Acids Res*
9. Cresawn, Bogel, Day et al. (2011) "Phamerator: a bioinformatic tool for comparative bacteriophage genomics" *BMC Bioinformatics*
10. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol*
11. Söding, Biegert, Lupas (2005) "The HHpred interactive server for protein homology detection and structure prediction" *Nucleic Acids Res*
12. Laslett, Canback (2004) "ARAGORN, a program to detect tRNA genes and tmRNA genes in nucleotide sequences" *Nucleic Acids Res*
13. Chan, Lowe (2019) "tRNAscan-SE: searching for tRNA genes in genomic sequences" *Methods Mol Biol*
14. Klyczek, Bonilla, Jacobs-Sera et al. (2017) "Tales of diversity: genomic and morphological characteristics of forty-six Arthrobacter phages" *Plos One*
15. Shi, Li, Yan et al. (2012) "Combined antibacterial activity of phage lytic proteins holin and lysin from Streptococcus suis bacteriophage SMP" *Curr Microbiol*
16. Demo, Kapinos, Bernardino et al. (2021) "BlueFeather, the singleton that wasn't: shared gene content analysis supports expansion of Arthrobacter phage Cluster FE" *PLoS One*
17. Ravenhall, Škunca, Lassalle et al. (2015) "Inferring horizontal gene transfer" *Plos Comput Biol*
18. Bohlin, Pettersson (2019) "Evolution of genomic base composition: from single cell microbes to multicellular animals" *Comput Struct Biotechnol J*
19. De Melo, Da Mata Gomes, Melo et al. (2019) "Characteri zation of a bacteriophage with broad host range against strains of Pseudomonas aeruginosa isolated from domestic animals" *BMC Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12153123&blobtype=pdf | # Correction: A full-length S1 gene sequencing of a novel emerged GI-19 and GI-23 lineages of Infectious bronchitis virus currently circulating in chicken flocks in upper Egypt reveals marked genetic diversity and recombination events
Eman Shosha, Sara Abdelnaser, Ali Zanaty
Correct Tables 4 and5 Table 4 Nucleotide identities and divergence of partially sequenced ACoV isolates comparable to other selected Egyptian and referential strains Amino acids and Nucleotide identities and divergence of our partially sequenced ACoV isolates comparable to other selected strains including vaccinal strains. The table utilizes a comparative alignment of the S1 gene in which, the S1 nucleotide identity percentage of our six Egyptian isolates (GI-23, GI-1, GI-12) ranges from 79 to 100% comparable to other referential strains. Besides, the amino acids identity percentage of these isolates ranges from 97 to 100% comparable to various referential strains
Table 5 Nucleotide and amino acids identities of full-length sequenced ACoV isolates comparable to other selected Egyptian and referential strains
Nucleotide and amino acids identities of our full-length sequenced ACoV isolates comparable to other selected strains. The table includes a comparative alignment of the S1gene in which, the S1 nucleotide and amino acids identity percentages of our five Egyptian isolates (GI-23, GI-19) range from 99 to 100% comparable to other referential strains
## References
1. Shosha, Abdelnaser, Zanaty (2025) "A full-length S1 gene sequencing of a novel emerged GI-19 and GI-23 lineages of Infectious bronchitis virus currently circulating in chicken flocks in upper Egypt reveals marked genetic diversity and recombination events" *Virol J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12446860&blobtype=pdf | # Summary of taxonomy changes ratified by the International Committee on Taxonomy of Viruses (ICTV) -General taxonomy proposals, 2025
F Murilo Zerbini, Anya Crane, Jens Kuhn, Peter Simmonds, Elliot Lefkowitz, Ictv Taxonomy, Summary Consortium, F Zerbini
## INTRODUCTION
Three taxonomy proposals of a general nature were approved by the International Committee on Taxonomy of Viruses (ICTV) in 2024. Two technical proposals were presented: the first proposed an updated and uniform system for naming proposals for error correction of accepted taxonomic proposals to ensure that they are readily accessible on the ICTV website (2024.001G). The second proposed to reformat the ICTV statutes and codes to produce a more unified text after numerous piecemeal changes to both documents made in previous years (2024.002G). The revised code and statutes are available at https:// ictv. global/ about/ code and https:// ictv. global/ about/ statutes.
The ICTV Executive Committee (EC) proposed that Professor Stuart Siddell be made a Life Member of the ICTV (2024.003G). Stuart received a B.Sc. degree in 1972, a B.Sc. degree (Hons, Class I) in Botany from the University of Liverpool, UK, in 1973, a Ph.D. in Biological Sciences from the University of Warwick, UK, in 1976, and a Dr. rer. nat. habil. degree (Biochemie) from the University of Würzburg, Germany. After positions at the Imperial Cancer Research Fund, UK, and the University of Würzburg, Germany, Stuart continued his career at the University of Bristol as a Professor of Virology from 1992 until his retirement in 2015, when he became an Emeritus Professor of Virology. His research focused on coronaviruses and played a role in the discovery of coronavirus discontinuous transcription; the development of reverse genetic systems for human, avian and murine coronaviruses; the elucidation of coronavirus protein structures; and constructing a genetic map of coronavirus replicase protein function [1,2].
Stuart began his work with the ICTV in 1982 on the Coronaviridae Study Group, serving as chair and subsequently as a member until 2000. He also served as an ICTV National Member for the United Kingdom from 1996 to 2011. Stuart served on the EC as an elected member from 2008 to 2014, Chair of the Animal dsRNA and ssRNA(-) Viruses Subcommittee from 2014 to 2017, and then as Vice-President until 2023. Stuart was also Editor-in-Chief of the ICTV 10th Report, helping to guide the transition of this key ICTV product from a hard-copy book to an online web-based resource [3]. Overall, Stuart provided over 40 years of service to the ICTV, including 16 years on the EC.
## Abstract
During the 56th annual meeting of the International Committee on Taxonomy of Viruses (ICTV), held in Bari, Italy, in August 2024, two technical proposals were presented. The first called for amended versions of accepted taxonomic proposals to be named in such a way to ensure that they are readily accessible on the ICTV website (2024.001G). The second proposed a substantial reformatting of the ICTV statutes and codes to produce a more unified text after the numerous changes made to both documents in previous years (2024.002G). Finally, the ICTV Executive Committee (EC) nominated Professor Stuart Siddell as a Life Member of the ICTV for his work over four decades on virus taxonomy, including 16 years as a member of the EC (2024.003G).
During his tenure on the EC, Stuart was a key participant in several significant changes to the processes, methods and intellectual underpinning of the approaches used by the ICTV to guide the classification of viruses and name the resulting taxa. These included the creation of higher rank taxonomic categories that support the deeper evolutionary classification of viruses [4], the establishment of evolutionary history as a key guiding principle underlying the ICTV taxonomy [5] and consultation on the use of binomial names for virus species [6] that paved the way for their recent adoption by the ICTV.
He was also responsible for creating and maintaining the Virus Metadata Resource (VMR), a spreadsheet that, for each species, lists an exemplar virus, along with a suggested abbreviation, a GenBank accession for the genome sequence, and general details about genome composition and type of host (https:// ictv. global/ vmr). The VMR is now a key resource in virology. More recently, Stuart has been responsible for collating information on the etymology of virus taxa for ranks at the family level and above (https:// ictv. global/ taxonomy/ etymology); curating a collection of virion diagrams organized by family (https:// ictv. global/ viriondiagrams); and assembling a table of virus genome, virion properties and type of host for all virus families (https:// ictv. global/ virus-properties).
## MAIN TEXT CONTENTS
2024.001G.Name_format_of_Taxonomy_Proposal_Corrections 2024.002G.ICVCN_and_Statutes_harmonization 2024.003G.Nomination_Stuart_Siddell_as_Life_Member
## 2024.001G.Name_format_of_Taxonomy_Proposal_Corrections
Title: Name format for expedited corrections of taxonomy proposals Authors: Simmonds P ( Peter. Simmonds@ ndm. ox. ac. uk), Zerbini M, Lefkowitz EJ
## Summary
## Brief description of current situation
The ratified taxonomy proposal (TP) 2020.002G.Expedited_error_correction describes how errors in the proposal spreadsheet may be corrected with approval from the ICTV President, Data Secretary and Proposals Secretary without the need to re-submit a corrected TP in the next ICTV cycle. However, it was not specified how the name used for the correction TP should be formatted.
## Proposed changes
We propose that the TP code will be suffixed with an X, and the name suffixed with the term '_Error_Correction' . Justification It is useful to use a standard format that unambiguously links the original and the correction TPs together. *Source/full text: https:// ictv. global/ ictv/ proposals/ 2024. 001G. Name_ format_ of_ Taxonomy_ Proposal_ Corrections. docx
## 2024.002G.ICVCN_and_Statutes_harmonization
Title: Revise the ICTV Statutes and the ICVCN Authors: Zerbini M ( zerbini@ ufv. br), Lefkowitz EJ ( elliotl@ uab. edu), Crane A, Kuhn J ( jenshkuhn@ comcast. net)
## Summary
## Brief description of current situation
In recent years, numerous individual and several bulk changes were made to the ICTV International Code of Virus Classification and Nomenclature (ICVCN) and the ICTV Statutes. These changes left both documents somewhat in disarray, including problems associated with terminology consistency, orthography and grammar, clarity and application examples.
## Proposed changes
We propose several changes to both documents to improve their qualities. *Source/full text: https:// ictv. global/ ictv/ proposals/ 2024. 002G. ICVCN_ and_ Statutes_ harmonization. docx
## References
1. Siddell, Anderson, Cavanagh et al. (1983) *Intervirology*
2. Sawicki, Sawicki, Siddell (2007) "A contemporary view of coronavirus transcription" *J Virol*
3. Lefkowitz, Dempsey, Hendrickson et al. (2018) "Virus taxonomy: the database of the International Committee on Taxonomy of Viruses (ICTV)" *Nucleic Acids Res*
4. Siddell, Walker, Lefkowitz et al. (2018) "Additional changes to taxonomy ratified in a special vote by the International Committee on Taxonomy of Viruses" *Arch Virol*
5. Simmonds, Adriaenssens, Zerbini et al. (2023) "Four principles to establish a universal virus taxonomy" *PLoS Biol*
6. Siddell, Walker, Lefkowitz et al. (2020) "Binomial nomenclature for virus species: a consultation" *Arch Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12783320&blobtype=pdf | # Evaluation of cellular immune response and biosafety of SV40 viruslike particle in tumor immunotherapy
Ting He, Ruoxuan Hei, Chong Liu, Huiping Wang, Zhaowei Gao, Ke Dong, Juan Zhang
## Abstract
Virus-like particle (VLP) holds great promise for applications in vaccines and tumor immunotherapy. However, their clinical translation has been limited by a lack of comprehensive in vivo studies on immune responses and antigenic toxicity. In this study, we systematically evaluated the efficacy and safety of VLP as immunological agents. We administered Simian Virus 40 (SV40) VLP through subcutaneous injection and analyzed their effects on immune cell populations in key organs. In vivo imaging of mice demonstrated the migration of SV40 VLP between lymph nodes. Flow cytometry revealed that SV40 VLP significantly increased the numbers of CD4 + T cells and NK cells in the spleen, along with elevated levels of CD4 + T cells in mesenteric lymph nodes. Moreover, SV40 VLP did not significantly affect immune cell populations in the lungs, liver, or kidneys, nor did they alter blood biochemistry or coagulation parameters. Although SV40 VLP alone did not exhibit tumortreating effects, in vitro imaging suggest that SV40 VLP can target tumor tissues and and quantitative analysis showed SV40 VLP significantly increased TNF-α expression in spleen. These findings suggest that SV40 VLP represent a promising tumor immunotherapy vector with potential for further modification.
## 1 Introduction
Virus-like particles are three-dimensional protein structures formed by the reversible self-assembly of capsid proteins of viruses, either in vivo or in vitro [1][2][3]. VLP cannot infect hosts because they do not contain genetic material [4,5], making them widely used in drug delivery [6], biosensing [7], diagnostics [8] and other applications [9]. The outer surface proteins of VLP are similar to the capsid proteins of viruses, thereby can induce the immune system to produce an antibody response [10]. This characteristic gives VLP significant potential for use in vaccine development [11] and tumor immunotherapy [12]. Although research interest based on VLP within the field of biomedicine continues to grow [13][14][15], the immune response and biosafety of VLP in the biological systems have not been fully elucidated. Therefore, there is an urgent need to investigate the cellular immune response and biosafety of VLP in vivo, particularly in major organs.
VLP come in various shapes and range in size from 20 nm to 200 nm. Studies have shown that VLP can be passively drained to the subcapsular sinus of lymph nodes through lymphatic vessels, where they are recognized and ingested by dendritic cells (DC). Then these VLP are presented via major histocompatibility complexes (MHC) to activate T cell-mediated immune responses [16,17]. Additionally, VLP can also promote the maturation of DC and stimulate the secretion of pro-inflammatory factors and chemokines, which helps mobilize more antigen-presenting cells to activate immune response [18]. For instance, engineered glycan-costumed VLP can reprogram DC and significantly inhibit tumor growth, as well as the reintroduced cancer cells [19]. And recently, Yang et al. reported a phage MS2-derived VLP-based nanotube hydrogel can induce T cell responses, leading to a dramatic increase in the production of Granzyme B and TNF-α in lymphocytes of spleen [20]. Therefore, the considerable potential of VLP in tumor immunotherapy should not be overlooked.
In addition, VLP are easily modified through bioengineering [21,22], have low production costs [23], and exhibit good stability [24], making them highly favored in the field of vaccine development. VLP-based vaccines can be enhanced immunogenicity and protective efficacy through reformation. Young-Tae Lee et al. engineered Sindbis-virus glycoprotein exhibited VLP with mRNA packaging, which could protect the mice from infections with SARS-CoV-2 or herpes simplex virus type 1 [25]. Furthermore, VLP can be engineered to carry multiple antigens for improving vaccine coverage. Jingdi Pan et al. constructed multivalent epitope-presenting VLP vaccine with broad protection against divergent influenza A and B viruses [26]. However, among these studies, there has been relatively little focus on the biosafety of VLP, especially regarding their cellular immune response and potential antigenic toxicity in major organs.
The natural capsid of simian virus 40 (SV40) is composed of three major structural proteins: VP1, VP2, and VP3. Therein, the main capsid protein VP1 can selfassemble into an icosahedral virus-like particle in vitro [27][28][29][30]. SV40 VLP is frequently used as a model due to their advantages of easy surface functionalization, biocompatibility, and biodegradability [31,32]. In this work, we investigated the influence of SV40 VLP on the immune cells quantity in major organs based on tumor immunotherapy. In vivo imaging revealed that SV40 VLP remained in the inguinal region one day after injection and significantly enriched in the inguinal lymph nodes for up to 5 days. Furthermore, we observed a significant increase in the number of CD4 + T cells and NK cells in the spleen, along with a rise of CD4 + T in the mesenteric lymph nodes. It indicated that SV40 VLP can migrate via lymphatic drainage and activate cellular immunity. Most importantly, SV40 VLP did not increase the number of immune cells in the lungs, liver, or kidneys, and blood biochemistry and coagulation function remained normal after multiple subcutaneous injections. This indicates the low antigenic toxicity and favorable biosafety profile of SV40 VLP. Although SV40 VLP failed in the lung cancer tumor suppression, the significant tumor-targeted enrichment and the increase of TNF-α in spleen suggest that SV40 VLP is a promising protein nanoparticle worthy of optimization and modification. Our study establishes a theoretical foundation for the clinical transformation of VLP in immunotherapy application, further emphasizing the significant potential of SV40 VLP in clinical settings.
## 2 Materials and methods
## 2.1 Materials
Rosstta-PET32a-5hcVP1 strain was provided by professor Feng Li, fluorescent dye Alexa 647 was purchased from Invitrogen (OR, USA), APC/Cy7 anti-mouse CD3, PE-Cy7 anti-mouse CD4, PE anti-mouse CD8, FITC anti-mouse CD45, APC/Cy7 anti-mouse CD19, Pe-Cy7 anti-mouse CD11b, FITC anti-mouse CD11c, PE antimouse NK1.1, PE anti-mouse MHC2 are all purchased from 4A-BIOTECH. Matrigel (Cultrex Pathclear BM) was purchased from R&D (USA). Lymphocyte Separation Kit and RIPA lysis buffer was purchased from Beyotime Biotechnology (Shanghai, China). Mouse TNF-alpha High Sensitivity ELISA Kit (RK04875) was purchased from ABclonal Biotechnology Co., Ltd. (Wuhan, China).
## 2.2 Cell and cell culture
## 2.4 Expression of SV40 Capsid Protein VP1
Rosstta-PET32a-5hcVP1 was cultured in 5 mL of LB medium containing ampicillin and chloramphenicol at 37 °C and 220 rpm for 9 h, then transferred to 500 mL of LB medium at the same temperature and speed, and cultured for an additional 4 hours until the OD 600 reached approximately 0.6. Afterward, IPTG (1 mM) was added to induce the protein expression at 25 °C and 160 rpm for 12 h. The bacteria were then harvested and lysed using ultrasonication. After centrifugation at 4 °C and 10,000 rpm for 30 min (using a Sorvall lynx 6000 Thermo), the supernatant underwent 80% ammonium sulfate precipitation. The precipitate was then centrifuged, and the entire precipitation was re-dissolved in the assembly buffer (1 M NaCl, 10 mM Tris-HCl, 1 mM CaCl 2 , 5% glycerol) with magnetic stirring overnight. Then the mixture was centrifuged again at 4 °C and 50,000 rpm (using an OPTIMA L-100XP Beckman) for 1 h to collect the precipitate, which was then resuspended in the depolymerization buffer (10 mM Tris-HCl, 200 mM NaCl, 10 mM EDTA, 5% glycerol) and centrifuged once more at 5000 rpm at 4 °C for 1 h. The supernatant contained SV40 VP1.
## 2.5 Self-assembly and purification of SV40 VLP
The concentration of SV40 VP1 was determined using gray analysis based on SDS-PAGE. Then 0.5 mg/mL of SV40 VP1(10 mM Tris-HCl, 200 mM NaCl, 10 mM EDTA, 5% glycerol, pH 8.9) was dialyzed (using an 8-14 KDa cutoff) in pre-cooled assembly buffer (1 M NaCl, 10 mM Tris-HCl, 1 mM CaCl 2 , 5% glycerol, pH 7.2) and stirred at 4 °C for at least 36 h. After dialysis, the SV40 VP1 solution was concentrated to 10 mL for sucrose density gradient centrifugation at 4 °C and 38,000 rpm for 4 hours. Then the selfassembled SV40 VLP were in the 4th to 5th layers of sucrose gradient. The purified SV40 VLP was then dialyzed against PBS to remove sucrose and concentration was measured for further use.
## 2.6 Transmission electron microscopy
Samples of SV40 VLP were adsorbed to carbon-coated copper grids for 5 min and stained with 2% phosphotungstic acid for 3 min. The grid was dried at room temperature and observed on a Thermofisher Talos L120C transmission electron microscope at 120 kV with an Olympus camera.
## 2.7 Dynamic Light Scattering Measurement of SV40
VLP Particle Size SV40 VLP (0.5 mg/mL, 200 μL) in PBS was loaded into a micro cell. Then the hydrodynamic diameter was measured by a Malvern Nanosizer Nano ZS DLS instrument (UK) with 10% power and sample was run three times.
## 2.8 Lung cancer model and animal experiments
Before tumor implantation, SV40 VLP was administered subcutaneously twice via the left lower inguinal region of about eight-week old female C57BL/6 mice. Then murine lung cancer Lewis cells were subcutaneously inoculated on the right hind limbs (1 × 10 6 cells in 100 μL Matrigel). After the tumors grew for 4 days, 80 μg of SV40 VLP was repeated injected through the left lower inguinal region for every 4 days and total 6 times. During the entire drug administration period, the tumor size was recorded with a vernier caliper. The tumor volume = 0.5 × length × width 2 .
After the sixth administration, animals were anesthetized with isoflurane and then euthanized. The main organs, lymphoid tissues and tumors were excised for imaging, flow cytometry and immunohistochemical analysis.
## 2.9 In vivo imaging following SV40 VLP injection
SV40 VLP were labeled with Alexa 647 fluorescent dye during the last injection, and the mice were imaged at 1 and 5 days post-injection after isoflurane anesthesia using an IVIS Spectrum In vivo Imaging System from Korea (with 650 nm excitation and 680 emission filters).
## 2.10 Biodistribution of SV40 VLP: subcutaneous vs. intravenous injection
Male C57BL/6 mice were administered a dose of 100 μg of Alexa 647 labeled SV40 VLP via either subcutaneous (inguinal region) or tail vein injection. At 24 h post-injection, the mice were anesthetized with isoflurane and euthanized. Major organs, including the heart, liver, spleen, lungs, and kidneys, were then harvested and rinsed with cold PBS. The biodistribution differences of VLP between the two injection routes were compared using a multimodal in vivo imaging system (AniView100 DXA, Guangzhou Biolight Biotechnology Co., Ltd.).
## 2.11 Flow cytometry analysis
The activation of immune cells in major organs was analyzed after multiple injections using flow cytometry. The axillary and inguinal lymph nodes, mesenteric lymph nodes, lung, liver, spleen and kidney were ground to prepare single-cell suspensions. Red blood cells were lysed using an NH 4 Cl solution and centrifuged at 4000 rpm for 5 min to collect the remaining cells.
After resuspension, the cells were labeled with FITC-conjugated anti-mouse CD45, APC-Cy7-conjugated anti-mouse CD3, PE-Cy7-conjugated anti-mouse CD4 and PE-conjugated antimouse CD8 to identify CD4 + and CD8 + T cells.
To identify NK and B cells, 100 μL of single-cell suspension was labeled with FITC-conjugated anti-mouse CD45, PE-conjugated anti-mouse NK1.1 and APC-Cy7conjugated anti-mouse CD19 antibodies. Dendritic cells were labeled with APC-conjugated anti-mouse CD11c antibody, and FITC-conjugated anti-mouse CD11b. After mixing the cells with antibodies, the mixture was incubated for 30 min in the dark. The samples were then centrifuged at 4000 rpm for 5 min, the supernatant was discarded, and cells were resuspended in saline. Finally, the samples were analyzed using a flow cytometry (BD Company, USA). CD4 + T cells, CD8 + T cells, NK cells, B cells and DC were gated using the following markers: CD45 + CD3 + CD4 + , CD45 + CD3 + CD8 + , CD45 + NK1.1 + , CD45 + CD19 + and CD11c + CD11b + .
## 2.12 Blood biochemical assays and blood coagulation test
80 μg of SV40 VLP were administered for six injections to each mouse, after which blood was collected through cardiac puncture following euthanasia. Serum was separated by centrifugation at 3000 g using EDTA as an anticoagulant. Serum biochemical parameters, including blood and urine nitrogen (BUN), alkaline phosphatase (ALP), alanine aminotransferase (ALT), and aspartate aminotransferase (AST), total bilirubin (TB) were measured using an automatic AU5800 analyzer (Beckman, USA). Prothrombin time (PT) and Prothrombin activity (PTA) were determined using an automatic coagulation CP2000 analyzer (Coapresta® 2000, Japan) following the manufacturer's protocol.
## 2.13 Western Blotting (WB) analysis of the anti-VLP antibody levels
After 8 dose treatment of SV40 VLP based on subcutaneous injection, the mice were anesthetized and euthanized. Blood was collected by cardiac puncture immediately and blood was centrifuged at 3000 rpm for 10 min. The supernatant was separated to prepare serum for WB analysis. Then same amount of SV40 VLP was subjected to SDS-PAGE and transferred to PVDF membrane. The serum of each mouse was separately incubated with the PVFD membrane as a primary antibody (4 mice per group), and the secondary antibody was HRP-labeled goat anti-mouse IgG.
## 2.14 Quantitative determination of TNF-α Levels in the Spleen
Both VLP and PBS were administered via subcutaneous injection into the right inguinal region every other day for a total of six injections. Two days after the final injection, the mice were euthanized, and the spleens were aseptically harvested. The spleens were gently homogenized to generate single-cell suspensions. Splenic lymphocytes were then isolated using a lymphocyte separation medium. Total protein was extracted from the lymphocytes with RIPA lysis buffer. The protein concentration was quantified using a Bradford assay kit and adjusted to a uniform concentration of 3 mg/ml. Finally, the concentration of TNF-α was quantitatively measured according to the manufacturer's instructions of an ELISA kit from ABclonal.
## 2.15 Statistical analyses
All statistical analyses were conducted using GraphPad Prism 9, and the data are presented as mean ± standard deviations (SD). Student's t-test was employed for statistical analyses. The threshold of statistical significance was set at P-values < 0.05.
## 3 Results and discussion
3.1 Characterization of SV40 VLP SV40 capsid protein VP1 was expressed by Escherichia coli, and the VP1 pentamer was obtained through ammonium sulfate precipitation followed by ultracentrifugation. VP1 has a molecular weight of 41 kDa, and SDS-PAGE analysis revealed that the purity of the VP1 was greater than 95% (Fig. 1A). The purified VP1 pentamer can selfassemble into an icosahedral cage structure upon induction with calcium ions and changes in pH in vitro. We isolated the assembled SV40 VLP using sucrose density gradient centrifuge. Transmission electron microscope (TEM) image demonstrated that the SV40 VLP were homogenized and well-dispersed (Fig. 1B). Dynamic light scattering measurements showed that the particle size of SV40 VLP was approximately 24.36 nm. (Fig. 1C). Studies have shown that the antigenicity of nanoparticles is significantly influenced by their size [33], with smaller nanoparticles being transported more rapidly than larger ones [34]. Consequently, SV40 VLP possesses a natural advantage in stimulating immune response.
## 3.2 In vivo imaging of SV40 VLP after subcutaneous inguinal injection
Determining whether SV40 VLP can drain to lymph nodes is crucial for their development as immune agents. We covalently labeled SV40 VLP with the fluorescent reagent Alaxe 647 and subsequently subcutaneously injected the labeled VLP into the left lower groin of mice (Fig. 2A). In vivo imaging revealed a strong fluorescence signal from the SV40 VLP in inguinal lymph node 1 (LN1) one day after injection (Fig. 2B), and the signal persisted until the day five post-injection (Fig. 2C). Our observations differ from previous findings using the J8-VLP formulation [35], in which subcutaneous administration resulted in rapid clearance within 24 h and no detectable fluorescent signal. This contrast underscores a potentially distinctive feature of the SV40 VLP platform, the in vivo stability. Furthermore, fluorescence signals of SV40 VLP began to appear at inguinal lymph nodes on the 5th day after injection (Fig. 2D). This suggests that SV40 VLP can migrate Enhancing immunogenicity relies on strategies that prolong antigen depot duration and promote lymphatic trafficking. For example, hepatitis B surface antigen VLP with lipid nanoparticle have been shown to extend antigen retention at the injection site for up to 72 h. In contrast to the short-lived exposure formulation of aluminum-VLP, this approach elicits a stronger cellular immune response [36]. Our observations indicate that SV40 VLP possess the fundamental conditions necessary for interacting with immune cells. However, more efficient lymphatic migration of SV40 VLPs may necessitate the incorporation of exogenous components. For instance, VLPs conjugated with BSA or iRGD demonstrated improved trafficking to lymph nodes compared to unmodified VLPs in a mouse model [37]. Su et al. reported that P22 VLP expressing cellpenetrating peptides could enhance the tissue accumulation and retention [38]. This finding suggests that the introduction of functional peptides represents an effective strategy for improving VLP retention in tissues. 3.3 Change in immune cell quantity in multiple organs induced by SV40 VLP SV40 VLP can drain to the contralateral lymph nodes, which is essential for antigen presentation and T cell activation. Therefore, we next analyzed the cellular immune response to SV40 VLP following repeated administration. Flow cytometry revealed that SV40 VLP significantly elevated the number of CD4 + T cells in mesenteric lymph nodes (Fig. 3A,B), while CD8 + T cells had no change.
Liang et al. reported that encapsulating the immunostimulatory CpG within the lumen of HBC VLPs only marginally increased CD4 + T cells, without eliciting a significant CD8 + T cell response. A substantial T cell activation was achieved only when HBC VLPs were surface-conjugated with an antibody targeting the immune checkpoint molecule B7-H3 [39]. These findings suggest that naked VLPs alone induce a limited immune response. Nevertheless, we also examined changes in immune cell populations in the spleen following SV40 VLP administration. We observed that SV40 VLP significantly increased the number of NK cells (Fig. 3C,D) and CD4 + T cells in the spleen (Fig. 3E,F). Theoretically, NK cells can directly kill tumor cells [40,41], while CD4 + T cells indirectly or directly kill tumor through MHCⅡ molecules on antigenpresenting cells [42][43][44][45], respectively. Therefore, our results reaffirm the significant potential of SV40 VLP as a novel tumor immunotherapy agent. However, similar to the immune response observed in lymph nodes, no significant increase in CD8 + T cells was significant in the spleen (Fig. S1). These results suggest that there remains considerable potential to further optimize SV40 VLPs for enhanced immunogenicity. For example, future modifications could include the incorporation of polyarginine/ cysteine-tagged antigens to promote [46], or the use of novel inorganic materials as adjuvants to improve the activation efficiency of both CD4⁺ and CD8⁺ T cells [47].
Owing to the absence of a viral genome and infectivity, VLPs are considered highly safe, making them promising platforms for vaccines [48,49], drug delivery [50], and immunotherapy fields [51]. However, the immune storm induced by immunotherapy agents is a limitation for their clinical translation [52][53][54]. Therefore, we further investigated the change of immune cell quantity in the lung, liver and kidney following SV40 VLP injections. Flow cytometry analysis revealed that subcutaneous injection of SV40 VLP did not influence CD4 + T cells, CD8 + T cells, B cells, NK cells or DCs in the liver (Fig. 4A), lung (Fig. 4B) or kidney (Fig. 4C). Taken together, SV40 VLP stimulates the proliferation of T cells and NK cells in the lymph nodes and spleen, without affecting the cell populations in the lungs, liver, or kidneys. This property avoids systemic immune storm and ensures the safety of SV40 VLP use as immune formulations. Statistical analysis showed no significant (ns) differences in immune cell activation among the three organs Next, we evaluated the blood safety of SV40 VLP after multiple subcutaneous injections. Prothrombin time (PT) and prothrombin activity (PTA) tests indicated that SV40 VLP do not significantly affect coagulation function (Fig. 5A). Additionally, SV40 VLP did not cause notable changes in the levels of various blood biochemical indicators, including blood urea nitrogen (BUN) (Fig. 5B), alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total bilirubin (TB) (Fig. 5C). These findings demonstrate that SV40 VLP have no apparent adverse effects and exhibit good biocompatibility, which is crucial for their application in immunotherapy in vivo.
Given these findings, can SV40 VLP be used as an agent to inhibit tumor growth? We administered SV40 VLP twice prior to subcutaneously inoculating Lewis lung carcinoma cells, followed by six additional VLP injections once every four days (Fig. 6A). After treatment, we first examined the distribution of SV40 VLP in tumor-bearing mice. Imaging results revealed no significant accumulation of VLPs in the heart, spleen, or lung compared to the PBS group (Fig. 6B), whereas a slightly higher level of VLP accumulation was observed in the kidneys compared to the PBS group (Fig. 6C). This suggests that subcutaneously administered SV40 VLP may be primarily cleared through renal metabolism.
Despite a visual impression of slight liver fluorescence in the VLP group, quantitative analysis across multiple mice confirmed no significant difference from PBS control (Fig. 6C), which also reflects the relatively high autofluorescence of liver tissue compared to other organs. Additionally, we confirmed that subcutaneous injection results in considerably lower liver accumulation of VLPs compared to intravenous injection (Fig. S4). Notably, tumor tissues exhibited a stronger fluorescence signal in the VLP group than in the PBS group (Fig. 6D,E), indicating that SV40 VLP may preferentially migrate to tumor tissues via lymphatic vessels. However, at the conclusion of the treatment, ex vivo imaging of the inguinal and axillary lymph nodes revealed that (Fig. 6F,G), with the exception of inguinal lymph node site 1 (LN1), which exhibited a significant difference in the VLP injection group, no significant differences in VLP enrichment were observed at the other three lymph node sites (LN2, LN3 and LN4). Western blot analysis also revealed a distinct SV40 VLP-specific antibody response in serum following repeated injections (Fig. 6H andI).
The weight monitoring process indicated that, although a significant weight loss on the 18th day of VLP injection, the weight eventually returned to a level comparable to that of the control group (Fig. 6J). The biological safety of VLP was clarified once again even function with cell response. Immunohistochemistry staining revealed a modest increase in TNFα expression in the spleen following SV40 VLP treatment (Fig. 6K). Consistent with this observation, ELISA quantification further demonstrated a significantly higher level of TNF-α in the spleens of the VLP-treated group (2.28 pg/mL) compared to the PBS control group (0.87 pg/mL) (Fig. S5). This finding is in line with previous reports showing that mannoside-engineered Qβ bacteriophage VLPs can enhance the proportion of TNF-α + CD4⁺ T cells [55]. As TNF-α is a key cytokine involved in T helper type 1 (TH1) responses, effective T helper type 1 (TH1)-like immune response is necessary for cancer treatment [56]. Notably, tumor growth in the SV40 VLP-treated mice was slower than in the PBS group up to day 31 after tumor implantation (Fig. 6L). However, the tumor was not completely suppressed by the end of the dosing regimen, which may be attributed to the lack of increased CD8 + T cell activation in the spleen or lymph nodes. Our results indicate that while VLP can function as delivery vehicles for tumor immunotherapy, their administration alone is insufficient to elicit specific tumor suppression or robust immune responses.
To fully exploit the therapeutic potential of SV40 VLP in cancer treatment, it will be necessary to integrate this approach with complementary strategies, such as incorporating a therapeutic agent that can synergize with the VLP to enhance antitumor efficacy, or embedding tumor-specific antigens and immunostimulatory adjuvants within the VLP structure to reduce immunosuppression and promote T lymphocyte infiltration into tumor sites.
## 4 Conclusion
SV40 VLP, formed from the main capsid protein VP1, is easy to assemble and disassemble, can be mass-produced and bioengineered, and serves as an excellent model for studying virus-like particles. It is known that various sources, structures and shapes of VLP can elicit different immune responses. Our results demonstrated that mammalian-derived SV40 VLP could effectively migrate to lymph nodes and significantly increase the numbers of CD4 + T cells and NK cells in the spleen, as well as CD4 + T cells in mesenteric lymph nodes. Most importantly, SV40 VLP did not induce the proliferation of immune cells in the lungs, liver, or kidneys, nor did they cause blood index disorders, demonstrating their biostability and safety, which is the essential prerequisite for its clinical application.
The main capsid protein SV40 VP1 exhibits a high tolerance for protein fusion or modification. Both the N-terminal region and the DE-loop as well as H1 loop of VP1 can be engineered to carry amino acids or even entire proteins without impairing VLP assembly. This allows for the incorporation of molecules such as PD-1 antibodies to counteract immune suppression [57], or IL-33 to promote the formation of tertiary lymphoid structures [58], or toll-like receptor agonists like resiquimod (R848) to enhance antitumor immune responses [59]. Although our results showed that VLP alone did not inhibit lung tumor growth, it can stimulate cellular immune responses in the spleen and lymph nodes. This suggests that, with rational modifications, SV40 VLP could be developed into effective novel strategies for tumor immunotherapy. Overall, our findings indicate that SV40 VLP have broad potential for immunotherapeutic applications, supported by their moderate organ-specific immune activation and favorable biosafety profile.
## References
1. Li, Wang (2014) "Fabrication of nanoarchitectures templated by virus-based nanoparticles: strategies and applications" *Small (Weinh der Bergstr, Ger)*
2. Li, Liang, Chen et al. (2022) "Virus-Like Particle-Templated Silica-Adjuvanted Nanovaccines with Enhanced Humoral and Cellular Immunity" *ACS nano*
3. Su, Huang, Teng et al. (2023) "Viruslike particles nanoreactors: from catalysis towards bioapplications" *J Mater Chem B*
4. Xue, Taning (2023) "CNT. Plant and insect virus-like particles: emerging nanoparticles for agricultural pest management" *Pest Manag Sci*
5. Li, Yang, Tian et al. (2024) "Unveiling the Dynamic Mechanism of SARS-CoV-2 Entry Host Cells at the Single-Particle Level" *ACS nano*
6. Hu, Steinmetz (2020) "Cisplatin Prodrug-Loaded Nanoparticles Based on Physalis Mottle Virus for Cancer Therapy" *Mol Pharm*
7. Hartzell, Lieser, Sullivan (2020) "Modular Hepatitis B Virus-like Particle Platform for Biosensing and Drug Delivery" *ACS Nano*
8. Chan, Du, Ignacio et al. (2021) "Biomimetic Virus-Like Particles as Severe Acute Respiratory Syndrome Coronavirus 2 Diagnostic Tools" *ACS nano*
9. Chung, Cai, Steinmetz (2020) "Viral nanoparticles for drug delivery, imaging, immunotherapy, and theranostic applications" *Adv Drug Deliv Rev*
10. Bessa, Schmitz, Hinton et al. (2008) "Efficient induction of mucosal and systemic immune responses by virus-like particles administered intranasally: implications for vaccine design" *Eur J Immunol*
11. Halfmann, Loeffler, Duffy et al. (2024) "Broad protection against clade 1 sarbecoviruses after a single immunization with cocktail spike-protein-nanoparticle vaccine" *Nat Commun*
12. Omole, Zhao, Chang-Liao et al. (2024) "Virus nanotechnology for intratumoural immunotherapy" *Nat Rev Bioeng*
13. Segel, Lash, Song et al. (2021) "Mammalian retrovirus-like protein PEG10 packages its own mRNA and can be pseudotyped for mRNA delivery" *Science*
14. Zhao, Li, Voyer et al. (2022) "Flagellin/Virus-like Particle Hybrid Platform with High Immunogenicity, Safety, and Versatility for Vaccine Development" *ACS Appl Mater Interfaces*
15. Tarrés-Freixas, Aguilar-Gurrieri, De La Concepción et al. (2023) "An engineered HIV-1 Gagbased VLP displaying high antigen density induces strong antibody-dependent functional immune responses. npj Vaccines"
16. Bachmann, Jennings (2010) "Vaccine delivery: a matter of size, geometry, kinetics and molecular patterns" *Nat Rev Immunol*
17. Donaldson, Lateef, Walker et al. (2018) "Viruslike particle vaccines: immunology and formulation for clinical translation" *Expert Rev Vaccines*
18. Gomes, Flace, Saudan et al. (2017) "Adjusted Particle Size Eliminates the Need of Linkage of Antigen and Adjuvants for Appropriated T Cell Responses in Virus-Like Particle-Based Vaccines" *Front Immunol*
19. Lensch, Hincapie, Bhagchandani et al. (2024) "Carbohydrate-Lectin Interactions Reprogram Dendritic Cells to Promote Type 1 Anti-Tumor Immunity" *ACS nano*
20. Yang, Rao, Li et al. (2024) "Transformation of a Viral Capsid from Nanocages to Nanotubes and Then to Hydrogels: Redirected Self-Assembly and Effects on Immunogenicity" *ACS nano*
21. Zhang, Falcone, Tsybovsky et al. (2023) "Increased neutralization potency and breadth elicited by a SARS-CoV-2 mRNA vaccine forming virus-like particles" *Proc Natl Acad Sci*
22. Panthi, Schmitt, Lorenz et al. (2021) "Paramyxovirus-Like Particles as Protein Delivery Vehicles" *J Virol*
23. Wang, Douglas (2023) "Tuning Multistep Biocatalysis through Enzyme and Cofactor Colocalization in Charged Porous Protein Macromolecular Frameworks" *ACS Appl Mater Interfaces*
24. Cai, Shukla, Wang et al. (2019) "Heterologous Prime-Boost Enhances the Antitumor Immune Response Elicited by Plant-Virus-Based Cancer Vaccine" *J Am Chem Soc*
25. Yin, Zhong, Ling et al. (2025) "Dendriticcell-targeting virus-like particles as potent mRNA vaccine carriers" *Nat Biomed Eng*
26. Pan, Wang, Qi et al. (2023) "An Intranasal Multivalent Epitope-Based Nanoparticle Vaccine Confers Broad Protection against Divergent Influenza Viruses" *ACS nano*
27. Liddington, Yan, Moulai et al. (1991) "Structure of simian virus 40 at 3.8-Å resolution" *Nature*
28. Stehle, Gamblin, Yan et al. (1993) "The structure of simian virus 40 refined at 3.1 A resolution" *Struct*
29. Zhang, Zhang, Li (2018) "Virus-Based Nanoparticles of Simian Virus 40 in the Field of Nanobiotechnology" *Biotechnol J*
30. Van Rosmalen, Kamsma, Biebricher et al. (2020) "Revealing in real-time a multistep assembly mechanism for SV40 virus-like particles" *Sci Adv*
31. Xu, Zhu, Mao et al. (2020) "Switch from Polymorphic to Homogenous Self-Assembly of Virus-Like Particles of Simian Virus 40 through Double-Cysteine Substitution" *Small (Weinh der Bergstr, Ger)*
32. Feng, Gao, Jing et al. (2022) "Intracellular Delivery of Micron-Sized Magnetic Particles through a Virus Infection Pathway" *ACS Appl Mater Interfaces*
33. Li, Yan, Xia et al. (2022) "Nanoparticle/ Nanocarrier Formulation as an Antigen: The Immunogenicity and Antigenicity of Itself" *Mol pharmaceutics*
34. Lu, Lee, Lee (2024) "Three-Dimensional Lymphatics-on-a-Chip Reveals Distinct, Size-Dependent Nanoparticle Transport Mechanisms in Lymphatic Drug Delivery" *ACS Biomater Sci Eng*
35. Seth, Kong, Lee et al. (2016) "Modular virus-like particles for sublingual vaccination against group A streptococcus" *Vaccine*
36. Liu, Song, Yue et al. (2023) "Potentiating humoral and cellular immunity using a novel hybrid polymer-lipid nanoparticle adjuvant for HBsAg-VLP vaccine" *J Nanobiotechnol*
37. Tan, Tang, Li et al. (2025) "Virus like nanoparticles of VLP-BSA-iRGD enhanced mucosal immune response by targeting M cells" *Colloids Surf B, Biointerfaces*
38. Su, Shen, Shi et al. (2024) "Cellpenetrating peptides TAT and 8R functionalize P22 virus-like particles to enhance tissue distribution and retention in vivo" *Front Vet Sci*
39. Liang, Chi, Wang et al. (2025) "Multifunctional HBc Virus-Like Particles Reprogram Immunosuppressive Macrophages and Potentiate CD8(+) T Cell Responses for Enhanced Cancer Immunotherapy" *Int J Nanomed*
40. Fernandez, Lozier, Flament et al. (1999) "Dendritic cells directly trigger NK cell functions: Cross-talk relevant in innate anti-tumor immune responses in vivo" *Nat Med*
41. Netskar, Pfefferle, Goodridge et al. (2024) "Pan-cancer profiling of tumor-infiltrating natural killer cells through transcriptional reference mapping" *Nat Immunol*
42. König, Huang, Germain (1992) "MHC class II interaction with CD4 mediated by a region analogous to the MHC class I binding site for CD8" *Nature*
43. Speiser, Chijioke, Schaeuble et al. (2023) "CD4+ T cells in cancer" *Nat Cancer*
44. Zheng, Si, Wang et al. (2021) "Pancancer single-cell landscape of tumor-infiltrating T cells" *Science*
45. Cardenas, Prokhnevska, Sobierajska et al. (2024) "Differentiation fate of a stem-like CD4 T cell controls immunity to cancer" *Nature*
46. Viscidi, Rowley, Bossis (2023) "Bioengineered Bovine Papillomavirus L1 Protein Virus-like Particle (VLP) Vaccines for Enhanced Induction of CD8 T Cell Responses through Cross-Priming" *Int J Mol Sci*
47. Li, Chen, Yu et al. (2022) "Virus-Like Particle-Templated Silica-Adjuvanted Nanovaccines with Enhanced Humoral and Cellular Immunity" *ACS nano*
48. Xiao, Chen, Wang et al. (2016) "Large-scale production of foot-and-mouth disease virus (serotype Asia1) VLP vaccine in Escherichia coli and protection potency evaluation in cattle" *BMC Biotechnol*
49. Kim, Kim, Cho et al. (2025) "Neutralizing Antibody Screening Using NanoBiT-Based Virus-like Particles of Foot-and-Mouth Disease Type Asia1 Enhances Biosafety and Sensitivity" *Viruses*
50. Li, Chen, Hu et al. (2016) "An Ebola Virus-Like Particle-Based Reporter System Enables Evaluation of Antiviral Drugs In Vivo under Non-Biosafety Level 4 Conditions" *J Virol*
51. Robilotti, Zeitouni, Orloff (2023) "Biosafety and biohazard considerations of HSV-1-based oncolytic viral immunotherapy" *Front Mol Biosci*
52. Hansel, Kropshofer, Singer et al. (2010) "The safety and side effects of monoclonal antibodies" *Nat Rev Drug Discov*
53. Maude, Barrett, Teachey et al. (2014) "Managing cytokine release syndrome associated with novel T cell-engaging therapies" *Cancer J*
54. Sr Ri (2018) "Adrenaline fuels a cytokine storm during immunotherapy" *Nature*
55. Alam, Hincapie, Mckay et al. (2021) "Glycan-Modified Virus-like Particles Evoke T Helper Type 1-like Immune Responses" *ACS nano*
56. Zhu, Ma, Shen et al. (2024) "Screening for lipid nanoparticles that modulate the immune activity of helper T cells towards enhanced antitumour activity" *Nat Biomed Eng*
57. Zou, Shen, Xue et al. (2024) "Anti-PD-1 cisdelivery of low-affinity IL-12 activates intratumoral CD8+T cells for systemic antitumor responses" *Nat Commun*
58. Amisaki, Zebboudj, Yano et al. (2025) "IL-33-activated ILC2s induce tertiary lymphoid structures in pancreatic cancer" *Nature*
59. Zhang, Wei, Yong et al. "Cell microparticles loaded with tumor antigen and resiquimod reprogram tumor-associated macrophages and promote stem-like CD8+"
60. (2023) "T cells to boost anti-PD-1 therapy" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12719972&blobtype=pdf | # Letter to the Editor Case report of a bloodstream infection due by Salmonella strathcona ST2559 in northeast Italy, 2025
Michela Bulfoni, Carlo Tascini, Paola Siega, Corrado Pipa, Silvio Brusaferro, Paolo Gaibani
A B S T R A C T Not applicable.
## Dear Editor,
Recently, a multi-country outbreak of Salmonella strathcona ST2559 has been reported in the European Union and the United Kingdom (UK) [1]. In particular, from January 2023 to November 2024, different countries reported a total of 232 confirmed cases of S. strathcona [2]. Epidemiological and traceback investigations demonstrated that the S. strathcona ST2559 widespread in different countries was vehiculated by small tomatoes originated from Sicily since 2011 [2][3][4]. Indeed, prior to 2010 no confirmed cases with S. strathcona were reported in the EU countries (https://atlas.ecdc.europa.eu/public/index.aspx). Here, we present the case of a bloodstream infection due to S. strathcona ST2559 isolated from a hospitalized patient in northeast Italy, 2025.
A young woman was admitted on November 2025 to the Azienda Sanitaria Universitaria Friuli Centrale for fever, vomiting, splenic abscess, and pleural effusion. During hospitalization patient developed fever and Blood culture yielded a positive result initially identified as Salmonella spp. by MALDI-TOF system (Bruker, Germany). Therefore, patient was initially treated with piperacillin-tazobactam, then switched to ceftriaxone 2 g once daily when the susceptibility test demonstrated a multi-susceptible Salmonella spp strain. The patient also had a positive stool sample for Salmonella (molecular test), which became negative after antibiotic therapy. Both the pleural effusion and the splenic abscess were drained; microbiological cultures from these samples were negative for Salmonella as well as for other pathogens. To achieve complete defervescence, a short tapering course of corticosteroids was administered.
In order to characterize the serovar of S. enterica, a whole-genome DNA sequencing was performed as previously described [5]. Briefly, genomic DNA was extracted using QIASymphony instrument (Qiagen, Germany) and library preparation was performed using the FX DNA Library Preparation Kit (Qiagen, Germany) and the Nextera™ XT Index Kit (Illumina, USA). Sequencing was carried out on an Illumina MiSeq System using a 2 × 300-bp and paired-reads quality was evaluated with FastQC v0.12.1 (https://www.bioinformatics.babraham.ac.uk/project s/fastqc/). Genome assembly was carried out using SPAdes v3.15.5 (https://github.com/ablab/spades) and quality was evaluated with QUAST v.5.3.0. The final assembled genome of the S. strathcona object of this study, named strain SAL-UD, produced a draft with a total size of 4.698.491 bp, composed of 78 contigs ranging from 378.300 to 500 bp in length. The genome had a 52.15 % G + C content, 173.434 N50, 45.456 N90 and 20x mean coverage. Species identification was determined using SpeciesFinder v.2.0 (https://cge.food.dtu.dk/services/ SpeciesFinder/) and results showed that strain SAL-UD belonged to the Salmonella enterica subsp. enterica. The MLST and antigenic profile identification were performed respectively using PubMLST (https:// pubmlst.org/bigsdb) and SeqSero (https://cge.food.dtu.dk/services /SeqSero/). Genome-based typing revealed that the strain SUL-UD belonged to ST2559 (aroC→481,dnaN→18,hemD→10,hisD→124, purE→5,sucA→10,thrA→14), with the following antigenic profile: O-Antigen:O-7; H1-Antigen:l,z13,z28; H2-Antigen:1,7. Analysis of antimicrobial resistance determinants showed that SAL-UD carried gene related to the aminoglycoside-resistance (aac(6′)-Iaa). Clonal relatedness of the strain SAL-UD with the genomes of S.strathcona available in GenBank was performed as previously described [5]. Genome comparison among S.strathcona strains displayed a wide range of nucleotide homology value (ranging from 85.36 % to 99.85 %) among isolates isolated in Europe (Fig. 1).
To evaluate the role of prophages among S.strathcona, we performed a pro-phage regions analysis in the genome of the strain SAL-UD. Our analysis demonstrated that SAL-UD harboured 2 complete (ranging from 31.8 Kb to 36.5 Kb), 3 incomplete (ranging from 7.4 Kb to 21 Kb) and 2 questionable pro-phage regions (ranging from 10.2 Kb to 42.2 Kb) (Fig. S1 in the Supplementary material). Blast analysis demonstrated that the intact prophage regions exhibited high homology (nucleotide identity of 100 % with a coverage of 100 %) with chromosome of S. strathcona strain N22-0456 (Acc.no CP179910) isolated in 2022 from a patient recovered in Switzerland (Fig. S2 and S3 in the Supplementary material). At the same time, analysis of prophage regions demonstrated that several salmonella loci were found (Table S1 in the Supplementary material).
In conclusion, here we reported the case of a bloodstream infection occurred in northeast Italy, 2025. Also, we described the genome of S. strathcona ST2559 strain by enlarging the acknowledgment of the diffusion of this emerging pathogen in Europe and analysing the role of the prophage regions in the diversification of S. strathcona. Based on these findings, we highlighted the importance of monitoring emerging strains which could pose a significant risk for public health also recommending a more widespread use of genome sequencing.
## Nucleotide sequence accession numbers
The draft genome assembly of the Salmonella strathcona strain SAL-UD as been deposited in the NCBI BioSample database under accession number SAMN53173069.
## CRediT authorship contribution statement
## References
1. Brait, Böff, Zmarlak-Feher et al. (2011) "Insights into recurring multi-country outbreaks of Salmonella strathcona associated with tomatoes" *Euro Surveill*
2. "European food safety authority"
3. Müller, Kjelsø, Frank et al. (2011) "Outbreak of Salmonella strathcona caused by datterino tomatoes, Denmark" *Epidemiol Infect*
4. Møller, Mølbak, Ethelberg (2018) "Analysis of consumer food purchase data used for outbreak investigations, a review" *Euro Surveill*
5. Gaibani, Latorre (2025) "Prophages in Bacteroides fragilis: distribution and genetic diversity" *Heliyon* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12502795&blobtype=pdf | # Positivity analysis of Bordetella spp. and SARS-CoV-2 coinfections in clinical samples from Colombia, 2021-2022
Valentina Bonilla-Bravo, Efrain Montilla-Escudero, Ximena Castro Martinez, Fabiola Rojas, Carolina Duarte, Sandra Aparicio, Sergio Yebrail Gómez-Rangel, Paula Romero, Sandra Lucero Bonilla, Franklyn Alvarado, Jaid Sotelo, Gloria Benito
## Abstract
Pertussis and COVID-19 are respiratory diseases with similar clinical manifestations, complicating diagnosis. In Colombia, pertussis surveillance is based on probable cases confirmed through laboratory detection of Bordetella pertussis. During the COVID-19 pandemic, pertussis incidence declined sharply. Given limited data on coinfections with COVID-19 and pertussis, this study aimed to estimate the positivity of Bordetella spp. in SARS-CoV-2-positive samples collected in Colombia during 2021-2022. A retrospective analysis was conducted on samples collected between January 2021 and December 2022 through simple random sampling from specimens stored at the National Reference Laboratory (NRL). Demographic data, symptoms, complications, and race were obtained from the National Surveillance System (SIVIGILA), and PCR results were used to identify Bordetella spp. coinfections. National and departmental positivity rates were estimated with Clopper-Pearson confidence intervals. PCR testing for Bordetella spp. was positive in 0.54% (6/1,102) of samples, with four cases involving B. pertussis and two involving B. parapertussis; no B. holmesii coinfections were detected. The national positivity rate of SARS-CoV-2 and Bordetella spp. coinfections was 1.92% in 2021, decreasing to 1.25% in 2022. Although coinfections with viral agents and B. pertussis are common, this study found low positivity of SARS-CoV-2 and Bordetella spp. coinfections in Colombia during 2021-2022, likely due to non-pharmacological interventions, the cyclical pattern of B. pertussis, and possible underestimation related to the geographic and demographic scope of the sample. This is the first report of such coinfections in Colombia, offering a reference for future studies in Latin America. IMPORTANCE This is the first report of SARS-CoV-2 and Bordetella spp. coinfections in Colombia, providing key evidence of their low occurrence during 2021-2022. Given that both diseases share clinical features and may complicate differential diagnosis, these findings underscore the importance of integrated surveillance for respiratory pathogens. The observed decrease in positivity may be linked to non-pharmaceutical interventions and the cyclical nature of pertussis, while also highlighting potential limitations in the geographic and demographic representativeness of the sample. This study offers a valuable baseline for future research in Latin America and reinforces the need to strengthen diagnostic capacity to detect coinfections, particularly in epidemic or pandemic contexts.
three doses with the pentavalent vaccine (diphtheria, pertussis, tetanus, hepatitis B, and Haemophilus influenzae type b [Hib]) administered at 2, 4, and 6 months of age, followed by two booster doses with the Diphtheria, Tetanus, and Pertussis (DPT) vaccine at 18 months and 5 years of age. In addition, since 2014, a booster dose of tetanus, diphtheria, acellular pertussis (Tdap) has been administered during the third trimester of pregnancy. The DTP3 vaccine coverage in Colombia was 86.5% in 2021 and 86.96% in 2022.
Clinical presentations of pertussis typically include prolonged cough, with episodes of paroxysmal cough, posttussive vomiting, apnea, or cyanosis. Symptoms can range from a relatively mild cough to severe respiratory illness and complications such as pneumonia, seizures, encephalopathy, respiratory failure, and death (1). While pertussis is caused by B. pertussis, other species, such as Bordetella parapertussis and Bordetella holmesii, are associated with milder respiratory infections in humans presenting as pertussis-like syndrome (2)(3)(4).
Pertussis and COVID-19 are respiratory diseases with overlapping clinical manifesta tions, making differential diagnosis challenging. Since 1997, pertussis surveillance in Colombia has been part of an epidemiological surveillance that integrates risk analysis, case reporting, and laboratory confirmation. The pertussis cases are reported to the National Surveillance System (SIVIGILA) from public and private healthcare institutions; meanwhile, laboratory confirmation of B. pertussis is performed by the Microbiology Group at the NRL and by trained network laboratories.
The strict non-pharmaceutical interventions implemented during the COVID-19 pandemic protected the population, especially children and older adults, by reducing virus transmission. However, a notable decline in the incidence of other respiratory pathogens was also observed (5), including B. pertussis (6,7). The global incidence of pertussis dropped significantly during the COVID-19 pandemic, decreasing from 23 cases per 1,000,000 inhabitants in 2019 to 9.9 in 2020 and 4.6 in 2021. However, after the pandemic, an increase in pertussis cases was observed in 2019 to a rise of 23.6 cases per 1,000,000 in 2023. This increase was particularly pronounced in the European region, where the incidence rose from 23.7 to 104 cases per 1,000,000 inhabitants during the same period. Similarly, in the Western Pacific region, cases increased from 6.1 to 26.4 per 1,000,000 (8), highlighting the need for more rigorous epidemiological surveillance due to the risk of complications in patients, coinfections, and mortality cases.
In this context, Colombia experienced a significant decrease in the number of laboratory-confirmed pertussis cases. Compared to 2019 (n = 302), confirmed cases decreased by 91.72% (n = 25; P < 0.05) in 2021 and by 86.75% (n = 40; P < 0.05) in 2022 (9).
As part of the country's response to the pandemic and the shifting focus of pub lic health surveillance, the Instituto Nacional de Salud (INS) assumed leadership in coordinating these efforts. During this period, SARS-CoV-2 samples were processed and stored at the NRL of Virology and at some Departmental Public Health Laboratories (DPHL) within the network. Currently, laboratories retain positive SARS-CoV-2 samples stored during this period, providing a valuable resource for research.
The relevance of this study lies in the context of the COVID-19 pandemic, during which the high transmissibility of SARS-CoV-2 highlighted the transmission dynamics of respiratory pathogens. In this scenario, it was considered pertinent to investigate the circulation of Bordetella pertussis, given the marked decline in reported cases observed during that period. Studies on coinfections between viral agents and B. pertussis are common, with reported coinfection rates reaching 69.7% (10), but, given the limited national and international data on COVID-19 and pertussis coinfections, as well as the availability of relevant samples for both pathogens, this study aimed to assess the positivity of Bordetella spp. coinfections in SARS-CoV-2-positive samples from laboratory surveillance at the INS in Colombia during 2021-2022.
## MATERIALS AND METHODS
## Sample selection
Of the samples available (n = 12,881), a simple random sample was carried out from which 1,102 (8.6%) samples were selected. Nasopharyngeal samples confirmed as SARS-CoV-2 positive by real-time reverse transcription PCR (11) were analyzed from the period January 2021 to December 2022. All samples were analyzed in the NRL's Microbiology Group.
## Molecular testing
All samples were stored at -80°C until analysis. Bacterial nucleic acids were extracted from 200 µL of nasopharyngeal samples using an automated method with the Mag MAX Viral/Pathogen Nucleic Acid Isolation Kit (Thermo Fisher Scientific, USA) on the KingFisher Flex system (Thermo Fisher Scientific, USA). Real-time PCR (qPCR) testing was duplicated to detect Bordetella species. DNA eluates were analyzed using a multitarget real-time PCR (qPCR) assay to identify the insertion sequences IS481, pIS1001, and hIS1001, along with a qPCR assay for RNase P as a human internal control. This was followed by a confirmatory singleplex assay targeting ptxS1 (12). The PCR algorithm interpretation is illustrated in Fig. 1. A PCR result was considered negative if the cycle threshold (CT) value was ≥40.
## Data collection
The sociodemographic, epidemiological, and clinical data of the analyzed samples were obtained by cross-referencing the variables recorded in SIVIGILA with the laboratory databases. The records from SIVIGILA and the Virology Group were anonymized and Protected Health Information, ensuring security and confidentiality in their management and analysis.
## Data analysis
Descriptive and bivariate analyses of sociodemographic characteristics, symptoms, and laboratory data were conducted for patients with and without coinfections. To compare the mean duration of clinical course between patients with SARS-CoV-2 infection alone and those with coinfections, a one-way analysis of variance (ANOVA) was performed. Data were analyzed by Epi Info 7 (13) and Microsoft Excel software, and P values < 0.05 were considered statistically significant.
To calculate positivity, the number of identified coinfection cases between COVID-19 and pertussis was used as the numerator, and the number of COVID-19-positive samples tested for Bordetella pertussis as the denominator. Positivity was expressed as a percentage by applying a multiplication factor of 100. This approach was applied both to estimate annual positivity (2021-2022) and to perform a stratified analysis by departments where coinfection was detected.
To estimate coinfection rates, the exact Clopper-Pearson method was used to calculate confidence intervals for binomial proportions. This method was chosen for its suitability in providing accurate interval estimates, particularly when sample sizes are small or observed proportions are near 0 or 1. Confidence intervals were computed using the OpenEpi software (https://www.openepi.com/BriefDoc/Licensing.htm).
## RESULTS
Of all the samples positive for SARS-CoV-2, 53.45% were female. Of all patients, 50.72% reported cough as the most frequent symptom of COVID-19. Records of pertussis symptoms were not available. The ages of all patients ranged from <1 to 80 years, but the most represented age group (77.58%) was between 20 and 69 years.
Of the samples analyzed, six (0.54%) tested positive for Bordetella spp. Among these, four (66.67%) were coinfected with Bordetella pertussis, and two (33.33%) with Bordetella parapertussis. Among these coinfections, 16.66% (1/6) were Indigenous and 33.3% (2/6) were Black race. The ages of these coinfections range from 20 to 70 years old, but the highest number of coinfections was identified in the age group 20-29 years. A significant difference in the prevalence of fatigue was observed between coinfected patients and those with only COVID-19 (P < 0.05), while 75% of coinfected patients reported fatigue, only 16.51% of patients with COVID-19 alone experienced this symptom. Table 1 details other symptoms registered in SARS-CoV-2 patients with and without coinfections.
A significant difference (P < 0.05) was observed in the mean duration of clinical course between patients with only SARS-CoV-2 (x̄ = 5.84 ± 23.63) and those with coinfections in 2021 (x̄ = 8.16 ± 5.42). Furthermore, there were no deaths or hospitalizations among patients with coinfections.
## Positivity of coinfections
The coinfection positivity of SARS-CoV-2 with Bordetella spp. in 2021 was 0.75% cases (four cases), while in 2022, it was 0.34% cases (two cases). In both years, the estimate of coinfection does not exceed the positivity of 2.0% (Table 2).
In 2021, positivity was confirmed in the departments of Chocó, Norte de Santander, and Guaviare. The highest positivity was identified in Guaviare, with 5.26% (1/19). The departments of Chocó and Guaviare exceeded the positivity of 1.0% in the estimation of coinfection between both diseases (Table 3).
For 2022, coinfection was identified in two departments: Magdalena and Amazonas. The highest positivity was observed in Magdalena, with 10.0% (1/10). The estimation of coinfection in Amazonas was 1.09% (1/91) (Table 3).
## DISCUSSION
This study identified laboratory-confirmed coinfections with Bordetella spp. in SARS-CoV-2 patients in Colombia during 2021-2022. Our findings are consistent with those of another study (14), which also reported a low proportion (0.78%) of SARS-CoV-2/ Bordetella spp. coinfections. Although the number of coinfections identified was low, this trend was expected due to the non-pharmacological measures adopted during the pandemic, such as social isolation, mask usage, and increased hygiene awareness, which contributed to a decrease in pertussis incidence (15)(16)(17)(18)(19). This reduction would limit the opportunities for coinfections, even if SARS-CoV-2 were widely circulating.
Coinfections involving not only B. pertussis but also B. parapertussis were identified. While B. parapertussis is not a major cause of mortality, sporadic cases unrelated to pertussis outbreaks have been reported (20)(21)(22). Notably, the diphtheria-tetanus-pertus sis (DTP) vaccine does not provide effective protection against B. parapertussis (23), and some cases manifest with like-pertussis symptoms. Given this, testing for both B. pertussis and B. parapertussis in patients with whooping cough symptoms is crucial. In response to this need, the Microbiology Group has been monitoring B. parapertussis using qPCR since 2014 (24).
Despite this, the vaccination history of the patients in the samples analyzed in this study was not available. However, 91.29% of the population was over 10 years old, likely lacking protection due to the waning immunity conferred by the pertussis vaccine (25).
In Colombia, the only documented report of SARS-CoV-2 and Bordetella spp. coinfections was reviewed by the INS. In 2020, four cases of coinfection between pertussis and COVID-19 were confirmed, of which 50% (2/4) affected children under 1 year of age. The predominant symptoms were cough and paroxysmal cough, with an average duration of 11 days, while 25% (1/4) presented complications such as The positivity estimation of SARS-CoV-2 with Bordetella spp. coinfection was obtained using a multiplication coefficient of 100, applying the confidence interval according to Fisher's exact method (Clopper-Pearson).
pneumonia (26). Our results suggest that patients with coinfections exhibited more clinical manifestations, such as fatigue, compared to those with SARS-CoV-2 alone.
In 2021, the INS notified four possible coinfections (4/75) involving B. pertussis and other respiratory viruses, one of which was a coinfection with SARS-CoV-2 (1/75). This coinfection occurred in the Department of Choco, where a large-scale pertussis outbreak affected an indigenous community (27). In 2022, a similar outbreak was reported in the Indigenous population of the Sierra Nevada de Santa Marta (28).
Our findings suggest that SARS-CoV-2 patients with coinfections were more frequently Indigenous or had epidemiological links to Indigenous communities. These populations are particularly vulnerable, as many are not enrolled in the Social Health Security System and often complete vaccination schedules (27). In Colombia, pertussis has predominantly affected Indigenous and rural communities. However, it is important to highlight that the samples analyzed in this study were drawn from an urban population, which may partially account for the low observed coinfection rate.
In addition, coinfections with Bordetella spp. were identified in the departments of Guaviare and Amazonas. However, the number of reported pertussis cases in these regions was very low (9), indicating an epidemiological silence surveillance system.
The coinfections were more frequent in young adult patients. In the study by Roh et al. (29), the median age of the SARS-CoV-2 patients with bacterial coinfections was reported as 32.0 ± 6.0 years, implying that young adults may be more susceptible to respiratory pathogen coinfections. Some patients with SARS-CoV-2, only in 2021, reported an unusually prolonged duration of symptoms. This suggests that they may have coinfections with other respiratory pathogens. Chen et al. (30) summarized the most common microbial coinfections with SARS-CoV-2, highlighting respiratory viruses such as coronavirus (non-COVID-19) (2.1%), Entero/rhinovirus (hRV) (6.9%), influenza A (4.57%), human metapneumovirus (hMPV) (1.98%), and respiratory syncytial virus (RSV) (3.46%). Bacterial coinfections accounted for 4.97% and included Acinetobacter baumannii, Actinomy ces spp., and Klebsiella pneumoniae. Fungal coinfections were observed in 3.16% of cases, involving Legionella pneumophila, Rothia spp., Streptococcus spp., Veillonella spp., Aspergillus spp., and Candida spp.
On the other hand, several studies reported coinfections in pertussis patients with other respiratory viruses. Ferronato et al. (31) identified coinfections in 7% of cases, with 4.9% specifically involving respiratory syncytial virus (RSV). Scutari et al. (10) found a high coinfection rate (69.76%) with respiratory viruses, including Human Rhinovirus/Enterovi rus (7/43), hMPV (3/43), Parainfluenza virus type 3 (PIV3) (1/43), coronavirus OC43 (1/43), and RSV (1/43), among others.
In addition, the INS documented a pertussis outbreak in 2024 among an Indigenous population in Urrao and Betulia, Antioquia, confirming 12 cases (32). Among these, coinfections with other respiratory agents were identified, including influenza A(H3) (25%), adenovirus (66.6%), and RSV (58.33%). This finding reinforces the hypothesis of potential interactions between pertussis and various respiratory pathogens and underscores the occurrence of such interactions across different geographic regions and population contexts. In the Americas, the incidence of pertussis declined significantly between 2020 and 2023, dropping from 20.1 to 6.1 cases per 1,000,000 inhabitants. This reduction was primarily due to the redirection of surveillance efforts toward SARS-CoV-2, which temporarily affected the detection and reporting of Bordetella pertussis cases. However, as of 2024, incidence rates are stabilizing and returning to pre-pandemic levels, reflecting the renewed focus on pertussis surveillance and its natural epidemiological trends. Several countries, including the United States, Mexico, Peru, and Brazil, are experiencing a resurgence, indicating that B. pertussis circulation was low during the analyzed period but is now reestablishing itself (33).
In Colombia, the lower coinfection rate observed during this period can also be attributed to the natural cyclical behavior of B. pertussis, which follows a 4-year pattern of fluctuating case numbers (34). In addition, protective measures implemented to mitigate the spread of SARS-CoV-2, such as mask-wearing, social distancing, and restricted mobility, likely contributed to reducing the transmission of B. pertussis and other respiratory pathogens.
The study also had some limitations. First, there was a low inclusion of samples from Antioquia, Bogotá D.C., and the Cundinamarca department that reported the highest number of pertussis cases. This may have impacted the representativeness by depart ment of our findings. Second, there was a lack of information on symptoms or medical records to associate pertussis symptoms in the analysis.
These results serve as a reference framework for the study of COVID-19 and pertussis coinfection in Latin America and are the first to confirm coinfections between SARS-CoV-2 with B. pertussis and B. parapertussis in Colombia.
A deeper follow-up through studies or syndromic surveillance is essential for detecting multiple pathogens and gaining a better understanding of coinfection dynamics. This approach would help identify the most affected populations and assess the potential public health impact. The need for such surveillance is particularly relevant for children under 5 and adults over 60, who are more vulnerable to severe complica tions. Due to their developing or aging immune systems, these populations face a higher risk of adverse outcomes, highlighting the importance of tailored prevention and treatment strategies to address their specific needs.
## References
1. Chamorro, Luca, Swaminathan et al. (2023) "Bordetella bronchiseptica and Bordetella pertussis: similarities and differences in infection, immuno-modulation, and vaccine considerations" *Clin Microbiol Rev*
2. Hiramatsu, Nishida, Nugraha et al. (2021) "RETRACTED: melanin produced by Bordetella parapertussis confers a survival advantage to the bacterium during host infection"
3. Shepard, Daneshvar, Kaiser et al. (2004) "Bordetella holmesii bacteremia: a newly recognized clinical entity among asplenic patients" *Clin Infect Dis*
4. Mooi, Bruisten, Linde et al. (2012) "VRS study group in Lyon. 2021. Low levels of respiratory syncytial virus activity in Europe during the 2020/21 season: what can we expect in the coming summer and autumn/winter?" *FEMS Immunol Med Microbiol*
5. Reicherz, Li, Watts et al. "Abu-Raya B. 2024. Bordetella pertussis infection following relaxation of COVID-19 nonpharmaceutical interventions in 2021-2023 in Vancouver metropolitan area" *Vaccine (Auckl)*
6. Sun, Xu, Zhu et al. (2021) "Impact of non-pharmaceutical interventions on the incidences of vaccine-preventable diseases during the COVID-19 pandemic in the eastern of China" *Hum Vaccines Immunother*
7. (2023) "World Health Organization. Portal de Datos"
8. "Instituto Nacional de Salud. Búsqueda de Microdatos"
9. Scutari, Linardos, Ranno et al. (2025) "A new epidemic wave of Bordetella pertussis in paediatric population: impact and role of coinfections in pertussis disease" *Ital J Pediatr*
10. Corman, Landt, Kaiser et al. (2020) "Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR"
11. Tatti, Sparks, Boney et al. (2011) "Novel multitarget real-time PCR assay for rapid detection of Bordetella species in clinical specimens" *J Clin Microbiol*
12. (2023)
13. Berry, Cole, Silk et al. (2020) "SARS-CoV-2 coinfections among pertussis cases identified through the enhanced pertussis surveillance system in the United States" *PLoS One*
14. Renzi, Clementi, Racca et al. (2021) "Is Bordetella pertussis co-infecting SARS-CoV-2 patients?" *Braz J Anesthesiol*
15. Tessier, Campbell, Ribeiro et al. (2022) "Impact of the COVID-19 pandemic on Bordetella pertussis infections in England" *BMC Public Health*
16. Matczak, Levy, Fortas et al. (2022) "Association between the COVID-19 pandemic and pertussis derived from multiple nationwide data sources, France, 2013 to 2020" *Euro Surveill*
17. Falkenstein-Hagander, Appelqvist, Cavefors et al. (2022) "Waning infant pertussis during COVID-19 pandemic" *Arch Dis Child*
18. Van Der Kooi, Beard, Dey et al. (2018) "Pertussis notifications decline in Australia during COVID-19 non-pharmaceutical interventions, 2020-2021" *Commun Dis Intell*
19. Montilla-Escudero, Rojas-Baquero, Bonilla-Bravo et al. (2023) "Draft genome sequences of Bordetella parapertussis clinical isolates during an increase in confirmed cases from Colombia" *Microbiol Resour Announc*
20. Bouchez, Toubiana, Guillot et al. (2024) "Transient reemergence of Bordetella parapertussis in France in 2022" *J Med Microbiol*
21. Karalius, Rucinski, Mandrekar et al. (2014) "Bordetella parapertussis outbreak in Southeastern Minnesota and the United States" *Medicine (Baltimore)*
22. Bouchez, Guiso (2015) "Bordetella pertussis, B. parapertussis, vaccines and cycles of whooping cough"
23. Montilla-Escudero, Rojas-Baquero, Virguez (2013) "Antioquia (Colombia) en" *Infectio*
24. Szwejser-Zawislak, Wilk, Piszczek et al. (2023) "Evaluation of whole-cell and acellular pertussis vaccines in the context of long-term herd immunity" *Vaccines (Basel)*
25. (2020) "Instituto Nacioinal de Salud. Informe de Evento Tosferina"
26. Ximena, María (2025) "Informe de Evento Tos Ferina 2021"
27. Ximena, María, Alberto et al. (2022) "Pertussis in an indigenous population in the Sierra" *Am J Field Epidemiol*
28. Roh, Kim, Kim et al. (2021) "Coinfections with respiratory pathogens among COVID-19 patients in Korea" *Can J Infect Dis Med Microbiol*
29. Chen, Liao, Cheng et al. (2020) "The microbial coinfection in COVID-19" *Appl Microbiol Biotechnol*
30. Ferronato, Leite, Vieira (2021) "The role of respiratory virus infection in suspected pertussis: a prospective study" *J Microbiol Immunol Infect*
31. (2025) "Boletín Epidemiológico Semanal (BES)"
32. De, Salud (2024) "Alerta epidemiológica -Tos ferina (coqueluche) en la Región de las Américas"
33. Broutin, Guégan, Elguero et al. (2005) "Largescale comparative analysis of pertussis population dynamics: periodicity, synchrony, and impact of vaccination" *Am J Epidemiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12768961&blobtype=pdf | # Corrections & amendments Author Correction: Single-particle genomics uncovers abundant non-canonical marine viruses from nanolitre volumes
Alaina Weinheimer, Julia Bro, Brian Thompson, Greta Leonaviciene, Vaidotas Kiseliovas, Simonas Joc, Jacob Munson-Mcgee, Gregory Gavelis, Corianna Mascena, Lina Mazutis, Nicole Poulton, Rapolas Zilionis, Ramunas Stepanauskas, Nature Microbiology
## Abstract
In the version of the article initially published, in Fig. 1, the panel labels "d" and "f" were switched and have now been corrected in the HTML and PDF versions of the article. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12720266&blobtype=pdf | # The influence of cell culture media and their additives on virus inactivation in vitro Der Einfluss von Zellkulturmedien und ihren Zusätzen auf die Virusinaktivierung in vitro
Florian Brill, Britta Becker, Anke Herrmann, Toni Meister, Eike Steinmann, Jochen Steinmann, Katharina Konrat, H Florian, Brill
## Abstract
Aim: Comparative inactivation studies with the murine norovirus (MNV) and different test substances showed considerable different results
## Introduction
Chemical disinfectants with proven virus inactivating properties are used in hospitals and other medical settings to interrupt infection chains. These disinfectants should be able to disinfect depending form the formulation e.g. hands, surfaces, instruments, and laundry. Their virucidal activity (reduction of virus titres by at least 4 log steps) is tested according to national and international guidelines with different test viruses with and without envelope in a quantitative suspension test, such as the guideline of DVV/RKI [1] or the EN 14476 standard [2], followed by tests under practical conditions, like EN 16777 [3] and EN 17111 [4]. In general, the different guidelines standardize many test parameters such as test viruses, test temperature, exposure time, soil load, and the internal validation controls. They also provide a detailed description of the testing procedure. At the same time, however, the standards allow a certain amount of flexibility e.g. in the production of the test virus suspension, the cell lines used for virus propagation and endpoint titration, and finally the respective cell culture media and their supplements used for the different systems. This raises the question of whether this flexibility could cause and/or explain different results in inactivation studies between different testing laboratories, when using the same test viruses and the same test substances. In this study, two independent laboratories (Lab 1 and Lab 2) compared the influence of different cell culture media, used for virus propagation and cell culture, on the results of suspension tests according to EN 14476:2019 with the murine norovirus (MNV) as non-enveloped test virus and different test substances. The results obtained showed large differences in virus reduction depending on the cell culture medium used and, in particular, its supplements.
## Materials and methods
## Cells and culture media
## Virus propagation
Virus suspensions were prepared by replicating the murine norovirus (MNV), strain S99 (RVB 0651, obtained from Friedrich-Loeffler-Institute, Greifswald, Insel Riems, Germany) in RAW 264.7 cells. The two laboratories used both DMEM as well as EMEM/MEM (1:1) with different supplements, depending on the test performed, as indicated in the results. A subconfluent monolayer of RAW cells was inoculated with MNV and incubated at 37°C in 5% CO 2 . After the cytopathic effect (CPE) became evident, cells were lysed by freeze-thaw cycle and virus suspension (virus pool) was harvested by removing cell debris by low speed centrifugation and stored in aliquots at -80°C.
## Test formulations
The formulations to be tested against MNV were prepared from commercially available products using Aqua dest., PBS (formaldehyde) or standardized hard water (propan-2-ol; Lab 2 only) as diluent at various concentrations (see Table 1).
## Determination of virucidal activity
In both laboratories tests were performed according to EN 14476:2019 [2] under clean conditions at 20°C. In general, 1 part of the prepared virus suspension was mixed with 1 part of 3 g/L bovine albumin (BSA) and 8 parts of the respective test formulation or 8 parts of aqua bidest (or hard water for propan-2-ol (Lab 2)) for the virus control (exception: 0.7% formaldehyde was prepared as described in the EN 14476). In some experi-Table 1: Overview of the test products and conditions used in the study ments, the virus suspension was mixed with varying amounts of HEPES or 3[N-morpholino]propane sulfonic acid (MOPS) buffer (ThermoFisher, art.-no. J61821) before preparing the actual inactivation mix. After the respective contact time, infectivity was stopped by immediate serial dilutions (1:10 dilutions) with ice-cold medium. The virus titres were calculated by applying 100 µL of each dilution in eight (Lab 1) or ten (Lab 2) wells of a 96-well microtitre plate containing 100 µL of cell suspension (endpoint dilution assay), followed by an incubation of the microtitre plate at 37°C in 5% CO 2 until a cytopathic effect could be detected. The respective virus titres were determined using the method of Spearman [5] and Kaerber [6] and expressed as lg TCID 50 /mL. The reduction factors (RF) were calculated as the difference between the virus titre after exposure to the formulation analyzed and the titre of the corresponding virus control (assay performed as described above but with water instead of test product).
## Results
## Influence of different culture media for virus propagation on the virucidal activity of 70% propan-2-ol
In the past, different results were reported from MNV inactivation studies according to EN 14476 (especially with alcoholic test substances). To find the reason, the culture media and all corresponding additives (including FCS) routinely used for this system were exchanged between the two laboratories involved in this study. Lab 1 normally used DMEM with 4.5 g/L glucose and its additives L-glutamine, NEA, sodium pyruvate, and HEPES and FCS for cell culture. Lab 2 commonly used EMEM/MEM with L-glutamine (1:1), NEA and sodium pyruvate as supplements and FCS. Both laboratories produced new virus pools under their own conditions and with their own RAW cells and stock virus and their own medium in parallel with the other laboratory's medium. For the first experiments, Lab 1 used the DMEM and the EMEM/MEM with all additives as described above and 2% of the respective FCS for virus propagation. Lab 2 prepared the MNV suspension without any supplements in both media. The virus pools produced were then used for an inactivation study in the suspension test with 70% propan-2-ol as the final concentration. The results are shown in Figure 1. Lab 1 achieved a reduction factor (RF) of 2.22 lg after 30 seconds exposure using DMEM (with all described additives including FCS) as the culture medium for virus propagation and inactivation tests. In contrast, with the EMEM/MEM medium (with the respective additives received from Lab 2) for virus propagation and titration, a RF of 5.44 lg after 30 s of exposure could be demonstrated (Figure 1A). In contrast, in Lab 2 only low differences in the reduction factors were determined, when using DMEM or EMEM/MEM without additives for virus propagation (Figure 1B). In summary, these results show that obviously specific additives are responsible for the differences obtained.
## Influence of FCS and HEPES in cell culture media for virus propagation on the virucidal activity of 70% propan-2-ol
Next, the effect of the FCS as an important and essential supplement in media during cell culture procedures was investigated as a possible cause. Therefore, pre-cultivation of the RAW cells, virus propagation, suspension test and titration were performed with both media as described above, but instead of the respective FCS for each medium, the same FCS was used for both media (Lab 1) or the medium with and without FCS (Lab 2) during virus pool production. Both laboratories achieved almost the same results as described above in the inactivation studies with 70% propan-2-ol with these new virus pools (data not shown), indicating, that FCS had no effect on the results. In the next step, the impact of HEPES, a commonly used buffer for cell culture medium was analysed. Suspension tests and titration of the respective dilutions were performed as in the previous studies, but virus production was carried out in both laboratories using only DMEM as the cell culture medium with or without HEPES. As shown in Figure 2, the addition or removal of the buffer during the virus propagation had a major impact on the results of the inactivation tests with propan-2-ol against MNV. The mean RFs for the systems using the medium with HEPES were 1.63 lg (Lab 1, Figure 2A) and 1.78 lg (Lab 2; Figure 2B) after 30 s of incubation and the mean RF for the systems without HEPES were 5.00 lg (Lab 1) and 4.76 lg (Lab 2) (Figure 2A and Figure 2B).
## Influence of HEPES in the culture media for virus propagation on the virucidal activity of additional active substances
To investigate whether HEPES could also affect the results of other active ingredients commonly used in disinfectants, further inactivation studies were performed according to EN 14476 with the same virus pools (generated in media with and without HEPES) and varied active formulations and concentrations. Beside the 70% v/v propan-2-ol solution tested so far, also significant differences in the results with a 60% v/v propan-2-ol solution as well as with other alcohols, as 50% v/v ethanol and 50% and 60% v/v propan-1-ol, could be shown in Lab 1 (Figure 3A). In addition, a slight effect with 100 ppm PAA as test substance could be detected, whereas no differences in the results could be shown testing formaldehyde or GDA (Figure 3B). Lab 2 produced similar results with significant differences in the reduction factors with 40% and 50% w/w ethanol as the test concentration, with no significant differences measured with GDA (Figure 3C) and 30% w/w ethanol (data not shown).
## Impact of the addition of HEPES or MOPS to the virus suspension
The final experiments tested whether adding a buffer directly to the virus suspension before an inactivation 4A). Similar results could be obtained with 40% w/w ethanol (Figure 4B) and no effect on the results with GDA could be observed (Figure 4C). Lab 1 performed these tests using only 70% v/v propan-2-ol and a virus pool propagated in DMEM without buffer (with FCS). HEPES and, in an additional approach, MOPS, which is also commonly used as a buffering agent in cell culture, were added to the test virus suspension at concentrations of 20 and 200 mM for 10 and 30 minutes before the inactivation assay. As a control, approaches were performed with the "pure" virus suspensions propagated in DMEM with or without HEPES. The mean reduction factor of the control (virus pool production without HEPES and the pure virus suspension) was 5.42 lg after 30 s of exposure (Figure 5A). In contrast, after 10 min of pre-incubation of the same virus suspension with HEPES this RF decreased to 4.08 lg (addition of 20 mM HEPES) and 3.38 lg (addition of 200 mM HEPES) (Figure 5B). Even after the addition of MOPS buffer, there was a strong decrease in the RF to 2.96 lg (20 mM) and 1.71 lg (200 mM). Prolonging the pre-incubation to 30 min with 20 mM buffer had no significant effect (Figure 5B and Figure 5C). However, not only the presence of HEPES during pool preparation itself, but also the subsequent addition of HEPES or MOPS to a pure virus suspension (propagated in medium without HEPES) can have a strong impact on the results of inactivation studies.
## Discussion
The various basic media routinely used for the cultivation of cell lines contain nutrients intended to ensure growth of the cells in culture. The ready-to-use composition of these media with their specific ingredients and individual additives (including FCS) is often cell type-specific and is also determined, among other things, by the type and manner of culture conditions (e.g. duration of cultivation or cultivation with and without CO 2 supply).
As the current standards for performing virucidal tests
do not yet contain any regulations on the medium used or the cells to be used for a specific test virus, the different systems consisting of test virus and cell line are generally established by the various test laboratories themselves on the basis of successful virus propagation to obtain high titres. This means that, when testing disinfectants for viral activity, different media and additives as well as different cells could be used with MNV as test virus in the different test laboratories. In this study, it could be shown for the first time that the choice and composition of cell culture media can have a significant HEPES is actually classified as a zwitterionic "GOOD" buffer and is routinely added to the medium at concentrations ranging from 10 to 25 mM if, for example, the buffer capacity of the bicarbonate in a medium is no longer sufficient (e.g. when cell cultures are cultivated without CO 2 in a closed system or stored outside the CO 2 incubator for a longer period of time) [7], [8]. However, if HEPES was contained in the cell culture medium (in this case DMEM) during virus propagation, this led to a drastic drop in the reduction factors after just 30 seconds in the inactivation assays with 70% v/v propan-2-ol. The difference in the mean RF without and with buffer in the medium amounted to Δ3.22 (Figure 1A) and Δ3.37 (Figure 2A) or Δ2.98 lg levels (Figure 2B). Such significant differences were not limited to results with propan-2-ol, but could also be obtained with other alcoholic formulations in the MNV inactivation assays, such as ethanol or propan-1-ol (Figure 3A and Figure 3C). Of note, some of the test concentrations of the alcoholic formulations were too low or too high so that the differences were difficult or impossible to visualize because of the excess or lack of activity. This may also apply to the inactivation studies with 100 ppm PAA against the MNV. Here, the RF was above 4 lg steps for all approaches and only slight differences could be detected when using virus suspensions from a culture with and without HEPES (Figure 3B). To detect a possible effect, further studies with PAA at a lower concentration (e.g. 50 ppm) would be necessary. In contrast, in the studies with GDA and formaldehyde, there were no visible effects (regardless of whether HEPES was added to the DMEM during virus cultivation or not, the reduction factors achieved were almost identical (Figure 3B).
In the past, a number of findings have already been made in connection with the use of HEPES in cell culture media, which describe further advantages and/or disadvantages in addition to its efficient buffering capacity. For instance, the presence of HEPES in cell culture medium promotes the uptake and transfection of proteins [9], [10]. The buffer can also be taken up by cells through endocytosis and thus influence various intracellular processes [11], [12], [13]. Hugel et al. [14] also showed the pH-dependent inhibition of GABAA receptors by HEPES, presumably through protonation of the buffer. This means that it is quite conceivable that the buffer substance could react directly with an individual active substance and thus inhibit it. It is also possible that HEPES interacts directly with the virus and that the virus is thus better protected against the attack by an active substance. The final inactivation studies carried out with HEPES and MOPS as a second buffer substance, in which only the virus (from propagation without buffer) was spiked with HEPES or MOPS [20 mM or 200 mM] shortly before the start of the experiment, do not provide a definitive explanation for the effects described. However, results show that significant differences in the reduction factors can occur not only after virus propagation in medium with and without buffer, but also after propagation of a MNV pool without HEPES and the subsequent addition of the buffer to the virus. In Lab 2, the RF after an exposure time of 30 seconds in the preparations without HEPES were 2.73 lg to 3.00 lg (70% v/v propan-2-ol) and 2.08 lg to 3.05 lg (40% w/w ethanol) lg levels higher than in the preparations in which the virus was mixed with the HEPES buffer (Figure 4). This means that the difference between the preparations with and without HEPES was close to the results achieved so far in the laboratory. In Lab 1, the results with 20 mM HEPES in the virus buffer mixture and 70% v/v propan-2-ol were not quite as clear, even if the effect was further enhanced by increasing the buffer concentration to 200 mM (Figure 5). However, it was interesting to note that after the addition of the MOPS buffer to the virus suspension, even higher effects were detected than with the use of the HEPES virus mixture (Figure 5C). If the buffer substance would be attached directly to the virus itself, then the direct addition of HEPES to the virus in both laboratories should have given roughly similar results, comparable to the results of the previous experiments with and without HEPES during virus propagation. However, this was not the case in Lab 1. Thus, the cause(s) or mechanism(s) ultimately responsible for the results of the MNV inactivation studies with or without HEPES (or MOPS) remain still unclear. It is possible, that the reported effect may be a combination of an interaction between the virus and the buffer, in which the virus replication itself also plays a role, and an interaction of the buffer with the test substance. But this should be the subject of further investigations. In our study, HEPES is the first cell culture additive to be identified that can significantly influence the results of inactivation studies with MNV, and thus the resulting activity of disinfectants, depending on the test substance and its concentration. At this time, it is not possible to assess which viruses this ultimately concerns or which specific factors in other systems may influence the results of virucidal tests individually. But the data obtained impressively illustrate the importance of better standardisation of the test specifications in guidelines and norms to ensure that results are comparable even between different test laboratories. Thereby, it is not only the standardisation of the used cell culture media and their additives that is crucial for valid virucidal inactivation tests, but it may also be necessary to determine the cell line for the respective replication system.
## Notes
Authors' ORCIDs
## References
1. Rabenau, Schwebke, Blümel et al. (2014) "Leitlinie der Deutschen Vereinigung zur Bekämpfung der Viruskrankheiten (DVV) e. V. und des Robert Koch-Instituts (RKI) zur Prüfung von chemischen Desinfektionsmitteln auf Wirksamkeit gegen Viren in der Humanmedizin : Fassung vom 1. Dezember 2014 [Guideline of the German Association for the Control of Viral Diseases (DVV) eV and the Robert Koch Institute (RKI) for testing chemical disinfectants for effectiveness against viruses in human medicine"
2. (2015) *Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz*
3. Din En 14476 (2013) "Chemical disinfectants and antiseptics -quantitative suspension test for the evaluation of virucidal activity in the medical area -test method and requirements (phase 2, step 1)" *+A*
4. "Chemical disinfectants and antiseptics -quantitative non-porous surface test without mechanical action for the evaluation of virucidal activity of chemical disinfectants used in the medical area -test method and requirements (phase 2, step 2)"
5. "Chemical disinfectants and antiseptics -quantitative carrier test for the evaluation of virucidal activity for instruments used in the medical area -test method and requirements (phase 2/step 2)"
6. Spearman (1908) "The method of "right or wrong cases" (constant stimuli) without Gauss's formulae" *Brit J Psychol*
7. Kärber (1931) "Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche" *Arch Exp Path Pharmak*
8. Good, Winget, Winter et al. (1966) "Hydrogen ion buffers for biological research" *Biochemistry*
9. Shipman (1973) "Control of Culture pH with Synthetic Buffers"
10. Luo, Pal, Shah et al. (2010) "Effect of HEPES buffer on the uptake and transport of P-glycoprotein substrates and large neutral amino acids" *Mol Pharm*
11. Chen, Chao, Tsai et al. (2019) "Utilization of HEPES for Enhancing Protein Transfection into Mammalian Cells" *Mol Ther Methods Clin Dev*
12. Depping, Seeger (2019) "H-NMR spectroscopy shows cellular uptake of HEPES buffer by human cell lines-an effect to be considered in cell culture experiments" *Anal Bioanal Chem*
13. Tol, Van Der Lienden, Gabriel et al. (2018) "HEPES activates a MiT/TFE-dependent lysosomal-autophagic gene network in cultured cells: A call for caution" *Autophagy*
14. Van Der Lienden, Aten, Boot et al. (2022) "HEPES-buffering of bicarbonate-containing culture medium perturbs lysosomal glucocerebrosidase activity" *J Cell Biochem*
15. Hugel, Kadiri, Rodeau et al. (2012) "pHdependent inhibition of native GABA(A) receptors by HEPES" *Br J Pharmacol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645911&blobtype=pdf | # Chronology of H3N2 human influenza virus surface glycoprotein adaptation from 1968 to 2019 reveals a surge of adaptation between 1997 and 2002
Hui Lei, Biying Xiao, Xia Lin, Zaolan Liang, Shiman Ling, Yaqin Bai, Vijaykrishna Dhanasekaran, Wenjun Song, Sook-San Wong, Mark Zanin
## Abstract
Subtype H3N2 influenza A viruses (IAVs), which emerged in 1968 to cause a pandemic, have shown continual circulation and adaptation that has necessitated frequent updates of candidate vaccine viruses. Here, we sought to determine how genetic changes in the hemagglutinin (HA) and neuraminidase of 21 antigenically distinct H3N2 IAVs isolated from 1968 to 2019 correlate with mammalian fitness and adaptation. We found a surge of adaptation between 1997 and 2002, resulting in the emergence of A/Fujian/411/2002 (H3N2) and poor vaccine efficacy, leading to an epidemic during the 2003-2004 season. This surge was characterized by a large reduction in binding to mammalian-type α2,6-linked sialic acids and increased infectivity and replication kinetics in humanized Madin-Darby canine kidney cells. HA glycosyla tion also increased most rapidly from 1968 to 2004 and then plateaued. Symptomatic infections were only evident in mice following inoculation with viruses isolated in the 1970s and A/Aichi/2/1968 (H3N2), which was the most pathogenic. More recent viruses did not cause any detectable symptoms, except for A/Sydney/5/1997 (H3N2), which caused some weight loss. The post-2002 shift to α2,6-linked sialic acid binding, coupled with reduced pathogenicity in mammalian models, underscores H3N2 adaptation to human circulation without affecting immunogenicity, which is a critical consideration for vaccine design. Overall, our data revealed that a surge of mammalian adaptation from 1997 to 2002 gave rise to A/Fujian/411/2002 (H3N2), with subsequent viruses showing more hallmarks of mammalian adaptation, such as increased binding to cells expressing α2,6-linked sialic acids and reduced mammalian pathogenicity.
IMPORTANCEThe continued endemicity of subtype H3N2 influenza A viruses (IAVs) in humans necessitates an understanding of the continuing accumulation of mammalian adaptations to inform public health countermeasures. We used a combined approach of studying the genetic, antigenic, and pathogenic adaptations of the surface glycoproteins of 21 H3N2 IAVs representing distinct antigenic groups from 1968 to 2019. We observed a loss of mammalian pathogenicity within 10 years of human circulation and a surge of adaptation between 1997 and 2002 that gave rise to A/Fujian/411/2002 (H3N2), which was a poor match for vaccine viruses. This surge was characterized by large shifts in glycan binding preferences, antigenicity, and genetic evolutionary distances. Overall, our study reveals novel insights into the chronology of the mammalian adaptation of H3N2 IAVs.
pathogen. The H3N2 subtype triggered the "Hong Kong flu" pandemic of 1968 to 1970, causing over a million deaths worldwide (1,2). These viruses originated through reassortment, combining avian-derived hemagglutinin (HA) and polymerase basic 1 (PB1) genes with six genes from the previously circulating H2N2 viruses, including neuraminidase (NA) (3). A key adaptation was the early gain of binding preference for α2,6-linked sialic acids, a prerequisite for efficient human-to-human transmission. Since then, H3N2 has remained a leading cause of seasonal influenza morbidity and mortality worldwide (4).
Compared to other human influenza viruses, such as H1N1 , H1N1pdm09 (2009-present), H2N2 (1957)(1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968), and influenza B, H3N2 exhibits exceptionally rapid antigenic evolution (5)(6)(7). Antigenic variants emerge every 2 to 5 years through HA substitutions that escape pre-existing immunity, necessitating frequent updates to the H3N2 vaccine component, 37 times to date (8). This antigenic instability underscores the public health risks posed by this subtype.
H3N2 adaptation has involved a gradual shift in receptor specificity and binding avidity. Between 1968 and the early 1990s, the virus transitioned from dual α2,3/α2,6 binding to exclusive α2,6 recognition (9,10), with biophysical analysis showing a gradual evolution from mixed α2,3/α2,6 binding to α2,6 specificity by 1979, followed by a decline in binding avidity after 1992 (11). By 1992, clinical isolates no longer aggluti nated chicken red blood cells, an indicator of diminished avian receptor binding (12). These changes correlate with reduced infectivity in standard Madin-Darby canine kidney (MDCK) cells, favoring infection in MDCK cells expressing the cDNA of human 2,6-sia lyltransferase (SIAT1) (MDCK-SIAT1 cells) engineered to express human-type receptors (13). Concurrently, H3N2 viruses evolved to preferentially bind elongated α2,6-linked sialic acids, likely under immune pressure (14)(15)(16). These receptor adaptations coincided with increasing glycosylation near the HA receptor-binding site, which both attenuates receptor engagement and masks antigenic epitopes (17,18). Given the overlap of receptor-binding and antigenic sites (A, B, D), mutations conferring immune escape often simultaneously affect receptor interactions (13,15,19,20).
H3N2 antigenic drift has practical implications for vaccine efficacy (VE). Notably, VE against H3N2 viruses has been suboptimal in multiple seasons, at only 22%-36% in the USA during the 2016-2017, 2017-2018, and 2021-2022 influenza seasons, despite antigenic similarity between vaccine and circulating strains (21)(22)(23). While HA is the focus, NA immunity is elicited by natural infection and vaccination, although the NA content of seasonal influenza vaccines is not standardized (24)(25)(26). NA antigenic drift has also been observed, exemplified by the emergence of S245N and S247T in 2016, which introduced an N-linked glycosylation site in a conserved epitope overlapping the active site (27,28). This led to reduced neuraminidase inhibition (NAI) titers, reduced binding of monoclonal antibodies raised against earlier strains, and reduced protection afforded by immunity generated by earlier viruses lacking this glycosylation site in vivo (27,28). Furthermore, reduced VE may be further compounded by adaptive mutations introduced during egg-based vaccine production (29,30).
Poor VE was particularly evident upon the emergence of the antigenically variant strain A/Fujian/411/2002 (H3N2), which was a poor antigenic match to the 2002-2003 vaccine strain A/Moscow/10/1999 (H3N2) (31,32). This led to an epidemic during the 2003-2004 influenza season (6,33,34).
While specific genetic or antigenic shifts in H3N2 strains have been documented, an integrative, longitudinal approach to understanding how key phenotypic traits, namely receptor binding, antigenicity, replication, and pathogenicity, co-evolve over time is needed. The persistent success of H3N2 viruses likely reflects the coordinated evolution of these traits. In this study, we reconstructed 5 decades of H3N2 evolution (1968-2019) using 21 recombinant viruses representing major antigenic clusters and vaccine strains. We assessed receptor specificity, glycosylation, hemagglutination, replication in mammalian cells, and pathogenicity in mice. These traits were selected based on their established roles in host adaptation, immune evasion, and disease severity. Our findings reveal a phased trajectory of adaptation, including early loss of avian features and a major evolutionary inflection between 1997 and 2002, marked by shifts in receptor binding and cell tropism. This longitudinal analysis provides an integrated perspective on the phenotypic strategies that have enabled H3N2 viruses to persist and adapt in humans over 5 decades.
## MATERIALS AND METHODS
## Virus selection, generation, and sequencing
Twenty-one human H3N2 viruses isolated between 1968 and 2019 were selected for this study (Table 1). Viruses from 1968 to 2003 were chosen to represent defined antigenic clusters, while those from 2004 to 2019 corresponded to WHO candidate vaccine viruses (5,8). Where possible, the HA and NA sequences from cell-grown viruses were used. However, the HA and NA sequences of A/Aichi/2/1968 (H3N2), A/England/42/1972 (H3N2), A/Victoria/3/1975 (H3N2), A/Sichuan/2/1987 (H3N2), and A/Wuhan/359/1995 (H3N2) were obtained from egg-passaged viruses, as sequences from cell-passaged viruses were not available. pHW2000 plasmids encoding the HA or NA genes listed in Table 1 were synthesized (Sangon Biotech) and used to generate recombinant viruses by eight-plasmid reverse genetics using the HA and NA genes from each strain with the six internal genes of A/Puerto Rico/8/1934 (H1N1) (35). Viruses were propagated in the allantoic cavity of 9-day-old specific-pathogen-free embryonated chicken eggs, as these viruses replicated to high titers in this background. We confirmed that no mutations were introduced following propagation in chicken eggs by Sanger sequencing using strain-specific primers and comparison with published sequences (Table 2). Viral RNA was extracted using the FastPure Viral DNA/RNA Mini Kit (Vazyme) and reverse transcribed using HiScript II One Step RT-PCR Kit (Vazyme).
## Genetic analysis and glycosylation site prediction
Maximum likelihood phylogenetic trees of HA and NA amino acid sequences were constructed using IQ-TREE (version 2.0.3), with final model parameter optimization, and visualized by FigTree (version 1.4.4). The clades were colored according to the antigenic map of subtype H3N2 influenza A viruses (IAVs) and rooted to A/Aichi/2/1968 (H3N2) (5). Genetic distances were calculated using MEGA 11 using amino acid sequences, the p-distance model, and the maximum likelihood method and 1,000 bootstrap replicates. Distances were visualized in GraphPad Prism 9.1.1. N-linked glycosylation sites in HA and NA were predicted using NetNGlyc 1.0 (https://services.healthtech.dtu.dk/services/NetN Glyc-1.0/).
## Cell culture
MDCK cells were maintained in minimal essential medium (MEM) supplemented with 10% fetal bovine serum (FBS), 100 units/mL penicillin, 100 µg/mL streptomycin, 1% L-glutamine, and 1% vitamins solution. Humanized MDCK (hCK) cells were maintained in MEM supplemented with 5% FBS, MEM amino acids, MEM vitamins solution, 1% L-glutamine, 100 units/mL penicillin, 100 µg/mL streptomycin, 10 µg/mL blasticidin, and 2 µg/mL of puromycin. Cells were incubated at 37°C in a humidified atmosphere with 5% CO 2 . Infection media contained MEM with 0.3% bovine serum albumin (BSA), 100 units/mL penicillin, 100 µg/mL streptomycin, and 1% L-glutamine. Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle medium containing 10% FBS, 100 units/mL penicillin, 100 µg/mL streptomycin, 1% L-glutamine, and 1% vitamins.
## Virus titration and replication kinetics
Virus titers were determined by plaque assay in MDCK or hCK cells. Replication kinetics were assessed in two independent experiments, each with four replicates. Cells were infected at a multiplicity of infection of 0.01, and culture supernatants were harvested at designated time points and stored at -80°C. Titers were quantified by tissue culture infectious dose 50% in MDCK or hCK cells using the Reed and Muench method (36).
## Hemagglutination and hemagglutination inhibition (HAI) assays
Hemagglutination (HA) assays were performed using 0.5% (vol/vol) chicken red blood cells (cRBCs) and 0.75% (vol/vol) guinea pig RBCs (gRBCs) (37). HA titers were read after incubation at room temperature for 30 min (cRBCs) or 60 min (gRBCs) and expressed as the reciprocal of the highest virus dilution showing complete agglutination. Samples below the detection limit (1:10) were assigned a titer of five. HAI assays were conducted using receptor-destroying enzyme (Accurate Chemical)treated serum in 96-well plates. Serial serum dilutions were incubated with four HA units of virus for 1 hour at room temperature, followed by the addition of 0.75% gRBCs. HAI titers were defined as the highest serum dilution that completely inhibited hemaggluti nation.
## Receptor-binding assay
Receptor-binding specificities were determined by solid-phase enzyme-linked recep tor-binding assay as previously described (38). Virus stocks were prepared by passage in eggs followed by purification and concentration over a cushion of 25% sucrose in STE buffer (0.1M NaCl, 10 mM Tris-HCl, 1 mM EDTA, pH 8.0) and ultracentrifugation at 25,000 rpm for 1 hour at 4°C. The biotinylated sialic acid receptor analogs used were; 3´-sialyllactose (3´SL, Neu5Acα2-3Galβ1-4Glcβ-PAA-biotin), 6´-sialyllactose (6´SL, Neu5Acα2-6Galβ1-4Glcβ-PAA-biotin), 3´-sialyl-N-ace tyllactosamine (3´SLN, Neu5Acα2-3Galβ1-4GlcNAc) and 6´-sialyllactosamine (6´SLN, Neu5Acα2-6Galβ1-4GlcNAcβ-PAA-biotin) (GlycoNZ).
Polyvinyl chloride enzyme immunoassay (EIA) microplates (Corning) were coated with 200 µL/well of bovine fetuin in phosphate-buffered saline (PBS) (10 µg/mL) overnight and then washed with distilled water. Plates were then blocked with 200 µL/well PBS containing 5% BSA at room temperature for 1 hour. Viruses, normalized by HA titer, were diluted to 128 HA units in 50 µL PBS and added to the fetuin-coated plates for incubation at 4°C overnight, followed by washing with an ice-cold PBS washing buffer. Serial twofold dilutions of respective sialylglycopolymers in reaction buffer (0.02% BSA, 0.01% Tween 80, 1 µM zanamivir in PBS) were then added at 100 µL/well and incubated at 4°C for 2 hours. Plates were then washed five times with PBS containing 0.05% Tween-20, then 100 µL of streptavidin-peroxidase conjugates (Invitrogen) diluted 1/2,000 in reaction buffer were added to each well and incubated for 1 hour at 4°C. Plates were then incubated with 3,3´,5,5´-Tetramethylbenzidine (Sigma) for 10 min at room temperature for color formation, and reactions were then stopped using 50 mM HCl. Optical densities were measured at 450 nm in a Synergy 2 multi-mode microplate reader (BioTek Instruments). The apparent association constant (K ass ) for each virus-receptor interaction was calculated using Scatchard plot analysis.
## Enzyme-linked lectin assay
NAI titers in mouse serum were determined using enzyme-linked lectin assays (ELLA). Recombinant viruses were incubated in fetuin-coated 96-well plates at 37°C for 18 hours. Concentration of viruses used in ELLA was standardized using a twofold dilution curve. The virus dilution used was within the linear portion of the titration curve, with OD 450nm equal to 95% of the maximum achievable OD and at least 10 times higher than the background (no virus) control. Plates were then washed and incubated with peanut agglutinin-horseradish peroxidase(PNA-HRP) for 2 hours at room temperature. After development with O-phenylenediamine dihydrochloride, absorbance was measured at 490 nm. For NAI titration, serum samples were serially diluted from 1:10, incubated with virus, and added to fetuin-coated plates overnight at 37°C. The highest serum dilution yielding ≥50% inhibition was recorded as the NAI titer.
## Neuraminidase activity
Neuraminidase enzymatic activity was measured using the fluorogenic substrate 2
´-(4-methylumbelliferyl)-α-D -N-acetylneuraminic acid (MUNANA; Sigma) (39,40). Viruses were normalized by HA titer, serially diluted, and incubated with 100 µM MUNANA at 37°C for 30 min. Reactions were stopped with a solution of 25% (vol/vol) ethanol and 12.5% (vol/vol) glycine (Fisher Scientific) in distilled water, and fluorescence was measured at excitation/emission wavelengths of 360/460 nm. Data represent the mean of three independent experiments.
## Mouse experiments
Female BALB/cJ mice (6 weeks old; Zhejiang Weitong Lihua Experimental Animal Technology Co.) were anesthetized with isoflurane and intranasally inoculated with 10 3 , 10 4 , or 10 5 PFU) of virus in 30 µL PBS, while control mice received PBS alone. Animals were monitored at least daily for 14 days for weight loss and clinical signs and euthanized upon reaching humane endpoints (≥30% wt loss or severe illness). Sera were collected 3 days pre-and 28 days post-inoculation.
## Statistical analyses
Statistical analyses were performed using GraphPad Prism 9.1.1 (GraphPad Software Inc.). Comparisons between two groups were made using unpaired t-tests; comparisons among multiple groups were conducted using one-way analysis of variance. Survival curves were compared using the Kaplan-Meier method with log-rank test. P-values <0.05 were considered statistically significant.
## RESULTS
## Emergence of A/Fujian/411/2002 (H3N2) was associated with progressive accumulation of predicted glycosylation sites on HA, but not NA, and greater specificity for α2,6-linked sialic acids
Homology modeling of genetic evolutionary distances between these 21 viruses spanning 1968 to 2019 revealed continual evolution, with the largest transition between SY97 and FU02, as shown previously (31,32,41). This has essentially given rise to two groups of IAVs: those isolated in 1968 to 1997 (Early Group) and in 2002 to 2019 (Recent Group) (Fig. 1A through D). The genetic distances of the HA and NA amino acid sequences between IAVs within these groups were of the ranges 83.2% to 98.7% and 85.6% to 99.4%, respectively. There were 12 stable amino acid changes differentiating the groups, at positions 25,50,57,75,137,172,192,202,222,386,452, and 530 (H3 numbering, Fig. 2; Table 3). Six of these amino acids are located in the HA globular head, with 57 and 75 in antigenic site E, 137 in the 130 ring of the receptor binding site (RBS) that is part of antigenic site A, 172 in antigenic site D, 192 in the 190 helix, and 222 in the 220 ring of the RBS (Fig. 2; Table 3) (42)(43)(44). Notably, FU02, the earliest Recent Group IAV, was the first with 156H and 225G, which are determinants of sialic acid binding specificity, with 225 additionally affecting HA stability and pH of activation (45)(46)(47)(48).
We next analyzed the predicted N-linked glycosylation sites on these HAs and NAs to assess their evolution over time. Glycosylation of HA, particularly on the globular head, can shield antigenic sites and influence immune evasion (17,18,49). At baseline, the prototype strain A/Aichi/1/1968 (AI68) carried seven glycosylation sites on HA (8NSTA, 22NGTL, 38NATE, 81NETW, 165NVTM, 285NGSI, and 483NGTY). Additional sites were acquired by 1987, and two more by 1997, reaching a peak of 11 sites in SY97 (Fig. 1E; Table 4). This was followed by modest fluctuations, with one site lost in 2005, two gained by 2011, and subsequent losses in 2013 and 2017. A single gain in 2019 restored the total to approximately 13 sites in recent viruses. Overall, HA glycosylation sites showed a dynamic pattern of accumulation and refinement, with major gains in the 1970s-1990s and a more variable profile in the post-2000 era.
In contrast, NA glycosylation site dynamics were more variable and showed no clear trend of progressive accumulation. NA initially had eight predicted sites in HK68, with two sites lost by 1972 and intermittent gains and losses over the next five decades. For example, glycosylation sites were gained in 1975, 1989, 2002 and 2005, while sites were lost in 2004, 2009, 2011, and 2017. Despite this fluctuation, the overall number of predicted NA glycosylation sites remained relatively stable, typically ranging between six and nine across decades (Fig. 1E Table 5). Together, these results reveal a gradual accumulation of glycosylation sites on HA during early decades of human circulation, followed by dynamic site turnover in recent years. In contrast, NA glycosylation sites exhibited a more stochastic pattern without a consistent directional trend. 4 and5). (F) Ratios of HA titers obtained using gRBCs over cRBCs of subtype H3N2 influenza A viruses. Virus strain name abbreviations can be found in Table 1. We next examined changes in receptor-binding preferences, which are known to evolve with prolonged human circulation (50). HA assays using red blood cells from chicken (α2,3-and α2,6-linked sialic acids) and guinea pig (predominantly α2,6-linked sialic acids) (51) revealed that the emergence of FU02 was associated with consistently higher gRBC:cRBCs HA titer ratios, indicating a stronger binding preference for α2,6linked receptors consistent with adaptation to α2,6-rich environments (Fig. 1F). To validate this, we performed solid-phase binding assays using biotinylated sialylglyco peptides. AI68, EN72, VI75, BK79, and SI87 all showed relatively strong binding to the α2,6-linked glycans 6´SL and 6´SLN (Fig. 3A through H). However, WI05, VI11, SWZ13, and SG16, all isolated after 2002, showed markedly diminished or undetectable binding to α2,6-linked glycans (Fig. 3I through L). AI68 and SI87 were the only viruses that showed appreciable binding to α2,3-linked glycans (Fig. 3A andE). Overall, these data reveal that viruses isolated after FU02 showed a large decrease in binding affinity for the α2,6-linked glycans used in this study.
## Emergence of A/Fujian/411/2002 (H3N2) was associated with greater replication in humanized MDCK cells
To assess functional implications of the observed shift in sialic acid specificity, we compared replication in MDCK cells and hCK cells. MDCK cells express both α2,3-and α2,6-linked sialic acids, while hCK cells have been engineered to express predominantly α2,6-linked sialic acids with extremely low expression of α2,3-linked sialic acids by CRISPR-Cas9 knockout of the genes encoding β-galactoside α-2,3 sialyltransferase (52). While all viruses replicated similarly in MDCK cells (Fig. 4A andC), recent viruses showed enhanced replication kinetics in hCK cells at 12 hours post-infection (Fig. 4B andD, P ≤ 0.05). These results indicate a temporal shift from mixed α2,3/α2,6 binding toward preferential, albeit weaker, α2,6 binding in more recent H3N2 viruses, likely shaped by increased HA glycosylation and human host adaptation.
## NA enzymatic activity was varied while predicted NA N-linked glycosylation increased over time
NA enzymatic activity is an important component of the influenza virus lifecycle and is critical for HA-NA functional balance (53)(54)(55). As such, we studied the NA activities of these viruses using the fluorogenic substrate 2´-(4-methylumbelliferyl)-α-D -N-acetyl neuraminic acid (MUNANA). AI68 demonstrated the highest NA activity, with activities subsequently decreasing until TX77, then plateauing until BE92, after which activities were more variable but generally lower than BE92, with the exception of VI11 and SWZ13 (Fig. 5A andB). Located in antigenic site C (colored green in Fig. 2). d Located in antigenic site A (colored red in Fig. 2).
e Located in antigenic site B (colored blue in Fig. 2).
f Located in antigenic site D (colored magenta in Fig. 2). Predicted NA N-linked glycosylation sites showed an increasing trend, from 7 on AI68 NA to 9 on SI87 to 12 on FU02 before plateauing, varying from 11 to 13 for the remaining viruses (Fig. 5B). Overall, there was no correlation between the number of predicted NA Non-glycosylated sites are indicated by; -: potential < 0.5, --: potential < 0.5 and jury agreement, whereby all nine neural networks > 0.5, ---: potential < 0.32 and jury agreement. HA numbering is mature-form (without signal peptide).
c Located in antigenic site C (colored green in Fig. 2).
d Located in antigenic site D (colored magenta in Fig. 2). e Located in antigenic site A (colored red in Fig. 2). N-linked glycosylation sites and NA enzymatic activity (R 2 = 0.2331), but NA enzymatic activity showed an overall decreasing trend over time (Fig. 5B).
$$A/Aichi/2/1968 (H3N2) AI68 L K R H N D T V W E R V A/England/42/1972 (H3N2) EN72 L K R H N D T V W E R V A/Victoria/3/1975 (H3N2) VI75 L K R H N D T V W E R V A/Texas/1/1977 (H3N2) TX77 L R R H Y D T V W E R V A/Bangkok/1/1979 (H3N2) BK79 L R R H Y G T V W E R V A/Sichuan/2/1987 (H3N2) SI87 L R R H Y G T V W E R V A/Beijing/352/1989 (H3N2) BE89 L R R H Y G T V W E R V A/Beijing/32/1992 (H3N2) BE92 L R R H Y G T V W E R V A/Wuhan/359/1995 (H3N2) WU95 L R R H Y D T V W E R V A/Sydney/5/1997 (H3N2) SY97 L R R H Y D T V W E R V A/Fujian/411/2002 (H3N2) FU02 I G Q Q S E I I R G K A A/California/7/2004 (H3N2) CA04 I G Q Q S E I I R G K A A/Wisconsin/67/2005 (H3N2) WI05 I G Q Q S E I I R G K A A/Perth/16/2009 (H3N2) PE05 I E Q Q S E I I R G K A A/Victoria/361/2011 (H3N2) VI11 I E Q Q S E I I R G K A A/Switzerland/9715293/2013 (H3N2) SWZ13 I E Q Q S E I I R G K A A/Hong Kong/4801/2014 (H3N2) HK14 I E Q Q S E I I R G K A A/Singapore/INFIMH-16-0019/2016 (H3N2) SG16 I E Q Q S E I I R G K A A/Kansas/14/2017 (H3N2) KA17 I E Q Q F E I I R G K A A/Hong Kong/2671/2019 (H3N2) HK19 I E Q Q S E I I R G K A A/South Australia/34/2019 (H3N2) SA19 I E Q Q S E I I R G K A a The$$
## Correlation between predicted NA N-linked glycosylation sites and HA affinity for 6´SLN sialic acids
We next studied the time course of HA affinity for α2,3-and α2,6-linked sialic acids. Affinities for 6´SL sialic acids were relatively stable, with the exception of SI87, which showed a much greater affinity compared to other viruses (Fig. 5C). Affinities for 3´SL and 3´SLN sialic acids were lower compared to 6´ sialic acids until FU02, whereby affinities for 6´ sialic acids decreased approximately 10-fold from 259.8 ± 15.4 to 26.4 ± 5.2 (Fig. 5C). Overall, there was also a negative correlation between the number of predicted NA N-linked glycosylation sites and HA affinity for 6´SLN sialic acids (R 2 = 0.719).
## Attenuated pathogenicity of more recent strains in the mouse model
We next used the mouse model to determine the impact of these adaptations on mammalian pathogenicity using 9 H3N2 viruses. AI68 and EN72 were the most pathogenic, with all titers causing weight loss. At 10 5 PFU, mean body weights of mice inoculated with AI68 or EN72 dropped to a minimum of 73.76 ± 0.41% and 80.20 ± 4.99% of starting body weights, respectively (Fig. 6A andB). VI75 and SY97 were also pathogenic, with all titers of VI75 causing weight loss and 10 4 and 10 5 PFU of SY97 causing weight loss. At 10 5 PFU, mean body weights of mice inoculated with VI75 or SY97 name abbreviations can be found in Table 1. dropped to 90.11 ± 0.79% and 88.11 ± 4.42% of starting body weights, respectively (Fig. 6C andH). The other IAVs tested, namely the Early Group BK79, SI87, BE92, and WU95 and the Recent Group WI05, did not cause any significant mean weight loss (Fig. 6D through G and I). AI68 was the only virus to cause mice to reach humane endpoints, with 0% and 20% survival evident in mice inoculated with 10 5 and 10 4 PFU, respectively (Fig. 6J). We next used mouse serum collected 28 days post-inoculation to determine hemagglutina tion inhibition (HAI) and NAI antibody titers. We detected HAI and NAI titers in all mice, indicating that mice were infected and mounted immune responses (Fig. 7). Overall, these data indicate that, apart from SY97, the earlier H3N2 viruses were the most pathogenic and that all viruses were immunogenic.
## DISCUSSION
The continuing evolution of H3N2 viruses since their introduction into humans in 1968 has revealed insights into several virus-host interactions driving the selective pressures on these IAVs. Our goal was to use multiple approaches to study the mammalian adaptation of the HAs and NAs of H3N2 IAVs selected to represent antigenic clusters from 1968 to 2003 and subsequent vaccine strains from 2004 to 2019. Our findings highlight divergent evolutionary strategies in HA and NA, whereby HA accumulates glycosylation for immune evasion and receptor adaptation, while NA maintains enzymatic function despite stochastic glycosylation changes. This would agree with previous studies that have identified discordant antigenic drift between HA and NA, particularly between Virus strain name abbreviations can be found in Table 1. *P ≤ 0.05. WU95 and SY97, whereby a major step in HA antigenic drift was observed with almost no changes in NA (56)(57)(58).
The post-2002 shift to α2,6-linked sialic acid binding, coupled with reduced path ogenicity in mammalian models, underscores H3N2 adaptation to human circulation influenza A viruses by year of isolation. Affinity is expressed as apparent association constants (K ass ) (C). Virus strain name abbreviations can be found in Table 1.
Full-Length Text without affecting immunogenicity, which is a critical consideration for vaccine design. More recently, H3N2 NAs appear to be undergoing incremental changes. Mutations have accumulated in exposed positions on the globular head domain, many of them near the active site and the central Ca 2+ binding site that is critical for NA activity in the 2020-2021 vaccine strains, compared to HK19 (59,60). However, these NAs were still antigenically similar to HK19 (59).
Our genetic data revealed that the transition from SY97 to FU02 delineated the Early and Recent Groups, as observed previously (31,32,41). The emergence of FU02 marked a departure from the A/Moscow/10/1999 (H3N2)-like or antigenically equivalent A/Panama/2007/1999 (H3N2)-like IAVs. This was evident in the reduced VE observed when A/Panama/2007/1999 (H3N2) was used as the vaccine strain for the 2003-2004 (E), BE89 (F), WU95 (G), WI05 (H), and SY97 (I) (n = 5 mice per group). Virus strain name abbreviations can be found in Table 1. Limit of detection for these assays was 1:10 (log 2 3.3). Graphs show the mean and the ± standard error of the mean. Serum samples were not collected from the group inoculated with 10 5 plaque-forming units of AI68 as all mice reached humane endpoints prior to 28 days post-inoculation. NAI titers against SY97 were not determined due to lack of serum (I). reductions in binding specificities for α2,6-linked sialic acids in H3N2 viruses isolated in 2006, 2008, and 2010 compared to an isolate from 1992 have been observed (13,20). Binding specificity changes have been observed, with the dual specificity of AI68 for α2,3-and α2,6-linked sialic acids changing to a predominant α2,6-linked sialic acid specificity that then declined after 1995 (50). Similarly, compared to an earlier H3N2 virus, more recent H3N2 viruses showed much more restricted binding to a subset of glycans with α2,6-linkages (63). Preferences for glycans with longer, linear chains have also been observed, with the longer glycans postulated to engage the binding sites of two HA monomers in the same trimer, increasing binding avidity (15,16). Here, we also observed binding to the longer 6´SLN sialylglycopeptide compared to the shorter 6´SL sialylglycopeptide. Interestingly, our data also indicate that these changes in binding largely differentiated the Early and Recent Group viruses, suggesting an evolutionary event differentiating SY97 and FU02. Studies of antigenic evolution describe a trend of incremental increases in N-linked glycosylation sites on the globular head domain of HA. This has the effect of shielding the RBS from antibodies and contributing to poor VE, despite vaccine strains being well matched to circulating strains (17,18). This pattern was also evident in the IAVs studied here, with increasing numbers of predicted HA glycosylation sites until SY97, after which the numbers remained relatively stable, although we did not study the impact of increasing glycosylation sites on VE and the immunogenicity of these viruses. Interest ingly, this also corresponded to the transition between the Early and Recent Groups. However, changes in binding preferences and glycosylation do not necessarily correlate with changes in infectivity or replication in vitro or pathogenicity in the mouse model (64). We also noted this here, with Recent Group viruses showing greater replication in hCK cells but reduced pathogenicity in the mouse model. This may reflect adaptation to the human host rather than mammalian hosts in general.
The pathogenicity of H3N2 viruses has been observed to decrease with continued human circulation (65,66). Here, we saw a similar trend, with AI68 and IAVs isolated in the 1970s causing symptomatic infections in mice, and AI68 being the only IAV to cause mice to reach humane endpoints. IAVs isolated after the 1970s did not lead to any observable symptoms in mice apart from the higher doses of SY97. Changes in pathogenicity did not appear to correlate with changes in genetic relatedness, receptorbinding specificity, or replication of these IAVs, with Early Group IAVs isolated from 1979 to 1995 inclusive and all Recent Group IAVs eliciting no detectable symptoms in mice. The reasons behind this, and the pathogenicity of SY97, are unclear based on the data obtained in this study but may warrant further investigation.
We used the mouse model to study mammalian pathogenicity, which is widely used but has limitations in the context of human pathogenicity. There are notable differences between the human and murine respiratory tracts in the context of influenza virus infection. The human and murine respiratory tracts show similar expression of α2,3linked sialic acids but not α2,6-linked sialic acids, which are present in the human but not the mouse respiratory tract (67,68). Human H3N2 and H1N1 IAVs, which are generally adapted to bind to α2,6-linked sialic acids, show binding to human trachea, bronchus, bronchioles, and alveoli but not to mouse trachea, bronchus, and bronchioles, and only rare binding to mouse alveoli (69). The ferret model represents a better approximation of the human respiratory tract, showing predominant influenza virus infection of the upper respiratory tract and expression of α2,6-linked sialic acids (70). However, the costs and husbandry of ferrets are prohibitive for such a study.
There were some limitations to this study. It should be noted that these IAVs were "6 + 2" reassortants, meaning they contained H3N2 HA and NA genes with the internal genes of PR8. As such, this study focused on the IAV surface glycoproteins and did not study other possible determinants of mammalian fitness related to the internal genes, which have been described (71). Furthermore, the use of 6 + 2 viruses meant that the binding specificities could not conclusively be attributed to HA and may be impacted by NA sialic acid binding, as observed previously (72,73). As this study was focused on binding preferences for α2,3-or α2,6-linked sialic acids, we used a limited number of glycans, which is restrictive to the depth of interpretation of binding specificities. Further studies using larger numbers of glycans would yield more information.
Where possible, the HA and NA sequences in this study were obtained from cellpropagated viruses. However, this was not possible for some older viruses, for which HA and NA sequences were only available from egg-propagated viruses (Table 1). While this may have implications for sialic acid binding preferences, we did not find obvious differences between viruses in the Early Group that were generated using sequences derived from egg-or cell-passaged viruses. IAVs can accumulate mutations during egg passage, particularly in the HA receptor-binding site, to facilitate binding to α2,3-linked sialic acids (74). These mutations may also affect antigenicity as the receptor-binding site overlaps with HA antigenic sites (75). This phenomenon has been observed particularly for egg-grown H3N2 vaccine strains, such as G186V and L194P, which affect antigenicity, and T160K, which causes the loss of a glycosylation site (76,77).
While we observed that SY97 and FU02 delineated the Early and Recent Groups, our study did not include IAVs isolated between 1998 and 2001. Further studies of these IAVs may give more insights into this transition. It should also be noted that MUNANA is a relatively small substrate for NA compared to glycans, meaning that NA activity measured by MUNANA should be taken as a proxy for NA activity using glycans as a substrate. Furthermore, our glycosylation data were also derived from prediction and not experimentally validated.
In summary, we found that the continuous genetic evolution of H3N2 IAV HA and NA since their 1968 introduction into humans showed a surge of mammalian adaptation between 1997 and 2002, leading to the emergence of FU02 viruses and subsequent viruses with more hallmarks of mammalian adaptation and reduced pathogenicity in mammalian models. The post-2002 shift to α2,6-linked sialic acid binding, coupled with reduced pathogenicity in mammalian models, underscores H3N2 virus adaptation to human circulation without affecting immunogenicity, which is a critical consideration for vaccine design. This study has, therefore, yielded new findings on the evolutionary time course of H3N2 viruses in humans.
## References
1. Jester, Uyeki, Jernigan (2020) "Fifty years of influenza A(H3N2) following the pandemic of 1968" *Am J Public Health*
2. Honigsbaum (2020) "Revisiting the 1957 and 1968 influenza pandemics" *Lancet*
3. Kawaoka, Krauss, Webster (1989) "Avian-to-human transmission of the PB1 gene of influenza A viruses in the 1957 and 1968 pandemics" *J Virol*
4. Malosh, Mcgovern, Monto (2023) "Influenza during the 2010-2020 decade in the United States: seasonal outbreaks and vaccine interven tions" *Clin Infect Dis*
5. Smith, Lapedes, De Jong et al. (2004) "Mapping the antigenic and genetic evolution of influenza virus" *Science*
6. Russell, Jones, Barr et al. (2008) "The global circulation of seasonal influenza A (H3N2) viruses" *Science*
7. Bedford, Suchard, Lemey et al. (2014) "Integrating influenza antigenic dynamics with molecular evolution"
8. (2025) "Candidate vaccine viruses and potency testing reagents"
9. Allen, Ross (2018) "H3N2 influenza viruses in humans: viral mechanisms, evolution, and evaluation" *Hum Vaccin Immunother*
10. Connor, Kawaoka, Webster et al. (1994) "Receptor specificity in human, avian, and equine H2 and H3 influenza virus isolates" *Virology (Auckl)*
11. Medeiros, Escriou, Naffakh et al. (2001) "Hemagglutinin residues of recent human A(H3N2) influenza viruses that contribute to the inability to agglutinate chicken erythrocytes" *Virology (Auckl)*
12. Nobusawa, Ishihara, Morishita et al. (2000) "): a single amino acid change in hemagglutinin altered its recognition of sialyloligosaccharides" *Virology (Auckl)*
13. Lin, Xiong, Wharton et al. (2012) "Evolution of the receptor binding properties of the influenza A(H3N2) hemagglutinin" *Proc Natl Acad Sci*
14. Thompson, Wu, Canales et al. (2024) "Evolution of human H3N2 influenza virus receptor specificity has substantially expanded the receptor-binding domain site" *Cell Host Microbe*
15. Peng, De Vries, Grant et al. (2017) "Recent H3N2 viruses have evolved specificity for extended, branched human-type receptors, conferring potential for increased avidity" *Cell Host Microbe*
16. Yang, Carney, Chang et al. (2015) "Structure and receptor binding preferences of recombinant human A(H3N2) virus hemagglutinins" *Virology (Auckl)*
17. Ohuchi, Ohuchi, Feldmann et al. (1997) "Regulation of receptor binding affinity of influenza virus hemagglutinin by its carbohydrate moiety" *J Virol*
18. Gallagher, Henneberry, Wilson et al. (1988) "Addition of carbohydrate side chains at novel sites on influenza virus hemagglutinin can modulate the folding, transport, and activity of the molecule" *J Cell Biol*
19. Wu, Wilson (2020) "Influenza hemagglutinin structures and antibody recognition" *Cold Spring Harb Perspect Med*
20. Gulati, Smith, Cummings et al. (2013) "Human H3N2 influenza viruses isolated from 1968 to 2012 show varying preference for receptor substructures with no apparent consequences for disease or spread" *PLoS One*
21. Flannery, Chung, Monto et al. (2019) "Influenza vaccine effectiveness in the United States during the 2016-2017 season" *Clin Infect Dis*
22. Rolfes, Flannery, Chung et al. (2019) "Effects of influenza vaccination in the United States during the 2017-2018 influenza season" *Clin Infect Dis*
23. Price, Flannery, Talbot et al. (2023) "Influenza vaccine effectiveness against influenza A(H3N2)-related illness in the United States during the 2021-2022 influenza season" *Clin Infect Dis*
24. Chen, Wohlbold, Zheng et al. (2018) "Influenza infection in humans induces broadly cross-reactive and protective neuraminidase-reactive antibodies" *Cell*
25. Couch, Atmar, Keitel et al. (2012) "Randomized comparative study of the serum antihemagglutinin and antineuraminidase antibody responses to six licensed trivalent influenza vaccines" *Vaccine (Auckl)*
26. Cate, Rayford, Niño et al. (2010) "A high dosage influenza vaccine induced significantly more neuraminidase antibody than standard vaccine among elderly subjects" *Vaccine (Auckl)*
27. (2009)
28. Stadlbauer, Mcmahon, Turner et al. (2025) *Full-Length Text Journal of Virology*
29. Wilson, Ward, Krammer (2022) "Antibodies targeting the neuraminidase active site inhibit influenza H3N2 viruses with an S245N glycosylation site" *Nat Commun*
30. Wan, Gao, Yang et al. (2019) "The neuraminidase of A(H3N2) influenza viruses circulating since 2016 is antigenically distinct from the A/Hong Kong/4801/2014 vaccine strain" *Nat Microbiol*
31. Chen, Alvarez, Ng et al. (2019) "Passage adaptation correlates with the reduced efficacy of the influenza vaccine" *Clin Infect Dis*
32. Levine, Martin, Petrie et al. (2019) "Antibodies against egg-and cell-grown influenza A(H3N2) viruses in adults hospitalized during the 2017-2018 influenza season" *J Infect Dis*
33. Holmes, Ghedin, Miller et al. (2005) "Wholegenome analysis of human influenza A virus reveals multiple persistent lineages and reassortment among recent H3N2 viruses" *PLoS Biol*
34. Kang, Yang, Lee et al. (2010) "Epidemio logic study of human influenza virus infection in South Korea from 1999 to 2007: origin and evolution of A/Fujian/411/2002-like strains" *J Clin Microbiol*
35. Besselaar, Botha, Mcanerney et al. (2004) "Antigenic and molecular analysis of influenza A (H3N2) virus strains isolated from a localised influenza outbreak in South Africa in 2003" *J Med Virol*
36. Jin, Zhou, Liu et al. (1999) "Two residues in the hemagglutinin of A/Fujian/411/02-like influenza viruses are responsible for antigenic drift from A/Panama" *Virology (Auckl)*
37. Hoffmann, Neumann, Kawaoka et al. (2000) "A DNA transfection system for generation of influenza A virus from eight plasmids" *Proc Natl Acad Sci*
38. Reed, Muench (1938) "A simple method of estimating fifty per cent endpoints" *Am J Hyg*
39. (2002) "WHO manual for animal influenza diagnosis and surveillance"
40. Matrosovich, Gambaryan (2012) "Solid-phase assays of receptorbinding specificity" *Methods Mol Biol*
41. Gubareva, Webster, Hayden (2002) "Detection of influenza virus resistance to neuraminidase inhibitors by an enzyme inhibition assay" *Antiviral Res*
42. Potier, Mameli, Bélisle et al. (1979) "Fluorometric assay of neuraminidase with a sodium (4-methylumbelliferyl-alpha-D-Nacetylneuraminate) substrate" *Anal Biochem*
43. Koelle, Cobey, Grenfell et al. (2006) "Epochal evolution shapes the phylodynamics of interpandemic influenza A (H3N2) in humans" *Science*
44. Broecker, Liu, Sun et al. (2018) "Immunodominance of antigenic site B in the hemagglutinin of the current H3N2 influenza virus in humans and mice" *J Virol*
45. Popova, Smith, West et al. (2012) "Immunodominance of antigenic site B over site A of hemaggluti nin of recent H3N2 influenza viruses" *PLoS One*
46. Webster, Laver (1980) "Determination of the number of nonoverlapping antigenic areas on Hong Kong (H3N2) influenza virus hemagglutinin with monoclonal antibodies and the selection of variants with potential epidemiological significance" *Virology (Auckl)*
47. Koel, Burke, Bestebroer et al. (2013) "Substitutions near the receptor binding site determine major antigenic change during influenza virus evolution" *Science*
48. Kong, Guan, Shi et al. (2021) "A single-amino-acid mutation at position 225 in hemagglutinin attenuates H5N6 influenza virus in mice" *Emerg Microbes Infect*
49. Matrosovich, Tuzikov, Bovin et al. (2000) "Early alterations of the receptorbinding properties of H1, H2, and H3 avian influenza virus hemaggluti nins after their introduction into mammals" *J Virol*
50. Pascua, Song, Lee et al. (2012) "Virulence and transmissibility of H1N2 influenza virus in ferrets imply the continuing threat of triplereassortant swine viruses" *Proc Natl Acad Sci*
51. Zeng, Yau, Lin et al. (1318) "Characterization and evolutionary analysis of a novel H3N2 influenza A virus glycosylation motif in southern China" *Front Microbiol*
52. Liu, Bakker, Narimatsu et al. (2023) "H3N2 influenza A virus gradually adapts to human-type receptor binding and entry specificity after the start of the 1968 pandemic" *Proc Natl Acad Sci*
53. Kiseleva, Larionova, Kuznetsov et al. (2010) "Phenotypic characteristics of novel swine-origin influenza A" *virus. Influenza Other Respir Viruses*
54. Takada, Kawakami, Fan et al. (2019) "A humanized MDCK cell line for the efficient isolation and propagation of human influenza viruses" *Nat Microbiol*
55. Yen, Liang, Wu et al. (2011) "Hemagglutinin-neuraminidase balance confers respiratory-droplet transmissibility of the pandemic H1N1 influenza virus in ferrets" *Proc Natl Acad Sci*
56. Zanin, Marathe, Wong et al. (2015) "Pandemic swine H1N1 influenza viruses with almost undetectable neuraminidase activity are not transmitted via aerosols in ferrets and are inhibited by human mucus but not swine mucus" *J Virol*
57. Vahey, Fletcher (2019) "Influenza A virus surface proteins are organized to help penetrate host mucus"
58. Sandbulte, Westgeest, Gao et al. (2011) "Discordant antigenic drift of neuraminidase and hemagglutinin in H1N1 and H3N2 influenza viruses" *Proc Natl Acad Sci*
59. Russell, Jones, Barr et al. (2008) "Influenza vaccine strain selection and recent studies on the global migration of seasonal influenza viruses" *Vaccine (Auckl)*
60. De Jong, Beyer, Palache et al. (2000) "Mismatch between the 1997/1998 influenza vaccine and the major epidemic A(H3N2) virus strain as the cause of an inadequate vaccine-induced antibody response to this strain in the elderly" *J Med Virol*
61. Gao, Li, Klenow et al. (2022) "Antigenic comparison of the neuraminidases from recent influenza A vaccine viruses and 2019-2020 circulating strains" *NPJ Vaccines*
62. (2025) *Full-Length Text Journal of Virology*
63. Wang, Dou, Östbye et al. (2019) "Structural restrictions for influenza neuraminidase activity promote adaptation and diversification" *Nat Microbiol*
64. (2003) "Assessment of the effectiveness of the 2003-04 influenza vaccine among children and adults--Colorado" *MMWR Morb Mortal Wkly Rep*
65. (2003) "Preliminary assessment of the effectiveness of the 2003-04 inactivated influenza vaccine--Colorado" *MMWR Morb Mortal Wkly Rep*
66. Kumari, Gulati, Smith et al. (2007) "Receptor binding specificity of recent human H3N2 influenza viruses" *Virol J*
67. Alymova, York, Air et al. (2016) "Glycosylation changes in the globular head of H3N2 influenza hemagglutinin modulate receptor binding without affecting virus virulence" *Sci Rep*
68. Fleming, Elliot (2008) "Lessons from 40 years' surveillance of influenza in England and Wales" *Epidemiol Infect*
69. Reichert, Simonsen, Sharma et al. (2004) "Influenza and the winter increase in mortality in the United States, 1959-1999" *Am J Epidemiol*
70. Ibricevic, Pekosz, Walter et al. (2006) "Influenza virus receptor specificity and cell tropism in mouse and human airway epithelial cells" *J Virol*
71. Shinya, Ebina, Yamada et al. (2006) "Influenza virus receptors in the human airway" *Nature*
72. Van Riel, Munster, De Wit et al. (2007) "Human and avian influenza viruses target different cells in the lower respiratory tract of humans and other mammals" *Am J Pathol*
73. Jayaraman, Chandrasekaran, Viswanathan et al. (2012) "Decoding the distribution of glycan receptors for human-adapted influenza A viruses in ferret respiratory tract" *PLoS One*
74. Vigeveno, Han, De Vries et al. (2024) "Long-term evolution of human seasonal influenza virus A(H3N2) is associated with an increase in polymerase complex activity" *Virus Evol*
75. Laver, Colman, Webster et al. (1984) "Influenza virus neuraminidase with hemagglutinin activity" *Virology (Auckl)*
76. Mögling, Richard, Vliet S Van Der et al. (2017) "Neuraminidasemediated haemagglutination of recent human influenza A(H3N2) viruses is determined by arginine 150 flanking the neuraminidase catalytic site" *J Gen Virol*
77. Liang, Tan, Wang et al. (2022) "Egg-adaptive mutations of human influenza H3N2 virus are contingent on natural evolution" *PLoS Pathog*
78. Liu, Chen, Duan et al. (2022) "Predicting egg passage adaptations to design better vaccines for the H3N2 influenza virus" *Viruses*
79. Wu, Lv, Thompson et al. (2019) "Preventing an antigenically disruptive mutation in egg-based H3N2 seasonal influenza vaccines by mutational incompatibility" *Cell Host Microbe*
80. Zost, Parkhouse, Gumina et al. (2017) "Contemporary H3N2 influenza viruses have a glycosylation site that alters binding of antibodies elicited by egg-adapted vaccine strains" *Proc Natl Acad Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724384&blobtype=pdf | # The viral proteins of influenza A virus competitively bind to TRIM31 with MAVS to fine-tune the antiviral innate immunity
Jiaxin Huang, Shuai Xu, Junwen Liu, Qian Wang, Lu Han, Mengyao Ji, Caoqi Lei, Qiyun Zhu, Hualan Chen
## Abstract
The influenza A virus (IAV) continues to pose a serious threat to animals and humans, making it urgent to reveal more about IAV-host interactions. Tripartite motif protein 31 (TRIM31), an E3 ubiquitin ligase, has been identified as an agonist of the type-I interferon (IFN-I) response against RNA viruses by targeting mitochon drial antiviral signaling protein (MAVS). Here, we demonstrated that TRIM31 plays critical and novel roles in the life cycle of IAV. TRIM31 promoted the IFN-I signaling induced by IAV; however, it was surprisingly found that TRIM31 does not affect IAV replication. Instead, IAV replication was significantly promoted by TRIM31 in MAVS-or interferon receptor-deficient cells, suggesting TRIM31 may facilitate IAV replication in an interferon-independent manner. Mechanistically, TRIM31 interacted specifically with the basic polymerase 1 (PB1), acidic polymerase (PA), and hemagglutinin (HA) proteins of different subtypes of IAV. The interaction between TRIM31 and the PB1, PA, and HA proteins enhances the stability and polymerase and membrane fusion activities of these viral proteins by catalyzing the K63-linked ubiquitination. Further, the PB1, PA, and HA proteins competitively bind to TRIM31 for IAV replication, leading to the attenuation of the TRIM31-MVAS complex-mediated IFN-I signaling activation. Therefore, the antiviral and proviral effects of TRIM31 reach a balance in IAV-infected cells, resulting in no significant impact on IAV replication. Our novel findings revealed an IAV-specific mechanism that IAV exploits TRIM31 to fine-tune the antiviral innate response and maintain the homeostasis of viral replication. IMPORTANCE During the long-term symbiosis with the host, IAVs have evolved a series of unique mechanisms to adapt to the host and support their own replication. The MAVS-mediated IFN-I signaling pathway is crucial for host cells to defend against RNA virus invasion, with TRIM31 functioning as a specific agonist for the activation of IFN-I antiviral response. In the present study, we demonstrated that IAV exploits TRIM31 to promote the stability and activity of viral proteins and reduces the positive effect of TRIM31 on the IFN-I response, thereby preventing TRIM31 from inhibiting IAV replication. Therefore, our results revealed a novel mechanism employed by IAV to adapt to host antiviral response and expanded our understanding of virus-host interactions. KEYWORDS influenza A virus, TRIM31, ubiquitination, viral protein, antiviral innate immunity I nfluenza A virus (IAV) is an enveloped, negative-sense, single-stranded RNA virus that causes respiratory diseases ranging from mild to severe. IAV poses a significant threat to both human and animal health, causing annual seasonal epidemics and sporadic pandemic outbreaks (1-3). According to reports from the World Health Organization (WHO), the influenza epidemics result in 3-5 million cases of severe illness and approximately 290,000-650,000 deaths worldwide each year (4). During the virus life cycle, IAV employs various strategies to facilitate viral replication and spread, including
counteracting host immune responses (5,6), hijacking energy and substance metabolism (7,8), and modulating cell death (9,10). The virus-host interactions serve as a critical foundation for the propagation, transmission, and adaptation of viruses. Consequently, elucidating the molecular mechanisms underlying IAV-host interactions is essential for developing both prophylactic and therapeutic strategies against IAV.
Ubiquitination is a vital post-translational modification (PTM) involved in various mechanisms that maintain intracellular homeostasis. The cellular ubiquitin ligases (E3s), in cooperation with ubiquitin-activating enzymes (E1s) and ubiquitin-conjugating enzymes (E2s), are responsible for the ubiquitination of substrate proteins (11). Ubiquitin chains are organized by the conjugation of different ubiquitin residues through a variety of lineages (12). The well-described proteasomal degradation machinery is activated by K48-linked polyubiquitination (13). K63-linked polyubiquitination has been demon strated to play a crucial role in activating and stabilizing proteins involved in vital cellular processes (14)(15)(16). Numerous studies have demonstrated the pivotal function of ubiquitination in virus-host interactions. For example, E3 ligases, such as tripartite motif 25 (TRIM25), TRIM9, and ring finger protein 5 (RNF5), have been reported to regulate the antiviral response by increasing the ubiquitination of key proteins in the interferon pathway (17)(18)(19). Further, several E3 ligases, including TRIM14, TRIM21, and Bcl2-associated athanogene 6 (BAG6), have been shown to directly target viral proteins, thereby regulating their stability and/or their interactions with other viral proteins (20)(21)(22).
The TRIM protein family is one of the largest subfamilies of E3 ubiquitin ligases, comprising over 80 members (23). As a member of the TRIM-containing proteins, tripartite motif 31 (TRIM31) has been implicated in a variety of pathological processes, including inflammatory diseases, protein quality control, autophagy, viral infection, and carcinoma development (24)(25)(26)(27). Upon RNA virus infection, TRIM31 promotes the K63-linked polyubiquitination of mitochondrial antiviral signaling protein (MAVS), thereby facilitating the formation of MAVS prion-like aggregates and MAVS-mediated type I IFN (IFN-I) signaling, and inhibiting the replication of RNA viruses, such as Sendai virus (SeV) and vesicular stomatitis virus (VSV) (26). Therefore, TRIM31 is a positive regulator of the MAVS-mediated IFN-I response.
In the present study, we revealed that TRIM31 potentiates the IFN-I response induced by IAV. Whereas, TRIM31 is exploited by IAV to enhance the stability of viral PB1, PA, and HA proteins by catalyzing K63-linked ubiquitination. The stabilized PB1, PA, and HA proteins of IAV competitively bind to TRIM31 with MAVS, thereby alleviating the innate immune response. Our data demonstrated that IAV exploits TRIM31 to fine-tune its positive effect on the IFN-I response for the homeostasis of viral replication.
## RESULTS
## TRIM31 potentiates the IFN-I response induced by IAV
Based on the promotion of TRIM31 to the aggregates of MAVS and SeV-or VSV-induced IFN-I responses (26), we hypothesized that TRIM31 may also promote the IAV-induced IFN-I response. As shown in Fig. 1A, overexpression of TRIM31 increased the transcrip tion of the interferon beta 1 (IFNB1), IFN-stimulated gene 15 (ISG15), oligoadenylate (OASL), and C-X-C motif chemokine 10 (CXCL10) genes in IAV-infected cells. Further, overexpressed TRIM31 enhanced IAV-induced activation of the IFN-β promoter in a dose-dependent manner (Fig. 1B). Similarly, TRIM31 overexpression led to an augmenta tion in the secretion of IFN-β following IAV, exhibiting a dose-dependent relationship (Fig. 1C). Conversely, silencing TRIM31 resulted in the downregulation of the transcrip tion of the IFNB1, ISG15, OASL, and CXCL10 genes induced by IAV (Fig. 1D andE). These data suggest that TRIM31 upregulates the IAV-triggered IFN-I signaling pathway.
We further detected the expression of TRIM31 following IAV infection. As shown in Fig. 2A andB, the mRNA and protein of TRIM31 were both upregulated upon IAV infection, which suggested that the transcription of the TRIM31 gene was induced by IAV infection. The online tools, such as hTFtarget (28) and JASPAR (29), were used to screen the transcription factors that bind to the TRIM31 promoter. The signal transducer and activator of transcription 1 (STAT1), as a well-known transcription factor within the IFN signaling pathway, was found to recognize several motifs within the TRIM31 promoter (data not shown), suggesting that TRIM31 may be an interferon-stimulated gene (ISG). To verify this, the mRNA expression of TRIM31 in cells was elevated following IFNα treatment, as well as the expression of TRIM31 protein (Fig. 2C andD). In addition, the IAV infection failed to enhance the expression of TRIM31 in interferon alpha and beta receptor subunit 1 (IFNAR1)-deficient cells (Fig. 2E andF). These data indicate that TRIM31 is a novel ISG that promotes the innate immune response.
## TRIM31 enhances IAV replication in VERO and MAVS-deficient cells
As TRIM31 upregulates the IFN-I response, it can be inferred that TRIM31 restricts IAV replication. Interestingly, TRIM31 overexpression did not impact the replication of H1N1 IAV in A549 cells (Fig. 3A), nor that of H3N2, H6N6, or H9N2 IAVs (Fig. 3B). In cells derived from different species (hamster BHK-21 cells, swine PK-15 cells, and human U2OS cells), IAV replication remained unaffected by TRIM31 overexpression (Fig. 3C). In addition, knockdown of TRIM31 had no effect on IAV replication (Fig. 3D). However, overexpressed TRIM31 significantly inhibited the replication of other RNA viruses, such as the Newcastle disease virus (NDV) and VSV (Fig. 3E), which is consistent with a previous study (26). As TRIM31 potentiates IFN induction by mediating the polyubiquitination and oligomerization of MAVS, we generated MAVS-deficient cells to investigate the impact of TRIM31 on IAV replication. In MAVS-deficient cells, TRIM31 overexpression promoted IAV replication, even with IFNα treatment restricting virus replication (Fig. 3F). In VERO cells (an interferon receptor-deficient cell line), TRIM31 overexpression promoted the replication of H1N1 and H9N2 viruses, whereas TRIM31 knockdown inhibited virus replication (Fig. 3G andH). These results indicate that TRIM31 promotes IAV replication in the absence of MAVS or the interferon pathway, suggesting that TRIM31 may facilitate IAV replication in an IFN-independent manner.
## TRIM31 interacts with and stabilizes PB1, PA, and HA of IAV
Given that TRIM31 promotes the IFN-I response induced by IAV but facilitates the replication of IAV in cells lacking interferon receptors, we hypothesized that TRIM31 might facilitate IAV replication through an alternative mechanism. As TRIM31 is an E3 ligase, we speculated that it may regulate the ubiquitination of viral proteins. An endogenous co-immunoprecipitation (co-IP) assay revealed that TRIM31 is associ ated with the PB1, PA, and HA proteins of the H1N1 virus but not with other viral proteins, such as PB2, NA, and M1 (Fig. 4A). In a confirmatory experiment, TRIM31 interacted with Flag-tagged PB1, PA, and HA (Fig. 4B). Furthermore, TRIM31 interacted with overexpressed PB1, PA, and HA proteins derived from H5N6 and H7N9 IAVs (Fig. 4C). Consistently, the PB1, PA, and HA proteins were found to specifically interact with endogenous TRIM31 in IAV-infected cells (Fig. 4D). In addition, a glutathione S-transfer ase (GST) pulldown assay demonstrated that TRIM31 directly interacts with PB1, PA, and HA in vitro (Fig. 4E). Moreover, TRIM31 co-localized with PB1, PA, and HA, but not NP, in the cytoplasm of virus-infected cells (Fig. 4F). These data demonstrate that TRIM31 specifically and directly associates with the PB1, PA, and HA proteins of IAV.
Studies have shown that TRIM31-mediated ubiquitination regulates the stability of substrate proteins (26,30,31). As shown in Fig. 5A, TRIM31 increased the expression of PB1, PA, and HA in a dose-dependent manner, while exerting no effect on NP expression. Knockdown of TRIM31 downregulated the expression of PB1, PA, and HA (Fig. 5B). Overexpression of TRIM31 upregulated the expression of PB1, PA, and HA in IAV-infected A549 cells, as well as in VERO and IFNAR1-knockout A549 cells (Fig. 5C). Using cyclohexi mide (CHX), an inhibitor of eukaryotic translation elongation, to block protein translation, it was confirmed that TRIM31 prolonged the half-lives of PB1, PA, and HA (Fig. 5D). PB1 and PA are vital subunits of the viral polymerase complex and contribute to the transcrip tion and replication of the viral genome. A minigenome assay showed that TRIM31 overexpression enhanced the viral polymerase activity in a dose-dependent manner, whereas knockdown of TRIM31 reduced the viral polymerase activity (Fig. 5E andF). One of the major functions of HA in the IAV life cycle is to mediate the fusion between the viral and cellular membranes. The membrane fusion assay demonstrated that overex pression of TRIM31 enhanced the HA-mediated cell-to-cell membrane fusion (Fig. 5G). These results indicate that TRIM31 strengthens the stability and activity of PB1, PA, and HA proteins in cells.
## TRIM31 catalyzes the K63-linked ubiquitination of PB1, PA, and HA
Next, we seek to demonstrate how TRIM31 contributes to the stabilization of PB1, PA, and HA. In a ubiquitination assay, TRIM31 overexpression was found to significantly increase the K63-linked ubiquitination of PB1, PA, and HA but not K27-or K48-linked ubiquitination (Fig. 6A). Furthermore, TRIM31 overexpression facilitated the ubiquitina tion of K63O, which contains only one lysine at position 63 of ubiquitin, to PB1, PA, and HA, but not K63R (in which the K63 lysine residue is mutated to arginine) (Fig. 6B). Knockdown of TRIM31 led to a decrease in the K63-linked ubiquitination of PB1, PA, and HA (Fig. 6C). Similarly, TRIM31 overexpression led to an enhancement of K63-linked ubiquitination of PB1, PA, and HA in IAV-infected cells (Fig. 6D). In addition, TRIM31 promoted the K63-linked ubiquitination of PB1, PA, and HA derived from H5N6 and H7N9 IAVs (Fig. S1). Taken together, these data indicate that TRIM31 increases the K63-linked ubiquitination of PB1, PA, and HA.
Prior studies have demonstrated that the RING domain of TRIM31 is indispensable for E3 ligase activity, with the critical cysteine residues at positions 53 and 56 (C53/56) located within this domain. TRIM31-ΔRING, which lacks the RING domain, and the C53/56A mutation, in which the critical cysteine residues were replaced with alanine, both lose the ubiquitin ligase activity of TRIM31 (26,30). To investigate whether TRIM31mediated stabilization of viral proteins and promotion of viral replication depends on E3 ligase activity, we constructed plasmids expressing TRIM31-ΔRING and TRIM31-C53/56A. To resist the interference effect of sh-TRIM31, the sh-TRIM31 off-target nonsense mutants of wild-type TRIM31, TRIM31-ΔRING, and TRIM31-C53/56A were constructed, i.e., TRIM31-OT, TRIM31-ΔRING-OT, and TRIM31-C53/56A-OT (Fig. S2A). As shown in Fig. S2B, the truncation of the RING domain and the C53/56A mutation did not affect the interaction of TRIM31 with PB1, PA, and HA. However, wild-type TRIM31 restored the decrease of PB1, PA, and HA proteins caused by TRIM31 knockdown, whereas TRIM31-ΔRING and TRIM31-C53/56A did not (Fig. 7A). Knockdown of TRIM31 led to a decrease in the K63-linked ubiquitination of PB1, PA, and HA in both HEK293T and MAVS-deficient HEK293T cells. This decrease was restored by wild-type TRIM31 but not by TRIM31-ΔRING or TRIM31-C53/56A (Fig. 7B; Fig. S2C). In VERO cells, overexpression of wild-type TRIM31 led to the promotion of IAV replication, while knockdown of TRIM31 led to the inhibition of IAV replication. TRIM31-ΔRING and TRIM31-C53/56A had no effect on IAV replication (Fig. 7C andD). Consistently, TRIM31-ΔRING and TRIM31-C53/56A failed to restore the inhibition of viral polymerase activity and HA-mediated membrane fusion caused by TRIM31 knockdown (Fig. 7E andF). These results suggest that TRIM31 catalyzes the K63linked ubiquitination of the PB1, PA, and HA proteins via E3 ligase activity and promotes the expression and function of the viral proteins.
## Influenza viral proteins attenuate the interaction between TRIM31 and MAVS
As TRIM31 binds to and enhances the ubiquitination of both MAVS and viral proteins, so we began to illustrate the impact of viral proteins on the TRIM31-MAVS interaction and MAVS-mediated interferon responses. As MAVS is a mitochondria-located protein, TRIM31 can be recruited to mitochondria by MAVS. SeV infection has been demonstrated to increase the co-localization of TRIM31 with mitochondria (26). Consistently, the fraction of TRIM31 on mitochondria was increased following SeV infection. However, in IAV-infected cells, the fraction of TRIM31 on mitochondria was not increased; rather, it was decreased following infection with IAV (Fig. 8A). These data suggest that less TRIM31 is recruited by MAVS in IAV-infected cells than in SeV-infected cells. As a verification, overexpression of PB1, PA, and HA reduced the interaction between TRIM31 and MAVS and the TRIM31-mediated, K63-linked ubiquitination of MAVS (Fig. 8B andC). A luciferase assay revealed that PB1, PA, and HA reduced the TRIM31-mediated upregulation of the IFN-β promoter, as well as the transcription of the IFNB1, ISG15, and regulated upon activation normal T cell expressed and secreted factor (RANTES) genes (Fig. 8D andE).
The above data revealed that IAV proteins competitively interact with TRIM31 from MAVS to weaken innate immunity and facilitate virus replication.
## DISCUSSION
A previous study revealed that TRIM31 catalyzes the K63-linked ubiquitination of MAVS and promotes the MAVS-mediated IFN-I response, thereby inhibiting the replication of model viruses, such as SeV and VSV (26). In the present study, we identified a new role of TRIM31 in the life cycle of IAV (Fig. 9). Upon IAV infection, the PB1, PA, and HA proteins of IAV exploit TRIM31 to increase their stability by catalyzing K63-linked ubiquitination. The stabilized PB1, PA, and HA proteins competitively bind to TRIM31 with MAVS to attenuate the TRIM31-elevated IFN-I signaling. Therefore, IAV proteins exploit TRIM31 to fine-tune the inhibitory effect of TRIM31 on IAV replication. Finally, TRIM31, which is a previously known agonist of the IFN-I response, does not inhibit IAV replication as it does with SeV and VSV infections. E3s constitute the central regulatory component of the ubiquitin-proteasome system and mediate the ubiquitination process through specific substrate recognition, thereby precisely controlling protein homeostasis by regulating stability, subcellular localization, and functional activity. Emerging evidence has demonstrated that certain E3s specifically target viral proteins, thereby modulating viral replication and pathogenicity through ubiquitination-dependent mechanisms. E3 TRIM21 has been shown to impede DNA replication of the hepatitis B virus by promoting the K48-linked ubiquitination and subsequent degradation of the viral DNA polymerase (32). S-phase kinase-associated protein 2 (Skp2) was recently identified as the probable E3 ubiquitin ligase responsible for the degradation of p24 and p55 of human immunodeficiency virus 1 (33). The E3 ligase RNF5 has been shown to restrict severe acute respiratory syndrome coronavirus 2 replication by catalyzing the ubiquitination and mediating the degradation of the E protein (34). In summary, these E3 ligases mostly act as restrictive factors against viral infections by catalyzing the ubiquitination and degradation of viral proteins. However, in the present study, we revealed that TRIM31 specifically catalyzes the K63-linked ubiquiti nation of the PB1, PA, and HA proteins of IAV, thereby strengthening the stability of the viral proteins. TRIM31, as an E3 ligase, has been reported to be involved in multiple cell processes by catalyzing various types of ubiquitin chains on different substrates. For example, TRIM31 has been shown to regulate the K48-linked ubiquitination and degradation of p53, thereby promoting cancer cell resistance to anoikis and facilitating hepatocellular carcinoma progression (35). Conversely, in breast cancer, TRIM31 has been observed to induce the K63-linked ubiquitination of p53 and thus inhibit the proliferation, colony formation, migration, and invasion of breast cancer cells (27). TRIM31 has been shown to facilitate the K27-linked polyubiquitination of the non-receptor tyrosine kinase SYK at K375/571, promote the binding of SYK with C-type lectin receptors, and enhance antifungal immunity (31). Liu et al. demonstrated that TRIM31 serves as a positive regulator of MAVS-mediated innate antiviral immunity through the direct conjugation of K63-linked ubiquitin chains to the K10/311/461 sites on MAVS (26). The present study revealed the dual roles of TRIM31 in antiviral immunity and IAV infection. Although most of the aforementioned studies were conducted under varying pathophysiological conditions, these findings collectively indicate that TRIM31 plays vital but distinct roles in inhibiting or promoting disease progression in different diseases. However, we failed to identify the key sites on the viral proteins that mediate the TRIM31-catalyzed ubiquitina tion. Further studies are necessary to determine the specific sites or motifs by which TRIM31 catalyzes ubiquitination on the various viral proteins.
Viruses need to utilize host systems to assist in their replication, and IAVs are no exception (36). During their long commensalism with hosts, IAVs have evolved a complex set of mechanisms to regulate or hijack host systems for viral replication. However, due to the small segmented genome and limited encoding ability of IAV, IAV proteins, in addition to their unique functions in the virus life cycle, need to work in concert with each other to ensure smooth replication in host cells. In some cases, the formation of complexes by distinct viral proteins is imperative for the execution of specific functions. For example, three polymerase subunits of IAVs collaborate with NP to form the viral ribonucleoprotein complex (vRNP) to accomplish the replication and transcription of the viral genome (36). For the nuclear export of progeny vRNPs, the viral M1 protein binds directly to the vRNPs via the NEP protein to form the viral nuclear export complex (37). In other cases, multiple IAV proteins need to function separately to achieve the same effect. To effectively replicate in host cells, different IAV proteins target various host proteins in the innate immune signaling pathway to antagonize the host antiviral response. The non-structural protein 1 of IAV has been shown to suppress IFN production by targeting multiple host factors, including TRIM25, interferon regulatory factor 3 (IRF3), and 5′-3′ exoribonuclease 1 (38)(39)(40). Our previous study revealed that both the NP and PB1 protein of IAV mediate the degradation of MAVS and block the MAVS-mediated innate signaling pathway (5,41). In the current study, we uncovered a novel pattern in which three different IAV proteins bind to the same target, TRIM31, in a synergistic manner, leading to the reduction of the MAVS-mediated IFN-I response elevated by TRIM31.
In summary, this study highlights the dual roles of TRIM31 in the life cycle of IAV. Upon IAV infection, TRIM31 catalyzes the K63-linked ubiquitination of MAVS and promotes the IFN-I response against virus infection. To counteract the antiviral effect of TRIM31, IAV exploits TRIM31 to catalyze the K63-linked ubiquitination of PB1, PA, and HA proteins, thereby enhancing the stability and functions of viral proteins. In addition, the stabilized viral proteins competitively interact with TRIM31 to inhibit the TRIM31-MAVS association and the MAVS-mediated IFN-I response. Taken together, IAV exploits TRIM31 to fine-tune its antiviral effect on the IFN-I response, thereby preventing TRIM31 from inhibiting IAV replication. Overall, our findings extend the knowledge of IAV-host interactions and reveal a key role of TRIM31 in the IAV life cycle.
## MATERIALS AND METHODS
## Cells, viruses, and plasmids
MDCK, HEK293T, BHK-21, PK-15, and U2OS were grown in DMEM (Gibco) supplemented with 10% (vol/vol) FBS (Gibco-BRL; 10099-141) and 1× penicillin/streptomycin (Gibco-BRL; 10378016). A549 cells were grown in Kaighn's modified Ham F-12 nutrient mixture medium (Gibco) supplemented with 10% FBS and penicillin/streptomycin. All cells were cultured and maintained at 37°C with 5% CO 2 . The H1N1 (A/Puerto Rico/8/1934, PR8), H1N1 (A/WSN/1933, WSN), and H6N6 (A/ duck/Hunan/4/2018, HN4) IAVs were stored in our laboratory. The H3N2 influenza virus (A/Swine/Shandong/TA05/2021, TA5) was kindly provided by Prof. Yihong Xiao (The Shandong Agricultural University, China). The H9N2 virus (A/chicken/Hunan/38/2018, HN38) was isolated from Gansu Province in China in 2018. Recombinant vesicular stomatitis virus expressing green fluorescent protein (VSV-GFP) was generated as described previously (42). The NDV (MG7 strain) was generated and stored in our laboratory (43). SeV (strain Fushimi) was kindly provided by Prof. Hongkui Deng (Peking University, China).
Human TRIM31, as well as TRIM31 truncations and C53/56A mutation, was construc ted into pRK vector and pET-28a prokaryotic expression vector by using standard molecular biology techniques. The sh-TRIM31 off-target nonsense mutants of wild-type TRIM31, TRIM31-ΔRING, and TRIM31-C53/56A were constructed as shown in Fig. S2A. The genes encoding PB2, PB1, PA, and NP of PR8 or other viruses were amplified and then cloned into the pCAGGS vector. Flag-tagged PB1, PA, and HA from PR8-H1N1, SC15-H5N6, and SZ19-H7N9 were constructed and stored in our laboratory. The plasmid pPol I-Luc, for the expression of a viral RNA-like firefly luciferase gene under the control of the human RNA polymerase I promoter, has been reported previously (44). The plasmids of the IFN-β-Luc and pRL-TK internal control luciferase reporter plasmids used in the study were described previously (6).
For the NDV and VSV-GFP, virus titers of virus stocks and cell culture supernatant were determined by end-point titration in MDCK cells. For end-point viral titration in MDCK cells, 10-fold serial dilutions of each sample were inoculated into MDCK cells. Infectious virus titers are reported as log 10 TCID 50 /0.1 mL and were calculated from three replicates by the method of Reed-Muench (47).
## RNA interference
The RNA interference was performed as previously described (48). SiRNA targeting TRIM31 and scrambled siRNA (si-NC) were purchased from RiboBio Co. (China); the sequences of the siRNAs are listed in Table S1. All transfections with siRNA were performed as per manufacturer's instructions using Lipofectamine RNAiMAX reagent.
Short hairpin RNA (shRNA) constructs were designed and cloned into the pLKO.1-EGFP-puro lentiviral backbone as per the manufacturer's protocol (Tsingke Biotechnol ogy, China). The target sequences for TRIM31 were listed in Table S1. The sequence of nonsense shRNA was provided by Tsingke. Lentivirus was produced in HEK293T cells transfected with viral constructs along with psPAX2 and pMD2G constructs. Viral supernatants were collected on days 2 and 3 after transfection and used to infect target cells.
## RNA isolation and qPCR
Total RNA from cells was extracted with TRIzol as previously described (48). For mRNAs, total RNA was subsequently transcribed into cDNA using M-MLV Reverse Transcriptase, according to the manufacturer's protocol (Promega). GAPDH and β-actin were used as control for the normalization of cellular mRNA. Real-time PCR was carried out using the LightCycler 480 II (Roche). The RNA level of each gene was shown as the fold induction (2 -ΔΔCT ) in the graph. The sequences of the primers used for qPCR are shown in Table S1.
## Dual-luciferase reporter assays
For the viral minigenome assay, HEK293T cells were transfected with pCAGGS constructs expressing viral PB2, PB1, PA, and NP from PR8 virus, the construct pPol I-Luc, and an internal control pRL-TK (Promega), along with other plasmids or siRNA. Cells were incubated at 37°C for 24 h, and cell lysates were subsequently prepared by using the Dual-Luciferase Reporter Assay System (Promega). To detect activation of the IFN-I pathway, HEK293T cells were transfected with luciferase reporter plasmids (IFN-β-Luc) and the pRL-TK plasmid. At 24 h after transfection, the cells were left uninfected or infected with different viruses for 12 h, and cell lysates were subsequently prepared by using the Dual-Luciferase Reporter Assay System (Promega). The luciferase activities were measured on a GloMax 96 microplate luminometer (Promega), as reported previously (6).
## Western blotting and co-IP assay
The Western blotting and co-IP assay were conducted, as previously described (44). For Western blotting, cells were lysed in RIPA buffer (Beyotime, China). Proteins were separated by 10% SDS-PAGE and transferred to a nitrocellulose membrane (Bio-Rad). The membrane was blocked for 1 h in TBST containing 5% milk and subsequently incubated with primary antibodies for 2 h. After 1-h incubation with HRP-conjugated secondary antibody. The immunoreactive bands were visualized using an e-BLOT system (e-BLOT Life Science, China). The intensities of the target bands were quantified by using the Image J program (NIH, USA).
HEK293T cells were co-transfected with the indicated plasmids with or without virus infection for 24 h. The transfected cells were then harvested and lysed in NP-40 lysis buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 1% NP-40, 1 mM EDTA with protease inhibitor cocktails). For each immunoprecipitation, 1 mL of lysate was incubated for 4 h at 4°C with 0.5 µg of the indicated antibody or control IgG and 30 µL of protein A/G-Sepharose (Sigma). The beads were washed three times with 1 mL of lysis buffer containing 500 mM NaCl. The precipitates were analyzed by using standard immunoblot ting procedures.
## GST pull-down assay
GST pull-down assays were conducted as previously described with slight modifications (49). Briefly, the GST-tagged fusion proteins and the control GST proteins were expressed in BL21 cells after induction with 0.1 mmol/L IPTG overnight at 18°C. Centrifuged cells were resuspended in lysis buffer (1× PBS, 0.2 mM PMSF, 1% Triton X-100) and sonicated for 15 min. After centrifugation, the supernatant was applied to a Glutathione-Sephar ose 4B bead column (GE Healthcare), in accordance with the manufacturers' instructions. Purified GST-tagged fusion proteins were diluted with 1× PBS and filtered through Amicon Ultra 0.5 mL filters (Millipore). Then, 1 µg of purified GST protein or GST fusion protein was captured by the Glutathione-Sepharose 4B beads (GE Healthcare). The beads were then washed three times with ice-cold PBS. The supernatant was loaded onto gels, followed by immunoblotting analysis.
## Confocal microscopy
Confocal microscopy was performed as previously described (49). Cells were seeded in 12-well plates (5 × 10 5 cells/well) on coverslips. After transfection or infection at indicated times, cells were then fixed with 4% paraformaldehyde for 20 min at room temperature and washed three times with PBS. Cells were permeabilized with 0.1% Triton X-100 in PBS for 10 min and blocked with 5% skimmed milk for 1 h. Cells were incubated with the indicated primary and secondary antibodies or not. Then, the cells were stained by DAPI before observation by ZEISS microscope (LSM 980) with a 100× oil objective. The colocalization analysis was conducted with ImageJ software. The data were further analyzed by using GraphPad Prism 8.
## Detection of ubiquitin-modified proteins
The experiments were performed as previously described (50). Briefly, the cells were lysed in lysis buffer containing 1% SDS and denatured by heating at 95°C for 10 min. After centrifugation, the supernatants were diluted with NP-40 lysis buffer until the concentration of SDS was 0.1%, and were then co-immunoprecipitated with the indicated antibodies. Ubiquitin-modified proteins were detected by immunoblotting with the indicated antibodies.
## Membrane fusion assay
The experiments were performed, as previously described (41). A549 cells were transfected with Flag-TRIM31 or EV for 24 h and then infected with PR8/H1N1 for 12 h. The cells were treated with trypsin in 2 µg/ mL for 15 min at 37°C to cleave HA into HA1 and HA2. After being washed, the cells were treated with citric acid (pH 5.0) for 15 min at 37°C. The acidic medium was replaced with DMEM supplemented with 10% FBS, and the cells were cultured for 3 h, followed by 4% polyoxymethylene fixation. Syncytium formation was observed by immunofluorescence with indicated antibodies and analyzed by counting the nuclei in syncytia in five random microscopic fields as previous study (51).
## Statistical analysis
Data are expressed as the mean ± standard deviation. Statistical significance was determined by using the Student's two-tailed unpaired t-test or analysis of variance with GraphPad Prism software (version 8.0, San Diego, CA, USA). Differences between groups were considered significant when the P-value was < 0.05 (*), <0.01 (**), <0.001 (***), and <0.0001 (****); ns indicates no significant difference.
## References
1. Muñoz-Moreno, Martínez-Romero, García-Sastre (2021) "Induction and evasion of type-I interferon responses during influenza A virus infection" *Cold Spring Harb Perspect Med*
2. Horimoto, Kawaoka (2005) "Influenza: lessons from past pandemics, warnings from current incidents" *Nat Rev Microbiol*
3. (2025) *Full-Length Text Journal of Virology*
4. Zhang, Liu, Huang et al. (2022) "H1N1 influenza A virus protein NS2 inhibits innate immune response by targeting IRF7" *Viruses*
5. Jiang, Chen, Li (2023) "Advances in deciphering the interactions between viral proteins of influenza A virus and host cellular proteins" *Cell Insight*
6. Zeng, Xu, Wei et al. (2021) "The PB1 protein of influenza A virus inhibits the innate immune response by targeting MAVS for NBR1-mediated selective autophagic degradation" *PLoS Pathog*
7. Xu, Han, Wei et al. (2022) "MicroRNA-200c-targeted contactin 1 facilitates the replication of influenza A virus by accelerating the degradation of MAVS" *PLoS Pathog*
8. Chen, Wang, Gan et al. (2023) "Comprehensive analysis of lipid metabolism in influenza virus infection" *Microb Pathog*
9. Zhou, Pu, Wu (2021) "The role of lipid metabolism in influenza A virus infection" *Pathogens*
10. Zhang, Yin, Boyd et al. (2020) "Influenza virus Z-RNAs induce ZBP1-mediated necroptosis" *Cell*
11. Sun, Ji, Liu et al. (2024) "Influenza virus infection activates TAK1 to suppress RIPK3-independent apoptosis and RIPK1-dependent necroptosis" *Cell Commun Signal*
12. Heaton, Borg, Dixit (2016) "Ubiquitin in the activation and attenuation of innate antiviral immunity" *J Exp Med*
13. Behrends, Harper (2011) "Constructing and decoding unconven tional ubiquitin chains" *Nat Struct Mol Biol*
14. (1038)
15. Zheng, Shabek (2017) "Ubiquitin ligases: structure, function, and regulation" *Annu Rev Biochem*
16. Park, Dezhbord, Lee et al. (2022) "The roles of ubiquitination in pathogenesis of influenza virus infection" *Int J Mol Sci*
17. Shmuel-Galia, Humphries, Lei et al. (2021) "Dysbiosis exacerbates colitis by promoting ubiquitination and accumulation of the innate immune adaptor STING in myeloid cells" *Immunity*
18. Li, Chen, Xie et al. (2023) "FBXO7 confers mesenchymal properties and chemoresistance in glioblastoma by controlling Rbfox2-mediated alternative splicing" *Adv Sci (Weinh)*
19. Gack, Albrecht, Urano et al. (2009) "Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I" *Cell Host Microbe*
20. Qin, Liu, Tian et al. (2016) "TRIM9 short isoform preferentially promotes DNA and RNA virus-induced production of type I interferon by recruiting GSK3β to TBK1" *Cell Res*
21. Zhong, Zhang, Lei et al. (2009) "The ubiquitin ligase RNF5 regulates antiviral responses by mediating degradation of the adaptor protein MITA" *Immunity*
22. Wu, Wang, Wang et al. (2019) "Inhibition of influenza A virus replication by TRIM14 via its multifaceted protein-protein interaction with NP" *Front Microbiol*
23. Song, Li, Wang et al. (2021) "E3 ubiquitin ligase TRIM21 restricts hepatitis B virus replication by targeting HBx for proteasomal degradation" *Antiviral Res*
24. Zhou, Li, Zhang et al. (2024) "BAG6 inhibits influenza A virus replication by inducing viral polymerase subunit PB2 degrada tion and perturbing RdRp complex assembly" *PLoS Pathog*
25. Jaworska, Wlodarczyk, Mackiewicz et al. (2020) "The role of TRIM family proteins in the regulation of cancer stem cell selfrenewal" *Stem Cells*
26. Song, Liu, Huai et al. (2016) "The E3 ubiquitin ligase TRIM31 attenuates NLRP3 inflammasome activation by promoting proteasomal degradation of NLRP3" *Nat Commun*
27. Ra, Lee, Kim et al. (2016) "TRIM31 promotes Atg5/Atg7independent autophagy in intestinal cells"
28. Liu, Zhang, Chu et al. (2017) "The ubiquitin E3 ligase TRIM31 promotes aggregation and activation of the signaling adaptor MAVS through Lys63-linked polyubiquitination" *Nat Immunol*
29. Guo, Li, Zhao et al. (2021) "Loss of TRIM31 promotes breast cancer progression through regulating K48-and K63-linked ubiquitination of p53" *Cell Death Dis*
30. Zhang, Liu, Zhang et al. (2020) "hTFtarget: a comprehensive database for regulations of human transcription factors and their targets" *Genomics Proteomics Bioinfor matics*
31. Rauluseviciute, Riudavets-Puig, Blanc-Mathieu et al. (2024) "JASPAR 2024: 20th anniversary of the open-access database of transcription factor binding profiles" *Nucleic Acids Res*
32. Zhang, Cao, Wang et al. (2022) "The E3 ubiquitin ligase TRIM31 plays a critical role in hyperten sive nephropathy by promoting proteasomal degradation of MAP3K7 in the TGF-β1 signaling pathway" *Cell Death Differ*
33. Wang, Zhang, Shao et al. (2021) "TRIM31 facilitates K27-linked polyubiquitination of SYK to regulate antifungal immunity" *Sig Transduct Target Ther*
34. Mu, Zhao, Zhu et al. (2020) "The E3 ubiquitin ligase TRIM21 promotes HBV DNA polymerase degradation" *Viruses*
35. He, Pang, Li et al. (2024) "IFI27 inhibits HIV-1 replication by degrading Gag protein through the ubiquitin-proteasome pathway" *J Virol*
36. Li, Hao, Zhao et al. (2023) "The E3 ligase RNF5 restricts SARS-CoV-2 replication by targeting its envelope protein for degradation" *Sig Transduct Target Ther*
37. Guo, Qiu, Ma et al. (2018) "Tripartite motif 31 promotes resistance to anoikis of hepatocarcinoma cells through regulation of p53-AMPK axis" *Exp Cell Res*
38. Staller, Barclay (2021) "Host cell factors that interact with influenza virus ribonucleoproteins. Cold Spring Harb Perspect Med 11:a038307"
39. Chase, Rameix-Welti, Zvirbliene et al. (2011) "Influenza virus ribonucleoprotein complexes gain preferential access to cellular export machinery through chromatin targeting" *PLoS Pathog*
40. Liu, Mok, Wang et al. (2021) "Cellular 5'-3' mRNA exoribonuclease XRN1 inhibits interferon beta activation and facilitates influenza A virus replication" *mBio*
41. (2025) *Full-Length Text Journal of Virology*
42. Kuo, Li, Lin et al. (2016) "Role of N terminus-truncated NS1 proteins of influenza a virus in inhibiting IRF3 activation" *J Virol*
43. Wang, Zeng, Xu et al. (2018) "A naturally occurring deletion in the effector domain of H5N1 swine influenza virus nonstructural protein 1 regulates viral fitness and host innate immunity" *J Virol*
44. Zhang, Xu, Liu et al. (2023) "The nucleoprotein of influenza A virus inhibits the innate immune response by inducing mitophagy" *Autophagy*
45. Ge, Deng, Wen et al. (2007) "Newcastle disease virus-based live attenuated vaccine completely protects chickens and mice from lethal challenge of homologous and heterologous H5N1 avian influenza viruses" *J Virol*
46. Ji, Liu, Du et al. (2018) "A novel genotype VII Newcastle disease virus vaccine candidate generated by mutation in the L and F genes confers improved protection in chickens" *Vet Microbiol*
47. Wang, Xu, Shen et al. (2025) "N 6 -methyladnosine of vRNA facilitates influenza A virus replication by promoting the interaction of vRNA with polymerase proteins" *Proc Natl Acad Sci*
48. Chen, Huang, Wang et al. (2025) "Influenza A Virus H7 nanobody recognizes a conserved immunodominant epitope on hemagglutinin head and confers heterosubtypic protection" *Nat Commun*
49. Huang, Liu, Xu et al. (2005) "SIKE is an IKKε/ TBK1-associated suppressor of TLR3-and virus-triggered IRF-3 activation pathways" *EMBO J*
50. Matumotom (1949) "A note on some points of calculation method of LD50 by Reed and Muench" *Jpn J Exp Med*
51. Yu, Wang, Wei et al. (2025) "Nucleophosmin 1 inhibits the replication of influenza A virus by competitively binding viral RNA with viral proteins" *Virol Sin*
52. Wang, Lei, Ji et al. (2016) "Duck tembusu virus nonstructural protein 1 antagonizes IFN-β signaling pathways by targeting VISA" *J Immunol*
53. Jin, Tian, Luo et al. (2017) "Tetherin suppresses type I interferon signaling by targeting MAVS for NDP52mediated selective autophagic degradation in human cells" *Mol Cell*
54. Zhao, Yang, Yang et al. (2021) "Cathepsin L plays a key role in SARS-CoV-2 infection in humans and humanized mice and is a promising target for new drug development" *Signal Transduct Target Ther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772302&blobtype=pdf | # Isolation and identification of a genotype F bovine enterovirus in western China
Kun Xu, Xiaohan Wang, Jie Yuan Guo, Yanpei Ku, Jiang Wang, Beibei Chu, Jiajia Pan, Guoyu Yang, Peer Reviewer, Eduardo Rodríguez-Román
## Abstract
This study successfully isolated a novel bovine enterovirus strain from a bovine fecal sample, which was designated as Sichuan/SQ/20. The isolate showed typical enterovirus morphology under electron microscopy. Phylogenetic analysis showed that this strain exhibits the closest genetic relationship with the HeN-YR91 and JPN/ TottoriU-31 strains, and all three belong to the BEV-F1 genosubtype. Subsequently, comprehensive investigations were conducted on the biological characteristics of this virus, both in vivo and in vitro. In vitro characterization revealed that viral replication commenced at 3 h post-infection (hpi) in Madin-Darby bovine kidney cells, reaching peak at 48 hpi with a virus titer of 1 × 10 8.73 TCID 50 /0.1 mL. Cytopathic effects ini tially appeared at 12 hpi. A 12-minute treatment at 55°C was sufficient to completely inactivate the virus. In vivo analysis revealed that significant pathological changes were specifically observed in the spleen, with no lesions observed in other organs. Immuno fluorescence assay detected specific fluorescent signals in the liver, spleen, and small intestine, which were consistent with the PCR results. These findings provide a scientific foundation for vaccine design and antiviral drug screening, as well as for the develop ment of effective prevention and control strategies. IMPORTANCE Bovine enterovirus (BEV) is an important pathogen causing calf diarrhea and has been detected in the feces of calves with diarrhea, although its pathogenicity remains unclear. This study systematically established an isolation and identification protocol for BEV, characterized its physicochemical properties, and further investiga ted the pathogenicity and tissue tropism of the isolated strain in mice. These find ings establish crucial baseline data for future vaccine development and therapeutic intervention strategies.
KEYWORDS bovine enterovirus, virus isolate, biological characterization, pathogenicityB ovine diarrhea syndrome remains a predominant cause of neonatal calf mor tality and morbidity, with pathogenesis linked to triadic interactions of patho gens, environment, and host susceptibility. Although calf diarrhea has multi-factorial origins, viral etiologies are garnering escalating research attention due to their epi demic potential and evolutionary dynamics (1). Bovine enteric viruses constitute major diarrheal pathogens whose clinical significance is heightened through frequent coinfections that potentiate disease severity and complicate outbreak control (2, 3). Member viruses of this genus exhibit conserved biological properties and can cause respiratory and digestive tract diseases in animal hosts (4-8). Clinical manifestations include diarrhea, pyrexia, respiratory distress, and abortion in cattle, with severe cases progressing to fatal outcomes due to dehydrating enteritis and bronchopneumonia (4, 9, 10).Bovine enterovirus (BEV) exhibits global distribution in cattle populations. BEV was detected at a rate of 14.5% in South American cattle herds surveyed between 2012
and 2016 (11). Epidemiological surveys conducted from 2021 to 2023 revealed a BEV prevalence range of 6.1%-10.7% in domestic cattle herds (3,12). Serological surveillance of BEV across Turkish regions revealed a 64.8% antibody prevalence in cattle herds, indicating a high level of virus circulation (13). BEV was detected in 20.1% and 28% of fecal samples collected from cattle in Tunisia and Romania, respectively (14). BEV is primarily transmitted among cattle herds via the fecal-oral route. This transmission mode not only enhances viral dissemination efficiency but also complicates epidemic control measures, ultimately inflicting substantial economic losses on the cattle industry (5). The environmental persistence of viral particles in contaminated feed and water sources perpetuates transmission cycles, necessitating enhanced biosecurity measures and continuous monitoring.
BEV is a member of the enterovirus species in the Enterovirus genus within the Picornaviridae family. BEV is a small, non-enveloped, icosahedral particle containing a single-stranded, positive-sense viral RNA genome of approximately 7.5 kb in length (15). Currently, the Enterovirus genus consists of 15 species. Among these, the enteric viruses infecting cattle are Enterovirus E and Enterovirus F. EV-E has five subtypes (E1-E5), and EV-F has eight subtypes (F1-F8). The single-stranded RNA genomes of BEV contain a single open reading frame (ORF) that encodes a polyprotein comprising four structural proteins (VP1-VP4) and seven non-structural proteins (2A-2C and 3A-3D). VP1, along with VP2 and VP3, is exposed on the surface of the capsid. VP1 not only serves as the receptor-binding protein but also as the main neutralizing antigen. Therefore, the VP1 gene exhibits the highest sequence variability compared with other parts of the genome, which determines the classification of the enterovirus. All the non-structural proteins, including some of their precursor proteins, may be involved in the replication of the viral RNA.
The virus was first isolated by American scientist Moll in 1955 (16). In 2011, Yingli Li' group identified and isolated the bovine enteric virus in China, classifying the strain as enterovirus species F (17). Subsequently, BEV strains of different subtypes were isolated from various cell lines using multiple methods, offering diverse research approaches for the isolation and infection (5,7,18). In addition, BEV has been detected in sheep (19), camels (20), and other animals, with seropositivity confirming exposure and suggesting potential cross-species transmission. An increasing number of BEVs have been isolated, although most infections are subclinical and exhibit limited pathoge nicity (21). Despite its characteristic high morbidity-low mortality profile, BEV demon strates pathogenic potential through associations with respiratory, gastrointestinal, and occasional neurological manifestations (4,22). Besides, as a member of the Picornaviri dae family, BEV exhibits high mutation rates due to inherent characteristics of its genetic material and replication mechanisms. These frequent mutations potentially compromise vaccine efficacy, thereby posing a substantial threat to the cattle industry.
In this study, we collected 108 batches of cattle fecal samples from Sichuan, Qinghai, and Inner Mongolia from June to October 2024. Using conventional PCR identifica tion and plaque purification methods, we isolated a bovine enteric virus F1 subtype strain, designated as Sichuan/SQ/20 (GenBank: PV290163.1). This study provides a comprehensive characterization of the Sichuan/SQ/20 strain. We investigated the growth kinetics, physicochemical properties, and genetic recombination patterns of this viral isolate. Concurrently, a mouse infection model was established to analyze BEV-induced pathological alterations in host organs and viral tissue distribution, thereby laying the groundwork for in-depth research on this pathogen. These findings establish crucial baseline data for future vaccine development and therapeutic intervention strategies.
## MATERIALS AND METHODS
## Sample collection and RT-PCR analysis
In this study, a total of 108 fecal and anal swab samples were collected from calves with diarrhea on large-scale cattle farms in three provinces, including Sichuan province (56), Qinghai province (21), and Inner Mongolia (31), from June to October 2024. All the samples were frozen in maintenance medium in virus-sampling tubes, transported to the laboratory within 2 days using cold chain transportation, and stored at -80℃.
A 20% fecal suspension was prepared and clarified in a sterile phosphatebuffered saline solution (PBS, pH = 7.2) by centrifugation of at 12,000 × g for 20 min, followed by filtration through a 0.22 µm filter. Total RNA from fecal samples was extracted accord ing to the manufacturer's instructions (the HiPure Viral RNA/DNA Midi Kit, Guangzhou Magen Biotechnology Co., Ltd.). The presence of the virus was confirmed by RT-PCR with the primers BEV-5′UTR-F and BEV-5′UTR-R (Table 1). The reaction system was performed using the Vazyme ClonExpress II One Step Cloning Kit (C112-01; Vazyme Biotech Co., Ltd.), following the manufacturer's instructions. The PCR was carried out under the following conditions: 50°C for 30 minutes, 95°C for 3 minutes, followed by 35 cycles of 95°C for 30 seconds, 55°C for 30 seconds, 72°C for 1 minute, and a final extension at 72°C for 10 minutes.
PCR products were subjected to electrophoresis with 1.0% agarose gel, visualized with a gel documentation system. The PCR product with 600 bp was confirmed by Sanger sequencing. The sequences were compared to existing database entries using the Basic Local Alignment Search Tool (BLAST).
## Cell culture
Madin-Darby bovine kidney cell (MDBK), Baby hamster kidney cell (BHK-21), and African green monkey kidney (Vero) cell were used to isolate the virus. These cells were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics at 37°C with 5% CO 2 . The three types of cells were cultured in a six-well tissue culture plate, and virus isolation was carried out once the confluence reached 90%.
## Virus isolation
Five BEV-positive samples, each from a different region, were selected. Their super natants were then reprepared with antibiotics (100 U/mL penicillin and 100 µg/mL streptomycin), filtered through a 0.22 µm filter. Exactly 0.1 mL of the filtrate was diluted (1:5) and then inoculated with MDBK, BHK-21, and Vero cells, respectively. The mono layers of these cells, prepared in advance, were washed with PBS. The inoculum was discarded after incubation with the three cells for 2 h. The cells were washed with PBS before adding DMEM supplemented with 2% FBS. DMEM was added to the first well of the plate as a negative control. The culture was incubated at 37°C with 5% CO 2 , and the cytopathic effect (CPE) on the cells was assessed daily. After five consecutive passages, viruses were harvested by three freeze-thaw cycles and were further confirmed by RT-PCR.
## Virus purification and identification
The isolated virus Sichuan/SQ/20 (PV290163) was purified by plaque purification. The MDBK cells were cultured in a six-well plate to form a monolayer with uniform distribu tion and a fusion rate of over 95%, washed with PBS, and then inoculated with the virus suspension that can cause cytopathic effects diluted to 10 -3 -10 -7 by DMEM without FBS.
The plate was incubated at 37°C for 3 h. Afterward, the virus suspension was discarded. The cells were washed three times with PBS and recovered using a mixture of sterilized low-melting point agarose (4%) and 2× DMEM at a ratio of 1:1. Then, the cell culture plate was inverted and incubated at 37°C in a 5% CO 2 . CPE were observed after 48 h and stained with neutral red. Single plaques were selected and placed in serum-free DMEM medium, then subjected to three freeze-thaw cycles. The mixture was then inoculated in the MDBK monolayer cultures in six-well tissue-culture plates. Finally, purified cultured virus samples were obtained after three rounds of plaque purification. To identify the purified strain, the cells were stained with 2 mL of a staining solution consisting of 0.5% crystal violet and 25% formaldehyde solution following the above procedure.
## Electron microscopy observation
MDBK cells were infected with the isolated Sichuan/SQ/20 strain and cultured. The virus supernatant was transferred into a dialysis bag (MD34), embedded in polyethylene glycol 8000, and concentrated at a ratio of 1:50. The concentrated and purified virus was obtained. The sample was examined by Transmission Electron Microscopy after being negatively stained with 2% phosphotungstic acid.
## Replication kinetics analysis of the isolates Sichuan/SQ/20
To determine the infectivity of the purified strain, titration of TCID 50 for Sichuan/SQ/20 isolates was performed using 96-well plates. The purified strain was serially diluted from 10 -1 to 10 -10 and used to infect the 96-well plates for each dilution at 37°C for 3 h after being washed with PBS. Forty-eight hours post-inoculation, record the cytopathic effects for each dilution and calculate the viral titer (TCID 50 ) using the Reed-Muench method. To detect virus growth kinetics, infected cells were harvested at 0, 3, 6, 9, 20, 24, 30, 48, and 60 h post-infection (hpi). The viral titer (TCID 50 ) was determined and calculated using the Reed-Muench method. To further examine temperature sensitivity, Sichuan/SQ/20 was heated at 4°C, 37°C, 42°C, 50°C, 54°C, 55°C, 56°C, and 60°C for 1 h. The viral titer (TCID 50 ) was determined and calculated using the Reed-Muench method. To establish the minimum inactivation time at 55°C, the virus was diluted 1:10 in DMEM and distributed into five tubes (1 mL/tube). Following vortex mixing, tubes were heated at 55°C for durations ranging from 3 to 15 minutes. Viral titers (TCID 50 ) were then assessed for all samples.
## Immunofluorescence assay
To characterize the antigenic properties of the isolated strain, the reactivity between the Sichuan/SQ/20 strain and mouse positive serum was evaluated using an immunofluores cence assay (IFA). MDBK cell monolayers were infected with the Sichuan/SQ/20 strain at a multiplicity of infection of 0.1. At 24 hpi, the cells were washed three times by PBS, fixed with 4% paraformaldehyde for 30 min, and incubated with 0.1% Triton X-100 for 10 min. After three PBS washes, cells were blocked with 5% bovine serum albumin for 1 h, incubated with mouse positive serum for 1 h at room temperature, and uninfected cells served as the negative control. After three washes with PBS, the cells were probed with AF488-labeled goat anti-mouse IgG (H + L; 1:500; Beyotime, Cat# A0428) for 1 h at room temperature. After washing three times with PBS again, the cells were incubated with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Solarbio, Cat#C0065) for 10 min. Fluorescence images were tested by a microscope (Olympus, IX73, Tokyo, Japan).
## Virus genome amplification
Total RNA from the purified strain was extracted with the HiPure Viral RNA/DNA Midi Kit (Guangzhou Magen Biotechnology Co., Ltd.) following the manufacturer's instructions. The nearly complete genome sequence gene of the Sichuan/SQ/20 strain was obtained by RT-PCR with four pairs of primers (Table 1) using the HiScript II One Step RT-PCR Kit (Dye Plus) (Vazyme Biotech Co., Ltd.) following the manufacturer's instructions. The PCR products of the expected size, according to each set of primers, were purified using a Gel Extraction kit (TaKaRa, Dalian, China) after electrophoresis. The purified DNA was cloned into pMD18-T vector (TaKaRa, China), and the resulting plasmid was used to transform competent E. coli cells. Positive inserts were confirmed by PCR and further sequenced by Sangon Biotechnology Company (Shanghai, China). To prevent contamination, the preparation of the PCR mix and the addition of the template DNA were performed in separate rooms using dedicated pipets and filtered tips. Sequence assembly and manual editing were performed using the SeqMan program (DNASTAR, Madison, WI). The nucleotide sequence identities were calculated by the Megalign program available within the Lasergene software package (version 7.1, DNAstar).
## Phylogenetic analysis
The phylogenetic analysis of the Sichuan/SQ/20 strain was performed based on the nucleotide sequences of P1 and VP1 using the maximum likelihood (ML) method. ML trees were constructed using MEGA version 7.0, employing the general time-reversi ble (GTR) nucleotide substitution model with optimized parameters for the gamma (Γ)-distribution and the proportion of invariable sites (i.e., GTR + Γ + I). Bootstrap support values were calculated from 1,000 replicates performed in MEGA version 7.0 (23).
## Mice
All animal experiments were conducted in strict compliance with the Guide for the Care and Use of Laboratory Animals and the ethical guidelines of Henan Agricultural University. Six-week-old female BALB/c mice were purchased from Liaoning Changsheng Biotechnology Co., Ltd. The pathogenicity of the isolated Sichuan/SQ/20 strain was evaluated by inoculating BALB/c mice. The mice were randomly assigned to two groups (groups A and B; four mice per group) and housed in separate cage litter. Following a 3-day adaptation period, Group A mice (n = 4) were administered 0.5 mL of the viral isolate (10 7.66 TCID 50 /0.1 mL) by intraperitoneal injection, and Group B mice (n = 4) served as negative controls and received an equivalent volume (0.2 mL) of sterile DMEM cell culture supernatant by intraperitoneal injection.
Clinical symptoms were monitored daily. During necropsy, spleen, liver, and small intestine tissues were collected and processed for dual purposes: one portion was snap-frozen at -80°C for viral RNA extraction and PCR detection, while the other portion was fixed in 4% paraformaldehyde for 24 h before undergoing paraffin embedding, sectioning, and staining for histological examination and IFA.
## RESULTS
## Detection and isolation of BEV in bovine stool samples
A total of 108 fecal and anal swab samples from three different provinces were detected by RT-PCR using the primer pair BEV-5′UTR_F and BEV-5′UTR_R (Table 1), which was designed based on the alignment results of the 5′UTR region of all BEVs available in the GenBank database. The results showed that BEV DNA was detected in six fecal samples. All BEVspecific RNA-positive fecal samples from three provinces were inoculated into MDBK, BHK-21, and Vero cells for BEV isolation. After five cycles of blind passage, a virus strain was isolated from positive fecal samples, which induced distinct CPE in MDBK cell but not in BHK-21 and Vero cells (Fig. 1A). It was characterized by RT-PCR (Fig. 1B), and sequencing of the PCR products followed by online BLAST analysis revealed that the nucleotide sequence of the isolated strain was 100% identical to that of the published sequence (HeN-YR91), confirming that the isolated virus was BEV. Subsequently, MDBK cells were chosen for plaque purification, and clear, uniform plaques were obtained (Fig. 1C) after three rounds of purification. This purified strain was named "Sichuan/SQ/20 strain. " The purified strain was observed using electron microscopy after negative staining, revealing that the virus particles were non-enveloped and spherical, with an average diameter of 30-40 nm (Fig. 1D), consistent with the size of picornaviruses.
## Biological characteristics of the purified strain
The biological characteristics of the Sichuan/SQ/20 strain were further analyzed using MDBK cells. The titration of the purified strain was 10 8.73 × TCID 50 /0.1 mL. Virus growth kinetics show that the virus titer presented a gradual upward tendency and peaked at 48 hpi (Fig. 2A). The Sichuan/SQ/20 strain exhibited thermal sensitivity with signifi cantly reduced viability at 50°C (Fig. 2B). At temperatures exceeding 55°C, complete viral inactivation occurred within 12 minutes (Fig. 2C). Moreover, the purified virus exhibited welldefined plaque morphology (Fig. 2D). Specific immunofluorescence was detected in BEV-infected MDBK cells using hyperimmune serum from inactivated whole-virus vaccinated mice as primary antibody, while no signal was observed in uninfected control cells (Fig. 2E).
## Analysis of genomic sequences
Full-length genomes of the isolated strains were amplified by RT-PCR with four pairs of primers (Table 1), and the PCR products of the expected size were obtained. The full-length genome sequence of the isolated strain (7,303 bp) was successfully assem bled from fragmented sequencing data. It contains a single large ORF spanning 6,501 nucleotides, predicted to encode a 2,167-amino acid polyprotein.
## Phylogenetic analysis of BEV
To further investigate the phylogenetic relationships of Sichuan/SQ/20 strain, phylo genetic analysis was reconstructed based on the P1 and VP1 coding sequence. We performed a comparative analysis of VP1 nucleotide sequences from 20 representative strains. The phylogenetic analysis revealed that Sichuan/SQ-20, HeN-YR91, and JPN/ TottoriU-31 clustered within the same branch with high bootstrap support, indicating their closest genetic relationship (Fig. 3). Phylogenetic analysis and amino acid sequence alignment of the P1 region from the Sichuan/SQ/20 strain further confirmed the aforementioned evolutionary relationships, showing consistent clustering patterns with other BEV-F1 subtype strains (Fig. 4).
To elucidate the genetic characteristics of the Sichuan SQ/20 strain, we performed a systematic phylogenetic analysis of the nucleotide and amino acid sequences of its P1 region in comparison with representative BEV reference strains. Nucleotide sequence similarity analysis revealed that Sichuan SQ/20 shares 61.8%-82.6% similarity with other strains in the P1 region, with the highest value (82.6%) observed against HeNYR91/China. At the amino acid level, the sequence identity between Sichuan SQ/20 and the reference strains ranged from 54.2% to 83.0% in the P1 region. Furthermore, Sichuan SQ/20 shares 55.6%-83.9% similarity with other strains in the VP1 region at the nucleotide level, with the highest value (83.9%) observed against HeNYR91/China (Table 2).
## Pathological lesions and viral tropism of Sichuan/SQ/20 in mice
Neither the experimental group nor the control group exhibited overt clinical signs during the study period. The tissue tropism of Sichuan/SQ/20 was determined by PCR analysis. Viral was detected in the spleen, liver, and small intestine at 3 days post-infec tion, with the highest viral load observed in the liver (as indicated by the brightest band on agarose gel electrophoresis; Fig. 5A). This pattern suggests that the liver may be a primary site of viral replication during early infection stages. However, at 7 days post-infection, the virus was detected exclusively in the spleen, while the liver and intestinal tissues tested negative (Fig. 5B), suggesting a clearance of the virus from peripheral sites and potential persistence within the splenic microenvironment. Tissue tropism was further confirmed by immunofluorescence detection. The results revealed specific fluorescent signals across all examined organ sections, indicating successful viral infection in these tissues (Fig. 5D), which was consistent with PCR results (Fig. 5A).
Necropsy findings revealed splenomegaly with dark red discoloration in mice, while no remarkable gross lesions were observed in other tissues. Histopathological exami nation of hematoxylin-eosin (H&E)-stained sections demonstrated red pulp atrophy, sinusoidal dilation, white pulp expansion, and lymphoid follicular hyperplasia in the spleen (Fig. 5C). In contrast, the liver and small intestine have no significant pathological alterations. These findings indicate that the spleen is the primary target organ of BEV infection.
Integrated analysis of histopathological lesions and viral distribution patterns indicates that this strain possesses limited pathogenicity, does not establish long-term latency or cause critical organ damage, undergoes a relatively short replication cycle, and can be effectively contained by the host immune system prior to widespread dissemi nation. These characteristics suggest that even though the virus can establish initial infection, it poses a relatively low risk for severe disease outcomes in immunocompetent hosts.
## DISCUSSION
Cell line susceptibility is crucial for virus isolation. Studies have reported that BEV can be isolated using both MDBK and Vero cells (18,21,24). Specifically, Chengyuan Ji et al. demonstrated efficient BEV replication in MDBK cells, with lower propagation efficiency observed in BHK-21 and Vero cell lines (7). In this study, the virus isolation process began with screening 108 bovine fecal samples from four regions by PCR to identify positive cases. Virus isolation was simultaneously attempted from PCR-positive samples using MDBK, Vero, and BHK-21 cell lines. After five blind passages in each cell type, a novel bovine enterovirus strain capable of inducing stable CPE was successfully isolated in MDBK cells, designated as Sichuan/SQ/20. The isolated virus was subsequently subjected to whole-genome sequencing, and the complete nucleotide sequence was deposited into GenBank under accession number PV290163.1.
The VP1 region, serving as the primary antigenic determinant of the viral capsid, is internationally recognized as the gold standard for enterovirus typing, with >25% divergence indicating distinct types (25,26). According to the International Committee on Taxonomy of Viruses standards, >60% amino acid sequence similarity in the P1 region is required for classification within the same enterovirus species (27). Phylogenetic analysis was performed following the classification criteria. Phylogenetic analysis of the amplified VP1 nucleotide and P1 amino acid sequences was conducted using the maximum likelihood method. The results demonstrated that this strain exhibits the closest genetic relationship with the HeN-YR91 and JPN/TottoriU-31 strains, and all three belong to the BEV-F1 genosubtype.
To elucidate pathogen characteristics, facilitate vaccine and diagnostic reagent development, and inform control strategies, the physicochemical and biological properties of the isolated strain were comprehensively analyzed. Analysis revealed that viral replication of the strain commenced at 3 hpi in MDBK cells, reaching peak titers at 48 hpi with a virus titer of 1 × 10 long-term viability at -80°C, analogous to entering a "dormant state. " This finding has significant implications for establishing biosample handling protocols and investigating viral structure. However, since such extremely low temperatures are uncommon in daily environments, greater attention should be given to viral behavior under ambient and elevated temperatures. Complete inactivation of this strain after 12-minute exposure to 55°C, indicating thermal susceptibility consistent with other isolated strains (7).
To characterize the pathogenicity of the Sichuan/SQ/20 strain in mice, 6-week-old mice were infected with the strain, and samples were collected on days 3 and 7 postinfection for analysis. PCR, H&E staining, and IFA revealed viral detection in the liver, spleen, and small intestine at day 3 post-infection. At day 7, the virus was only detected in the spleen, indicating a gradual decline in viral load over time. H&E staining revealed significant pathological changes specifically in the spleen, with no lesions observed in other organs. IFA detected specific fluorescent signals in the liver, spleen, and small intestine, confirming viral infection in these tissues, which was consistent with PCR results. These findings indicate low susceptibility of mice to this virus, a self-limiting infection pattern, and distinct splenic tropism-collectively explaining the absence of notable clinical manifestations. Due to limitations of the current murine model system, further infection studies in other species are required to delineate the viral host range and infection mechanisms.
## Conclusion
In summary, this study successfully isolated a novel bovine enterovirus strain and conducted a comprehensive investigation into the physicochemical and biological properties of the Sichuan/SQ/20 isolate. These findings provide a scientific foundation for vaccine design and antiviral drug screening, as well as the development of effective prevention and control strategies.
## References
1. Gomez, Weese (2017) "Viral enteritis in calves" *Can Vet J*
2. Medina, Castañeda, Páez-Triana et al. (2024) "High prevalence of enterovirus E, bovine kobuvirus, and astrovirus revealed by viral metagenomics in fecal samples from cattle in central Colombia" *Infect Genet Evol*
3. Chen, Meng, Zeng et al. (2024) "Epidemiological survey of calf diarrhea related viruses in several areas of Guangdong Province" *Front Microbiol*
4. Zhu, Xing, Gai et al. (2014) "Identification of a Novel Enterovirus E Isolates HY12 from Cattle with Severe Respiratory and enteric diseases" *PLoS One*
5. He, Tang, Chen et al. (2017) "Isolation and characterization of a new enterovirus F in yak feces in the Qinghai-Tibetan Plateau" *Arch Virol*
6. Weldon, Blue, Wooley et al. (1979) "Isolation of picornavirus from feces and semen from an infertile bull" *J Am Vet Med Assoc*
7. Ji, Zhang, Sun et al. (2022) "Isolation and identification of two clinical strains of the novel genotype Enterovirus E5 in China" *Microbiol Spectr*
8. Blas-Machado, Saliki, Boileau et al. (2007) "Fatal ulcerative and hemorrhagic typhlocolitis in a pregnant heifer associated with natural bovine enterovirus type-1 infection" *Vet Pathol*
9. Omatsu, Tsuchiaka, Hirata et al. (2014) "Detection of enterovirus genome sequence from diarrheal feces of goat" *Virus Genes*
10. Dunne, Ajinkya, Bubash et al. (1973) "Parainfluenza3 and bovine enteroviruses as possible important causative factors in bovine abortion" *Am J Vet Res*
11. Candido, Almeida-Queiroz, Buzinaro et al. (2019) "Detection and molecular characterisation of bovine Enterovirus in Brazil: four decades since the first report" *Epidemiol Infect*
12. Cao, Liu, Li et al. (2023) "Biological characteristics of feline calicivirus epidemic strains in China and screening of broad-spectrum protective vaccine strains" *Vaccines (Basel)*
13. Gür, Yapkiç, Yilmaz (2008) "Serological survey of bovine enterovirus type 1 in different mammalian species in Turkey" *Zoonoses Public Health*
14. Rahmani, Hmaied, Matei et al. (2020) "Occurrence of Staphylococcus spp. and investigation of fecal and animal viral contaminations in livestock, river water, and sewage from Tunisia and Romania" *Environ Monit Assess*
15. Hyypiä, Hovi, Knowles et al. (1997) "Classification of enteroviruses based on molecular and biological properties" *J Gen Virol*
16. Moll, Ad (1959) "Isolation and characterization of cytopatho genic enteroviruses from cattle with respiratory disease" *Am J Vet Res*
17. Li, Chang, Wang et al. (2012) "Isolation of two Chinese bovine enteroviruses and sequence analysis of their complete genomes" *Arch Virol*
18. Tsuchiaka, Rahpaya, Otomaru et al. (2017) "Identification of a novel bovine enterovirus possessing highly divergent amino acid sequences in capsid protein" *BMC Microbiol*
19. Shi, Wang, Hu et al. (2025) "Enterovirus E infections in goats with respiratory disease" *BMC Vet Res*
20. Woo, Lau, Li et al. (2015) "A novel dromedary camel enterovirus in the family Picornaviridae from dromedaries in the Middle East" *J Gen Virol*
21. Blas-Machado, Saliki, Sánchez et al. (2011) "Pathogenesis of a bovine enterovirus-1 isolate in experimentally infected calves" *Vet Pathol*
22. Simmonds, Gorbalenya, Harvala et al. (2020) "Recommendations for the nomenclature of enteroviruses and rhinoviruses" *Arch Virol*
24. Kumar, Stecher, Li et al. (2018) "MEGA X: molecular evolutionary genetics analysis across computing platforms" *Mol Biol Evol*
25. Beato, Marcacci, Schiavon et al. (2018) "Identification and genetic characterization of bovine enterovirus by combination of two next generation sequencing platforms" *J Virol Methods*
26. Earle, Skuce, Fleming et al. (1988) "The complete nucleotide sequence of a bovine enterovirus" *J Gen Virol*
27. Oberste, Maher, Flemister et al. (2000) "Comparison of classic and molecular approaches for the identification of untypeable enteroviruses" *J Clin Microbiol*
28. Zerbini, Siddell, Lefkowitz et al. (2023) "Changes to virus taxonomy and the ICTV statutes ratified by the international committee on taxonomy of viruses" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584624&blobtype=pdf | # HCMV infection of terminally differentiated neurons disrupts microtubule organization, resulting in neurite retraction
Jacob Adelman, Andrew Sukowaty, Kaitlyn Partridge, Jessica Gawrys, Allison Akins, Scott Terhune, Allison Ebert
## Abstract
Human cytomegalovirus (HCMV) is a prolific human herpesvirus that infects most individuals by adulthood. While typically asymptomatic in adults, congenital infection can induce serious neurological symptoms, including hearing loss, visual deficits, cognitive impairment, and microcephaly in 10%-15% of cases. HCMV has been shown to infect most neural cells, with our group recently demonstrating this capacity in stem cell-derived forebrain neurons. Infection of neurons induces deleterious effects on calcium dynamics and electrophysiological function paired with gross restructuring of neuronal morphology. Here, we utilize an induced pluripotent stem cell-derived model of the human forebrain to demonstrate how HCMV infection induces syncytia, drives neurite retraction, and remodels microtubule networks to promote viral production and release. We establish that HCMV downregulates microtubule-associated proteins while largely sparing other cytoskeletal elements. Furthermore, we pharmacologically modulate microtubule dynamics using paclitaxel (stabilize) and colchicine (destabilize) to examine the effects on neurite structure, syncytial morphology, and viral release. With paclitaxel, we found improvement of neurite outgrowth, but neither paclitaxel nor colchicine impacted viral titers. Together, these data suggest that HCMV infectioninduced disruption of microtubules in human cortical neurons can be partially mitigated with microtubule stabilization, suggesting a potential avenue for future neuroprotective strategies.IMPORTANCE Infection by human cytomegalovirus (HCMV) continues to cause significant damage to human health. In the absence of a vaccine, vertical transmission from mother to fetus can result in profound neurological damage impacting quality of life. These studies focus on understanding the impact of HCMV infection on forebrain cortical neurons derived from induced pluripotent stem cells (iPSCs). We show that infection results in loss of neurite extension accompanied by cell-to-cell fusion. These pathogenic changes involve HCMV infection-mediated disruption of the microtubule network in iPSCs from different patient backgrounds. The microtubule stabilization agent paclitaxel partially protected neurite length and altered syncytia morphology without impacting viral replication. This work is part of our continued efforts to define putative strategies to limit HCMV-induced neurological damage. KEYWORDS HCMV, induced pluripotent stem cells, tubulin, paclitaxel, colchicine H uman cytomegalovirus (HCMV) is a pervasive pathogen that is estimated to infect between 40% and 90% of adults worldwide (1, 2). As a human beta-herpesvirus, HCMV infection is lifelong, occurring in waves of dormancy (latency) and reactivation. Furthermore, transfer of infection can occur either through post-natal exposure to infected bodily fluids (horizontal transfer) or vertically from parent to fetus in utero. HCMV infects a wide range of cell types, including fibroblasts, various subtypes of
endothelial and epithelial cells, placental trophoblasts, and hematopoietic precursors (3)(4)(5)(6)(7). Neural cells have also been established as a site of infection. Neural progenitor cells (NPCs) are central nervous system-specific stem cells and a key site of infection within the human brain (8)(9)(10). Other cell types demonstrated to sustain infection include astrocytes (11), microglia (12), oligodendrocytes (13), and ependymal cells (14). The potential for neurons to become infected has, however, been debated within the field (11,(15)(16)(17). Recently, both our group and others have demonstrated that terminally differentiated neurons are susceptible to infection regardless of whether their progeni tors were induced pluripotent stem cell (iPSC) or fetal stem cells (18)(19)(20)(21). Furthermore, infection has been demonstrated to have functional impacts on neurons, including alterations to calcium signaling and reduced action potential generation (18,21). While these studies have been helpful in characterizing HCMV-induced functional changes in neurons, further evaluation is needed to understand how these cells change structurally.
The neuronal cytoskeleton is composed of three primary elements: actin micro filaments, neurofilaments, and microtubules. Together, these structures function to resist structural deformation, maintain intracellular organization, and allow the cell to interact with its surrounding environment. Actin microfilaments are composed of polar, filamentous actin (F-actin) strands that offer a quick means for neurite expansion and retraction in neurons. Typically, actin structures are found near the cell membrane and enable dynamic interaction with the surrounding environment via filopodia and lamellipodia formation (22)(23)(24). Synaptic structures are also organized specifically by actin microfilaments (23). Finally, actin filaments are also capable of inducing cell movement via the associated motor protein myosin. Neurofilaments are the neuronspecific members of the intermediate filament protein family. Compared with actin microfilaments and microtubules, neurofilaments are not polar and therefore have no associated motor proteins. Neurofilaments primarily serve to maintain cell structure and, secondarily, facilitate neuronal function. At axonal projections, neurofilaments arrange to increase radial diameter, improving the conductivity of electrical depolarizations (action potentials) (25)(26)(27)(28). Finally, microtubules compose the third branch of the neuronal cytoskeleton and perhaps serve the most varied roles in neuronal physiology.
Microtubules are hollow, polar, cylindrical structures that are composed of tubulin proteins. Tubulins are a protein superfamily with three primary variants: alpha, beta, and gamma. α-and β-tubulins form heterodimers that are the primary components of microtubules with a structure arising from the repeating, organized assembly of the dimerized proteins (29). αβ dimers bind guanosine diphosphate or triphosphate functional groups that dictate microtubule assembly (30,31). Microtubules are further strengthened through the binding of microtubule-associated proteins (MAPs) such as Tau and MAP2. MAPs bind microtubules longitudinally, acting both to crosslink individual heterodimers and connect cytoskeletal elements (32,33). At the negative end of the microtubules, αβ dimers are stabilized using a combination of γ-tubulin and negativeend binding proteins (34). γ-tubulin also plays a key role in nucleation for microtubules at the centrosome (or, to a lesser degree, at the Golgi Apparatus), coordinating with other gamma-tubulin complex components (35,36). Microtubule nucleation events occur at a microtubule-organizing center. These structures act as a hub for microtubule extension throughout the whole cell and serve a key role in coordinating microtubule function.
Microtubules form an intracellular "highway" system, facilitating the movement of cellular components to their proper destinations. This process is mediated by microtu bule-associated motor proteins such as dynein and kinesin to move cargo. This cargo can wildly range in size, from objects as small as viral particles to those as large as entire organelles (37). In neurons, microtubules are indispensable for the trafficking of neurotransmitter-containing synaptic vesicles (38) and for facilitating energy distribution via the trafficking of mitochondria (39). Additionally, microtubules are key to maintaining neuronal morphology, whether semi-statically at the soma and axon or dynamically within dendrites (40). Outside of these roles, microtubules also have a putative role in intracellular signaling cascades (41).
Here, we examined how HCMV infection alters the neuronal cytoskeleton in the context of iPSC-derived forebrain neurons from apparently healthy individuals. We observed significant structural alterations in neuronal morphology occurring between 2 and 14 days post-infection (dpi) with a significant impact on MAP expression. Low-dose pharmacological microtubule modulation impacted neurite length, but it had no impact on viral titers. Taken together, these data demonstrate that a low concentration of microtubule modulation does not alter HCMV infection in iPSC-derived neurons but does help neurons maintain their structure, suggesting that microtubules may be valuable targets for mitigating the neuronal effects of HCMV infection.
## RESULTS
## HCMV infection alters neurite length, localization of tubulin, and expression of microtubule-associated genes
We used four independent and unrelated iPSC lines with a minimum of three differentiations for each line to assess the impact of HCMV infection on human neuron structure. To generate forebrain neurons, we differentiated iPSCs into an NPC stage and subsequently patterned NPCs toward a forebrain neuron lineage (Fig. 1A). Consistent with our previous work (18), neurons were infected with bacterial artificial chromosome (BAC)-derived HCMV TB40/E-eGFP (42,43) at a multiplicity of infection (MOI) of 3 infectious units per cell at day 51 of differentiation (Fig. 1A). To validate the maturity of our culture system, we evaluated expression of the early neuronal marker doublecortin (DCX; Fig. 1B). Further differentiation demonstrates that DCX expression is lost as the neurons mature to express neuron-specific β III tubulin (TUJ1) and exhibit the expected morphology with long thin processes (Fig. 1B). As we have shown previously (18), cultures also contain an astrocyte population (glial fibrillary acidic protein [GFAP], Fig. 1B). Together, these data demonstrate consistent differentiation of mature neural populations. Finally, using a genetically encoded eGFP reporter HCMV, we determined sites indicative of infection at 7 dpi (Fig. 1C).
Initial observation of infected cultures revealed significant alteration to neuronal structure, including formation of syncytia and a visual change to neurites (Fig. 1C; Video S1). Using neurite tracing analysis in combination with MAP2 staining, we found that neurite quantity per nucleus was significantly decreased when neurons were infected with HCMV compared to mock conditions (Fig. 1D andE). Neurite tracing also revealed a trend toward decreased neurite length in infected cultures (Fig. 1D andF). Furthermore, when longer neurites (those equal to or greater than 75 µm; bottom threshold indicated by the red line in Fig. 1F) were isolated from the total population, infection significantly decreased neurite length (Fig. 1G). Additionally, lengthy neurites occurred more frequently in mock-treated cells (n = 18) relative to their infected counterparts (n = 5; Fig. 1G). Notably, neurites ≥75 µm were only present in one of the two lines post-infection (Fig. 1G). To confirm that the decrease in neurite length was not due to overt axonal degeneration, we measured axonal varicosities (e.g., beads on a string [44]), a phenotype commonly associated with axonal breakdown and degeneration, in observable neurites in mock and HCMV conditions. We did not observe a significant difference, suggesting that HCMV infection is not causing overt neurodegeneration (Fig. 1H). These data are consistent with our previous findings showing no significant increase in cell death in HCMV-infected cortical neuron cultures (18).
To further investigate the neurite retraction phenomenon, we sought to evaluate the effects of HCMV on various elements of the neuronal cytoskeleton, including actin filaments, neurofilaments, and microtubules (Fig. 2A). Assessment of gene expression for these structural components in neurons from multiple iPSC lines was first conducted using RT-qPCR. The effects of HCMV on β-actin (ACTB) transcripts showed no significant change at 7 dpi but a significant decrease at 14 dpi (41%; Fig. 2B andC). Neurofilament subunits showed a non-significant HCMV-induced reduction in both light (NEFL) and medium (NEFM) chain mRNAs at 7 dpi and no consistent change at 14 dpi, but there was an HCMV-induced upregulation of heavy chain (NEFH) transcripts at both timepoints (fivefold at 7 dpi; 17-fold at 14 dpi; Fig. 2D andE). Finally, mRNAs for neuron-specific beta-III tubulin (TUBB3) were found to be significantly downregulated at both 7 (49%) and 14 dpi (62%; Fig. 2F andG). These findings indicate downregulation of both ACTB and TUBB3 gene expressions, but upregulation of NEFH in neurons from multiple iPSC lines. We next examined the temporal changes in cytoskeletal elements at a protein level. We noted a consistent trend toward downregulation across all timepoints for ACTB expression (Fig. 2H through J), which mirrored a similar trend to the mRNA levels. Neurofilament 68 (NF68, NEFL gene product) was selected to represent the effects of HCMV on neurofilament expression. There was no significant effect of HCMV on NF68 protein expression across all tested timepoints (2, 4, 7, and 14 dpi; Fig. 2H, I andK). We were unable to identify a working antibody to NEFH. Finally, TUJ1 was examined to assess the effects of infection on microtubules. No significant change in TUJ1 expression was noted at any timepoint, though a modest decrease in tubulin was present at 14 dpi (Fig. 2H, I andL). As we see HCMV-induced structural deformations of the neuronal cytoskeleton at 7 dpi (Fig. 1C), our findings indicate that HCMV modulation of cytoske letal elements likely does not involve altered expression levels of the tested structural proteins. Therefore, we next sought to evaluate proteins that modulate cytoskeletal stability.
Using neurons derived from multiple iPSC lines, we assessed two key MAP genes, MAPT and MAP2, using RT-qPCR. MAPT transcripts were found to be downregulated at both 7 (48% decrease) and 14 dpi (62% decrease; Fig. 3A andB). Likewise, MAP2 transcripts were decreased at both timepoints (72% at 7 dpi and 53% at 14 dpi; Fig. 3C andD). Assessment of total tau protein (MAPT-encoded isoforms) by western blot found significant downregulation at both 7 (24%) and 14 dpi (50%; Fig. 3E through H). MAP2 protein expression was not changed at 7 dpi (Fig. 3I andJ) but was modestly decreased at 14 dpi by 18%; Fig. 3K andL). Together, these data show decreased MAP expression levels, suggesting that neuronal microtubule stability may be impacted by HCMV infection.
Although there were no significant changes noted in total tubulin protein expression through the first 7 days of infection, neuron structure was altered. We examined neuronal infection at 2, 4, and 7 dpi using immunofluorescence and observed changes in tubulin and MAP distribution within infected cells. First, a subclass of neurites associated with eGFP-labeled cell bodies appeared thickened relative to their mock-treated counterparts (Fig. 4A). This structural change was noticed as early as 4 dpi and continued to be observed within 7 dpi cultures (Fig. 4A, white arrowheads). Furthermore, at 7 dpi, TUJ1 staining within syncytia revealed non-filamentous, punctate signal within the syncytial core (Fig. 4A, blue arrowheads). Finally, we noted areas devoid of neurite processes within infected cultures, especially distal to syncytia (Fig. 4A, orange arrow heads). These data suggest that infection induces a structural phenotype by utilizing the existing tubulin present within a cell. Therefore, we hypothesized that HCMV infection alters microtubule stability rather than affecting overall microtubule abundance.
## Stabilization of microtubules with paclitaxel alters syncytial morphology
We next utilized two compounds known to impact the formation and stability of microtubules to determine the role of microtubule stability in forebrain neuron infection: paclitaxel (Pac), a microtubule-stabilizing agent, and colchicine (Col), a microtubuledestabilizing compound. As both compounds are associated with cell death at high concentrations (45)(46)(47)(48), we treated uninfected neurons with varying amounts of Pac and Col for 7 days and evaluated varicosities, neurites per nucleus, and neurite length.
Treatment with 20 nM Pac significantly increased varicosity formation, neurite loss, and neurite retraction compared to mock (Fig. 5A through C). A total of 10 nM Pac also significantly induced varicosity formation, although it did not significantly impact neurite number or length (Fig. 5A through C). Neither 1 nM nor 5 nM treatment of Pac significantly altered neurite phenotypes compared to mock (Fig. 5A through C). Therefore, we selected 5 nM Pac for subsequent studies. Col treatment revealed few negative effects other than increased varicosities with the 10 nM concentration (Fig. 5D). Col treatment did not have any effect on neurite number at any concentration tested (Fig. 5E) but surprisingly increased neurite length at all concentrations with the greatest increase occurring with 5 nM Col (Fig. 5F). Based on these data, we chose 2 nM Col for the remaining studies.
To test the effect of Pac and Col on neuron morphology in the context of HCMV infection, neurons were infected with HCMV and then treated with Pac (5 nM), Col (2 nM), or DMSO (vehicle, 5 nM) for 6 days before collection of cultures at 7 dpi (Fig. 5G). As microtubules are required for HCMV entry and pathogenesis, treatment of the neurons was delayed until 24 hpi. An initial assessment of TUJ1 was conducted at 7 dpi to determine if drug application altered expression. At these treatment concentrations, no significant differences in TUJ1 expression levels were detected between the various treatment groups (Fig. 5H andI). Following this, we investigated tubulin structure via immunofluorescence. Within mock-treated cultures, we detected no significant differences in neurite quantity or length (Fig. 6A andC). We then assessed neurites ≥75 µm in length (above the red line in Fig. 6C) and found no significant differences in neurite length between uninfected untreated (utx) control, DMSO control, Pac, and Col-treated conditions. Likewise, all treatments produced relatively similar amounts of long neurites (utx: n = 15, DMSO: n = 23, Pac: n = 18, and Col: n = 18), and there were no significant differences in the number of neurites per nucleus amongst the experimental groups in mock-treated cells (Fig. 6G). For ease of comparisons across treatment groups in Fig. 6C through H, data for the untreated condition are repeated from Fig. 1. Additionally, consistent with Fig. 1F, the DMSO HCMV infection condition showed reduced neurite length compared to the DMSO mock condition (Fig. 6C andE, gray plots). In the HCMVinfected conditions, immunofluorescence demonstrated that syncytia formed in the untreated, DMSO-treated, and Col-treated infected cultures and were similar in shape and size (Fig. 6A, B, I andJ), and neurite number and neurite length were not significantly different among these groups (Fig. 6C through H). Interestingly, Pac-treated HCMVinfected neurons formed more linearly organized syncytia than those found within utx, DMSO control, and Col-treated cultures (Fig. 6A, B andI). Furthermore, infected neurons treated with Pac trended toward having longer neurite lengths overall (Fig. 6E), with significantly longer processes ≥75 µm in length compared to HCMV-untreated, HCMV + DMSO and HCMV + Col (Fig. 6F) and with lengths similar to the Pac-treated mock condition (Fig. 6C through F, blue plots). Long neurites were also more abundant among the Pac-treated cells (n = 10) relative to utx neurons (n = 5), DMSO controls (n = 3), and Col-treated neurons (n = 3). Interestingly, while longer neurites were still predominantly found in infected cultures derived from iPSC Line 2 (reminiscent of Fig. 1G), we did observe that Pac treatment induced a slight preservation of longer neurites in cells derived from iPSC Line 1, as well (Fig. 6F). Interestingly, Col treatment did not increase neurite length in the presence of HCMV infection as it had done in control conditions (Fig. 6F vs Fig. 5F). Overall neurite counts were not affected (Fig. 6H). These data indicate that microtubule stabilization helps maintain neurite extension and may partially disrupt typical syncytia organization induced during HCMV infection. Considering this, we hypothesized that microtubule modulation could also disrupt viral production. Furthermore, neurite changes can be observed as early as 4 dpi relative to mock cultures. These findings include thickened processes (denoted by white arrows).
At 7 dpi, tubulin and MAP2 staining appeared punctate in the syncytia core (blue arrows), rather than filamentous as examined elsewhere. Furthermore, regions lacking neurites were more common in the areas adjacent to the syncytia (orange arrows). Scale bars = 40 µm.
## Markers of viral production are not significantly altered by modulating microtubule stability
Finally, we sought to examine the subsequent impact of microtubule modulation on viral output from infected neurons. Live imaging was conducted for ~7 dpi, with frequent monitoring of eGFP expression within the culture. These traces are depicted in Fig. 7A and did not demonstrate any significant deviations by Pac or Col treatment relative to DMSO-treated and untreated controls. These findings were confirmed upon evaluation of total eGFP expression in 7 dpi cell lysates (Fig. 7B andC). Cell lysates were then probed for relative expression of HCMV immediate early protein 1 (IE1) and late protein pp28, but there were no significant differences between any of the treatment groups for expression of these viral targets (Fig. 7D through F). Finally, viral titers were analyzed using the conditioned medium (CM) from HCMV-infected neurons. Despite changes in syncytia morphology of Pac-treated cells (Fig. 6), there were no significant differences in viral titers between treatment groups (Fig. 7G). Taken together, these data suggest that HCMV infection-induced structural changes can be partially abated by microtubulemodulating treatments without a corresponding impact on viral production.
## DISCUSSION
Microtubules have been known to be associated with HCMV infection for decades. Their relevance in virion entry, capsid trafficking, viral assembly, and, ultimately, egress processes has been thoroughly documented in various cell models (49)(50)(51)(52)(53)(54). However, the recent groundswell of interest regarding the effects of HCMV infection in nervous tissue (19,21,(55)(56)(57)(58)(59)(60)(61) requires a reassessment of its impact on microtubules in the context of neurites (axons and dendrites). Due to substantially different structures and functions, the application of findings from fibroblasts onto neurons and NPCs is likely to be problematic. Fortunately, since the discovery of iPSCs, access to human-derived neural cells has increased exponentially, allowing for easier evaluation of HCMV within neurons. Here, we sought to better describe neuron-specific structural phenomena relating to HCMV infection. Specifically, we analyzed effects on the expression of neuronal cytoske letal components and demonstrated that stabilization of microtubules can mitigate some of the morphological aberrancies induced by infection.
HCMV infection induced significant alterations to neuronal morphology within the first 7 days of infection (Fig. 1C and4A). Neurites have been described by both our group and others as a focal site of these changes, with projections becoming dissociated from their underlying matrix (21) and ultimately retracting into the neuronal soma (18, 21) (Fig. 1D through G). The other significant morphological change that occurs in HCMV-infected neurons is the formation of multinucleated syncytia (Fig. 1C; Video S1) (18). While the mechanism of syncytia formation is likely conserved from other cell types -viral proteins on the cell surface driving membrane fusion-little is understood about the way in which retraction of cellular projections occurs. Our assessment of cytoskeletal elements indicated that neurite loss is likely not due to altered gene expression, as neurofilaments, actin, and tubulin protein levels are not differentially expressed at 7 or 14 dpi (Fig. 2H through L) despite reductions in mRNA (Fig. 2B through L). In contrast, we detected significant decreases in mRNA and protein expressions of MAP2 and Tau (Fig. 3). Together, these findings suggest that the infection-induced observed reorganization of the neuronal cytoskeleton more likely involves altered microtubule-associated proteins than expression levels of microtubule subunits.
Utilizing both Pac and Col, we sought to examine how pharmacological modulation of microtubule stability would impact HCMV-induced cytoskeletal rearrangement. Both compounds are clinically approved pharmaceuticals and have been proven efficacious for the stabilization (Pac) and destabilization (Col) of microtubules in both in vivo and in vitro systems (62)(63)(64)(65). When deciding our concentrations for each drug, we wanted to ensure an appropriate balance between the compound's intended effects and its potential to induce cytotoxicity. The upper limits for Pac and Col were 20 nM and 10 nM, respectively, due to documented cell toxicity above that range (45,48). We found similar results for Pac (10-20nM) and Col (10 nM), resulting in increased neuronal varicosities (Fig. 5A andD). As we intended to incubate neurons with each for a minimum of 6 days, we decided upon concentrations below those utilized in short-term studies (Fig. 5G). However, it remains possible that the modest effects we describe are due to these low dosages. For this reason, additional follow-up studies are necessary to better titer each drug for peak effectiveness. However, Pac treatment did induce some maintenance of neurite structure in treated cultures relative to the untreated infection group (Fig. 6A,B, E andF). Quantification of neurite length in Pac-treated cells demonstrated a trend toward increased neurite length relative to infected neurons treated with DMSO (Fig. 6E). In longer neurites-specifically those ≥75 µm-these changes were significant and indicate Pac's potential for preserving neuron structure during infection (Fig. 6F). This is particularly apparent when considering the presence of longer neurites from iPSC line 1, which are not present in non-Pac-treated data sets. Together, these data suggest that there may be a potential benefit of Pac treatment in the context of HCMV neuronal infection.
In addition to the structural damage induced by infection, neurons undergo severe functional disruptions, including reductions in calcium dynamics and elimination of action potential generation (18,55,56). Additional research is needed to determine the impact of microtubule-modulating treatments on calcium dynamics and electrophy siological activity in HCMV-infected neurons to fully understand the potential benefits or consequences of microtubule stability. However, considering the importance of neurite structure in neuronal signaling (66,67), we would hypothesize that even modest maintenance of structure could improve overall synaptic activity. In support of this, others have found that low-dose Pac positively impacts neuronal function in models of Alzheimer's Disease (68), tauopathy (69), and spinal cord injury (70).
To our surprise, there was no significant disruption in virus production. Neither live imaging for eGFP expression (Fig. 7A) nor expression of the viral proteins IE1 and pp28 (Fig. 7D through F) showed a substantial reduction with Pac treatment. It is possible that with higher Pac concentration or a longer duration treatment paradigm, there would be a more obvious decrease in viral titers. However, this may come at the expense of overall neuron health. Additionally, the culture system does not contain brain-resident macrophage cells (microglia), which could alter the effects of neuronal HCMV infection (71,72) and/or the effect of Pac and Col treatment (73,74). As such, additional research is needed to fully understand the role that the immune system plays in HCMV-induced neuronal damage. Nevertheless, even a modest improvement in neurite outgrowth could have subsequent downstream beneficial effects on neuron health, survival, and function. Together, these data show microtubule stability as a potential modifier of HCMV neuronal pathology in neurons and may be a novel target of therapeutic intervention.
## MATERIALS AND METHODS
## Cell culture and differentiation
iPSC lines were derived from either reprogrammed skin fibroblasts or patient blood cells (Line 1: Coriell GM03814 [Fibroblast line reprogramed to iPSCs]; Line 2: WiCell PENN022i-89-1 [purchased as iPSCs; derived from blood cells]; Line 3: Coriell AG25370D [purchased as iPSCs; derived from fibroblasts]; Line 4: Coriell AG27607D [purchased as iPSCs; derived from fibroblasts]; Line 5: HB53, kind gift of Dr. Ivor Benjamin [received as iPSCs; derived from blood cells]). The utilized cell lines were chosen due to our extensive use of each line, availability, consistency of differentiation, and varying patient backgrounds. Lines 1, 3, and 4 were derived from white females at ages of 30, 81, and 69, respectively. Line 2 was obtained from a 28-year-old African American male. Line 5 was derived from a 25-year-old Caucasian male (75). Neither we nor previous studies have evaluated HCMV infection status (latency) in these individuals prior to utilizing the above cell lines. Stem cell cultures were expanded under feeder-free culture conditions using Geltrex (Gibco) basement membrane matrix. iPSCs were plated onto six-well plates and maintained with complete, daily replacements of EssentialEight medium (Thermo Fisher Scientific). Post-thaw, all iPSC cultures were passaged a minimum of three times before differentiation. All cultures were subject to quarterly mycoplasma testing to ensure freedom from contamination.
Patterning of iPSC colonies toward an NPC fate was completed utilizing dual-SMAD inhibition, as described by Chambers et al. (76). For this step, STEMdiff SMADi Neural Induction Kit (STEMCELL Technologies) and lab-prepared neural progenitor medium
100 ng/mL Laminin [Sigma], 40 µM SB431542 [Biogems], and 0.2 µM LDN193189 [Biogems]) were used interchangeably. NPCs were cultured for a minimum of three passages in SMAD-inhibition medium (with daily changes) prior to differentiation to ensure complete and consistent generation of neural progenitors. For each passage, NPCs were dissociated using accutase (STEMCELL Technologies) to generate a single-cell suspension. Cells were then replated at a density of 2 × 10 5 cells/cm 2 onto Geltrex-coated six-well plates. NPCs were passaged once every 6-7 days. Between 18 and 21 days of differentiation, NPCs were dissociated and plated at a density of 1.25 × 10 5 cells/cm 2 in preparation for patterning toward immature forebrain neurons.
Using the STEMdiff Forebrain Neuron Differentiation Kit (STEMCELL Technologies), NPCs were directed toward a forebrain neuron cell fate. Cells were maintained under daily media changes for 6-7 days prior to final accutase dissociation and plating onto either Poly-L-Lysine-and Laminin-glassware or Geltrex-coated plasticware at varying densities (six-well plate: 5.2 × 10 4 cells/cm 2 , 24-well plate: 1.32 × 10 5 cells/cm 2 , and coverslips: 3.1 × 10 5 cells/cm 2 ). Upon plating, immature forebrain neurons were cultured in forebrain neuron maturation medium (BrainPhys Neuronal Medium
$$(Neurobasal, 10% [vol/vol] Knockout Serum Replacement [Gibco], 1% [vol/vol] 100× Non-essential Amino Acids [Gibco], 2% [vol/vol] 50× B27 Supplement [Gibco], 1% [vol/ vol] 100× N2 Supplement [Gibco], 1% [vol/vol] 100× Antibiotic-Antimycotic [Gibco],$$
## Viruses
Engineered to express eGFP, BAC-TB40/E (TB40/E-eGFP) was generously provided by Felicia Goodrum (University of Arizona). This HCMV variant was generated via trans fection of MRC-5 fibroblasts with both a BAC containing the HCMV genome and a UL82-encoding plasmid, as previously described (42,43,77). Utilizing the stocks of TB40/ E-eGFP prepared in fibroblasts, ARPE-19 epithelial cells were infected and utilized to produce epithelial cell-derived TB40/E-eGFP. Viral titers were determined for epithelialderived stocks using the method previously reported by our group (18,60,78) and are expressed in infectious units per milliliter (IU/mL). All neuronal infections were conduc ted at an MOI of 3. Viral inoculum (or phosphate buffered sailine (PBS) for mock-treated) was applied to the cell medium for 2 hours at 37°C, with constant agitation (rotary shaker plate). After exposure, cells were washed once with 1× Dulbecco's PBS (dPBS, Gibco) to remove residual viral particles, and fresh medium was added.
## Microtubule modulation treatments
Colchicine (Col, Tocris) and paclitaxel (Pac, MilliporeSigma) working stocks were created by dissolving each compound in dimethyl sulfoxide (DMSO) to obtain concentrations ranging from 0.5 to 10 µM and 1 to 20 µM, respectively. Stocks were kept frozen at 4°C between administrations. At 21 days of differentiation, neurons were treated with 0.5-10 nM of Col or 1-20 nM Pac for 7 days. DMSO was used as a diluent control. Fresh compound was added with medium replacement at day 3. Cells were fixed for analysis 7 days post-treatment. Beginning at 51 days of differentiation, Col (2 nM), Pac (5 nM), or DMSO was added to the culture medium at a dilution of 1:1,000. When media change was required (every 3-4 days), fresh Col, Pac, or DMSO was added to the neuronal medium at 1:1,000. Cultures were treated with each compound until 58 days of differentiation, at which point they were collected for analysis.
## Immunofluorescence and live imaging
Forebrain neuron cultures were plated onto 12 mm Poly-L-Lysine (PLL, Sigma)/Laminin (Sigma)-coated coverslips at a density of 35,000 per coverslip (3.09 × 10 5 cells/cm 2 ). Cells were fixed in 4% paraformaldehyde for 15 minutes, washed once with dPBS, and stored in fresh dPBS at 4°C until use. Cell permeabilization was conducted by applying Triton X-100 solution (0.2% [vol/vol], in PBS) to each coverslip and incubating for 10 minutes at room temperature (RT). After 1× PBS wash, cells were treated with blocking buffer (5% [vol/vol] normal goat/donkey serum [NGS/NDS; dictated by secondary] in PBS) for 1 hour at RT. Primary antibody solution (primary antibodies, 5% [vol/vol] serum [NDS/NGS], 0.1% [vol/vol] Triton X-100, in PBS) was then applied to coverslips overnight at 4°C. After 4× PBS washes, cells were treated with secondary antibody solution (secondary antibodies, 5% [vol/vol] serum, 0.1% [vol/vol] Triton X-100, in PBS) and allowed to incubate for an hour at RT. The secondary antibody solution was removed, and coverslips were washed 4× with PBS to clear non-specific binding. Coverslips were mounted onto slides using Fluoromount-G Mounting Medium with DAPI (SouthernBiotech) and sealed using clear nail polish (Ted Pella Inc.). Coverslip imaging was conducted using a Zeiss LSM980 confocal microscope at various magnifications (20×-63×). Image analysis was conducted using NIS Elements (Nikon), Zen Blue (Zeiss), and ImageJ.
The following antibodies and dilutions were used for immunofluorescence during this study: chicken anti-TUJ1 (1:250-300; GeneTex), chicken anti-MAP2 (1:250-500; Invi trogen), mouse anti-MAP2 (1:250-500; Invitrogen), rabbit anti-TGN46 (1:500; Invitro gen), mouse anti-Doublecortin/DCX (1:50; Santa Cruz Biotechnology), rabbit anti-GFAP (1:1,000; Dako), goat anti-chicken IgY (H + L) AF647 (1:250, Invitrogen), goat anti-rabbit IgG (H + L) AF568 (1:500, Invitrogen), donkey anti-mouse IgG (H + L; (1:250, Invitrogen), and goat anti-chicken IgG (H + L) AF568 (1:500, Invitrogen).
Single-instance live cell imaging of virally encoded eGFP was collected using a Nikon TS100 inverted microscope. Additionally, timelapse live imaging was conducted using an IncuCyte S3 in-incubator system (Sartorius AG). Images were collected every 2 hours for 7 days, using settings to capture brightfield and green fluorescence images (ex. 440-480 nm; em. 504-544 nm). Incucyte imaging was conducted using a 20× objective, and analysis was conducted using the Incucyte 2022B revision 2 software package (video stitching, background subtraction, and eGFP quantification) and ImageJ (neurite retraction).
## Neurite and syncytia measurements
The ImageJ plugin NeuronJ was for neurite tracing and analysis. Each image consisted of a DAPI and MAP2 channel; in infected cultures, the GFP channel was referenced to determine areas of high infection. Prior to tracing, three regions of interest (ROI; 2,000 × 2,000 pixels) were overlayed onto each image. For ROIs without syncytia, five nuclei were chosen for neurite analysis. Neurites were traced from the soma to the longest branch. For ROIs containing syncytia, neurites were traced from the edge of the syncytia to the longest branch (79)(80)(81). To examine the effects of drug treatment and infection on long neurites (≥75 µm), which we identified as the median length, the total data were filtered by length using Microsoft Excel. The ImageJ freeform line tool was used to trace around the general shape of nuclei forming each syncytium. Each image consisted of a DAPI and MAP2 channel, and the GFP channel was referenced to verify the presence of syncytia. The measurement functions, shape descriptors, and area were used to obtain the measurements for circularity and area, respectively. To measure axon swellings, each image consisted of a DAPI and MAP2 channel. Prior to data collection, ROIs of 500 × 500 pixels were overlaid in areas without nuclear staining and low neurite density. Axon swellings were then counted using the ImageJ multi-point tool within each ROI.
Varicosity measures were conducted using dual-colored images labeling DAPI and MAP2. Prior to data collection, ROIs of 500 × 500 pixels were overlaid in areas without nuclear staining and low neurite density to limit intermingling of neurites. Varicosities were counted using the ImageJ multi-point tool within each ROI, and measurements were transposed into GraphPad Prism for further analysis.
To determine syncytial area and circularity, the ImageJ freeform line tool was used to trace around the general shape of nuclei forming each syncytium. As with varicos ity measures, dual-channel images labeling DAPI and MAP2 were used to ascertain structure, and the GFP channel was referenced to verify that each putative syncytium was exposed to HCMV. The measurement functions "shape descriptors" and "area" were used to obtain the measurements for circularity and area, respectively, and data were analyzed within Graphpad Prism.
## Viral titers
To ensure adequate cell conditioning of all medium samples, collections occurred after a minimum of 48 hours in culture. Neuronal CM was harvested at 7 dpi for both compound-and untreated-neurons and subsequently was stored at -80°C until use. To determine the viral titers of each CM sample, stocks were serially diluted, and the resulting dilutions were applied to HCMV-naïve ARPE-19 cells. After 2 hours of exposure to undiluted inoculum, fresh medium was added to dilute the drug remaining in the conditioned medium. At 24 hours post-infection (hpi), all inoculum-containing medium was removed from the cells and replaced with fresh medium. Cells were allowed to incubate in conditioned medium for 72 hours prior to being fixed and stained for HCMV IE1 (mouse α-IE1, added 1:500 was generously provided by Tom Shenk [Princeton University]; Goat α-mouse AF488 [1:1,000]). As above, results are reported as infectious units per milliliter (IU/mL).
## Protein and DNA analyses
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR) was conducted to analyze relative amounts of various elements of the neuronal cytoskeleton. Using pelleted cells from 2, 4, 7, and 14 DPI, mRNA transcripts were isolated using the standard protocol for the RNeasy extraction kit (Qiagen). Isolated RNA was treated with RQ1 RNA-free DNase (Promega) to ensure removal of contaminating genomic DNA prior to amplification. RNA transcript amplification was conducted using random hexamers and the AMV Reverse Transcriptase System (Promega). Cytoskeletal elements were assessed using the following primer sets recommended by the PrimerBank tool (82, 83): TUBB3 (5′-ATCAGCAAGGTGCGTGAGGAG-3′ and 5′-TCGTTGTCGATGCAGTAGGTC-3′), MAP2 (5′-CG AAGCGCCAATGGATTCC-3′ and 5′-TGAACTATCCTTGCAGACACCT-3′), ACTB (5′-CATGTACG TTGCTATCCAGGC-3′ and 5′-CTCCTTAATGTCACGCACGAT-3′), NEFH (5′-CCGTCATCAGGCCG ACATT-3′ and 5′-GTTTTCTGTAAGCGGCTATCTCT-3′), NEFM (5′-AGGCCCTGACAGCCATTAC -3′ and 5′-CTCTTCGGCTTGGTCTGACTT-3′), NEFL (5′-ATGAGTTCCTTCAGCTACGAGC-3′ and 5′-CTGGGCATCAACGATCCAGA-3′), MAPT (5′-CCAAGTGTGGCTCATTAGGCA-3′ and 5′-CCA ATCTTCGACTGGACTCTGT-3′), and GAPDH (5′-GTGGACCTGACCTGCCGTCT-3′ and 5′-GG AGGAGTGGGTGTCGCTGT-3′; Integrated DNA Technologies). Sequences were amplified using 2× SYBR Green Master Mix (Bio-Rad, Thermo Fisher). Data collection was accom plished using a Quantstudio 6 Flex real-time PCR machine (Thermo Fisher). Results for each gene were standardized against the expression of GAPDH transcripts.
Neurons intended for protein analysis were plated at a density of 5.0 × 10 6 cells per well of a Geltrex-coated (Thermo Fisher) six-well plate (5.2 × 10 4 cells/cm 2 ). At 51 days of differentiation, cells were infected with TB40/E-eGFP and allowed to persist for another 2-7 dpi. For 14 dpi samples, cells were grown for 84 days prior to infection. Collected cells were dislodged from their basement membrane using a P1000 pipette, pelleted by centrifugation (300 × g; 5 minutes), and frozen at -20°C until use. Cell pellets were treated with 50-150 µL of cold, pH 7.4 lysis buffer (150 mM NaCl, 50 mM Tris-HCl [pH 8.0], 1 mM EDTA, 1% Triton X-100 [vol/vol], 1% protease inhibitor [vol/vol], and 1% phosphatase inhibitor [vol/vol]) for 20 minutes and lysed via sonication (2 × 3 s pulses; 30% amplitude). Sample concentrations were determined using the Pierce bicinchoninic acid Assay Kit (Thermo Fisher) and a Glomax Microplate Reader (Promega). Utilizing 15-30 µg of protein (consistent within blots), sample contents were resolved by SDS-PAGE on a 4%-20% acrylamide gradient gel (Bio-Rad). Separated protein bands were transferred to a polyvinylidene difluoride membrane (Millipore) using a standard wet transfer procedure and a Mini TransBlot Cell (Bio-Rad). Using Intercept (PBS) Blocking Buffer (LI-COR), membranes were blotted to reduce the possibility of non-specific binding. Next, primary antibody solution (primary antibody, Intercept Blocking Buffer, 0.2% Tween-20 [vol/vol]) was applied, and membranes were incubated overnight at 4°C with agitation. Subsequently, membranes underwent three 5-minute washes with TBS-T (tris-buffered saline, 0.1% Tween-20) prior to applying secondary antibody solution (secondary antibody, Intercept Blocking Buffer, 0.2% Tween-20 [vol/vol], and 0.02% SDS [vol/vol]) for 30 minutes at room temperature. Following several washes (3× TBS-T and 1× TBS), blots were visualized using an Odyssey CLx fluorescent imaging system (LI-COR). Primary antibodies used in this study include chicken anti-TUJ1 (1:250-500; GeneTex), chicken anti-MAP2 (1:500-1,000; Invitrogen), mouse anti-MAP2 (1:1,000, Invitrogen), mouse anti-NF68 (1:500, Sigma), rabbit anti-beta actin (1:500, Pierce), mouse anti-Tau46
(1:250-500, Cell Signaling Technologies), rabbit anti-GFP (1:1,000, Invitrogen), mouse anti-IE1 (1:1,000), and mouse anti-pp28 (1:400). Anti-HCMV antibodies were generously provided by Tom Shenk. Secondary antibodies used in this study include donkey anti-rabbit 680RD (1:2,000-3,000, LI-COR), donkey anti-mouse 800CW (1:2,000-3,000, LI-COR), donkey anti-chicken 680RD (1:3,000, LI-COR), and donkey anti-chicken 800CW (1:3,000, LI-COR).
Protein loaded into each well for western blot analysis was normalized using total protein staining. In some instances, blots were stripped and re-probed, but the total protein stain was unchanged. Reused blots are indicated as such in the appropriate figure legend. Subsequently, for blots comparing mock to HCMV at multiple timepoints (Fig. 2J through L and 3F H, J and L), HCMV-treated samples were normalized to their mock counterpart within each timepoint. Due to this, no comparisons are made between timepoints. For blots comparing the effects of drug treatments on neurons, all protein values are normalized to the average of untreated, mock-infected values after total protein normalization.
## Statistical analysis
Data were generated from a minimum of three independent differentiations from each iPSC line, with the observer blinded to the treatment condition. All statistical analyses contained within this study were completed using the GraphPad Prism software suite. For all data sets, normality of distribution was established using the "Normality and Lognormality Tests" function, thereby allowing for appropriate selection of parametric vs nonparametric statistical tests. Should the data set fail any of the four applied assessments (D'Agostino and Pearson test, Anderson-Darling test, Shapiro-Wilk test, and Kolmogorov-Smirnov test), data were assessed in a nonparametric manner.
## References
1. Fowler, Stagno, Pass et al. (1992) "The outcome of congenital cytomegalovirus infection in relation to maternal antibody status" *N Engl J Med*
2. Zuhair, Smit, Wallis et al. (2019) "Estimation of the worldwide seroprevalence of cytomegalovirus: a systematic review and meta-analysis" *Rev Med Virol*
3. Sinzger, Grefte, Plachter et al. (1995) "Fibroblasts, epithelial cells, endothelial cells and smooth muscle cells are major targets of human cytomegalovirus infection in lung and gastrointestinal tissues" *J Gen Virol*
4. Mendelson, Monard, Sissons et al. (1996) "Detection of endogenous human cytomegalovirus in CD34+ bone marrow progenitors" *J Gen Virol*
5. Sindre, Tjøonnfjord, Rollag et al. (1996) "Human cytomegalovirus suppression of and latency in early hematopoietic progenitor cells" *Blood*
6. Zhuravskaya, Maciejewski, Netski et al. (1997) "Spread of human cytomegalovirus (HCMV) after infection of human hematopoietic progenitor cells: model of HCMV latency" *Blood*
7. Hahn, Jores, Mocarski (1998) "Cytomegalovirus remains latent in a common precursor of dendritic and myeloid cells" *Proc Natl Acad Sci*
8. Perlman, Argyle (1992) "Lethal cytomegalovirus infection in preterm infants: clinical, radiological, and neuropathological findings" *Ann Neurol*
9. Boppana, Ross, Shimamura et al. (2025)
10. Disorders, Study (2011) "Saliva polymerase-chain-reaction assay for cytomegalovirus screening in newborns" *N Engl J Med*
11. Barkovich, Lindan (1994) "Congenital cytomegalovirus infection of the brain: imaging analysis and embryologic considerations" *AJNR Am J Neuroradiol*
12. Lokensgard, Cheeran, Gekker et al. (1999) "Human cytomegalo virus replication and modulation of apoptosis in astrocytes" *J Hum Virol*
13. Lecointe, Héry, Janabi et al. (1998) "Productive replication of human cytomegalovirus in primary microglial cells" *Pathol Biol*
14. Spiller, Borysiewicz, Morgan (1997) "Development of a model for cytomegalovirus infection of oligodendrocytes" *J Gen Virol*
15. Wiley, Schrier, Denaro et al. (1986) "Localization of cytomegalovirus proteins and genome during fulminant central nervous system infection in an AIDS patient" *J Neuropathol Exp Neurol*
16. Cheeran, Hu, Ni et al. (2005) "Neural precursor cell susceptibility to human cytomegalovirus diverges along glial or neuronal differentiation pathways" *J Neurosci Res*
17. Teissier, Fallet-Bianco, Delezoide et al. (2014) "Cytomegalovi rus-induced brain malformations in fetuses" *J Neuropathol Exp Neurol*
18. Piccirilli, Gabrielli, Bonasoni et al. (2023) "Fetal brain damage in human fetuses with congenital cytomegalovirus infection: histological features and viral tropism" *Cell Mol Neurobiol*
19. Adelman, Rosas-Rogers, Schumacher et al. (2023) "Human cytomegalovirus induces significant structural and functional changes in terminally differentiated human cortical neurons" *mBio*
20. Luo, Schwartz, Fortunato (2008) "Neonatal neural progenitor cells and their neuronal and glial cell derivatives are fully permissive for human cytomegalovirus infection" *J Virol*
21. Egilmezer, Hamilton, Foster et al. (2024) "Human cytomegalovirus (CMV) dysregulates neurodevelopmental pathways in cerebral organoids" *Commun Biol*
22. D'aiuto, Maio, Heath et al. (2012) "Human induced pluripotent stem cell-derived models to investigate human cytomegalo virus infection in neural cells" *PLoS ONE*
23. Amin, Ercolini, Shahapure et al. (2011) "The elementary events underlying force generation in neuronal lamellipodia" *Sci Rep*
24. Gentile, Carrizales, Koleske (2022) "Control of synapse structure and function by actin and its regulators" *Cells*
25. Konietzny, Bär, Mikhaylova (2017) "Dendritic actin cytoskeleton: structure, functions, and regulations" *Front Cell Neurosci*
26. Hoffman, Cleveland, Griffin et al. (1987) "Neurofilament gene expression: a major determinant of axonal caliber" *Proc Natl Acad Sci*
27. Eyer, Peterson (1994) "Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-betagalactosidase fusion protein" *Neuron*
28. Sakaguchi, Okada, Kitamura et al. (1993) "Reduced diameter and conduction velocity of myelinated fibers in the sciatic nerve of a neurofilament-deficient mutant quail" *Neurosci Lett*
29. Yuan, Rao, Nixon (2012) "Neurofilaments at a glance" *J Cell Sci*
30. Binarová, Tuszynski (2019) "Tubulin: structure, functions and roles in disease" *Cells*
31. Knossow, Campanacci, Khodja et al. (2020) "The mechanism of tubulin assembly into microtubules: insights from structural studies"
32. Gardner, Hunt, Goodson et al. (2008) "Microtubule assembly dynamics: new insights at the nanoscale" *Curr Opin Cell Biol*
33. Al-Bassam, Ozer, Safer et al. (2002) "MAP2 and tau bind longitudinally along the outer ridges of microtubule protofilaments" *J Cell Biol*
34. Nishida, Matsumura, Tamura et al. (2023) "Effects of three microtubule-associated proteins (MAP2, MAP4, and Tau) on microtubules' physical properties and neurite morphology" *Sci Rep*
35. Akhmanova, Hoogenraad (2015) "Microtubule minus-end-targeting proteins" *Curr Biol*
36. Oakley, Paolillo, Zheng (2015) "Tubulin complexes in microtubule nucleation and beyond"
37. Tovey, Conduit (2018) "Microtubule nucleation by γ-tubulin complexes and beyond" *Essays Biochem*
38. Cooper (2000) "The Cell: A Molecular Approach"
39. Waites, Qu, Bartolini (2021) "The synaptic life of microtubules" *Curr Opin Neurobiol*
40. Barbosa, Serrat, Ferreira et al. (2014) "Neuronal mitochondrial trafficking impairment: the cause or a consequence of neuronal dysfunction caused by amphetamine-like drugs" *J Drug Alcohol Res*
41. Baas, Rao, Matamoros et al. (2016) "Stability properties of neuronal microtubules" *Cytoskeleton (Hoboken)*
42. Dent, Baas (2014) "Microtubules in neurons as information carriers" *J Neurochem*
43. Sinzger, Hahn, Digel et al. (2008) "Cloning and sequencing of a highly productive, endotheliotropic virus strain derived from human cytomegalovirus TB40/E" *J Gen Virol*
44. Umashankar, Petrucelli, Cicchini et al. (2011) "A novel human cytomegalovirus locus modulates cell type-specific outcomes of infection" *PLoS Pathog*
45. Yong, Hunter-Chang, Stepanova et al. (2021) "Axonal spheroids in neurodegeneration" *Mol Cell Neurosci*
46. Figueroa-Masot, Hetman, Higgins et al. (2001) "Taxol induces apoptosis in cortical neurons by a mechanism independent of Bcl-2 phosphorylation" *J Neurosci*
47. Ren, Zhao, Chang et al. (2018) "Paclitaxel suppresses proliferation and induces apoptosis through regulation of ROS and the AKT/MAPK signaling pathway in canine mammary gland tumor cells" *Mol Med Rep*
48. Takadera, Nakajima, Kanai (2010) "Colchicine-induced apoptosis was prevented by glycogen synthase kinase-3 inhibitors in PC12 cells" *Cell Mol Neurobiol*
49. Kristensen, Noer, Gramsbergen et al. (2003) "Colchicine induces apoptosis in organotypic hippocampal slice cultures" *Brain Res*
50. Furey, Astar, Walsh (2021) "Human cytomegalovirus exploits TACC3 to control microtubule dynamics and late stages of infection" *J Virol*
51. Wofford, Mccusker, Green et al. (2020) "Betaherpesvirus assembly and egress: Recent advances illuminate the path"
52. Long, Qiu, Zhang et al. (2021) "Insight for immunotherapy of HCMV infection" *Int J Biol Sci*
53. Ogawa-Goto, Tanaka, Gibson et al. (2003) "Microtubule network facilitates nuclear targeting of human cytomegalovirus capsid" *J Virol*
54. Lyon, Kalejta (2018) "HCMV assembly is totally tubular" *Dev Cell*
55. Procter, Banerjee, Nukui et al. (2018) "The HCMV assembly compartment is a dynamic golgi-derived MTOC that controls nuclear rotation and virus spread" *Dev Cell*
56. Sison, Brien, Johnson et al. (2019) "Human cytomegalovirus disruption of calcium signaling in neural progenitor cells and organoids" *J Virol*
57. Sun, Chiuppesi, Chen et al. (2020) "Modeling human cytomegalovirus-induced microcephaly in human iPSC-derived brain organoids" *Cell Rep Med*
58. Mokry, Brien, Adelman et al. (2022) "Nitric oxide attenuates human cytomegalovirus infection yet disrupts neural cell differentiation and tissue organization" *J Virol*
59. Mody, Marvin, Hynds et al. (2023) "Cytomegalovirus infection induces alzheimer's disease-associated alterations in tau" *J Neurovirol*
60. Lurain, Hanson, Martinson et al. (2013) "Virological and immunological characteristics of human cytomegalovirus infection associated with alzheimer disease" *J Infect Dis*
61. O'brien, Mokry, Schumacher et al. (2023) "Neutralizing antibodies with neurotropic factor treatment maintain neurodevelopmental gene expression upon exposure to human cytomegalovirus" *J Virol*
62. Ijezie, Dowd, Kuan et al. (2023) "HCMV infection reduces nidogen-1 expression, contributing to impaired neural rosette development in brain organoids" *J Virol*
63. Leung, Hui, Kraus (2015) "Colchicine--Update on mecha nisms of action and therapeutic uses" *Semin Arthritis Rheum*
64. Foland, Dentler, Suprenant et al. (2005) "Paclitaxel-induced microtubule stabilization causes mitotic block and apoptotic-like cell death in a paclitaxel-sensitive strain of Saccharomyces cerevisiae" *Yeast*
65. Yang, Horwitz (2017) "Taxol: the first microtubule stabilizing agent" *IJMS*
66. Leung, Cassimeris (2019) "Reorganization of paclitaxel-stabilized microtubule arrays at mitotic entry: roles of depolymerizing kinesins and severing proteins" *Cancer Biol Ther*
67. Flynn (2013) "The cytoskeleton and neurite initiation" *Bioarchitecture*
68. He, Ferguson, Patterson et al. (2018) "Chapter 3 -Neurite Development and Neurotoxicity"
69. Cross, Huber, Silverman et al. (2021) "Intranasal paclitaxel alters alzheimer's disease phenotypic features in 3xTg-AD mice" *J Alzheimers Dis*
70. Zhang, Maiti, Shively et al. (2005) "Microtubule-binding drugs offset tau sequestration by stabilizing microtubules and reversing fast axonal transport deficits in a tauopathy model" *Proc Natl Acad Sci*
71. Liu, Liu, Sun (2023) "Low-dose taxol promotes neuronal axons extension and functional recovery after spinal cord injury" *Mediators Inflamm*
72. Brizić, Hiršl, Britt et al. (2018) "Immune responses to congenital cytomegalovirus infection" *Microbes Infect*
73. Guaita, Jimoh, Barrall et al. (2025) "Characteri zation of human cytomegalovirus infection dynamics in human microglia" *J Gen Virol*
74. Klein, Boenert, Lange et al. (2022) "Glia from the central and peripheral nervous system are differentially affected by paclitaxel chemotherapy via modulating their neuroinflammatory and neuroregenerative properties" *Front Pharmacol*
75. Grant, Sullivan, Wentworth et al. (2023) "Microglia are implicated in the development of paclitaxel chemotherapy-associated cognitive impairment in female mice" *Brain Behav Immun*
76. Riedel, Jou, Lai et al. (2014) "Functional and pharmacological analysis of cardiomyocytes differentiated from human peripheral blood mononuclear-derived pluripotent stem cells" *Stem Cell Reports*
77. Chambers, Fasano, Papapetrou et al. (2009) "Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling" *Nat Biotechnol*
78. Baldick, Marchini, Patterson et al. (1997) "Human cytomegalo virus tegument protein pp71 (ppUL82) enhances the infectivity of viral DNA and accelerates the infectious cycle" *J Virol*
79. Mokry, Brien, Adelman et al. (2021) "Nitric oxide attenuates human cytomegalovirus infection yet disrupts neural cell differentiation and tissue organization" *bioRxiv*
80. Meijering (2010) "Neuron tracing in perspective" *Cytometry A*
81. Abramoff (2004) "Image Processing Image Processing with ImageJ"
82. Meijering, Jacob, Sarria et al. (2004) "Design and validation of a tool for neurite tracing and analysis in fluorescence microscopy images" *Cytometry A*
83. Wang, Seed (2003) "A PCR primer bank for quantitative gene expression analysis" *Nucleic Acids Res*
84. Wang, Spandidos, Wang et al. (2012) "PrimerBank: a PCR primer database for quantitative gene expression analysis, 2012 update" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671142&blobtype=pdf | # Genetic structure and phylogenetic analysis of grapevine leafroll-associated virus-1 (Ampelovirus univitis) in different grape-producing regions of Iran
Nesa Razavi, Davoud Koolivand, Masoud Naderpour, Milad Yousefi
## Abstract
Grapevine leafroll-associated virus 1 (GLRaV-1) is a major viral pathogen within the genus Ampelovirus and a key contributor to grapevine leafroll disease, posing significant threats to viticulture worldwide. Despite its importance, the genetic structure and diversity of GLRaV-1 in Iran have remained largely unexplored. In this study, a total of 325 symptomatic grapevine samples were collected from 5 major grape-produc ing provinces of Iran and screened using immunoassay and reverse transcription-PCR (RT-PCR) for GLRaV-1. Nine positive samples were subjected to sequencing of the coat protein (CP) gene. Phylogenetic analyses of CP gene sequences and publicly available full-genome sequences, performed using MEGA 11, revealed three distinct phylogenetic clades among Iranian isolates. Recombinant sequences were identified and excluded to avoid phylogenetic bias. Population genetic analyses indicated high haplotype diversity and strong purifying selection across the genome. Significant population structure and restricted gene flow were observed, consistent with geographic isolation of viral populations. Molecular clock analysis suggested recent divergence of GLRaV-1 lineages within Iran. These findings provide new insights into the evolutionary dynamics of GLRaV-1 and its population structure in Iranian vineyards, with implications for virus spread, surveillance, and disease management.IMPORTANCE Grapevine leafroll-associated virus 1 (GLRaV-1) is a widespread and economically important virus affecting grapevines. This study is the first to investigate the genetic diversity and population structure of GLRaV-1 across major grape-growing regions of Iran. We discovered a high level of genetic variation and geographically structured virus populations, which may reflect localized transmission and limited movement of infected material. Understanding this diversity is crucial for improving diagnostic strategies and managing the spread of grapevine leafroll disease. Our findings support the need for region-specific disease control efforts and contribute to the global understanding of Ampelovirus evolution.
Despite its economic importance, grapevine production faces numerous challenges, particularly from biotic stressors such as fungi, bacteria, nematodes, and viruses. Among these, viral infections represent a major threat to vineyard health, affecting vine vigor, reducing yield, and compromising fruit quality. More than 80 viruses have been reported to infect grapevines, with some of the most economically devastating species including Grapevine fanleaf virus (GFLV), Arabis mosaic virus (ArMV), and the Grapevine leafroll-asso ciated viruses (GLRaVs) (2). Of these, grapevine leafroll disease (GLRD) is one of the most destructive viral diseases, first reported in the mid-19th century in California (3). It is estimated that GLRD can cause a 30%-50% reduction in grape yield, significantly impacting fruit color, sugar content, and overall marketability (4,5).
The causative agents of GLRD, known as Grapevine leafroll-associated viruses (GLRaVs), belong to the family Closteroviridae. This family includes several genera such as Ampelovirus, Closterovirus, Crinivirus, and Velarivirus (6). To date, at least 13 GLRaV species have been identified, with ongoing research indicating that this number may continue to increase as advanced molecular detection techniques become more widely implemen ted (7). Among these, Grapevine leafroll-associated virus-1 (GLRaV-1) is considered a major pathogen in viticulture, ranking second in economic importance after GLRaV-3 (8).
GLRaV-1 possesses a positive-sense single-stranded RNA genome of approximately 18.7-18.9 kb, encapsidated within flexuous, filamentous virions (9). The viral genome comprises nine open reading frames (ORFs) that encode proteins responsible for replication, movement, encapsidation, and vector transmission (10). The virus spreads predominantly through vegetative propagation, including grafting, and is transmitted by mealybug species such as Planococcus ficus and Pseudococcus longispinus in a semi-per sistent manner (11). These transmission routes facilitate its rapid dissemination within vineyards, leading to significant economic losses in both table grape and wine grape industries.
GLRaV-1 has been reported in major grape-producing countries, including Italy, France, Spain, Greece, Germany, Switzerland, Slovakia, Tunisia, and Turkey (12)(13)(14)(15)(16)(17). In Iran, extensive surveys and molecular studies have confirmed its widespread occur rence, particularly in the Fars province, one of the country's key viticultural regions (2,18). Phylogenetic analyses have revealed substantial genetic diversity among GLRaV-1 isolates, with multiple lineages circulating globally.
Recent research from the United States, China, Italy, Turkey, and Russia have demonstrated that GLRaV-1 isolates cluster into distinct phylogenetic groups (16). However, studies have not established a clear correlation between genetic diversity and geographic distribution, suggesting that other factors, such as vector specificity and vineyard management practices, may influence viral diversity (4,8,(19)(20)(21). Despite multiple reports on GLRaV-1 worldwide, comprehensive studies on its genetic diversity and population structure in Iran remain limited, underscoring the need for further investigation.
Given the significant impact of GLRaV-1 on grapevine yield and quality, and the lack of phylogenetic data from Iran, this study investigates the genetic diversity of GLRaV-1 isolates from Iranian vineyards. By sequencing key genomic regions and comparing them with global isolates, we aim to clarify the virus's evolutionary patterns and inform more effective management strategies.
## MATERIALS AND METHODS
## Sampling
During the 2023 growing season, a total of 325 samples were collected from symp tomatic grapevines across five Iranian provinces: Kohgiluyeh and Boyer-Ahmad (136 samples), West Azerbaijan (68 samples), Khorasan Razavi (59 samples), Qazvin (34 samples), and Fars (28 samples) (Tables 1 and2). From each vine, one fully developed compound leaf was sampled from the mid-canopy. Leaf samples were collected from grapevines showing symptoms such as mosaic patterns, leaf rolling, chlorosis, reddening, deformation, stunted growth, and uneven berry size. Samples were transported in cool boxes and stored at 4°C prior to analysis. Each sample was assigned a unique number based on the time and date of collection and immediately frozen at -80°C for subse quent total RNA extraction. Details of the collection locations, including GPS coordinates and number of samples per region, are provided in Table 2.
## Serological test and RNA extraction and RT-PCR
Double-antibody sandwich enzyme-linked immunosorbent assay (DAS-ELISA) was performed using virus-specific antibodies against GLRaV-1 and GLRaV-3 (Bioreba AG, Reinach, Switzerland) according to the manufacturer's instructions. Leaf samples (0.5 g) were ground in extraction buffer (PBST containing 2% PVP-40 and 0.05% Tween-20) at a ratio of 1:10 (wt/vol) using a chilled mortar and pestle. Microtiter plates were coated with 200 µL of IgG antibody diluted 1:200 in carbonate-bicarbonate buffer (pH 9.6) and incubated overnight at 4°C. After three washes with PBST, 200 µL of plant extract was added to each well and incubated for 2 h at 37°C. Plates were washed again before adding 200 µL of alkaline phosphatase-conjugated antibody diluted 1:200 in PBST, followed by incubation for 2 h at 37°C. After a final wash, 200 µL of p-nitro phenyl phosphate (pNPP, 1 mg/mL in diethanolamine buffer, pH 9.8) was added, and the reaction was developed for 60 min at room temperature in the dark. Absorbance was measured at 405 nm using a Microplate Reader. Samples with absorbance values exceeding the mean of healthy controls by at least three times the standard deviation were considered positive. Total RNA was extracted from symptomatic grapevine leaf tissues using the GeneAll Plant RNA Extraction Kit (GeneAll Biotechnology, South Korea) according to the manufacturer's protocol. The quantity and purity of the extracted RNA were assessed using a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). The RNA concen tration and absorbance ratios at 260/280 nm were recorded to evaluate RNA quality. Samples with A260/A280 ratios between 1.8 and 2.0 were considered suitable for downstream applications.
## cDNA synthesis
Reverse transcription (RT) was carried out using the Easy cDNA Synthesis Kit (Parstous, Iran) following the manufacturer's instructions. The total reaction volume was 10 µL, and the thermal cycling conditions for RT included 25°C for 10 min primer annealing, 47°C for 60 min reverse transcriptase enzyme activity, and 85°C for 5 min enzyme inactivation. The reaction was performed in a T100 Thermal Cycler (Bio-Rad, USA).
## Polymerase chain reaction amplification
The incidence of GLRaV-1 was determined using a pair of primers (22), to amplify a 232 bp DNA fragment. Subsequently, PCR amplification of the GLRaV-1 coat protein (CP) gene was performed using specific primers (5,19): Forward primer (GLRaV-1 CP/F): CG CGCTTGCAGAGTTTAAGTGGTT and Reverse primer (GLRaV-1 CP/R): TCCGTGCTGCATTGC AACTTTCTC. PCRs were performed in a total volume of 25 µL, consisting of 1 µL cDNA template, 12.5 µL 2 × PCR Master Mix (Ampliqon, Denmark), 1 µL of each primer (10 µM), and 9.5 µL nuclease-free water. The thermal cycling conditions for PCR were as follows: initial denaturation at 94°C for 3 min, 40 cycles of: 94°C for 30 s (denaturation), 58°C for 30 s (annealing), 72°C for 45 s (extension), final extension at 72°C for 5 min PCR products were analyzed by 1% agarose gel electrophoresis in 1 × TAE buffer and visualized under UV light after staining with GelRed (Biotium, USA). Amplified products were subsequently purified and sent for sequencing to assess the genetic diversity of GLRaV-1 isolates.
## Phylogenetic analysis
Phylogenetic analyses of 24 complete GLRaV-1 sequences available in the GenBank database worldwide were carried out using NCBI-BLAST (http://www.ncbi.nlm.nih.gov/ BLAST/) (Table 1). Sequence alignment was performed with MEGA 11 software (23), utilizing the Clustal W program (24). In addition, the obtained nucleotide sequences of PCR products were compared with reference sequences available in the NCBI GenBank database using the BLAST algorithm (http://www.ncbi.nlm.nih.gov/BLAST/). The nine newly obtained nucleotide sequences were aligned with 61 previously reported GLRaV-1 sequences available in the NCBI GenBank database (Table 1). The complete genome alignment was 17,649 nt in length after trimming. For the coat protein (CP) data set, the same 9 new isolates and 61 previously reported isolates from different geo graphical origins were included, resulting in an alignment length of 732 nt. Multiple sequence alignment was performed using MEGA 11 software (25) and the ClustalW algorithm (24). After alignment, necessary edits were applied for further analysis. The evolutionary distances were computed using the Jukes-Cantor model. The best-fit nucleotide substitution model for phylogenetic tree construction was selected using MEGA 11. Phylogenetic relationships among isolates were inferred using the Neighbor-Joining method ( 26) with the Tamura 3-parameter model (25), implemented in MEGA 11. Bootstrap resampling with 1,000 replicates was applied to assess the robustness of the phylogenetic relationships, and branches with bootstrap values below 50% were collapsed to improve clarity. The phylogenetic tree was generated based on the coat protein (CP) and complete genome sequences. Recombinant groups and isolates with <0.5% genetic distance were omitted to enhance the resolution of clustering patterns. The final phylogenetic tree provided insights into the genetic diversity of GLRaV-1 isolates from Iran in comparison to global isolates, and detailed information on Iranian isolates is presented in Table 1. Furthermore, a pairwise nucleotide sequence identity matrix was generated using MEGA 11 to quantify the genetic similarity among the studied isolates.
## Recombination analysis and phylogeny
To detect potential recombination events and phylogenetic anomalies, the Recombina tion Detection Program 4 (RDP4) was used with its comprehensive suite of analytical tools, including RDP, Chimaera, MaxChi, Bootscan, Siscan, GENECONV, and 3Seq. The analysis was conducted using default parameters, applying a P-value threshold of 0.05 to identify significant recombination signals. Any anomalies that were supported by fewer than five of these detection methods and had a Bonferroni-corrected P-value below 0.05 were excluded from further consideration. To validate the parent/recombinant relationships of the two recombinant strain clusters, the sequence alignments were divided into regions that either contained or excluded the recombinant segments.
## Population genetic parameters
The genetic differentiation of the coat protein (CP) gene and complete genomes across diverse populations was analyzed using DnaSP v6.10.01. Several genetic parameters were examined, including haplotype diversity (Hd), the number of polymorphic sites (S), total number of mutations (η), average nucleotide differences (k), pairwise nucleotide diversity (π), and the ratio of non-synonymous to synonymous substitutions (dN/dS). To assess neutral selection, Tajima's D (27) and Fu and Li's D and F tests (28) were applied. To investigate population differentiation, statistical tests such as KS, Z, Snn, and FST (fixation index) were performed using 1,000 replicates (29). The pairwise nucleotide sequence identity was calculated in MEGA 11 to evaluate genetic similarity between isolates.
## Molecular dating analysis
Divergence times of the GLRaV-1 population along with four closely related Ampe lovirus species, including grapevine leafroll-associated virus-3 (GLRaV-3), grapevine leafroll-associated virus-13 (GLRaV-13), grapevine leafroll-associated virus-4 (GLRaV-4), and little cherry virus 2 (LChV2) was estimated (Table 2). Additionally, grapevine leafroll-associ ated virus-2 (GLRaV-2), a Closterovirus species, was used as an outgroup for comparative analysis. To estimate divergence times, internal node ages were analyzed, and a TimeTree was reconstructed using the RelTime-ML computational method under the Tamura 3-parameter model (30) within MEGA 11 software. The time to the most recent common ancestor (TMRCA) was calculated using the default calibration method (31). A discrete Gamma distribution with five categories (+G, parameter = 2.2907) was applied to model evolutionary rate variation across different sites, while 8.65% of sites were classified as evolutionarily invariable ([+I]). This analysis included 35 nucleotide sequences, incorpo rating 1st, 2nd, and 3rd codon positions, as well as noncoding regions, resulting in a final data set of 6,302 positions. The RelTime method estimated divergence times exclusively for the ingroup clade, as it does not assume that the evolutionary rates of the ingroup apply to the outgroup. Therefore, divergence times for outgroup nodes were not calculated. All estimated divergence times are relative since no external calibrations were applied.
## RESULTS
## Field observation
Observations from vineyard surveys in West Azerbaijan, Fars, Qazvin, Kohgiluyeh and Boyer-Ahmad, and Razavi Khorasan provinces (2019-2023) confirmed the presence of symptoms consistent with grapevine leafroll disease, including leaf reddening, chlorotic spotting, curling, deformation, and reduced leaf size. In some vineyards, leaf curling was particularly prevalent. Table 2 provides the detailed mapping of newly identified GLRaV-1 haplotypes to their respective collection sites, GPS coordinates, and collection dates. Other plant diseases, such as powdery mildew, were also observed in several vineyards.
## ELISA detection
Out of the 325 grapevine samples collected across West Azerbaijan, Fars, Qazvin, Kohgiluyeh and Boyer-Ahmad, and Razavi Khorasan provinces (2019-2023), all samples were tested for GLRaV-1+3 infection using DAS-ELISA with virus-specific antibodies (Bioreba AG, Switzerland) according to the manufacturer's instructions. Among these, 59 samples (18.2%) tested positive for GLRaV-
## PCR amplification
PCR reactions were performed using two specific forward and reverse primer pairs (R) and (F) targeting the coat protein gene of GLRaV-1, the virus associated with grapevine leafroll disease. This process successfully amplified and detected DNA fragments of 734 base pairs. Nine positive samples were sequenced based on regions and symptoms, and the sequences were uploaded to GenBank NCBI, under the accession numbers PP213141 to PP213148 and PP182259. The results of recombinant event analysis using the RDP4 program, which includes several recombinant detection algorithms such as RDP, Chimaera, Maxchi, Bootscan, GENECONV, and SISCAN, revealed the presence of recombinant strains among the studied isolates. Among the newly identified isolates in this study, isolate PP182259, detected in the Kohgiluyeh and Boyer-Ahmad province, was found to be a recombinant virus associated with grapevine leafroll associated virus-1, with mutations having occurred. In other words, the recombination regions in this isolate were found to be swapped. The parental strain of the newly identified isolate PP213143 from Kohgiluyeh and Boyer-Ahmad is of unknown origin, and recombinant events in this isolate were detected as positive using three recombinant detection methods (M, C, T), increasing the likelihood of true recombination. Consequently, isolate PP182259, identified as a recombinant sequence, should be excluded from phylogenetic analysis to avoid distortion of the results (32).
## Phylogenetic analysis
The phylogenetic tree constructed from the analyzed sequences based on the complete genome revealed three distinct major clades, each strongly supported by high bootstrap values (98-100), indicating robust phylogenetic relationships and genetic differentiation among isolates (Fig. 1). Clade I, the first major clade consists of isolates such as KY827404, OQ029645, MT953195, MT953193, OQ029678, PQ521015, and PQ521017, all forming a well-supported monophyletic group with a bootstrap value of 100 (Fig. 1; Table 1). Clade II, the second major clade encompasses a diverse set of isolates, including MZ344577, OP718744, MG925332, KU674797, MT953194, MH545961, and KU674796 (Fig. 1; Table 1).
The phylogenetic analysis of GLRaV-1 based on the CP gene sequences revealed three well-supported clades, indicating significant genetic diversity among the isolates (Fig. 2; Table 1). The newly identified Iranian isolates (bold accession numbers) clustered within different groups, suggesting multiple evolutionary origins of GLRaV-1 in this region (Fig. 2; Table 1). The first major clade (Clade I) contains the Iranian isolates PP213141-PP213148 and PP182259, which are closely related and form a distinct monophyletic group (Fig. 2). While these isolates do not all belong to the same subgroup, their placement within Clade I suggests they share a broader evolutionary lineage. This pattern may reflect multiple introductions, local divergence, or incomplete sampling from other regions (Fig. 2). All Iranian isolates were grouped within the first major clade (Clade I), which contains several sub-clades. Within Clade I, isolate PP213147 grouped in a distinct sub-clade, isolate PP213148 is placed in another separate sub-clade, and the remaining Iranian isolates are grouped in a different sub-clade within Clade I (Fig. 2; Table 1). This sub-clade structure indicates that, although all Iranian isolates belong to the same major clade, they exhibit genetic differentiation that may reflect localized divergence or other evolutionary processes (Fig. 2 and3; Table 1).
## Genetic diversity and population structure analysis
Genetic diversity analysis of the complete genome sequences revealed substantial variation among the three clades. Haplotype diversity was high in all clades, with Clades II and III showing Hd = 1 and Clade I slightly lower (Hd = 0.972). Polymorphic sites (S), nucleotide diversity (π), and average nucleotide differences (k) were highest in Clade III (S = 3,562, π = 0.09334, k = 1,523.429), indicating greater genetic variation, followed by Clade I and Clade II. Selection pressure analysis (ω = dN/dS) showed values below 1 for all clades (Clade I = 0.49, Clade II = 0.54, Clade III = 0.58), consistent with predominant purifying selection, while Clade III had the highest dN (0.079) and dS (0.135), suggesting stronger evolutionary diversification compared to the more conserved Clade II (dN = 0.043, dS = 0.079) (Table 3).
Population genetic analysis of GLRaV-1 based on CP gene sequences revealed high genetic diversity and strong purifying selection across three clades. Clade I (N = 34) exhibited the highest nucleotide diversity (π = 0.0712), average nucleotide differences (k = 28.41), segregating sites (S = 135), and total mutations (η = 159), with Hd = 0.996 and ω = 0.108, indicating a highly diverse yet evolutionarily constrained population (Tables 1 and4). Clade II (N = 8) showed lower nucleotide diversity (π = 0.0422) and fewer mutations (S = 53, η = 60), but the highest haplotype diversity (Hd = 1.00) and ω = 0.21, suggesting potential relaxed purifying or episodic positive selection in certain CP regions (Table 3). Clade III (N = 11) displayed moderate diversity (Hd = 0.982, π = 0.0628, S = 83, η = 94) with the lowest ω (0.064), reflecting strong purifying selec tion and evolutionary conservation of the CP gene (Table 3). Overall, ω values < 1 in all clades indicate that purifying selection predominates, removing deleterious amino acid changes, while differences among clades suggest ongoing evolutionary adaptation influenced by host interactions and vector-mediated transmission. Clade I represents a highly diverse lineage comprising all newly characterized Iranian isolates, highlighting the importance of continuous genetic surveillance for effective virus management.
## Population differentiation and genetic structure analysis
Analysis of neutrality and population differentiation of GLRaV-1 revealed insights into evolutionary dynamics among the three clades. Tajima's D values were negative in all clades, indicating an excess of low-frequency polymorphisms consistent with past population expansion or purifying selection, with Clade I showing the most negative value (-1.44256) though none were statistically significant (Table 4). Fu and Li's D and F** values showed similar trends, with Clade I and II exhibiting more negative values, suggesting historical expansion or selective constraints, while Clade III appeared relatively stable (Table 4). Genetic differentiation analysis based on the complete genome indicated substantial divergence among clades, with the highest FST between Clade I and II (0.547), followed by Clade I/III (0.457) and Clade II/III (0.389), reflecting limited gene flow (Table 5). Additional tests (KS, KST, Z, Snn, FST) confirmed significant population structure, with KS and Z values high for all comparisons (Clade I/II: KS = 6.79, Z = 3.19; Clade I/III: KS = 6.94, Z = 3.11; Clade II/III: KS = 6.84, Z = 3.12) and all S nn = 1.00, with highly significant P-values (P = 0.0000), supporting strong genetic differentiation and limited gene flow among the clades (Table 5).
All pairwise comparisons based on the CP sequences yielded highly significant P-values (P = 0.000) across multiple differentiation metrics, indicating strong genetic divergence between the clades. KS and KST values indicated significant differentiation in sequence composition, with the highest observed between Clade 1 and Clade 3 (KS = 3.22, KST = 0.08), followed by Clade 1 vs Clade 2 (KS = 3.19, KST = 0.06). The Z statistic was consistently high across all comparisons, particularly between Clade 1 and Clade 3 (Z = 5.40), indicating strong genetic separation. The S nn test yielded values of 1.00 for all comparisons, confirming nearly complete genetic differentiation, where sequences within each clade are more closely related to each other than to sequences in other clades. FST values were highest between Clade 2 and Clade 3 (FST = 0.622), followed by
## Sliding-window analysis of polymorphism
A sliding-window analysis of nucleotide diversity (π) across the complete genome of GLRaV-1 revealed fluctuations in genetic variability, with distinct peaks and valleys indicating regions of high and low polymorphism (Fig. 4). Overall, π remained relatively low across most of the genome, with moderate variability in the 5′ and 3′ regions, suggesting conserved sequences likely due to functional or structural constraints. Notably, a peak around nucleotide position 10,000 indicates a region of heightened polymorphism, potentially representing a recombination hotspot, positive selection, or relaxed evolutionary constraints. Beyond position 15,000, diversity levels show an upward trend, highlighting another region with increased genetic variation (Fig. 4). A sliding-window analysis of nucleotide diversity (π) across the GLRaV-1 CP gene revealed fluctuations in sequence variability along the gene (Fig. 4). Nucleotide diversity ranged from approximately 0.05 to 0.18, with the highest variability observed between 50 and 150 bp, suggesting potential hotspots of polymorphism, while a conserved region was detected between 250 and 350 bp, likely reflecting functional or structural constraints. A slight increase in diversity after 400 bp indicates minor accumulation of polymorphisms. High-variability regions may be under selective pressure related to host adaptation or immune evasion, whereas low-diversity regions likely maintain essential viral structure and replication efficiency. Overall, this analysis highlights a balance between evolutionary constraints and adaptation, providing insights for molecular diagnostics, antiviral strategies, and resistance breeding programs (Fig. 4).
## Time to the most recent common ancestor estimation
A time-calibrated phylogenetic tree of Ampelovirus members was reconstructed using the RelTime method in MEGA 11, based on complete or near-complete genome sequences, with GLRaV-2 (JX559644) as the outgroup (Fig. 5; Table 2). The analysis resolved five major monophyletic clades corresponding to GLRaV-1, GLRaV-3, GLRaV-4, GLRaV-13, and LChV-2. The GLRaV-1 clade, including isolates such as OQ029645, KY827404, NC_016509, and MH545961, exhibited a relative divergence time of ~0.20, indicating recent diversification (Tables 6 and7). In contrast, GLRaV-3 and GLRaV-13 showed deeper divergence (~0.26-0.32 and ~0.30, respectively), while GLRaV-4 formed a compact cluster with shallow divergence (≤ 0.10), suggesting low sequence variabil ity or recent expansion. LChV-2 diverged at ~0.45, and the most basal split occurred between GLRaV-2 and all other Ampelovirus taxa (~0.80). Bootstrap support values were generally high (>0.90), confirming the robustness of the inferred phylogeny and supporting current taxonomic groupings within the genus Ampelovirus.
To estimate the evolutionary age of GLRaV-1, the time to its most recent common ancestor (TMRCA) was calculated using the RelTime-derived relative divergence (D = 0.20) and a reported Closteroviridae substitution rate (μ = 1 × 10⁻⁴ substitutions/site/ year), suggesting that GLRaV-1 likely emerged approximately 1,000 years ago (~1,025 C.E. CE), potentially coinciding with historical grapevine domestication and viticulture expansion. GLRaV-1 exhibited the lowest mean patristic distance (0.20), serving as a baseline for comparison. GLRaV-2, used as the outgroup, had the highest mean patristic distance (0.80), indicating greater evolutionary divergence from GLRaV-1. Among other strains, GLRaV-4 showed the smallest distance (0.10), while GLRaV-13 (0.30) and GLRaV-3 (0.32) were slightly more distant, and LChV-2 had the largest divergence (0.45), with ratios of 1.50, 1.60, and 2.25 relative to GLRaV-1, respectively (Table 6). These results indicate varying levels of genetic divergence among the strains, with GLRaV-1 being most closely related to the other lineages, while GLRaV-2 represents a more distant evolutionary lineage.
## DISCUSSION
Ampeloviruses, particularly GLRaV-1, represent a major challenge to global viticulture due to their wide distribution, significant economic impact, symptom overlap with nutrient deficiencies, and other important factors such as vector transmission and vegetative propagation. The precise identification of GLRaV-1 and the study of its genetic variability are essential for effective disease management and epidemiological surveil lance. Understanding viral population structure and evolutionary dynamics also aids in tracing epidemic pathways and anticipating the emergence of novel variants (33). In this study, symptomatic grapevine samples were collected from five major grapegrowing provinces in Iran. Consistent with previous reports, visual symptoms varied between red and white cultivars and changed seasonally. Molecular detection confirmed GLRaV-1 in several regions, with Khorasan Razavi showing the highest infection rate, while no virus was detected in Fars and Qazvin, possibly due to primer mismatch from sequence divergence in the targeted region. Phylogenetic analysis of coat protein (CP) and complete genome sequences revealed three major clades, indicating a complex phylogeographic structure. Several Iranian isolates formed a distinct lineage, highlight ing substantial regional diversity and the likelihood of localized viral evolution. This pattern is consistent with earlier studies reporting limited gene flow and geographically structured GLRaV-1 populations (19,34,35).
Genetic diversity analysis based on the complete genome sequences showed that Clade III exhibited the highest nucleotide and haplotype diversity, as well as signs of diversifying selection (ω = 0.58). In contrast, Clade I was under stronger purifying selection (ω = 0.49), reflecting evolutionary constraints on essential viral genes. High F ST values between clades (0.389-0.547) confirmed significant genetic differentiation and low inter-population gene flow. Statistical measures of genetic differentiation (K S , Z, S nn , and F ST ) confirmed the presence of three genetically isolated GLRaV-1 popula tions. These clades appear to have followed distinct evolutionary trajectories, shaped by geographic separation, host genotype preferences, and vector specificity. Recombination was limited, reinforcing the clade structure and supporting the observed patterns of genetic partitioning. Population genetic parameters derived from CP sequences further corroborated these trends. The Iranian group displayed the highest mutation rate, number of polymorphic sites, and mean nucleotide divergence. Such variability may increase The TimeTree was constructed using the RelTime-ML method under the Tamura 3-parameter model (30) in MEGA 11 software. The analysis included GLRaV-1 and four closely related Ampelovirus species:
GLRaV-3, GLRaV-13, GLRaV-4, and LChV-2, with GLRaV-2 (Closterovirus species) used as an outgroup.
Divergence times were estimated based on internal node ages with the time to the most recent common ancestor (TMRCA) calculated using the default calibration method (31). A discrete Gamma distribution (+G, parameter = 2.2907) was applied to model rate variation across sites, and 8.65% of sites were considered invariable (+I). The analysis involved 35 nucleotide sequences, including codon positions 1st, 2nd, 3rd, and noncoding regions, with a total of 6,302 positions. Divergence times for outgroup nodes were not calculated. All estimates are relative, as no external calibrations were applied. the likelihood of recombination and contribute to rapid virus evolution. In fact, one isolate (PP182259) from Kohgiluyeh and Boyer-Ahmad was identified as a recombinant, highlighting the dynamic nature of GLRaV-1 evolution in Iran. This finding is consistent with the known high mutation and recombination rates in positive-sense RNA viruses (36,37).
Neutrality tests (Tajima's D, Fu and Li's D*/F*) yielded negative values across all Iranian clades, indicative of recent population expansion or bottlenecks. These may be attributed to vector transmission, propagation practices, or environmental stress. In contrast, positive neutrality values in European and East Asian populations point to evolutionary stability and reduced selection pressure.
Sliding-window analysis identified nucleotide diversity hotspots within the GLRaV-1 genome, indicating regions potentially under selection pressure. These variable regions may contribute to host adaptation, immune evasion, or vector interactions and merit further functional characterization. The time-calibrated phylogenetic reconstruc tion provided new insights into the evolutionary dynamics of Ampelovirus members, particularly GLRaV-1. The topology and divergence estimates indicate that GLRaV-1 is a recently diversified lineage, with shallow node heights and tight clustering sug gesting rapid evolutionary radiation, likely driven by host adaptation, regional spread, or vector-mediated transmission. In contrast, GLRaV-3 exhibited deeper divergence, consistent with long-term co-evolution with grapevine or historical geographic dispersal.
GLRaV-13 and LChV-2 displayed intermediate divergence, while GLRaV-4 formed a compact cluster indicative of low genetic variability, possibly reflecting a narrow host range, recent emergence, or limited geographic sampling. The RelTime-derived divergence estimates, though relative, offer a temporal framework for evaluating Ampelovirus evolution in the absence of absolute calibration points. These patterns may reflect historical drivers such as clonal grapevine propaga tion, international plant trade, and the spread of mealybug and scale insect vectors. The high genetic variability within GLRaV-1 underscores the importance of genomic surveillance and molecular epidemiology, as distinct genotypes may differ in pathoge nicity or transmission efficiency. Demographic analyses suggested historical population expansions, particularly in Clade I, potentially linked to grapevine domestication and anthropogenic dissemination. Although not statistically significant, these patterns align with known viticultural history. Overall, this study reveals strong genetic structuring, predominant purifying selection with localized diversifying pressure, and limited gene flow among GLRaV-1 populations in Iran. The findings highlight the need for continu ous surveillance, region-specific management strategies, and further studies integrating broader sampling, vector identification, and host-virus interaction analyses to better anticipate the evolutionary potential of this economically important pathogen.
## References
1. Creasy, Creasy (2018) *Grapes. Cabi*
2. Moradi, Koolivand, Eini et al. (2018) "Molecular identification of four important nepovirus from vineyards of Zanjan province" *Iran J Plant Prot Sci*
3. Goheen, Harmon, Weinberger (1958) "Leafroll (white Emperor disease) of grapes in California" *Phytopathology*
4. Porotikova, Terehova, Volodin et al. (2021) "Distribution and genetic diversity of grapevine viruses in Russia" *Plants (Basel)*
5. Caruso, Bertacca, Ragona et al. (2022) "Epidemiological survey of grapevine leafroll-associated virus 1 and 3 in Sicily (Italy): genetic structure and molecular variability" *Agriculture*
6. Lu, Wang, Yang et al. (2023) "Discovery of a Closterovirus infecting jujube plants grown at aksu area in Xinjiang of China" *Viruses*
7. Ito, Nakaune (2016) "Molecular characterization of a novel putative ampelovirus tentatively named grapevine leafroll-associated virus 13" *Arch Virol*
8. Fan, Hong, Dong et al. (2015) "Genetic diversity and recombination analysis of grapevine leafroll-associated virus 1 from China" *Arch Virol*
9. Donda, Jarugula, Naidu (2017) "An analysis of the complete genome sequence and subgenomic RNAs reveals unique features of the Ampelovirus, Grapevine leafroll-associated virus 1" *Phytopathology*
10. Maree, Almeida, Bester et al. (2009) "Grapevine leafroll-associated virus 3" *Front Microbiol*
11. Fuchs, Martinson, Loeb et al. (2009) "Survey for the three major leafroll disease-associated viruses in Finger Lakes Vineyards in New York" *Plant Dis*
12. Fortusini, Scattini, Cinquanta et al. (1996) "Natural spread of grapevine leafroll virus 1 (GLRV-1), grapevine leafroll virus 3 (GLRV-3) and grapevine fleck virus (GFkV)" *Informatore Fitopatologico*
13. Bertolini, García, Yuste et al. (2010) "High prevalence of viruses in table grape from Spain detected by real-time RT-PCR" *Eur J Plant Pathol*
14. Messmer, Bohnert, Schumacher et al. (2021) "Studies on the occurrence of viruses in planting material of grapevines in Southwestern Germany" *Viruses*
15. Güvenateş, Topkaya (2023) "Molecular determination of the presence of some grapevine viruses and phylogenetic analyses of grapevine virus A in Tokat Province Grapevine area" *Tekirdağ Ziraat Fakültesi Dergisi*
16. Sarıtaş, Topkaya (2021) "Genetic variability of Grapevine Leafroll-Associated Virus-1, 2, 3, 4 infecting vineyards in Tokat Province of Turkey" *Uluslararası Tarım ve Yaban Hayatı Bilimleri Dergisi*
17. Karadeniz, Yağcı, Topkaya et al. (2018) "Tokat İli ve İlçelerinde Bazı Bağ Virüs Hastalıklarının Serolojik Yöntemlerle Belirlenmesi" *Bitki Koruma Bülteni*
18. Naderpour, Shahbazi, Alizadeh et al. (2020) "The status of Grapevine leafroll-associated viruses in Iran" *Acta Hortic*
19. Alabi, Rwahnih, Karthikeyan et al. (2011) "Grapevine leafroll-associated virus 1 occurs as genetically diverse populations" *Phytopathology*
20. Buja, Sabella, Monteduro et al. (2022) "Detection of Ampelovirus and Nepovirus by Lab-on-a-Chip: a promising alternative to ELISA test for large scale health screening of Grapevine" *Biosensors (Basel)*
21. Elçi (2019) "Genomic variability and recombination analysis of Grapevine leafroll-associated virus-1 Isolates from Turkey" *Tarım Bilimleri Dergisi*
22. Gambino, Perrone, Gribaudo (2008) "A Rapid and effective method for RNA extraction from different tissues of grapevine and other woody plants" *Phytochem Anal*
23. Tamura, Stecher, Kumar (2021) "MEGA11: molecular evolutionary genetics analysis version 11" *Mol Biol Evol*
24. Larkin, Blackshields, Brown et al. (2007) "Clustal W and Clustal X version 2.0"
25. Tamura (1992) "Estimation of the number of nucleotide substitutions when there are strong transition-transversion and G+C-content biases" *Mol Biol Evol*
26. Saitou, Nei (1987) "The neighbor-joining method: a new method for reconstructing phylogenetic trees" *Mol Biol Evol*
27. Tajima (1989) "Statistical method for testing the neutral mutation hypothesis by DNA polymorphism" *Genetics*
28. Fu, Li (1993) "Statistical tests of neutrality of mutations" *Genetics*
29. Hudson, Slatkin, Maddison (1992) "Estimation of levels of gene flow from DNA sequence data" *Genetics*
30. Kumar, Stecher, Li et al. (2018) "MEGA X: molecular evolutionary genetics analysis across computing platforms" *Mol Biol Evol*
31. Mello (2018) "Estimating timetrees with MEGA and the timetree resource" *Mol Biol Evol*
32. Chare, Holmes (2006) "A phylogenetic survey of recombination frequency in plant RNA viruses" *Arch Virol*
34. Crnogorac, Panno, Mandić et al. (2021) "Survey of five major grapevine viruses infecting Blatina and Žilavka cultivars in Bosnia and Herzegovina" *PLoS One*
35. Gao, Lin, Shen et al. (2016) "Genetic diversity and molecular evolution of arabis mosaic virus based on the CP gene sequence" *Arch Virol*
36. Abadkhah, Hajizadeh, Koolivand (2020) "Global population genetic structure of Bean common mosaic virus" *Arch Phytopathol Plant Prot*
37. Drake, Holland (1999) "Mutation rates among RNA viruses" *Proc Natl Acad Sci*
38. Steinhauer, Domingo, Holland (1992) "Lack of evidence for proofreading mechanisms associated with an RNA virus polymerase" *Gene* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12567398&blobtype=pdf | # New Assay Systems to Characterize the Broad-Spectrum Antiherpesviral and Non-Herpesviral Activity of Cyclin-Dependent Kinase (CDK) 8 Inhibitors
Debora Obergfäll, Friedrich Hahn, Jintawee Kicuntod, Christina Wangen, Melanie Kögler, Sabrina Wagner, Benedikt Kaufer, Manfred Marschall
## Abstract
Background. To date, a number of human pathogenic viruses are still unaddressed by the current repertoire of approved antiviral drugs. In order to widen this spectrum of preventive measures against virus infections, we have focused on additional host targets that exert interesting virus-supportive functions. Inhibitors of cyclin-dependent kinase 8 (CDK8) have been found to exhibit highly pronounced and relatively broad antiviral activity. Objectives. The current research question concerning the potential for broad-spectrum antiviral drug activity should be addressed in detail to understand the mechanistic basis of the antiviral target function of CDK8. Materials and Methods. We established and specifically customized six assay systems, three of these newly developed for the present study, to corroborate the range of CDK8 inhibitors' antiviral activity against four α-, β-, and γ-herpesviruses as well as two non-herpesviruses. Results. Similar to our earlier analysis of CDK7 and CDK9 inhibitors, the clinically relevant CDK8 inhibitors currently in use demonstrated antiherpesviral activity in cell-culture-based infection models. Interestingly, the antiviral efficacy against various human and animal cytomegaloviruses was particularly strong at nanomolar concentrations, whereas other herpesviruses or nonherpesviruses showed an intermediate or low sensitivity to CDK8 inhibitors. Thus, this approach provided novel insights into the inhibitory potential of the CDK8 inhibitors, such as CCT-251921, MSC-2530818, and BI-1347, when analyzed against equine herpesvirus 1 (EHV-1, α-herpesvirus), human herpesvirus 6A (HHV-6A, β), Epstein-Barr virus (EBV, γ), murine herpesvirus 68 (MHV-68, γ), vaccinia virus (VV, non-herpes DNA virus), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2, non-herpes RNA virus). Conclusions. Our results confirm that drug sensitivity to CDK8 inhibitors, on the one hand, is very strong for certain viruses and, on the other hand, varies widely within the spectrum of viruses and host cell types analyzed. This suggests that CDK8 may play several different roles in viral replication. The option of a refined CDK8-specific antiviral drug targeting is discussed.
## 1. Introduction
Antiviral drug research has become an actively processed field. Not only through the driving force of COVID-19 pandemic, but also through decades of intense viral and antiviral research, the repertoire of clinically available drugs and vaccines against human pathogenic viruses has markedly increased. Nevertheless, still only a small number of all relevant human viruses can be addressed by preventive and curative measures, so that additional novel approaches with so-far-unexploited mechanisms of inhibitors and innovative targeting strategies are urgently needed. This attempt may include both the conventional antiviral targeting with direct-acting antiviral drugs (DAAs) and the newly recognized chances with host-directed antiviral strategies (HDAs). In areas spanning the treatments of tumor, metabolic, infectious, and inflammatory diseases [1][2][3][4][5][6][7][8][9][10][11][12][13][14], HDAs already proved great success and strongly suggested their extended application into antiviral treatment. In this field of antiviral therapy, however, the topic misses in-detail experience, since only a low number of HDAs have been clinically approved and are in practical use (such as maraviroc, ribavirin, and bulevirtide), mostly against human immunodeficiency and hepatitis viruses, so far (references for HIV/maraviroc [15][16][17]; HCV/ribavirin [18][19][20][21][22][23]; HDV and HBV/bulevirtide [24][25][26][27][28][29]).
Our recent achievements in the field pointed to the very promising antiviral activity of protein kinase inhibitors. This new gain of knowledge did not only concern the special situation that herpesviruses encode their own protein kinases (HvPKs; [30]) and that the respective inhibitors comprise a pronounced antiviral potency, such as the clinically approved inhibitor of cytomegalovirus kinase pUL97, maribavir (MBV; [31][32][33]). In addition, this aspect also includes virus-supportive functions of cyclin-dependent kinases of the host (CDKs) and the previously reported potency of antiviral CDK inhibitors [34][35][36][37][38]. Moreover, based on the recognition that specific HvPKs can represent viral orthologs of host CDKs, i.e., vCDK/pUL97 [33,39,40], we fostered the understanding that inhibitors of host kinases, in ideal terms, CDKs and vCDKs, actually bear a huge potential of antiherpesviral cotreatment synergy [3,10,11]. Moreover, a specific point of relevance may be presented by the demonstration of a broad-spectrum antiviral activity of certain selective CDK inhibitors, such as those targeting CDKs 2, 7, 8, and 9 [35,[41][42][43][44][45][46][47][48]. In particular, CDK8 is a complex regulator with manifold activities, as CDK8 recruits the core Mediator and super elongation complex (SEC) and is able to deactivate the CDK7-cyclin H-MAT1 transcription complex by phosphorylation [49]. Interestingly, these activities have recently been considered as ratelimiting steps of herpesviral replication efficiency, particularly of HCMV replication [38], but this CDK8 linkage also involves a number of other human and animal pathogenic viruses [35,46,50]. In the present study, our experimentation focused on the role of CDK8 as both a virus-supportive regulator and a putative novel antiviral drug target. To this end, we established new assay systems to assess the antiviral properties of clinically relevant CDK8 inhibitors.
As previously reported by other researchers, the three current inhibitors of interest exert a strong activity against human CDK8 [51][52][53]. They represent candidate drugs for ongoing clinical development. It should also be noted that secondary, lower-affinity targets of these CDK8 inhibitors have been identified, including CDK19 and potentially other CDKs [51][52][53]. As this is a general phenomenon of many CDK inhibitors that partly exert cross-talk with related kinases, the antiviral efficacy of these inhibitors may be based on a mode of action that goes beyond CDK8 mono-selectivity. Nevertheless, apart from this mechanistic aspect, the main target CDK8 has already been shown to be highly relevant for herpesvirus replication [46]. Thus, our current data contribute to specifying the broadspectrum antiherpesviral and non-herpesviral activity of this type of host-directed antiviral.
## 2. Results and Discussion
## 2.1. Establishment of New Antiviral Assay Systems for a Selection of Human and Animal Pathogenic Viruses
In order to evaluate the postulated broadness on the antiviral efficacy of CDK8 inhibitors, new antiviral assay systems were established, including a selection of human and animal pathogenic viruses. Thus, α-, β-, and γ-herpesviruses as well as non-herpesvirus DNA and RNA viruses were utilized in parallel settings. These quantitative assay systems were either newly established or specifically optimized and provided a valuable platform for comparative antiviral drug analysis (Table 1). Notably, these assays, which were established and customized for EHV-1, HHV-6A, and VV, were applied for the first time in this study. The SARS-CoV-2 [54], MHV-68, and EBV assays were further developed in this study. The latter two, MHV-68 and EBV, have recently been used in brief form [35,48,55]. Of note is the fact that the various readout systems, which are mostly based on reporter expression, were applied using a number of different host cell types to assess the specific drug sensitivity of viral replication to CDK8 inhibitors and reference compounds. Such references, all possessing a previously characterized antiviral activity, were particularly useful as tools to optimize the respective virus replication models. The main focus was directed to the antiviral activity of three clinically relevant CDK8 inhibitors. Notably, the possibility that secondary targets recognized by these inhibitors (such as CDK19 [51][52][53]) play an accessory role cannot be fully ruled out. However, it is unknown whether the putative secondary targets of the present CDK8 inhibitors are relevant to herpesviral replication. Apart from this open question, the antiviral efficacy of these compounds, based on the main target CDK8, is undoubtedly strong. Thus, the sensitivity of the selected viruses to the inhibitors in these cell types could be categorized as strong, intermediate, low, or variable (Table 1).
## 2.2. Assessment of Antiviral Activity of Three Selected CDK8 Inhibitors Against α-, β-, and γ-Herpesviruses
Concerning the α-herpesviruses, equine herpesvirus 1 (EHV-1) was chosen as a candidate for comparing inhibitory drug profiles between infected-cell cultures because it can infect a number of different animal hosts, including horses, cattle, and rodents. It may also be useful as an in vivo mouse model in the future [56,57]. For our analyses, we established an EHV-1 GFP-based assay. To this end, we infected 96-well plate cultures of Vero, COS-7, MDCKII, HFFs, or MEFs with a titer-determined stock of EHV-1-GFP. The first readout was GFP-specific automated fluorometry (VictorX4; Table 1). Reliable levels of GFP signals were obtained for EHV-1-GFP infection of Vero, COS-7 cells, or HFFs, whereas only very low signal levels were obtained for MEFs or MDCKII cells. To confirm sufficient EHV-1 permissiveness, a virus-specific qPCR was performed on samples of infected cell culture media. Based on the intensity of these signals, the permissiveness of the cells to EHV-1 could be ranked as follows: COS-7 > HFF > Vero > MDCKII > MEF (MDCKII and MEF were basically not susceptible to EHV-1 infection). Therefore, COS-7 cells were primarily selected for detailed analysis (Figure 1). The antiviral system was adjusted using reference drugs, i.e., GCV, CDV, and BCV (Figure 1C), and readouts were performed using reporter GFP measurements with infected cell layers and EHV-1-specific qPCR based on virus release into the culture medium (Figure 1D). Concentration-specific curves of antiviral activity were obtained in all cases, with half-maximal effective mean concentrations (EC 50 values) ranging from 0.002 ± 0.003 µM to 4.50 ± 2.40 µM. The anti-EHV-1 efficacies of CDK8 inhibitors were pronounced, with submicromolar concentrations (Figure 1B). The three compounds BI (BI-1347), CCT (CCT-251921), and MSC (MSC-2530818) were compared based on GFP and qPCR measurements (Figure 1E). The strongest inhibitory efficacy was found for BI in EHV-1-infected COS-7 cells, with EC 50 values down to the nanomolar range. Interestingly, the EHV-1-directed efficacy of the three CDK8 inhibitors showed some quantitative variation in HFFs (GFP measurement; Figure 1E), with EC 50 values still in a submicromolar range. These differences in the efficacy of identical compounds against one specific virus in different host cell types (see further examples below) may reflect the host-directed antiviral mode of these compounds. This may point to a cell-type-specific inhibitory mode, such as virus activation, maturation, release, etc. Specifically, the virussupportive CDK8 target may exhibit differential expression levels in various cell types, which may be functionally compensated, at least in part, by cell-type-specific expression of related CDKs.
Concerning the β-herpesviruses, human herpesvirus 6A (HHV-6A) was used because it is a close relative of cytomegaloviruses (CMVs) that is moderately pathogenic. It should be compared to CMVs in terms of sensitivity to antiviral CDK inhibitors, which had been analyzed against various CMV strains and reporter recombinants in detail before [11,35,43]. To this end, HHV-6A-GFP was used to infect the primary fibroblast model of HCMV infection (HFFs), and an unachieved readout was established in these cells using a plaque reduction assay (PRA; Table 1). This newly established PRA-based measurement of anti-HHV-6A drug activity in HFFs is specifically improved and useful because it allows for fresh infection with an HHV-6A virus stock instead of using HHV-6A-positive J-Jhan carrier cells, which carry HHV-6A genomes in an integrated form and continuously produce progeny virus. Based on our experience with this method of drug assessment in HFFs, we observed that HHV-6A is more sensitive with this fresh infection and PRA readout (Figure 2). We adjusted the assay conditions using antiherpesviral reference drugs, i.e., BCV, CDV, and GCV (Figure 2C). Regarding the CDK8 inhibitors, we reconsidered our previous statement that HHV-6A has very limited sensitivity to these types of HDAs [35]. Instead, we recognized the clear anti-HHV-6A efficacy of all three analyzed compounds under these assay conditions (Figure 2B). The mean EC 50 values of BI, CCT, and MSC in the HFF infection system were 0.2 ± 0.2, 0.2 ± 0.4, and 0.04 ± 0.2 µM, respectively. Concerning the γ-herpesviruses, Epstein-Barr virus (EBV) was chosen, as it represents a major human oncogenic viral pathogen with worldwide clinical importance. EBV was specifically included in the comparative CDK inhibitor analysis because well-established quantitative anti-EBV drug research systems are rare. Therefore, we further developed the lytic EBV system with TPA-induced P3HR-1 cells. The productive viral replication and release could be quantified by qPCR detection of EBV genomic loads in culture media samples of a 96-well plate format (Table 1). The antiviral data indicate an intermediate sensitivity of EBV replication to CDK8 inhibitors. As based on the P3HR-1 model, EC 50 values were obtained in the range of 5.5 ± 4.2, 1.4 ± 0.6, and 2.5 ± 1.0 µM for the three compounds (Figure 3).
## GFP
In addition to EBV, MHV-68-Luc, a murine model virus, was applied. The Luc reporter signal, provided by MHV-68-Luc infection, proved to be a reliable quantitative marker of virus replication and release (Figure S1). During the refinement of this virus system, MHV-68-Luc demonstrated an enhanced ability to undergo productive replication in various host cell types. Consequently, the analysis of host-directed antiviral CDK8 inhibitors yielded valuable insights. Here, we compared anti-MHV-68 drug efficacy in four different cell types, i.e., Vero, MEF, HFF, and COS-7 cells, to verify and specify our previous finding of a particularly strong anti-MHV-68 activity of BI, CCT, and MSC (Figure 4) [35]. Our current findings confirm the nanomolar-range activity of CDK8 inhibitors against MHV-68-Luc and provide additional data for MHV-68-Luc in independent cell systems. Finally, our findings specify the antiviral potency as directly dependent on the infected host cell type. In particular, regarding the latter aspect, we substantiated our statement about the importance of cell type for the antiviral CDK8 targeting. Here, three factors may be relevant for MHV-68 in cell-type comparison, namely, the species varieties of CDK8, the differential CDK8 expression levels in cell types, and the functional complementation between related CDKs.
## 2.3. Antiviral Activity of CDK8 Inhibitors Across Herpesviral Subfamilies
In a previous report, we could already identify a marked antiviral efficacy of CDK8 inhibitors against several, but not all, of the investigated herpesviruses [35]. In the present study, we used additional α-, β-, and γ-herpesviruses to address their sensitivity towards the three CDK8 inhibitors of current interest. Findings from the newly adopted assay systems for EHV-1, HHV-6A, EBV, and MHV-68 (see Tables 1 andS1) revealed that the drugs BI, CCT, and MSC have broad antiherpesviral activity. In general, CDK8 appears to be an important virus-supportive host factor; however, the magnitude of CDK8 inhibitor sensitivity varied significantly among the herpesvirus species. A novel finding of this study is that the impact of host cell types is at least as important as the analyzed virus species. A novel finding of this study is that the impact of host cell types is at least as important as the analyzed virus species. This was evident when three or four different host cells were used for EHV-1 and MHV-68, respectively. For both viruses, the EC 50 values of the three drugs varied depending on the infected cell type. This may be due to several reasons, as discussed briefly in the previous section. The main reason may be related to cell-type-specific expression characteristics of CDK8 and other CDKs that can provide cross-complementing virus-supportive functionality. A very marked example was given by MHV-68 because, here, the difference in sensitivity to compound BI (CCT/MSC) was specifically high when comparing the infection of Vero, COS-7, HFF, or MEF, i.e., primate, human, and murine cell types (Figure 4F). Such a difference in cell-type-specific CDK8 expression levels could be illustrated by semi-quantitative Wb analysis (Figure 5). These results indicated that the expression levels of CDK8 varied substantially in the MHV-68permissive host cell types (as based on the cross-species reactivity of the CDK8-specific antibody used for all Wb stainings). Thus, variability in anti-viral CDK8 inhibitor sensitivity may reflect differences in virus-host interactions. Among these differences, quantitative and qualitative differential levels of host CDKs appear relevant. Apart from individual variations, we confirmed the broad-spectrum activity tendency of CDK8 inhibitors across representatives of α-, β-, and γ-herpesviruses (see summarizing Table S1, comparing data of the present study with those previously reported). Therefore, CDK8 could be considered a target for the next generation of host-directed antiviral drug development. (C) Total protein lysates (30 µg/mL) were prepared from uninfected (mock/-) or MHV-68-infected (MHV-68-Luc/+) cells and subjected to standard SDS-PAGE/WB procedures. CDK8 levels were detected by Wb staining using mAb-CDK8; β-actin was stained as a loading control. Based on the WB analysis, a densiometric determination, using AIDA, was performed for CDK8 levels relative to actin. Analysis was performed in quadruplicate measurements (SDS-PAGE/Wbs in duplicate, densitometry in duplicate), and statistical evaluation was carried out using ANOVA followed by Bonferoni's correction for multiple comparisons. *, p ≤ 0.05; ***, p ≤ 0.001; ****, p ≤ 0.0001.
## 2.4. Addressing the Characteristics of Antiviral MoA Displayed by CDK8 Inhibitors Against Three Strains of HCMV: Time-of-Addition Experimentation
The first characteristics of the anti-HCMV-directed mode-of-action (MoA) of CDK8 inhibitors have been addressed by previous investigations by our group [35,38,58]. Thereby, a marked late-phase inhibitory activity of pharmacological CDK8 inhibition by CCT as well as siRNA-mediated CDK8 knock-down could be identified [35]. In the current context of analyses, we addressed the question of antiviral efficacy of CCT upon delayed drug addition to a multi-round infection setting. To his end, HFFs were infected with strains of HCMV before the CDK8 inhibitor CCT was applied for antiviral treatment (serial 5-fold dilutions at concentrations of 50 nM to 0.00064 nM) at the time points indicated (0 d p.i. to 4 d p.i.). Three different viral strains (expressing YFP or GFP reporter as indicated) were used for infection, i.e., the tropism-adaptable HCMV TB40, the genetically intact clinical isolate HCMV Merlin, and the laboratory HCMV strain AD169 (Figure 6). The strains Merlin and AD169 showed a particularly strong sensitivity to CCT inhibition (at EC 50 concentrations compatible with earlier investigations; [35]), and strain TB40 showed a lower level of drug sensitivity (Figure 6A), which may be explained by the generally prolonged replication behavior of TB40 in this antiviral system [58]. The time-of-addition experiment in this multi-round replication assay indicated a requirement of early onset of drug treatment (starting at 0 d, 1 d, or 2 d p.i.), as clearly seen for the strains TB40 and AD169, respectively (Figure 6A,B, left andright). An addition of the drug at days 3 or 4 p.i. did no longer exert measurable antiviral activity for these two strains. The result suggests that CCT starts its host-directed antiviral effect at early time points of viral replication. The strain Merlin showed a slightly different temporal course of drug sensitivity, as here, the lowest EC 50 values were also obtained under condition 0 d p.i., but even later time points of drug addition (up to 4 d p.i.) produced measurable antiviral efficacy (Figure 6A,B, central; quantitative variations in EC 50 values might refer to an onset of viral cytopathic effects ≥ 2 d p.i.). This may indicate a specifically strong CDK8 dependency of the clinically relevant, genetically intact strain Merlin. Combined, the time-of-addition experiment provided further information on the antiviral MoA of the CDK8 inhibitor, in that, on the one hand, inhibitory activity is provided for various HCMV strains. On the other hand, drug addition must be ensured during the early period (days 0-2 p.i.) of the first HCMV replication cycle, which lasts 3-4 days within a multi-round replication assay (terminated at day 7 p.i.).
## A
## 2.5. Non-Herpesviral Activities of CDK8 Inhibitors
In order to further address the question of a broader antiviral potency of CDK8 inhibitors, we also analyzed two non-herpesviruses, i.e., VV (DNA virus) and SARS-CoV-2 (RNA virus), respectively. Using the assay systems developed and refined in this study, additional data could be collected for antiviral reference compounds in general and for CDK8 in particular. For SARS-CoV-2, the recently established reporter virus d6-YFP [54] was used for infection of Caco-2 cells in a 96-well plate (Figure 7A). Antiviral-drug-specific inhibition of lytic virus replication was quantified at 30 h p.i. by automated YFP-based fluorometry. Data indicate a very limited efficacy of CDK8 inhibitors against SARS-CoV-2 in this infection model (comprising low drug sensitivity expressed by the high micromolar EC 50 values in a range between 15.4 and 62.4 µM, Figure 7D). Thus, our data indicated a lack of CDK8 sensitivity (CCT) or a low level of sensitivity (BI, MSC) of SARS-CoV-2 in Caco-2 cells. None of the three compounds, i.e., BI, MSC, or CCT, produced an EC 50 value lower than 15.4 µM (Figure 7B; compared to the EC 50 value of 4.1 ± 1.4 µM for the control drug EIDD-1931/MPV, Figure 7C,D). However, it should be emphasized that the anti-SARS-CoV-2 activity of host-directed compounds may be virus-strain-or cell-type-specific [59]. This aspect may be relevant to our experimental system since the virus strain used to introduce the YFP reporter module was originally isolated from a patient in 2021 [54,60] A highly sensitive reporter assay was developed for VV analysis, which detects VVmediated shutdown of reporter protein expression. A 293T cell line stably expressing tdTom-Luc [61] served as a reporter model in a 96-well format to utilize VV-induced shutdown for monitoring antiviral drug efficacy VV infection correlated with complete luciferase activity in uninfected cells (Table 1). EC50 values reflected the antiviral rescue of VV-induced shutdown. The inhibitory effect of the reference compound BCV demonstrated the antiviral efficacy of VV replication at an EC50 of 1.9 ± 0.4 µM (Figure 8). Similar EC 50 values for BCV in a low micromolar range were previously reported [62]. However, none of the three analyzed CDK8 inhibitors produced a measurable anti-VV activity in this system. This point adds to our earlier understanding that the antiviral efficacy of CDK8 inhibitors can be very high depending on the virus species and cellular environment. Thus, the findings of Figures 7 and8 support the earlier statement that the CDK8-controlled Mediator transcription complex can act as a virus-supportive factor for several human viruses [46], which is particularly relevant for human and animal herpesviruses but to a lesser extent for unrelated viruses.
## 3. Materials and Methods
## 3.1. Cells and Viruses
293T, Vero, COS-7, MEF, and MDCKII cells, all derived from ATCC (Manassas, VA, USA), were cultivated in Dulbecco's modified Eagle's medium (DMEM) (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Capricorn, Ebsdorfergrund, Germany), 1× GlutaMAX™ (Thermo Fisher Scientific, Waltham, MA, USA), and 10 µg/mL gentamicin (Thermo Fisher Scientific, Waltham, MA, USA). For Caco-2 cells (ATCC; Manassas, VA, USA), 1× non-essential amino acids (Thermo Fisher Scientific, Waltham, MA, USA) were added to the DMEM medium described above. Primary human foreskin fibroblasts (HFFs, clinical samples from the Children's Hospital Erlangen, Germany) and murine embryonic fibroblasts (MEFs; ATCC, Manassas, VA, USA) were cultured in Eagle's Minimal Essential Medium (MEM), and J-Jhan (kindly provided by Benedikt Kaufer, FU Berlin, Germany) and P3HR-1 B cells containing a lytic mutant of the EBV genome (kindly provided by Dr. Susanne Delecluse, DKFZ, Heidelberg, Germany, and Hans Helmut Niller, Virology, Univ. Regensburg, Germany) were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Thermo Fisher Scientific, Waltham, MA, USA). All media were supplemented with 10% FBS, 1× GlutaMAX™, and 10 µg/mL gentamicin. Cells were maintained at 37 • C, 5% CO 2 , and 80% humidity and were regularly monitored for the absence of mycoplasma contamination using a Lonza™ My-coalert™ kit (Thermo Fisher Scientific, Waltham, MA, USA). Viruses used in this study were equine herpesvirus 1 [63], human herpesvirus 6A (HHV-6A; infected J-Jhan/HHV-6A producer culture or BACmid-derived reconstituted virus stock; [64]), Epstein-Barr virus (EBV, strain P3HR-1, in immortalized P3HR-1 B cell producer culture), vaccinia virus (VV, strain IHD-5; American Type Culture Collection; [43]), and severe acute respiratory syndrome coronavirus 2 [54].
## 3.2. Antiviral Compounds
BI-1347, CCT-251921, MSC-2530818 targeting CDK8, adefovir (AFV), brincidofovir (BCV), cidofovir (CDV), EIDD-1931, ganciclovir (GCV), tenofovir alafenamide (TFA), and tenofovir disoproxil (TDP) as reference compounds were obtained from MedChemExpress, Monmouth Junction, NJ, USA. Stock solutions were prepared in aliquots using sterile DMSO (Sigma Aldrich, St. Louis, MO, USA) and stored at -20 • C.
## 3.3. Virus Infection of Cultured Cells and Quantitative Readouts of Viral Replication
Equine herpesvirus 1 (EHV-1). Various cell types were used for infection with EHV-1-GFP and analyzed for their virus-productive permissiveness. After 90 min of viral adsorption, the antiviral compounds were added to the infected cells without removing the viral inoculum. Cells were cultivated at 37 • C for 3 to 5 d depending on their permissiveness. Thereafter, cells were fixed with 10% formalin at room temperature for 10 min. The EHV-1-GFP signals were measured utilizing a Victor X4 microplate reader (VictorX4; PerkinElmer, Waltham, MA, USA).
Human herpesvirus 6A (HHV-6A). The HHV-6A-GFP genomic BACmid was purified with a NucleoBond Xtra BAC Kit (Macherey-Nagel, Düren, Germany) and used for virus reconstitution from transfected HFFs (Lipofectamine 3000; Thermo Fisher Scientific, Waltham, MA, USA) to obtain a cell-free HHV-6A stock. Based on this virus stock used as an inoculum, a classical plaque reduction assay was newly established for lytic HHV-6 infection of HFFs. For this, 1.5 × 10 4 HFFs were seeded per well (24-well plates) 1 d prior to infection (MOI of 0.0005). After 90 min of virus adsorption, medium supernatants were removed, and the cells were treated with antiviral compounds or solvent control using 2× MEM (Thermo Fisher Scientific, Waltham, MA, USA). These samples were added to a heat-dissolved agarose 0.6% solution in 1:1 mixtures of medium-agarose (final 0.3%, partly cooled when adding) for further cultivation of cells. At 10 to 11 d p.i., the agarose overlay was removed before cells were stained with 1% crystal violet in 20% EtOH to count virus-induced plaque formation under a microscope.
Human cytomegalovirus (HCMV). Antiviral replication assays for HCMV reporter viruses were performed as described previously [35,65]. Briefly, HFFs were seeded in the 96-well format and infected with either AD169-GFP, TB40-YFP, or Merlin-GFP. To address the question of a time-dependent effect of CDK8 inhibitors, CCT-251921 was added at different time points post-infection (at 0 d, 1 d, 2 d, 3 d, or 4 d p.i.). Cells were harvested 7 d p.i. (or 12 d p.i. for Merlin-GFP) before GFP or YFP expression was measured using the Victor X4.
Epstein-Barr virus (EBV). The determination of anti-EBV drug activities was briefly described by our group previously in principle terms [43,55,66]. Here, we optimized the setup by using the EBV-positive immortalized B cell line P3HR-1, which is sensitive to lyticcycle EBV production, as chemically induced with 40 ng/mL of 12-O-tetradecanoylphorbol-13-acetate (TPA) [55]. P3HR-1 cells were seeded in 96-well plates and were simultaneously induced with TPA and treated with antiviral compounds for 10 d before cells were subjected to quantitative assessment of EBV lytic productivity in the culture medium samples. Thus, at 10 d post-treatment, samples were exposed to proteinase K digestion, followed by the EBV/BGLF5-specific quantitative polymerase chain-reaction (qPCR) analysis (primers 5 ′ -TGA CCT CTT GCA TGG CCT CT-3 ′ and 5 ′ -CCT CTT TTC CAA GTC AGA ATT TGA C-3 ′ ; FAM probe 5 ′ -CCA TCT ACC CAT CCT ACA CTG CGC TTT ACA-3 ′ ).
Murine gammaherpesvirus 68 (MHV-68). The reporter virus MHV-68-Luc was multiplied in Vero cells, and stock virus was used for antiviral drug assessment in a 96-well plate setting, as described before [35]. In brief, MHV-68-Luc-infected cells were incubated for the cell-type-specific durations indicated, lysed using 100 µL of lysis buffer per well, and applied to reporter luminescence analysis with sample volumes of 50 µL per measurement (0.1 M KH 2 PO 4 , 15 mM MgSO 4 , 5 mM ATP, 1 mM D-luciferin), as measured in the Perkin Elmer Victor X4 reader.
Vaccinia virus (VV). The VV-mediated host shut-off was determined using a specific reporter plasmid (kindly provided by Kazuhiro Oka [61]). For this, 293T cells stably expressing Luc2-P2A-tdTom were prepared by using a lentiviral gene transfer approach utilizing the reporter plasmid pCDH-EF1-Luc2-P2A-tdTom [67]. Positive transfected cells were selected with FACS-based cell sorting according to the tdTom expression. For the VV host shut-off assay, 2 × 10 4 293T-tdTom-Luc cells were seeded per well (96-well plate), before virus-medium inocula as well as antiviral compounds were gently added on the next day (to minimize 293T cell detachment). At 2 d p.i., cells were lysed in 100 µL of standard assay buffer (100 mM phosphate buffer (a combination of 1 M KH 2 PO 4 and 1 M K 2 HPO 4 to reach pH 7.8) and 15 mM MgSO 4 supplemented with 1% Triton X-100) per well [58]. To quantify the luciferase (Luc) signal, 10 mM rATP (Abcam, Cambridge, UK) and 1 mM luciferin (PJK GmbH, Kleinbittersdorf, Germany) were added to the standard assay buffer. A volume of 50 µL of freshly prepared buffer was added to measure Luc signals using the Victor X4 microplate reader (VictorX4). Wells infected with VV and solvent-treated DMSO were used as a background (indicating a maximal VV-induced Luc signal shut-off), where mock-infected 293T-tdTom-Luc cells defined the 100% Luc level.
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). A SARS-CoV-2 replication assay was performed as previously described [54]. In brief, SARS-CoV-2 infection of Caco-2 cells was performed with the d6-YFP reporter virus in a 96-well plate format. Cells were harvested at 30 h p.i., and the quantities of viral replication were assessed by YFP quantitation in the Victor X4 microplate reader. Antiviral efficacy (mean of biological quadruplicates) was expressed as the percentage of the mock-treated control.
Concerning the fluorescence-based readouts, for EHV-1-GFP, HCMV AD169-GFP, and SARS-CoV-2 d6-YFP [35,48,54,58,68], the automated fluorometry measurements using the Victor X4 microplate reader (VictorX4) were compared with automated microscopy measurements using a Pico ImageXpress ® Device (PicoMD; Molecular Devices LLC, San Jose, CA, USA). Parallel measurements in the experiments provided confirmation of the data; however, in the context of this study, VictorX4 runs proved to be more facile and reliable, so these data are presented exclusively. was added to a final concentration of 40 µg/mL and incubated for 2-4 h at 37 • C. The NR-containing supernatant was finally removed, and cells were destained using a solution of 50% ethanol, 1% acetic acid, and 49% distilled water. For 293T cells, an Alamar Blue (AB) assay was alternatively used to assess cell viability at 2 d. Here, cells were seeded similarly and incubated with 25 ng/mL resazurin (AB; Ann Arbor, MI, USA) for 6 h at 37 • C before fluorescence was measured at 560/630 nm (both AB or NR) using the VictorX4.
## 3.5. Western Blot Analysis of MHV-68-Infected Cells
HFFs, Vero cells, COS-7, and MEFs were seeded in a 6-well plate. For each cell type, the cells were either left mock-infected or infected with MHV-68-Luc. Four d p.i., the cells were lysed in 100 µL, and two times 10 µL of each lysate was used to analyze the luciferase expression. The luciferase assay was performed as described for the MHV-68-Luc antiviral assay. To estimate the total protein amount per sample, a bicinchoninic acid assay (BCA) was performed by using a Pierce™ BCA Protein Assay Kit (Thermofisher, Rockford, IL, USA). For each sample, two times 10 µL of total protein lysate or a 1:10 dilution was used to determine the total protein amount. All steps were performed following the manufacturer's protocol, and the colorimetric change was measured at 562 nm using the VictorX4. From the remaining lysate, a protein amount of 30 µg/mL was used for Western blot analysis of CDK8 expression, as previously described [35]. Western blot analysis was performed using a CDK8-specific antibody (A302-501A, Bethyl Laboratories, Montgomery, TX, USA) and a monoclonal β-actin antibody (A5441, Sigma-Aldrich, St. Louis, MO, USA) as a loading control. Detection was carried out using the appropriate HRP-conjugated secondary antibodies.
## 4. Conclusions
In the present study, we focused our analysis on specific assay systems to apply innovative readout measurements. This approach was intended to provide insight into the broad-spectrum antiherpesviral and non-herpesviral activity of recently identified CDK8 inhibitors. The assays comprised a variety of reporter-driven quantitative assessments, in particular, for four herpesviruses, i.e., EHV-1, HHV-6A, EBV, and MHV-68, and two non-herpesviruses, i.e., VV, and SARS-CoV-2. Interestingly, the results of the present study, together with the data published by our group before [35,38], provide a relatively wide spectrum of viruses analyzed for CDK8 inhibitor sensitivity so far (Table S1). The viruses showing antiviral effects against CDK8 inhibitors, as expressed by defined micromolar-to-nanomolar EC 50 values, span members of the families Herpesviridae (species of α-, β-, and γ-herpesviruses), Adenoviridae, and Polyomaviridae and may include even more. On the other hand, two examples of Coronaviridae and Poxviridae did not show CDK8 inhibitor sensitivity. Concerning the possible role of CDK8 in viral replication, the current knowledge is rather limited, so that on the basis of our analyses, two branches of virus-supportive functions of CDK8 appear plausible. On the one hand, the fine modulation of transcription mediated by CDK activity in virus-infected cells is a very profound regulatory mechanism. In this regard, the replication efficiency of most human and animal viruses, if not all, behaves sensitive towards changes in the host transcription machinery, and obviously number of viruses are dependent on the transcription factor CDK8. On the other hand, viral proteins may represent direct substrates of the CDK8 kinase and may thereby become regulated in a phosphorylation-mediated manner. This has been demonstrated for CDK7-specific phosphorylation of cytomegalovirus proteins before, and a similar situation may contribute to viral sensitivity towards CDK8 activity. Beyond these two regulatory mechanisms, even more CDK8-specific processes might have relevance for individual viruses, such as a virusinduced modulation of CDK8 activity. This may be conferred through the cyclin-binding property of viral proteins, as recently specified for HCMV [36,[69][70][71], but this aspect still remains speculative. Thus, our experimental achievements were as follows: (i) CDK8 inhibitors proved to possess a very pronounced and relatively broad antiviral activity; (ii) the antiviral spectrum spanned in particular across certain subfamily members of α-, β-, and γ-herpesviruses; (iii) non-herpesviruses showed a relatively weak or no sensitivity; and thus (iv), CDK8 may represent an interesting target of host-directed drug candidates. Combined, the novel antiviral data substantiate the concept of a broad herpesvirus-supportive role of human CDK8 and its potential as a new target of refined pharmacological strategies.
## References
1. Buxmann, Hamprecht, Meyer-Wittkopf et al. (2017) "Primary Human Cytomegalovirus (HCMV) Infection in Pregnancy" *Dtsch. Arztebl. Int*
2. Chen, Wang, Chen et al. "Recent Advances and Perspectives in the Treatment of Human Cytomegalovirus Infections" *Trop. Med. Infect. Dis*
3. Chou, Ercolani, Derakhchan (2018) "Antiviral activity of maribavir in combination with other drugs active against human cytomegalovirus" *Antivir. Res*
4. Griffiths, Reeves (2021) "Pathogenesis of human cytomegalovirus in the immunocompromised host" *Nat. Rev. Microbiol*
5. Gugliesi, Coscia, Griffante et al. (2020) "Where do we Stand after Decades of Studying Human Cytomegalovirus? Microorganisms"
6. Tsutsui (2009) "Effects of cytomegalovirus infection on embryogenesis and brain development" *Congenit. Anom*
7. Fröhlich, Hahn, Belmudes et al. (2018) "Synthesis of Artemisinin-Derived Dimers, Trimers and Dendrimers: Investigation of Their Antimalarial and Antiviral Activities Including Putative Mechanisms of Action" *Chemistry*
8. Oiknine-Djian, Bar-On, Laskov et al. (2019) "Artemisone demonstrates synergistic antiviral activity in combination with approved and experimental drugs active against human cytomegalovirus" *Antivir. Res*
9. Wild, Hahn, Grau et al. (2020) "The Artemisinin-Derived Autofluorescent Compound BG95 Exerts Strong Anticytomegaloviral Activity Based on a Mitochondrial Targeting Mechanism" *Int. J. Mol. Sci*
10. Wild, Karner, Eickhoff et al. (2023) "Combined Treatment with Host-Directed and Anticytomegaloviral Kinase Inhibitors: Mechanisms, Synergisms and Drug Resistance Barriers" *Pharmaceutics*
11. Wild, Kicuntod, Seyler et al. (2021) "Combinatorial Drug Treatments Reveal Promising Anticytomegaloviral Profiles for Clinically Relevant Pharmaceutical Kinase Inhibitors (PKIs)" *Int. J. Mol. Sci*
12. Bachman, Zwezdaryk (1083) "Targeting the Host Mitochondria as a Novel Human Cytomegalovirus Antiviral Strategy" *Viruses*
13. Bogner, Egorova, Makarov (2021) "Small Molecules-Prospective Novel HCMV Inhibitors" *Viruses*
14. Krishna, Wills, Sinclair (2019) "Advances in the treatment of cytomegalovirus" *Br. Med. Bull*
15. Yost, Pasquale, Sahloff (2009) "Maraviroc: A coreceptor CCR5 antagonist for management of HIV infection" *Am. J. Health-Syst. Pharm*
16. De Luca, Pezzotti, Boucher et al. (2019) "Clinical use, efficacy, and durability of maraviroc for antiretroviral therapy in routine care: A European survey" *PLoS ONE*
17. Abel, Back, Vourvahis (2024) "Maraviroc: Pharmacokinetics and drug interactions" *Antivir. Ther*
18. Loustaud-Ratti, Debette-Gratien, Jacques et al. (2016) "Ribavirin: Past, present and future" *World J. Hepatol*
19. Crotty, Cameron, Andino (2022) "Ribavirin's antiviral mechanism of action: Lethal mutagenesis?" *J. Mol. Med*
20. Feld, Jacobson, Sulkowski et al. (2017) "Ribavirin revisited in the era of direct-acting antiviral therapy for hepatitis C virus infection" *Liver Int*
21. Janowski, Dudley, Wang (2020) "Antiviral activity of ribavirin and favipiravir against human astroviruses" *J. Clin. Virol*
22. Dixit, Perelson (2006) "The metabolism, pharmacokinetics and mechanisms of antiviral activity of ribavirin against hepatitis C virus" *Cell. Mol. Life Sci*
23. Snell (2001) "Ribavirin--current status of a broad spectrum antiviral agent" *Expert Opin. Pharmacother*
24. Degasperi, Anolli, Lampertico (2023) "Bulevirtide-based treatment strategies for chronic hepatitis delta: A review" *J. Viral. Hepat*
25. Kang, Syed (2020) "Bulevirtide: First Approval" *Drugs*
26. Wedemeyer, Aleman, Brunetto et al. (2023) "A Phase 3, Randomized Trial of Bulevirtide in Chronic Hepatitis D" *N. Engl. J. Med*
27. Wedemeyer, Schöneweis, Bogomolov et al. (2023) "Safety and efficacy of bulevirtide in combination with tenofovir disoproxil fumarate in patients with hepatitis B virus and hepatitis D virus coinfection (MYR202): A multicentre, randomised, parallel-group, open-label, phase 2 trial" *Lancet Infect. Dis*
28. Dietz-Fricke, Tacke, Zollner et al. (2023) "Treating hepatitis D with bulevirtide-Real-world experience from 114 patients" *JHEP Rep*
29. Brillanti (1587) "Management of Delta Hepatitis 45 Years after the Discovery of HDV" *J. Clin. Med*
30. Marschall, Feichtinger, Milbradt (2011) "Regulatory roles of protein kinases in cytomegalovirus replication" *Adv. Virus Res*
31. Avery, Alain, Alexander et al. (2022) "Maribavir for Refractory Cytomegalovirus Infections with or Without Resistance Post-Transplant: Results From a Phase 3 Randomized Clinical Trial" *Clin. Infect. Dis*
32. Maertens, Cordonnier, Jaksch et al. (2019) "Maribavir for Preemptive Treatment of Cytomegalovirus Reactivation" *N. Engl. J. Med*
33. Marschall, Schütz, Wild et al. (1338) "Understanding the Cytomegalovirus Cyclin-Dependent Kinase Ortholog pUL97 as a Multifaceted Regulator and an Antiviral Drug Target" *Cells*
34. Hutterer, Niemann, Milbradt et al. (2015) "The broad-spectrum antiinfective drug artesunate interferes with the canonical nuclear factor kappa B (NF-kappaB) pathway by targeting RelA/p65" *Antivir. Res*
35. Obergfäll, Wild, Sommerer et al. (1238) "Cyclin-Dependent Kinase 8 Represents a Positive Regulator of Cytomegalovirus Replication and a Novel Host Target for Antiviral Strategies" *Pharmaceutics*
36. Rechter, Scott, Eickhoff et al. (2009) "Cyclindependent Kinases Phosphorylate the Cytomegalovirus RNA Export Protein pUL69 and Modulate Its Nuclear Localization and Activity" *J. Biol. Chem*
37. Schang (2006) "First demonstration of the effectiveness of inhibitors of cellular protein kinases in antiviral therapy" *Expert Rev. Anti Infect. Ther*
38. Wild, Hahn, Bruckner et al. (2022) "Cyclin-Dependent Kinases (CDKs) and the Human Cytomegalovirus-Encoded CDK Ortholog pUL97 Represent Highly Attractive Targets for Synergistic Drug Combinations" *Int. J. Mol. Sci*
39. Hume, Finkel, Kamil et al. (2008) "Phosphorylation of Retinoblastoma Protein by Viral Protein with Cyclin-Dependent Kinase Function" *Science*
40. Kuny, Chinchilla, Culbertson et al. (2010) "Cyclin-dependent kinase-like function is shared by the beta-and gammasubset of the conserved herpesvirus protein kinases" *PLoS Pathog*
41. Feichtinger, Stamminger, Muller et al. (2011) "Recruitment of cyclin-dependent kinase 9 to nuclear compartments during cytomegalovirus late replication: Importance of an interaction between viral pUL69 and cyclin T1" *J. Gen. Virol*
42. Hahn, Hamilton, Wangen et al. (2021) "Development of a PROTAC-Based Targeting Strategy Provides a Mechanistically Unique Mode of Anti-Cytomegalovirus Activity" *Int. J. Mol. Sci*
43. Hutterer, Eickhoff, Milbradt et al. (2015) "A novel CDK7 inhibitor of the Pyrazolotriazine class exerts broad-spectrum antiviral activity at nanomolar concentrations" *Antimicrob. Agents Chemother*
44. Kelso, Baumgart, Eickhoff et al. (2014) "Cyclin-dependent kinase 7 controls mRNA synthesis by affecting stability of preinitiation complexes, leading to altered gene expression, cell cycle progression, and survival of tumor cells" *Mol. Cell. Biol*
45. Kicuntod, Häge, Hahn et al. (1021) "The Oligomeric Assemblies of Cytomegalovirus Core Nuclear Egress Proteins Are Associated with Host Kinases and Show Sensitivity to Antiviral Kinase Inhibitors" *Viruses*
46. Rovnak, Quackenbush (2022) "Exploitation of the Mediator complex by viruses" *PLoS Pathog*
47. Schang (2001) "Cellular proteins (cyclin dependent kinases) as potential targets for antiviral drugs" *Antivir. Chem. Chemother*
48. Yu, Wagner, Schütz et al. "An Antiherpesviral Host-Directed Strategy Based on CDK7 Covalently Binding Drugs: Target-Selective, Picomolar-Dose, Cross-Virus Reactivity" *Pharmaceutics*
49. Akoulitchev, Chuikov, Reinberg (2000) "TFIIH is negatively regulated by cdk8-containing mediator complexes" *Nature*
50. Rice, Kim, Rodriguez et al. (2020) "Suppression of a Subset of Interferon-Induced Genes by Human Papillomavirus Type 16 E7 via a Cyclin Dependent Kinase 8-Dependent Mechanism" *Viruses*
51. Ingelheim, Opnme (2017) "com-Boehringer Ingelheim Open Innovation Portal"
52. Czodrowski, Mallinger, Wienke et al. (2016) "Structure-Based Optimization of Potent, Selective, and Orally Bioavailable CDK8 Inhibitors Discovered by High-Throughput Screening" *J. Med. Chem*
53. Mallinger, Schiemann, Rink et al. (2016) "Discovery of Potent, Selective, and Orally Bioavailable Small-Molecule Modulators of the Mediator Complex-Associated Kinases CDK8 and CDK19" *J. Med. Chem*
54. Hahn, Häge, Herrmann et al. (1076) "Methodological Development of a Multi-Readout Assay for the Assessment of Antiviral Drugs against SARS-CoV-2. Pathogens 2021"
55. Tillmanns, Battisti, Kicuntod et al. (2025) "The conserved core nuclear egress complex (NEC) as an antiherpesviral drug target: Pharmacophore-based identification of NEC-specific inhibitors" *Antivir. Res*
56. Mesquita, Costa, Zanatto et al. (1556) "Equine herpesvirus 1 elicits a strong pro-inflammatory response in the brain of mice" *J. Gen. Virol*
57. *EFSA Panel on Animal Health and Welfare*
58. Nielsen, Alvarez, Bicout et al. (2016) "Assessment of listing and categorisation of animal diseases within the framework of the Animal Health Law (Regulation (EU) No"
59. Tillmanns, Kicuntod, Ehring et al. "Establishment of a Luciferase-Based Reporter System to Study Aspects of Human Cytomegalovirus Infection"
60. Hoffmann, Kleine-Weber, Schroeder et al. (2020) "SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor" *Cell*
61. Hahn, Niesar, Wangen et al. (2020) "Target verification of artesunate-related antiviral drugs: Assessing the role of mitochondrial and regulatory proteins by click chemistry and fluorescence labeling" *Antivir. Res*
62. Kazuhiro (2025) "Addgene Plasmid # 72486"
63. Kern, Hartline, Harden et al. (2002) "Enhanced inhibition of orthopoxvirus replication in vitro by alkoxyalkyl esters of cidofovir and cyclic cidofovir" *Antimicrob. Agents Chemother*
64. Neubauer, Rudolph, Brandmuller et al. (2002) "The equine herpesvirus 1 UL34 gene product is involved in an early step in virus egress and can be efficiently replaced by a UL34-GFP fusion protein" *Virology*
65. Gravel, Dubuc, Wallaschek et al. (2017) "Cell Culture Systems to Study Human Herpesvirus 6A/B Chromosomal Integration" *J. Virol*
66. Marschall, Freitag, Weiler et al. (2000) "Recombinant green fluorescent protein-expressing human cytomegalovirus as a tool for screening antiviral agents" *Antimicrob. Agents Chemother*
67. Auerochs, Korn, Marschall (2011) "A reporter system for Epstein-Barr virus (EBV) lytic replication: Anti-EBV activity of the broad anti-herpesviral drug artesunate" *J. Virol. Methods*
68. Häge, Horsch, Stilp et al. (2020) "A quantitative nuclear egress assay to investigate the nucleocytoplasmic capsid release of human cytomegalovirus" *J. Virol. Methods*
69. Lösing, Häge, Schütz et al. "Shared-Hook' and 'Changed-Hook' Binding Activities of Herpesviral Core Nuclear Egress Complexes Identified by Random Mutagenesis" *Cells*
70. Graf, Feichtinger, Naing et al. (2016) "New insight into the phosphorylation-regulated intranuclear localization of human cytomegalovirus pUL69 mediated by cyclin-dependent kinases (CDKs) and viral CDK orthologue pUL97" *J. Gen. Virol*
71. König, Buscher, Steingruber et al. (2017) "Dynamic regulatory interaction between cytomegalovirus major tegument protein pp65 and protein kinase pUL97 in intracellular compartments, dense bodies and virions" *J. Gen. Virol*
72. Steingruber, Marschall (2020) "The Cytomegalovirus Protein Kinase pUL97:Host Interactions, Regulatory Mechanisms and Antiviral Drug Targeting. Microorganisms"
73. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12730590&blobtype=pdf | # Importance of Early Recognition and Initiation of Management in Haemophagocytic Lymphohistiocytosis: A Case Report
Sreemallika Ramireddy, Sarah Jones
## Abstract
We present the case of an 18-year-old previously healthy male who was admitted with severe abdominal pain and subsequently found to have concurrent Epstein-Barr virus (EBV) and cytomegalovirus (CMV) infections. Due to concern over potential deterioration, he was transferred to critical care, where progressive cytopenias and deranged liver function tests were noted. The combination of acute liver failure, pancytopenia, hyperferritinaemia, and coagulopathy raised suspicion for haemophagocytic lymphohistiocytosis (HLH). This was supported by bone marrow aspirate findings demonstrating megakaryocytic proliferation and elevated soluble CD25 levels on further testing performed at the Great Ormond Street Hospital (GOSH).High-dose methylprednisolone and anakinra were initiated with subsequent clinical and biochemical improvement. Targeted antiviral therapy with ganciclovir and rituximab was commenced for CMV and EBV, respectively. The patient's haematological parameters and liver function were normalised, and he was discharged on a tapering regimen of prednisolone and anakinra, alongside oral valganciclovir. This case highlights the importance of early recognition of HLH in the context of viral coinfection, particularly EBV and CMV. Prompt multidisciplinary involvement and initiation of immunosuppressive and antiviral therapy are crucial to improving outcomes in this potentially fatal condition.
## Introduction
Haemophagocytic lymphohistiocytosis (HLH) is a potentially fatal hyperinflammatory syndrome defined by a disproportionate immune response involving natural killer cells and T cells [1][2][3]. Organ failure, leading to intensive care admissions and death are potential consequence of missed diagnosis [1,4], highlighting the importance of early recognition and initiation of treatment. There are two categories of HLH: primary, which usually presents in childhood with a familial genetic component, and secondary, which arises following an acute immune response, such as in malignancy, sepsis, or immunological conditions, more common in the adult population [1,2,4].
Diagnosis is often difficult due to the presentation being similar to that of haematological malignancies or infections causing sepsis [1,4]. The diagnosis of HLH was previously considered if the patient met the HLH-94 or the subsequent HLH-2004 criteria [3]. However, the probability of an HLH diagnosis can now be performed using the HScore, which looks for physical signs, for example, splenomegaly or pyrexia, presence of biochemical markers, such as cytopenias, and the presence of haemophagocytosis on bone marrow aspirate [4,5]. Once a diagnosis is made, treatment can be initiated with the aid of the multidisciplinary team (MDT), involving haematologists and, if appropriate, microbiology specialists.
Treatment of HLH involves management that would be counterproductive and dangerous in cases that present similarly, for example, in severe sepsis, as the mainstay is immunosuppression and cytotoxic medications [1]. In this case, treatment involved high-dose corticosteroids for immunosuppression and anakinra, leading to significant clinical improvement.
We present the case of an 18-year-old previously healthy male who was admitted with severe abdominal pain and subsequently found to have concurrent Epstein-Barr virus (EBV) and cytomegalovirus (CMV) infections, leading to a diagnosis of HLH.
## Case Presentation
Over the course of one week (from 10th to 16th August), an 18-year-old male presented to the Emergency Department (ED) twice with a two-day history of sore throat, reduced oral intake, and a sensation of facial and neck swelling. He had no significant past medical history and no known consanguinity. On both occasions, he was reviewed by the Ear, Nose, and Throat (ENT) team, received intravenous dexamethasone, fluids, and analgesia, and was discharged once able to tolerate oral intake with a prescription for oral phenoxymethylpenicillin. The patient's first admission lasted 10 hours and 43 minutes, with a working diagnosis of presumed tonsillitis. Examination findings were documented as "bilateral lymphadenopathy, grade 3 tonsils with exudate, full neck range of motion, no trismus."
Due to persisting symptoms and being unable to maintain good oral intake, the patient returned to the ED 14 hours later. During this second admission (lasting five days, from 11th to 16th August), a glandular fever screen returned positive. Blood tests showed an elevated white cell count of 26.5 × 10⁹/L, lymphocytes 11.4 × 10⁹/L, and atypical lymphocytes on blood film. Immunophenotyping was negative for lymphoma.
Five days after the second discharge, the patient re-presented to the ED with sudden-onset severe abdominal pain, hypotension, and tachycardia. Examination revealed splenomegaly, raising the suspicion of spontaneous splenic rupture. A CT abdomen and pelvis (CTAP) demonstrated splenomegaly without evidence of rupture or haematoma, mild hepatomegaly, reactive gallbladder wall oedema, and mesenteric invagination into the small bowel suggestive of intussusception. He was taken for emergency diagnostic laparoscopy on the 21st of August, which revealed no evidence of current or recent intussusception.
Intraoperatively, the surgeons noted a diffusely "oozing" omentum, deranged coagulation on rotational thromboelastometry (ROTEM), and a markedly enlarged liver and spleen. He was transferred to the Intensive Therapy Unit (ITU) on the 22nd of August, for monitoring in view of possible acute liver failure and anticipated post-operative deterioration.
On the day of ITU admission, the patient developed high fever, anaemia, and thrombocytopenia. A calculated HLH probability score indicated >99% likelihood of HLH (Table 1). Following haematology review, high-dose methylprednisolone and anakinra were commenced. The patient's liver function improved with supportive management. Virology studies revealed high levels of both CMV and EBV PCR positivity, suggesting either CMV reactivation secondary to immunosuppression or a dual infection. On virology advice, ganciclovir was initiated for CMV, and rituximab for EBV. Bone marrow biopsy demonstrated increased histiocytic activity with evidence of haemophagocytosis, consistent with HLH (Figure 1). The case was discussed at the national HLH MDT meeting (held at University College London Hospital), and functional studies were sent to GOSH, alongside an R15 genetic panel to the local genetics team. Genetic testing yielded a negative result, with no detectable variants to indicate the presence of a hereditary disorder.
## HScore Criteria Our Patient Data
## Underlying Immunosuppression Yes
## FIGURE 1: Histopathology images from the patient's bone marrow biopsy specimens
Bone marrow aspiration was performed on 28th August. The aspirate was stained with May-Grunwald Giemsa (MGG). Original magnification used, x50 with immersion oil. Increased histiocytic activity is seen in patients with HLH, with histiocytes ingesting and destroying bone marrow and blood cells. This is called haemophagocytosis and is demonstrated in these images (A-C).
At his most recent outpatient review (on the 1st of October), HLH markers had normalised. The patient is being gradually weaned off prednisolone and anakinra. He was discharged on oral Valganciclovir, which is being continued as CMV remains detectable (2564 copies/mL). The virology team will be consulted regarding the duration of antiviral therapy.
## Discussion
Distinguishing HLH from other causes of systemic inflammation, such as sepsis or severe viral infection, remains challenging, often delaying diagnosis and treatment initiation. EBV is the most frequently implicated viral trigger of HLH, with pathogenesis linked to excessive activation of infected cytotoxic T cells and macrophages. CMV-related HLH is less common and typically occurs in immunocompromised individuals [7,8]. This unusual combination of EBV and CMV infection in an immunocompetent individual highlights how concurrent viral immune activation can overwhelm normal regulatory mechanisms and perhaps precipitate HLH.
In this case, the combination of progressive cytopenias, hyperferritinaemia, and coagulopathy prompted calculation of the HScore, which indicated a >99% probability of HLH. The diagnosis was further supported by elevated soluble CD25 levels and characteristic bone marrow findings. This highlights the importance of maintaining a high index of suspicion and applying standardised diagnostic tools when laboratory features are disproportionate to the initial presumed infection.
Management of infection-associated HLH requires prompt initiation of immunosuppressive therapy to mitigate the hyperinflammatory response, while concurrently addressing the underlying trigger [9,10]. In this case, high-dose corticosteroids and anakinra led to rapid clinical improvement, with concurrent antiviral therapy (ganciclovir and rituximab) targeting CMV and EBV replication (Table 2). Long-term follow-up remains essential in HLH, even in apparently idiopathic or infection-triggered cases.
## Laboratory Tests
Relapse can occur months after apparent remission, particularly if an underlying genetic predisposition is present. Functional studies and a targeted gene panel (R15) were therefore performed to exclude primary or familial HLH. The absence of recurrent symptoms and normalisation of HLH markers are reassuring in this patient; however, continued surveillance remains important, given the risk of HLH recurrence.
Early multidisciplinary input contributed significantly to this patient's favourable outcome. As the HScore was first calculated during his third admission, it is reasonable to consider whether earlier application of standardised assessment tools may have expedited diagnosis and reduced the overall length of hospitalisation.
## Conclusions
In conclusion, this case represents an unusual example of HLH secondary to concurrent EBV and CMV infection in an immunocompetent young adult. It highlights the diagnostic complexity of HLH, the importance of prompt immunosuppressive and antiviral therapy, and the critical role of multidisciplinary management. Increasing awareness of HLH among frontline clinicians may facilitate earlier recognition and intervention, ultimately improving survival in this highly treatable but potentially fatal syndrome. Detailing cases where the underlying driver remains unclear contributes to the broader evidence base needed to better characterise atypical or multifactorial HLH presentations.
## References
1. Konkol, Killeen, Rai *Lymphohistiocytosis. StatPearls*
2. El-Mallawany, Curry, Allen (2022) "Haemophagocytic lymphohistiocytosis and Epstein-Barr virus: A complex relationship with diverse origins, expression and outcomes" *Br J Haematol*
3. Ponnatt, Lilley, Mirza (2022) "Hemophagocytic lymphohistiocytosis" *Arch Pathol Lab Med*
4. Cox, Mackenzie, Low (2024) "Diagnosis and investigation of suspected haemophagocytic lymphohistiocytosis in adults: 2023 Hyperinflammation and HLH Across Speciality Collaboration (HiHASC) consensus guideline" *Lancet Rheumatol*
5. Fardet, Galicier, Lambotte (2014) "Development and validation of the HScore, a score for the diagnosis of reactive hemophagocytic syndrome" *Arthritis Rheumatol*
6. Khare, Jinkala, Kanungo (2021) "Performance of HScore in reactive hemophagocytic lymphohistiocytosis" *Indian J Hematol Blood Transfus*
7. Matias-Lopes, Atalaia-Barbacena, Guiomar (2024) "Cytomegalovirus infection in an immunocompetent host presenting as hemophagocytic lymphohistiocytosis" *Eur J Case Rep Intern Med*
8. Bami, Vagrecha, Soberman (2020) "The use of anakinra in the treatment of secondary hemophagocytic lymphohistiocytosis. Pediatr Blood Cancer"
9. Hines, Bhatt, Talano et al. (2019) "Diagnosis, treatment, and management of hemophagocytic lymphohistiocytosis in the critical care unit. Critical Care of the Pediatric Immunocompromised Hematology/Oncology Patient"
10. Rosée, Horne, Hines (2019) "Recommendations for the management of hemophagocytic lymphohistiocytosis in adults" *Blood*
11. (2025) *Ramireddy et al. Cureus* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724204&blobtype=pdf | # Characterization of full-length and cytoplasmic tailtruncated envelope glycoproteins incorporated into human immunodeficiency virus (HIV-1) virions and virus-like particles
Saumya Anang, Shijian Zhang, Amanda Ennis, Haitao Ding, Ashlesha Deshpande, Hanh Nguyen, John Kappes, Joseph Sodroski
## Abstract
During transport to the surface of infected cells, the human immunodefi ciency virus (HIV-1) envelope glycoprotein (Env) trimer is cleaved to produce the mature functional Env trimer [(gp120/gp41) 3 ]. Env cleavage stabilizes the pretriggered Env conformation (PTC), the major target for broadly neutralizing antibodies. Although the mature Env is relatively enriched in virions and virus-like particles (VLPs), conformation ally flexible uncleaved Envs typically contaminate preparations of these particles. In non-permissive cells, the long ~149-residue gp41 cytoplasmic tail (CT) is necessary for Env incorporation into virions. In a minority of HIV-1 strains, the gp41 CT is clipped in virions by the viral protease. Here, we compare Envs with CT truncations and CT alterations that increase or decrease protease clipping in permissive cells. Changes in protease clipping affected amphotericin B sensitivity but did not alter other viral phenotypes. By contrast, a corresponding CT truncation (L748STOP) increased cell-sur face and virion Env levels, cell-cell fusion, and virus infectivity and cytotoxicity. Notably, in diverse HIV-1 strains, the ratio of cleaved/uncleaved Envs in preparations of virions and extracellular vesicles was increased by this CT truncation. ESCRT and ALIX-binding region (EABR) vesicles incorporated significantly more uncleaved CT-truncated Env than HIV-1 VLPs. Env CT deletion/truncation did not qualitatively alter the viral neutralization profile; however, increased antibody concentrations were required to neutralize viruses with the higher levels of cleaved Env that resulted from CT truncation. Specific CT truncations provide a means of enriching the PTC and limiting the incorporation of nonfunctional and conformationally heterogeneous uncleaved Envs into preparations of virions and VLPs.
IMPORTANCEThe human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer mediates entry of the virus into host cells. The pretriggered conformation (PTC) of Env is the major target for protective broadly neutralizing antibodies, but the PTC is unstable and therefore difficult to study. The cleavage of the flexible Env precursor stabilizes the PTC. Therefore, the presence of uncleaved Env compromises the purity of the PTC in Env preparations. We found that certain truncations of the Env cytoplasmic tail resulted in improved ratios of cleaved:uncleaved Env in preparations of HIV-1 viruses or virus-like particles. In some contexts, cytoplasmic tail truncation increased the level of Env in virus preparations. Although higher concentrations of antibodies were required to neutralize these viruses, Envs with specific truncations of the cytoplasmic tail retained the PTC. Thus, cytoplasmic tail truncation could assist efforts to purify and characterize the Env PTC on the viral membrane.
T he human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer [(gp120/gp41) 3 ] mediates virus entry and cytopathic effects (1)(2)(3)(4)(5). In the infected cell, a fraction of the Env precursor in the endoplasmic reticulum passes through the Golgi, where it is cleaved and modified by complex glycans on the way to the cell surface and virions (6)(7)(8)(9)(10)(11)(12). Prior to receptor binding, virion Envs largely occupy the pretriggered conformation (PTC), which is a major target for small-molecule entry inhibitors and broadly neutralizing antibodies (bNAbs) (13)(14)(15)(16). Uncleaved Env, on the other hand, is conformationally flexible, recognized by poorly neutralizing antibodies (pNAbs) and antigenically distinct from the PTC (17)(18)(19)(20)(21)(22). Therefore, uncleaved Env is an undesirable component in preparations used to study the structure and immunogenicity of the pretriggered Env.
The Envs of HIV-1 and other lentiviruses have unusually long cytoplasmic tails (CTs) (12,23,24). The ~149-residue HIV-1 Env CT consists of a relatively unstructured CT N region (residues 707-752) and a CT C region (residues 753-856) that contains membraneinteractive amphipathic helices (12,(23)(24)(25)(26). The HIV-1 Env CT is palmitoylated and has endocytic motifs ( 712 YXXL 715 and a C-terminal dileucine motif [ 855 LL 856 ]), which interact with clathrin adaptor proteins (27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37). In non-permissive cells, which include primary target cells, the CT is important for the recruitment of Env to virus assembly sites and Env incorporation into HIV-1 virions (38)(39)(40)(41)(42)(43)(44)(45)(46). This process has been reported to involve the interaction of the Env CT with Rab 11 family-interacting protein 1C (FIP1C), a member of a family of proteins that mediate sorting of cargo from the endosomal recycling compartment to the plasma membrane (47)(48)(49). However, a more recent study used cells in which FIP1C expression was knocked out to show that virion incorporation in primary CD4 + T cells and in some cell lines is not dependent on FIP1C (50). In permissive cells, the CT is not required for Env incorporation into virions (38)(39)(40)(41)(42)(43)(44)46). Compatibility between the HIV-1 matrix (MA) protein and the Env CT can influence the efficiency of Env incorporation into virions (51,52). For example, some MA changes that exclude Env from virus particles can be compensated in permissive cells by CT truncation (53,54). Thus, both cell-and virusspecific factors influence the phenotypes of Env CT variants with respect to virion Env incorporation.
Alterations of the Env CT have been reported to affect HIV-1 sensitivity to neutralizing antibodies, host cell factors, and antiviral agents. For some HIV-1 strains, CT truncation has been reported to alter Env antigenicity and decrease virus sensitivity to broadly neutralizing antibodies directed against quaternary V2 gp120 epitopes near the Env trimer apex (55,56). In other cases, however, the absence of the CT exerted little or no influence on susceptibility to neutralization by antibodies (17). Truncation of the CT has been reported to alter the immunogenicity of HIV-1 Env coexpressed with Gag in an mRNA vaccine (57). CT deletions have also been reported to allow HIV-1 to escape from the host restriction factors SERINC5 and IFITMs (58). In the virions of some HIV-1 strains, clipping of the gp41 CT by the viral protease occurs; CT clipping, as well as truncation of the CT, leads to resistance to the cholesterol-binding fungal antibiotic amphotericin B (59)(60)(61)(62). Finally, certain truncations of the CT have been suggested to increase the triggerability of HIV-1 Env, rendering Env more resistant to T20 peptides that recognize gp41 intermediates in the virus entry process (63)(64)(65). The mechanisms underlying the observed effects of Env CT changes on HIV-1 sensitivity to inhibition are not well understood.
Here, we compare the phenotypes of HIV-1 Env mutants with two types of CT alterations: (i) CT changes that increase or decrease CT clipping by the viral protease, and (ii) introduction of stop codons in env that truncate the CT near the membrane or at the viral protease clip site (between residues 747 and 748) (60,62). Except for altered sensitivity to amphotericin B, CT changes affecting viral protease clipping did not significantly alter the viral phenotypes. By contrast, in permissive virus-producing cells, CT-truncated Envs differed from the corresponding full-length Envs in several pheno types. Env variants with CT truncations were efficiently expressed on the cell surface and exhibited increases in the ratio of cleaved:uncleaved Env in preparations of virions, virus-like particles (VLPs), and extracellular vesicles (EVs). The antigenicity and glycosy lation of the CT-truncated Envs were comparable with those of the full-length Envs, consistent with the qualitatively similar neutralization profiles of viruses with these Envs. Correlating with the increased levels of cleaved Env on the virions of some CT-truncated Env mutants, higher concentrations of entry blockers and antibodies were required to achieve neutralization. Our results suggest that specific CT truncations can be a useful means to increase the level of cleaved pretriggered HIV-1 Env in virion/VLP preparations.
## RESULTS
## HIV-1 Env cytoplasmic tail (CT) constructs
In this study, we compared the phenotypes of the full-length HIV-1 AD8 Env with those of two types of Env CT mutants: (i) mutants with alterations in the CT site of viral protease clipping (between residues 747 and 748), and (ii) CT-truncated Envs resulting from the introduction of stop codons in env (Fig. 1A). The Env of HIV-1 AD8 , a primary Tier-2 strain, is efficiently cleaved and therefore was included in most of these studies (11,17,21). The R747L and L748A changes increase and decrease, respectively, the clipping of the CT by the viral protease (60,62). The R747L and L748A changes should manifest their phenotypes primarily after Env has been incorporated into virions. The CT-truncated Envs were produced by introducing stop codons in place of the codons for Tyr 712, Leu 715, and Leu 748. The L748STOP Env has a CT that is truncated at a position corresponding to the site at which clipping of the full-length HIV-1 AD8 Env by the viral protease occurs (60). The phenotypic effects of the CT truncations could potentially manifest in the Env-expressing cells as well as in the virions. The designed set of HIV-1 AD8 Env mutants allows a direct comparison of the phenotypic effects of CT clipping by the viral protease in virions and CT truncation.
## Characterization of Env mutants on pseudovirus VLPs
The Env variants with CT alterations and truncations were expressed with Rev ("Env only") and in recombinant HIV-1 VLP pseudotypes ("Pseudovirus") (Fig. 1B through E). Env expression was evaluated by cotransfection of HEK 293T cells with the pSVIIIenv expressor plasmids and a plasmid encoding the HIV-1 Tat protein (8:1 Env:Tat weight ratio) (Fig. 1B). Forty-eight hours later, Envs in cell lysates were analyzed by western blotting. Envs on the cell surface were precipitated with a mixture of the PGT121 and PGT151 antibodies. The PGT121 antibody recognizes a glycan-dependent gp120 epitope near the third variable (V3) loop, and the PGT151 antibody recognizes a glycan-depend ent epitope spanning the gp120-gp41 interface (67,68). The CT-truncated L715STOP and L748STOP mutants were expressed at the highest levels in cell lysates and on the cell surface (Fig. 1B). The ratio of cleaved:uncleaved Env was higher for the CT-truncated Envs (Y712STOP, L715STOP, and L748STOP) than for the wild-type AD8, R747L, and L748A Envs.
We next evaluated the processing and incorporation of these Envs into pseudotyped VLPs. Pseudovirus VLPs were produced from HEK 293T cells cotransfected with pSVIIIenv, the psPAX2 packaging plasmid, and a luciferase-expressing HIV-1 vector (Fig. 1C). Although all of the Env variants were efficiently expressed, the overall Env level in both the cell lysates and the VLPs was greatest for the L748STOP Env. Most of the Envs in the cell lysates were uncleaved, whereas the cleaved Envs were enriched in the VLPs. The ratio of cleaved:uncleaved Env in the VLPs was similar for the wild-type AD8, R747L, and L748A Envs. By contrast, the CT-truncated Envs exhibited much higher ratios of cleaved:uncleaved Env in the VLPs (see Fig. 1E below for quantitation). Clipping of the gp41 CT in the VLPs was increased by the R747L change and decreased by the L748A change, relative to that of the wild-type AD8 Env, as expected (62). Thus, CT truncations can increase the processing and incorporation of Env into pseudovirus VLPs.
We evaluated the glycosylation of the HIV-1 AD8 Env CT variants on pseudovirus VLPs. We treated lysates of VLPs with protein N-glycosidase F (PNGase F), which removes all Nlinked glycans, or endoglycosidase Hf (Endo Hf ), which removes high-mannose and hybrid glycans but not complex glycans (69), and then analyzed the samples by western blotting (Fig. 1D). The PNGase F digests confirmed that the CT truncations increased Env levels and also improved the ratio of cleaved:uncleaved Env on pseudovirus VLPs (Fig. 1E). The glycosylation patterns of the wild-type HIV-1 AD8 Env and the CT variants were similar. The cleaved Envs on VLPs were all Endo Hf-resistant, indicating that these Envs had passed through the Golgi compartment and had been modified by complex glycans. The VLPs also contained Endo Hf-resistant gp160 that apparently represents Envs that have passed through the Golgi compartment but escaped furin cleavage. In addition, the pseudovirus VLP preparations with the wild-type AD8, R747L, and L748A Envs contained Endo Hf-sensitive gp160 that had not been modified by complex glycans (Fig. 1D, lower right panel). This high-mannose/hybrid glycan-containing gp160 may be a component of extracellular vesicles that contaminate the pseudovirus VLP preparations (see below).
In summary, all of the cleaved and most of the uncleaved Envs on virions, regardless of Env CT structure, have acquired complex carbohydrates and therefore have passed through the Golgi network.
## Characterization of Env CT mutants on virions
Next, we examined the effects of CT changes on the quantity and quality of Env on virions produced by an infectious molecular proviral clone (IMC). The migration of the wild-type AD8, R747L, and L748A gp41 glycoproteins in virions was consistent with partial viral protease clipping for the wild-type AD8 Env, nearly complete clipping for the R747L Env, and minimal clipping for the L748A Env (Fig. 2A). For all the Env variants, a higher ratio of cleaved:uncleaved Env was observed on virions than in cell lysates, consistent with previous observations that the processing of virion/VLP Envs produced from IMCs or defective proviruses is efficient (70)(71)(72). These results indicate that even in the case of virions produced by IMCs, which typically are enriched in mature Envs (70)(71)(72), CT-truncated Envs are efficiently expressed and maintain a high ratio of cleaved:uncleaved Env. We evaluated the glycosylation of the HIV-1 AD8 Env CT variants on virions produced from proviral IMCs. Virion lysates were treated with PNGase F or Endo Hf and then analyzed by western blotting (Fig. 2B). The glycosidase digestions confirmed that CT truncations increased Env levels on virions. The cleaved Envs on virions were all Endo Hfresistant, indicating that these Envs had passed through the Golgi compartment and had been modified by complex glycans. The small amount of gp160 on virions produced from IMCs was mostly Endo Hf-resistant and therefore modified by complex carbohy drates. This gp160 fraction has apparently passed through the Golgi compartment but escaped furin cleavage. Thus, regardless of Env CT structure, the Envs on virions pro duced by proviruses have acquired complex carbohydrates and therefore have passed through the Golgi network.
We confirmed the mechanism of gp41 CT clipping (60,62) by treating HEK 293T cells transfected with an IMC proviral plasmid expressing the wild-type HIV-1 AD8 Env with ritonavir, an inhibitor of the HIV-1 protease (73). Cells transfected with the proviral plasmid that were not treated with ritonavir were studied in parallel. The virions from the untreated and ritonavir-treated cells were pelleted and lysed for western blotting. The later, the cell supernatants were clarified by a low-speed spin, filtered through a 0.45 µm membrane, and centrifuged at 14,000 × g for 1 h at 4°C. In parallel, the cells were lysed. The pelleted VLPs and clarified cell lysates were western blotted with a goat anti-gp120 antibody (upper panels), the 4E10 anti-gp41 antibody (middle panels) or antibodies against Gag p55/p24/p17 or hsp70 (lower panels). (D) Pseudovirus VLPs produced in the experiment shown in C were lysed and treated with PNGase F or Endo Hf. The deglycosylated Envs were analyzed by western blotting with a goat anti-gp120 antibody (upper panels) and with the 4E10 anti-gp41 antibody (lower panels). Deglycosylated forms of Env are designated with a "D" (e.g., dgp160). In C and D, CT-truncated or CT-clipped forms of gp160 and gp41 are asterisked. The results shown in B-D are representative of those obtained in more than two independent experiments. (E) Western blots of the PNGase F digests in D and the blots from repeat experiments (not shown) were used to measure the ratio of gp120/gp160 on VLPs pseudotyped by the indicated Env variants. The statistical significance of the observed differences was evaluated with an unpaired t-test. Significant differences are indicated (**, P < 0.01). Forty-eight hours later, the cell supernatants were clarified by a low-speed spin, filtered through a 0.45 mm membrane, and centrifuged at 14,000 x g for 1 hour at 4 degrees Celsius. In parallel, the cells were lysed. The pelleted virions and clarified cell lysates were western blotted with a goat anti-gp120 antibody (upper panels), the 4E10 anti-gp41 antibody (middle panels) or antibodies against Gag p55/p24/p17 or hsp70 (lower panels). partial clipping of the gp41 glycoprotein observed for the wild-type HIV-1 AD8 Env was not observed after ritonavir treatment (Fig. 2C). These results are consistent with the clipping of the HIV-1 AD8 gp41 CT in virions being mediated by the viral protease (60,62).
## Characterization of Env CT mutants on VLPs, virions, and extracellular vesicles
The above studies suggested that truncations of the Env CT could potentially be useful in preparing VLPs and virions with either increased Env levels or improved cleaved:uncleaved Env ratios. We compared the expression and processing of the wild-type AD8 and CT-truncated Envs in pseudovirus VLPs, provirus-produced virions, and extracellular vesicles (EVs). EVs were prepared from the medium of cells expressing Env but no Gag proteins. For all three systems, the expression and processing of the full-length AD8 and CT-truncated Envs in the cell lysates were comparable (Fig. 2D, left panel). The EVs incorporated more wild-type AD8 Env than the L715STOP or L748STOP Envs (Fig. 2D, right panel). Most of the AD8 Env in EVs was uncleaved. Although the CT-truncated Envs were incorporated into EVs very inefficiently, these Envs were mostly cleaved in the EV preparations. In the pseudovirus VLPs, the L748STOP Env exhibited higher levels and a better cleaved:uncleaved Env ratio compared with the wild-type AD8 Env. Both the wild-type AD8 and L748STOP Envs in virions produced from a provirus were efficiently cleaved. Our results underscore the advantage of the L748STOP Env in producing pseudotyped VLP preparations with higher levels of mature Env from transiently expressing HEK 293T cells. The quality of the L748STOP pseudovirus VLPs may be improved by the negligible incorporation of CT-truncated Envs into EVs that potentially contaminate these VLP preparations.
## Phenotypes of CT truncation in Envs of other HIV-1 strains
To determine if the phenotypes observed for the HIV-1 AD8 L748STOP Env hold true for other primary HIV-1 Envs, we introduced this change into the Envs of six HIV-1 strains from multiple clades. To evaluate Env expression and processing, the Env expressor plasmids were transfected into HEK 293T cells along with the psPAX2 plasmid expressing HIV-1 packaging proteins. The cell lysates and VLPs were western blotted (Fig. 3A).
The expression levels of the full-length Envs and their L748STOP Env counterparts in cell lysates were comparable. As seen in the case of the HIV-1 AD8 Env, different gp41 glycoforms were observed. Unlike the case for the HIV-1 AD8 Env, the L748STOP truncation of the CT did not increase the levels on VLPs of Envs from other HIV-1 strains. However, in every HIV-1 strain examined, the cleaved:uncleaved Env ratio on VLPs was greater for the L748STOP Env compared with the full-length Env (Fig. 3B). Thus, in multiple diverse HIV-1 strains, the L748STOP CT truncation can enrich the proportion of mature Env in pseudovirus VLP preparations. We studied the phenotypes resulting from CT modifications of Envs on virions produced by IMCs (Fig. 3C). The L748STOP change was introduced into the Envs encoded by the HIV-1 NL4-3 , HIV-1 JR-FL , and HIV-1 ADA Env IMCs. The HIV-1 ADA Env is derived from the same virus isolate as the HIV-1 AD8 Env, and these Envs are thus related, but not identical (74,75). In addition, the L715STOP and R747L changes were introduced into the HIV-1 NL4-3 Env encoded by the IMC. The IMCs were transfected into HEK 293T cells, and the virions produced in the cell medium were analyzed. The full-length HIV-1 NL4-3 Env and CT variants were expressed at similar levels on virions. The HIV-1 NL4-3 R747L Env was packaging proteins ("Pseudovirus"); or (iii) pNL4-3 proviral plasmids expressing the indicated Envs ("provirus"). Forty-eight hours later, the cell supernatants were clarified by low-speed centrifugation. The pseudovirus and provirus-produced virions were filtered through a 0.45 µm membrane. The cell supernatants were all centrifuged at 14,000 × g for 1 h at 4°C. In parallel, the cells were lysed. Clarified cell lysates and the pellets from the cell supernatants (containing extracellular vesicles, VLPs, and virions) were western blotted as described in the Fig. 1D clipped in the gp41 CT, as expected (60). Compared with the other HIV-1 NL4-3 Env variants, the HIV-1 NL4-3 L748STOP Env was processed efficiently, and gp160 was relatively excluded from virions. The cleaved:uncleaved ratios of the HIV-1 JR-FL L748STOP and HIV-1 ADA L748STOP Envs on virions were increased compared with those of the respec tive full-length Envs. These results confirm that the L748STOP CT truncation can reduce the relative level of uncleaved Env on virion preparations from multiple HIV-1 strains. The Envs of primary HIV-1 strains mediated cell-cell fusion more efficiently when the CT was truncated by the L748STOP change (Fig. 3D). At least some of this increased syncytium-forming ability results from increased cell-surface expression of the CT-trunca ted Envs (Fig. 1B).
## Infectivity and cytopathic effects of Env CT mutants
To measure the ability of the Envs to support virus entry and cytopathic effects, IMCs encoding the Env variants were transfected into HEK 293T cells. The virus-containing cell supernatants were incubated with TZM-bl target cells. Forty-eight hours later, luciferase activity in the TZM-bl cells, as well as the cell viability, was measured. The infectivity of the Y712STOP and L748STOP viruses was higher than that of the viruses with the wild-type AD8 and other mutants Envs (Fig. 4A, upper left panel). When higher amounts of the Y712STOP and L748STOP viruses were added to the TZM-bl cells, the measured luciferase activity decreased. This decrease was related to the greater induction of cytopathic effects by the Y712STOP and L748STOP viruses (Fig. 4A, lower left panel). As expected, an alteration of the PTAP motif of the Gag polyprotein that reduces ESCRTmediated virion release (76-78) reduced virus infectivity and cytopathic effects (Fig. 4A, middle panels). Regardless of this alteration in the Gag PTAP motif, viruses with the L748STOP Env killed the target cells more efficiently than the corresponding viruses with the wild-type AD8 Env. We wished to address whether the expression of HIV-1 proteins, including Env, in the target cells is necessary for the induction of cytopathic effects in this experimental system. To this end, we introduced the wild-type AD8 and the L748STOP env genes into the def5 provirus. The def5 provirus is replication-incompetent due to defective genes for reverse transcriptase, RNase H, integrase, Vif, and Vpr; nonetheless, transfection of the def5 plasmid results in the production of VLPs containing Env (72). We also introduced the PTAP gagLIRL changes into the def5 proviruses. The infectivity of the def5 VLPs was very low, at the background of the assay, as expected (Fig. 4A, upper right panel). Nonetheless, VLPs with the L748STOP Envs induced greater cytopathic effects than VLPs with the wild-type AD8 Env (Fig. 4A, lower right panel). These results indicate that VLPs with the L748STOP Env can kill the TZM-bl cells even without initiating a productive infection. These results also suggest that the L748STOP Env is more cytotoxic than the wild-type AD8 Env.
## Susceptibility of viruses with Env CT variants to inhibition
Changes in the HIV-1 Env CT have been suggested to alter the sensitivity of viruses to small-molecule inhibitors and neutralizing antibodies (55,56,(58)(59)(60)(61)(62)(63)(64)(65). We examined the effect of different inhibitors on the infectivity of recombinant viruses pseudotyped by the wild-type AD8 and CT mutant Envs. We first evaluated the effects of cholesterolmodifying agents on the infectivity of the pseudotyped viruses. Methyl-β-cyclodextrin (MBCD), which depletes cholesterol from membranes (80-82), and cholesterol oxidase, which modifies membrane cholesterol (83), inhibited the viruses with the wild-type AD8 Env and CT mutant Envs comparably (data not shown). Next, we investigated the susceptibility of the viruses to amphotericin B, using the less toxic amphotericin B methyl ester (AME). We tested the hypothesis that the antiviral activity of AME depends upon cholesterol in the target membrane (60,(84)(85)(86)(87)(88). For this purpose, we incubated the pseudotyped viruses with a subneutralizing concentration (3 mM) of MBCD or control buffer for 1 h at 37°C. The viruses were then pelleted to remove MBCD, resuspended, and incubated with various concentrations of AME for 1 h at 37°C. The viruses were added to target cells, and after 2 days of culture, luciferase activity in the cells was measured. For The pseudoviruses were then pelleted, resuspended, and incubated with amphotericin B methyl ester at different concentrations for 1 h at 37°C, and added to Cf2Th-CD4/CCR5 cells. After 2 days, the cells were lysed, and luciferase activity was measured. The reported IC 50 value of the antibody/inhibitor was calculated by fitting the data in four-parameter dose-response curves using GraphPad Prism 9. The half-life of infectivity of the pseudotyped viruses incubated on ice (0°C) was also measured. Relative to viruses with wild-type HIV-1 AD8 Env, greater than 2-fold increases (green) or decreases (red) in virus sensitivity to inhibition are indicated. The results shown are representative of those obtained in more than three independent experiments. the control viruses not treated with MBCD, the viruses with the wild-type AD8 Env and L748A Env were efficiently inhibited by AME (Fig. 4B). The viruses with the R747L and L748STOP Envs were relatively resistant to AME, consistent with the proposed role of the Env CT as a determinant of amphotericin B sensitivity (59,60,89). The viruses treated with 3 mM MBCD were completely resistant to AME (Fig. 4B). These results support a model for amphotericin B antiviral activity that depends upon cholesterol in the viral membrane and upon the HIV-1 Env CT (60,89).
The sensitivity of HIV-1 infection to inhibition by specific Env ligands or by exposure to cold can provide an indication of changes in Env conformation (14,15,70,71,(90)(91)(92)(93)(94)(95). We examined the susceptibility of recombinant viruses pseudotyped with the HIV-1 AD8 Env CT variants to inhibition by conformation-sensitive ligands and by cold exposure. The conformation-sensitive ligands included broadly and poorly neutralizing antibodies, sCD4-Ig, the T20 gp41 heptad repeat (HR2) region peptide, BNM-III-170 (a CD4-mimetic compound), and BMS-806. The overall patterns of sensitivity of the viruses with the wild-type AD8 and CT variant Envs to these treatments were qualitatively similar (Fig. 4B). However, in several cases, higher concentrations of antibodies/compounds were required to neutralize the viruses pseudotyped by the L712STOP, L715STOP, and L748STOP Envs. Viruses with the L712STOP, L715STOP, and L748STOP Envs also exhibited longer infectious half-lives after incubation on ice. Dilution of the CTmodified pseudo virus preparations reduced basal infectivity but did not change the higher bNAb IC 50 values for the L715STOP and L748STOP pseudoviruses compared with those of the wild-type AD8 pseudoviruses (data not shown).
To investigate the mechanism of the relative resistance of viruses with CT-truncated Envs, we examined the composition of the pseudovirus particles. Consistent with the results shown in Fig. 1, the L715STOP and L748STOP pseudoviruses exhibited a higher level of cleaved Env than the pseudoviruses with the wild-type AD8, R747L, and L748A Envs (Fig. 5A). We hypothesized that an increased level of functional Env on the viral surface results in a requirement for higher concentrations of bNAbs to neutralize the L715STOP and L748STOP pseudoviruses. To test this hypothesis, we generated pseudo viruses with lower levels of the L715STOP and L748STOP Envs by transfecting smaller amounts of the plasmids encoding these CT-truncated Envs. The level of cleaved Env (gp120/Gag p24 ratio) on the pseudoviruses was directly related to the amount of the Env-expressing plasmid transfected (Fig. 5B andC, left panels). The infectivity of the pseudoviruses increased and then reached a saturating level as the amount of cleaved Env on the virus increased (Fig. 5C, right panels). We tested the sensitivity of the L715STOP and L748STOP viruses with different levels of cleaved Env to neutralization by the PGT151 and 35O22 bNAbs. As the amount of cleaved Env on the L715STOP and L748STOP viruses decreased below the maximum, the IC 50 values of both bNAbs decreased dramatically, approximating that seen for the wild-type AD8 virus (Fig. 5D). These results support the hypothesis that the number of functional, cleaved Envs on the viral surface can influence the concentration of antibody required for neutralization. The non-linear relationship between the Env content of the pseudovirus preparations and the IC 50 values of the bNAbs has additional implications that will be addressed in the Discussion. The viruses pseudotyped by the wild-type HIV-1 AD8 Env and the CTmodified Envs exhibited similar patterns of resistance to pNAbs and, after adjustment of virion Env levels, sensitivity to bNAbs. These observations are consistent with these functional Env variants mainly occupying the PTC.
## Effect of host restriction factors on the infectivity of HIV-1 with CT-truncated Envs
We asked if the host restriction factors SERINC5 and IFITM (96)(97)(98) inhibited the infectiv ity of pseudoviruses with the wild-type HIV-1 AD8 Env and CT-truncated Envs equivalently. We produced virions by transfecting IMCs encoding the wild-type HIV-1 AD8 Env, the L715STOP Env, or the L748STOP Env with a plasmid expressing human SERINC5. In parallel negative-control experiments, we cotransfected a plasmid, ΔSERINC5, that contains a stop codon eliminating SERINC5 expression. The virions were added to TZM-bl cells, and the infectivity was evaluated by measuring luciferase activity in the cells. The viruses with the wild-type HIV-1 AD8 Env and the CT-truncated Envs were not inhibited by SERINC5 (Fig. S1A). We also tested viruses with the laboratory-adapted HIV-1 NL4-3 Env and its CT-truncated version NL4-3 L748STOP; these viruses were inhibited equivalently by SERINC5 (Fig. S1B), in contrast with a previous report (58). We also deleted nef from the IMCs producing the wild-type HIV-1 AD8 and L748STOP Env variant and tested the viruses for sensitivity to SERINC5; the nef deletion did not render either virus sensitive to SERINC5 (data not shown). The expression of SERINC5 in the above studies was docu mented by western blot (Fig. S1C). The wild-type HIV-1 AD8 and L715STOP viruses were inhibited by another host restriction factor, IFITM (Fig. S1D). These results indicate that truncation of the Env CT had little effect on the inhibition of the tested HIV-1 variants by SERINC5 or IFITM.
## Inducible expression of full-length and CT-truncated Envs alone and in VLPs
To evaluate the Env phenotypes associated with CT truncation in another context, we studied A549 cell lines that inducibly express the full-length TFAR, TFAR-L715STOP, and TFAR-L748STOP Envs. As mentioned above, the TFAR Env is a PTC-stabilized derivative of the HIV-1 AD8 Env (71,72). The doxycycline-inducible polyclonal A549 cell lines were enriched for Env expression by selection with the PGT145 bNAb, which preferentially recognizes cleaved Env (67,99,100). In preliminary experiments, we found that the overall expression of the CT-truncated TFAR Envs was higher than that of the full-length TFAR Env; therefore, we used half the concentration of doxycycline to induce the cell lines expressing CT-truncated Envs as that used to induce the cell lines expressing the TFAR Env. With this adjustment, roughly equivalent levels of uncleaved and cleaved Env expression were attained in the lysates of cells expressing the TFAR, TFAR-L715STOP, and TFAR-L748STOP Envs (Fig. 6A). The TFAR-L715STOP and TFAR-L748STOP Env levels on the cell surface and in EVs prepared from the cell medium were higher than those of the TFAR Env. Most of the cell surface and EV Envs were cleaved. The amount of uncleaved TFAR-L748STOP Env on the cell surface was less than that of the TFAR-L715STOP Env. Nearly all of the Envs in EVs contained complex carbohydrates (Fig. 6B). Thus, in the A549 cells, the increased levels of cell-surface and EV Envs observed for the CT-truncated Envs, relative to the full-length Env, derive from Envs that have trafficked through the Golgi.
We also studied A549 cell lines that inducibly express replication-defective def5 HIV-1 proviruses with the TFAR and TFAR-L748STOP Envs; these cell lines, which produce defective VLPs, are designated D1712 (with the TFAR Env) and D1758 (with the TFAR-L748STOP Env) (72). Both cell lines were induced with 2 µg/mL doxycycline. The total levels of TFAR Env and TFAR-L748STOP Env expressed in the cell lysates and VLPs were similar (Fig. 6A). The two Envs were also processed comparably.
Nearly all the TFAR and TFAR-L748STOP Envs in the VLPs contained complex glycans (Fig. 6B andC). The L748STOP truncation resulted in decreased incorporation of the uncleaved Env into the VLPs (Fig. 6A through C). Thus, even in VLPs produced from HIV-1 proviruses, which generally exhibit high ratios of cleaved:uncleaved Env (11,70,72), the L748STOP truncation can help decrease the amount of uncleaved Env.
We compared the antigenic profile of the TFAR and TFAR-L748STOP Envs on the surface of the A549 cells transduced with def5 proviruses and on VLPs (Fig. 7). On the cell surface, both cleaved Envs were recognized by bNAbs but not by pNAbs, as expected for Envs in a PTC. The cleaved TFAR-L748STOP Env on the cell surface bound the b12 CD4binding site (CD4BS) bNAb better than the cleaved TFAR Env; another CD4BS bNAb, 3BNC117, bound the TFAR-L748STOP Env slightly worse than the TFAR Env (Fig. 7A andB, left panels). The b12 bNAb has been suggested to bind an occluded intermediate conformation distinct from the PTC, which is recognized by other CD4BS bNAbs (101)(102)(103)(104)(105)(106). The antigenic profiles of the TFAR and TFAR-L748STOP Envs on VLPs were indistin guishable (Fig. 7A andB, right panels). Thus, the VLPs produced by the D1758 cell line represent a source of mature, PTC-stabilized HIV-1 Env for further study.
## Incorporation of CT-truncated Envs and ESCRT-and ALIX-binding region (EABR)-tagged Envs into pseudovirus VLPs and EVs
Membrane proteins can be induced to self-assemble into EVs that bud from the cell membrane by directly recruiting proteins from the Endosomal Sorting Complex Required for Transport (ESCRT) (107). This recruitment is mediated by a short peptide called the FIG 6 The effect of CT truncations on a PTC-stabilized Env expressed in stable, inducible cell lines. We established stable, inducible A549 cell lines expressing a PTC-stabilized Env (TFAR) and TFAR-L715STOP and TFAR-L748STOP variants. We also established A549 cell lines (designated D1712 and D1758) that inducibly express replication-defective VLPs containing the full-length TFAR Env and the TFAR-L748STOP Env, respectively. (A) These A549 cell lines were induced with 2 µg/mL doxycycline (for the full-length Envs) and 1 µg/mL doxycycline (for the CT-truncated Envs) to achieve roughly comparable Env expression in the cells.
Forty-eight hours after induction, clarified cell supernatants containing VLPs were filtered (0.45 µm), and the cell supernatants containing EVs and/or VLPs were centrifuged at 14,000 × g for 1 h at 4°C. In parallel, cell lysates were prepared. Cell-surface Envs were analyzed by incubation of intact cells with a mixture of the PGT121 and PGT151 antibodies, as described in Materials and Methods. The cell-supernatant pellets, cell lysates, and immunoprecipitated cell-surface Envs were analyzed by western blotting with a goat anti-gp120 antibody, the 4E10 anti-gp41 antibody, an anti-Gag antibody, and an antibody against hsp70.
(B) Extracellular vesicles and VLPs produced from the induced A549 cells in A were treated with PNGase F or Endo Hf, as described in Materials and Methods. The deglycosylated Envs were analyzed by western blotting with a goat anti-gp120 antibody and the 4E10 anti-gp41 antibody. The results shown in panels A and B are representative of those obtained in three independent experiments. (C) To achieve comparable levels of the TFAR and TFAR-L748STOP Envs in VLPs, the D1712 and D1758 cell lines were both induced with 2 µg/mL doxycycline. Lysates of cells and VLPs were prepared and analyzed by western blotting as in A (left panel). The samples were also digested with PNGase F and Endo Hf and analyzed by western blotting with a goat anti-gp120 antibody (middle and right panels).
The Env bands resulting from PNGase F treatment are labeled red, and the Env bands from Endo Hf treatment are labeled green. ESCRT-and ALIX-binding region (EABR) positioned within the cytoplasmic tail of the membrane protein. We compared the quantity and quality of CT-truncated HIV-1 AD8 Env variants incorporated into EABR EVs and pseudovirus VLPs. We added the EABR sequence to the HIV-1 AD8 Env CT at positions 715 and 753 to allow comparison with the HIV-1 AD8 L715STOP and L748STOP Envs, respectively. We also added the EABR sequence at CT position 753 of the TFAR Env, which is an HIV-1 AD8 derivative stabilized in the PTC (71, prepared, and cell supernatants were cleared by low-speed centrifugation and then centrifuged at 14,000 × g for 1 h at 4°C. Lysates of cells and pellets prepared from the supernatants were analyzed by western blotting, either directly (Fig. 8A) or after PNGase F/Endo Hf digestion (Fig. 8B).
The EABR sequences dramatically increased the incorporation of the AD8 and TFAR Envs into EVs, compared with the AD8-L748STOP and TFAR-L748STOP Envs ("Env only" lanes in Fig. 8A andB). Although the AD8-L748STOP and TFAR-L748STOP Envs in EVs were mostly cleaved, the AD8-753EABR and TFAR-753EABR Envs in the EVs consisted of a mixture of cleaved and uncleaved Envs. Substantial fractions of the AD8-753EABR and TFAR-753EABR Envs that were uncleaved at the gp120-gp41 junction underwent another cleavage to produce fragments that migrated around 54 kDa after PNGase F digestion (labeled dgp[V3 +gp41] in Fig. 8B). The dgp[V3 +gp41] fragment contains both gp120 and gp41 components and likely results from adventitious proteolytic cleavage in the gp120 V3 loop (108)(109)(110)(111). The AD8-753EABR and TFAR-753EABR Envs consisted of Endo Hf-resistant and Endo Hf-sensitive fractions, indicating that at least some of these Envs were modified by complex carbohydrates (Fig. 8B). Similar results were seen for the AD8-715EABR Env (data not shown). Even on the surface of expressing cells, the AD8-753EABR and TFAR-753EABR Envs are cleaved less than their respective AD8-L748STOP and TFAR-L748STOP Env counterparts (Fig. 8C). As expected (17)(18)(19)(20)(21)(22), gp120-gp41 cleavage resulted in better recognition of the AD8-L748STOP and TFAR-L748STOP Envs by several bNAbs. By contrast, even the PTC-stabilizing changes in the TFAR-753EABR Env do not preserve the PTC, consistent with earlier suggestions that these changes stabilize the PTC much more efficiently in the context of cleaved Env (71). In summary, the large amount of uncleaved EABR-tagged Envs and the adventitious cleavage of gp120 in the EVs make them undesirable for studies of the Env PTC.
We analyzed the particles in the supernatants of the cells expressing the L748STOP and 753EABR Envs, HIV-1 packaging proteins, and an HIV-1 luciferase-expressing vector (Pseudovirus lanes in Fig. 8A andB). The AD8-753EABR and TFAR-753EABR particu late fractions included substantial amounts of Env lacking gp120-gp41 cleavage and exhibiting evidence of adventitiously cleaved gp120 (Fig. 8A andB). As these particulate fractions likely include EVs, this was expected. This fraction also includes recombinant viruses pseudotyped with the AD8-753EABR and TFAR-753EABR Envs that can infect TZM-bl cells (Fig. 8D). As expected, the EVs containing these EABR-tagged Envs did not detectably infect the TZM-bl cells in this assay; incubation of the TZM-bl cells with these EVs did not result in detectable cytotoxicity (data not shown).
In contrast to the pseudovirus preparations with the EABR-tagged Envs, the VLPs pseudotyped with the AD8-L748STOP and TFAR-L748STOP Envs contained very high percentages of cleaved Envs (Fig. 8A andB). Comparison of the pseudovirus samples digested with PNGase F and Endo Hf indicated that nearly all the AD8-L748STOP and TFAR-L748STOP Envs on the VLPs were modified by complex glycans. As previously observed (11,70,72), Env passage through the Golgi appears to be required for efficient incorporation into VLPs/virions. The pseudoviruses with the AD8-L748STOP and TFAR-L748STOP Envs infected TZM-bl cells with comparable efficiencies (Fig. 8D). Our results indicate that the proteolytic processing of the L748STOP Envs on HIV-1 VLPs is superior to that of EABR-tagged Envs on VLPs or EVs. Thus, we expect that VLPs and virions with CT truncations, when produced in permissive cells, represent a potentially useful source of mature and functional Envs, including PTC-stabilized Envs.
## DISCUSSION
The identification of homogeneous sources of the PTC of membrane HIV-1 Envs would assist studies of the structure and immunogenicity of this metastable state. In cells producing viruses or VLPs, HIV-1 Envs are transported to the cell surface by two pathways (11) (Fig. 9). In the conventional secretory pathway, Envs in the Golgi are modified by complex glycans and cleaved on their way to the cell surface. Golgi-processed Envs are preferentially incorporated into virions via a process that, in non-permissive cells, is dependent upon the CT (23,24,(39)(40)(41)(42)(43)(44)(45)(46)(47)(48)(49). A fraction of Env bypasses the Golgi and is transported to the cell surface in a form that is uncleaved and lacks complex carbo hydrates (11). Although largely excluded from virions, these Envs can be incorporated into EVs, a common contaminant of virion/VLP preparations. Envs that are uncleaved as a result of either bypassing the Golgi or overwhelming the Golgi furin protease are conformationally flexible and recognized by pNAbs (17)(18)(19)(20)(21)(22)112). Therefore, minimizing or eliminating uncleaved Env in preparations of virions/VLPs would enhance Env PTC homogeneity.
We investigated the potential effect of modification of the Env CT on the incorpora tion of cleaved and uncleaved Envs into virions, VLPs, and EVs in permissive cells. The CT was completely removed from the Y712STOP and L715STOP Envs. For some natural HIV-1 Env variants, partial clipping of the Env CT by the viral protease occurs (60). Based on the site in the CT at which this clipping occurs, we studied the effects of a corresponding Env truncation (L748STOP) on Env phenotypes (Fig. 9). Compared with the full-length Env, in permissive HEK 293T cells, the CT truncations resulted in higher Env levels on the cell surface and increased ratios of cleaved:uncleaved Env in VLPs pseudotyped with Envs from multiple HIV-1 strains. In virions produced in HEK 293T cells from proviral IMCs, in which Env processing is generally efficient (70,72), CT truncation allowed efficient virion Env incorporation while maintaining a high cleaved:uncleaved Env ratio. Although the different CT-truncated Envs shared many phenotypes, we found that the L748STOP mutant consistently exhibited efficient expression and processing and therefore included this mutant in most of the experiments. In HEK 293T cells, the high levels of cleaved L748STOP Env on the cell surface are apparently available for incorporation into virions and VLPs but are not efficiently incorporated into EVs (Fig. 9). In these cells, the endocytosis signals on the CT likely influence the efficiency with which cell-surface Envs are recycled into endosomes/multivesicular bodies (23,24,(33)(34)(35)(36)(37). Exosomes, a major class of EVs, are generated by inward budding of the late endosomes/multive sicular bodies (113)(114)(115)(116). At diameters of 30-150 nm, exosomes overlap HIV-1 virions in size and thus are frequent contaminants of virion/VLP preparations (72). Although the full-length Env in EVs exhibits more uncleaved than cleaved Envs, resembling the proportion on the HEK 293T cell surface, the small amount of the L748STOP Env in EVs is completely cleaved. Therefore, contaminating EVs contribute less uncleaved Env to VLP preparations in the case of the L748STOP Env. For both full-length and CT-truncated Envs, the cleaved Envs on virions/VLPs are modified by complex glycans. In preparations of VLPs produced in HEK 293T cells, a fraction of the full-length uncleaved Env is Endo Hf-sensitive, suggesting that these Envs may have bypassed the Golgi. On the other hand, the uncleaved L748STOP Env in VLPs is Endo Hf-resistant, consistent with passage through the Golgi. Thus, compared with the full-length Env expressed in HEK 293T cells, a higher proportion of the L748STOP Env appears to traffic through the Golgi, undergo cleavage, populate the cell surface, and become incorporated into virions/VLPs. At the same time, L748STOP Envs on the HEK 293T cell surface, some of which are uncleaved, largely avoid incorporation into exosomes/EVs (Fig. 9).
We recently reported the inducible production from A549 cells of replication-defec tive VLPs containing Envs stabilized in a PTC (72). To evaluate whether the observations made in HEK 293T cells might apply to another, more practical VLP production sys tem, we compared A549 cells producing VLPs with either the full-length or L748STOP PTC-stabilized TFAR Env. During their establishment (72), the A549 cell lines had been enriched for cell-surface expression of cleaved Env by selection with the PGT145 bNAb, which preferentially recognizes the mature Env (99,100). As in the HEK 293T cells, compared with the full-length TFAR Env, the TFAR L748STOP Env exhibited a higher overall level of expression and much higher cell-surface expression of cleaved Envs (Fig. 9). Perhaps because of the selection with the PGT145 bNAb, nearly all the full-length and L748STOP TFAR Envs on VLPs and EVs were cleaved. A small amount of Endo Hf-resistant full-length TFAR Env on VLPs was uncleaved; relatively little uncleaved TFAR L748STOP Env was detected on VLPs but was also Endo Hf-resistant. These low levels of uncleaved Env have apparently passed through the Golgi and acquired some complex glycans but escaped furin processing. More cleaved TFAR L748STOP Env was observed on EVs compared with the full-length TFAR Env. Unlike the case in HEK 293T cells, in A549 cells, the CT does not appear to be important for efficient Env incorporation into EVs. The levels of full-length Env and L748STOP Env in the EVs from A549 cells reflect the levels of those Envs on the cell surface; these EVs may represent microvesicles, which bud directly from the plasma membrane (113)(114)(115)(116). In summary, even in cells selected for expression of cleaved HIV-1 Envs, the L748STOP truncation can decrease the small amount of uncleaved Env incorporated into VLP preparations.
We also evaluated HIV-1 Env-EABR fusion proteins as a means of producing EVs that efficiently incorporate oriented Env trimers, an approach that produced immunogenic SARS-CoV-2 S glycoprotein EVs (74). Unfortunately, a substantial fraction of the HIV-1 Env-EABR fusion proteins was not proteolytically cleaved at the gp120-gp41 junction. This was the case regardless of the placement of the EABR sequence at residue 715 or 753 of the CT or whether the wild-type HIV-1 AD8 Env or the more efficiently processed TFAR Env was used. Adventitious cleavage of the HIV-1 Env-EABR fusion proteins in gp120, presumably in the V3 loop (108)(109)(110)(111), further suggested that the Env compo nents of these EVs are conformationally heterogeneous. The cell-surface Env-EABR fusion proteins were largely uncleaved and recognized by pNAbs but not the PGT145 bNAb. At least a fraction of the HIV-1 Env-EABR fusion proteins is functional, as these proteins supported the entry of a luciferase-expressing pseudovirus. Based on these results, HIV-1 VLP preparations containing the CT-truncated Envs appear to be a much better source of homogeneous, cleaved, and PTC-stabilized Env than the Env-EABR EVs.
Preserving the native structure of membrane HIV-1 Env in VLP preparations is important to the relevance and reliability of studies utilizing these particles. As some truncations of the CT have been reported to affect the conformation of the Env ectodomain (55,56), we evaluated the susceptibility to inhibition of recombinant viruses with full-length and CT-truncated Envs. Env CT truncations had no significant effect on the susceptibility of the tested HIV-1 strains to the host restriction factors SERINC5 and IFITM, a result apparently at odds with those reported in reference 58. Viruses with the L715STOP and L748STOP Envs exhibited increases in the IC 50 values for multiple bNAbs, sCD4-Ig, T20, the CD4mc BNM-III-170, and BMS-806. However, viruses with these CT-truncated Envs remained resistant to neutralization by pNAbs. Thus, although the qualitative pattern of sensitivity is similar for the viruses with the wild-type HIV-1 AD8 and CT-truncated Envs, quantitative differences in the IC 50 values of multiple ligands resulted from the increased concentrations of CT-truncated Envs on virions. When the virion levels of the L715STOP and L748STOP Envs were reduced, the bNAb IC 50 values were similar for viruses with the wild-type HIV-1 AD8 Env and the CT-truncated Envs. These results suggest that the wild-type HIV-1 AD8 Env and the Envs with CT truncations beginning at residues 715 and 748 have similar ectodomain conformations.
The observed relationships between virion Env levels, virus infectivity, and bNAb concentrations required for neutralization provide information relevant to studies of HIV-1 entry stoichiometry. Definition of T, the number of Env trimers required for entry, is complicated by the low infectivity:particle ratio of HIV-1 preparations (117)(118)(119)(120). This leads to uncertainties about the functional virion Env number, distribution, mobility, and ability to participate in productive entry events; all of these parameters can potentially influence the estimation of T (121)(122)(123)(124)(125)(126). We encompass these multiple parameters in the variable n t , which describes the number of functional Env trimers on a virus particle that can participate in the virus entry process. On an infectious virion, n t must be greater than or equal to T. At lower Env levels on the virions, over a 40-fold range of gp120:Gag p24 ratios, there is a direct relationship between the virion Env levels and infectivity (Fig. 5C). In this range of gp120:Gag p24 values, infectivity is apparently limited by n t . Under these circumstances, if Env trimers are incorporated independently into virus particles, then n t will be only marginally greater than T on most infectious virions (Fig. S2). Over this 40-fold range of virion Env levels, the neutralization of viruses with the wild-type HIV-1 AD8 Env and CT-truncated Envs by the PGT151 and 35O22 bNAbs was comparable (Fig. 5D). The highest level of CT-truncated Envs on virions achieved in our experimental system was beyond the Env level required to saturate virus infectivity. This redundancy in the amount of virion Env needed for entry likely contributes to the relative inefficiency with which these viruses were neutralized by antibodies, entry inhibitors, and cold exposure. We note that neutralization of viruses with the highest levels of CT-truncated Envs by the 35O22 bNAb was particularly inefficient; additional factors, such as steric restrictions on accessing its epitope at the gp120-gp41 interface, may limit 35O22 efficacy when virion Env levels are high. In summary, in our experimental system, the results are consistent with a model in which (i) Env is distributed among the virus particles randomly; (ii) for most of the range of Env levels per virion achieved, n t is less than T (on non-infectious virions) and equal to or slightly greater than T on infectious virions; and (iii) only at the highest levels of Env achieved is there significant redundancy in the entry process (Fig. S2). Thus, even in systems using optimally expressed CT-trunca ted Envs, n t is only marginally greater than T on most infectious viruses. This perspective should inform estimates of n t made in systems using full-length Envs, where the virion Env concentrations are typically lower. Indeed, several studies suggesting low values of T for primary HIV-1 Envs have been based on estimates of n t that are equal to or only slightly greater than T (120,123,124).
We evaluated the infectivity and induction of cytopathic effects by recombinant viruses with full-length and CT-truncated Envs. The relative increases in the levels of CT-truncated Envs on the surface of expressing cells were associated with increased syncytium-forming ability. The relative increases in the levels of CT-truncated Envs on virions were associated with increased infectivity and induction of cytopathic effects in the target cells. This cytotoxicity was retained by the replication-defective def5 viruses and therefore did not require infection of the target cells. Changes in the Gag PTAP motif significantly reduced the cytotoxicity, suggesting that VLPs mediate the cytopathic effects. Incubation of CD4+/CCR5+ cells with sufficiently high concentrations of virions results in "fusion from without, " a cytotoxic process that has been reported to be enhanced by deletion of the Env CT (127,128).
In contrast to the CT truncations, the CT changes (R747L and L748A) near the viral protease cleavage site exhibited limited phenotypes. As this proteolytic clipping of the CT occurs only after the HIV-1 protease dimerizes and becomes active during virus assembly and release, these changes do not influence Env intracellular transport, processing, cell-surface expression, or virion incorporation. With one exception, increases (R747L) and decreases (L748A) in the degree of protease clipping of gp41 in pseudovi rus VLPs exhibited no detectable phenotype. This implies that the major phenotypes associated with the L748STOP Env arise because of effects on one or more of the above processes, rather than an effect of the truncated virion Env CT per se. The one phenotype associated with the truncated Envs in the R747L and L748STOP mutant virions is resistance to amphotericin B methyl ester (AME). Our results suggest that both the Env CT and cholesterol contribute to the antiviral activity of AME, supporting models in which AME and cholesterol create pores in the viral membrane (59,60,89). A cholesterol-binding element has been mapped to the HIV-1 gp41 CT (129). Also, the Env CT has been implicated in the permeabilization of the HIV-1 membrane to dNTPs during natural endogenous reverse transcription (79,130).
These studies can guide efforts to utilize more homogeneous sources of HIV-1 Envs to characterize the structure and immunogenicity of the elusive PTC. The availability of more homogeneous preparations of the Env PTC would expedite biophysical characteri zation as well as the studies of immunogenicity. The conformationally flexible uncleaved Env on cell surfaces, extracellular vesicles, or VLPs may be immunodominant, eliciting undesirable responses that misdirect the immune system. Awareness of these potential pitfalls may facilitate efforts to focus the antibody response on the functional pretrig gered Env conformation.
## MATERIALS AND METHODS
## Plasmids
The full-length HIV-1 AD8 Env and CTmodified Envs were expressed using the pSVIIIenv plasmid. The env genes encoding the L715STOP and L748STOP Envs in the pSVIIIenv plasmids have stop codons replacing the codons for Leu 715 and Leu 748, respectively. The Y712STOP Env encoded by the pSVIIIenv plasmid has a carboxy-terminal Gly 3 His 6 tag; in the env gene encoding the Y712 Env, a stop codon replaces the codon for Tyr 712. The primary HIV-1 Envs [191955-A4 (Clade A), BG505 (Clade A), 1054_07_TC4_1499 (Clade B/TF), WITO4160.33 (Clade B), CAP45.200.63 (Clade C), and 191821 (Clade D)] were expressed using pcDNA3.1 (119). For experiments in which the Envs were expressed in the context of an HIV-1 provirus, the env genes were cloned into the pNL4-3 IMC, obtained from the NIH HIV Reagent Program (131). The env genes encoding the Y712STOP, L715STOP, and L748STOP Envs in the pNL4-3 IMCs have stop codons replacing the codons for Tyr 712, Leu 715, and Leu 748, respectively. The construction of the JR-FL IMC was described in reference 68. The AD8 and ADA env genes were cloned into the pNL4-3 proviral plasmid between 5′ cagaagacagtggca and 3′ gatgggtggcaagtg sequences. Def5 was created by introducing multiple stop codons into the AD8 IMC that eliminate the expression of reverse transcriptase (RT), RNase H, integrase (IN), Vif, and Vpr (72). The codon-optimized sequence encoding EABR (FNSSINNIHEMEIQLKDA LEKNQQWLVYDQQREVYVKGLLAKIFELEKKTETAAHSLP), followed by two stop codons, was inserted after the codon for Ser 752 in the pSVIIIenv plasmids expressing the HIV-1 AD8 and TFAR Envs. The EABR sequence was inserted after the codon for Pro 714 to make the pSVIIIenv plasmid expressing the AD8-715EABR Env.
The PTAP motif within the p6 domain of the HIV-1 Gag protein (76-78, 132) was changed to LIRL using Q5 site-directed mutagenesis (New England BioLabs). The pBJ5 plasmid expressing SERINC5 was generously provided by Dr. Heinrich Göttlinger (UMass Worcester) (97). In the control ΔSERINC5 plasmid, two stop codons were introduced by mutating the initiation codon ATG of SERINC5. The pQCXIP plasmid expressing IFITM1 was obtained from Addgene.
The DNA sequences of all constructs were confirmed.
## Cell lines
HEK 293T, TZM-bl, and A549 cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 100 µg/mL penicillin-streptomycin (Life Technologies). Cf2Th-CD4/CCR5 cells stably expressing the human CD4 and CCR5 coreceptors for HIV-1 were grown in the same medium supple mented with 0.4 mg/mL of G418 and 0.2 mg/mL of hygromycin. A549 cell lines inducibly expressing the HIV-1 Rev protein and the full-length TFAR, TFAR-L715STOP, and TFAR-L748STOP Envs were established as previously described (133). The TFAR Env is a PTC-stabilized derivative of the HIV-1 AD8 Env (71).
The D1712 and D1758 A549 cell lines contain the doxycycline-inducible def5 provirus with the TFAR and TFAR-L748STOP Envs, respectively, and were established as described (72). The def5 proviruses are defective in the expression of reverse transcriptase, RNase H, integrase, Vif, and Vpr. In addition, the transactivation response (TAR) element in the def5 long terminal repeat (LTR) is mutated to reduce Tat-mediated transactivation. Two tet operator sequences have been inserted into the U3 regions of the LTRs to upregulate transcription in the presence of doxycycline. The D1712 and D1758 cells were maintained as polyclonal cell lines.
## Small-molecule compounds
The CD4-mimetic compound (CD4mc) BNM-III-170 was synthesized as described previously (37,57). The compound was dissolved in dimethyl sulfoxide (DMSO) at a stock concentration of 10 mM and diluted to the appropriate concentration in cell culture medium for antiviral assays. BMS-806 was purchased from Selleckchem. Ritonavir and methyl-β-cyclodextrin were purchased from Sigma-Aldrich and amphotericin B methyl ester from Santa Cruz Biotechnology.
## Enzymes
Protein N-glycosidase F (PNGase F) and endoglycosidase Hf (Endo Hf ) were purchased from New England BioLabs. Cholesterol oxidase was purchased from Sigma-Aldrich.
## Antibodies
Poorly neutralizing antibodies (F105, 19b, 447-52D, 17b, and F240) and broadly neutralizing antibodies (VRC01, VRC03, b12, b6, PGT145, PG9, PGT151, 35O22, PGT121, 3BNC117, 2F5, 4E10, and 10E8.v4) against the HIV-1 Env were obtained through the NIH HIV Reagent Program, Division of AIDS, NIAID, NIH. The bNAbs included PGT121 against V3 glycans; VRC03, VRC01, and b12 against the gp120 CD4-binding site (CD4BS); PG9 and PGT145 against quaternary V2 epitopes at the Env trimer apex; PGT151 and 35O22 against the gp120-gp41 interface; and 2F5, 4E10, and 10E8.v4 against the gp41 membrane-proximal external region (MPER). The pNAbs included F105 against the gp120 CD4BS; 19b, 39F, and 447-52D against the gp120 V3 region; 17b against gp120 CD4i epitopes; and F240 against a cluster I epitope on gp41. Western blots were developed with 1:2,000 goat anti-gp120 polyclonal antibody (Invitrogen), 1:2,000 4E10 anti-gp41 antibody, 1:3,000 rabbit anti-gag (p55/p24/p17) antibody (Abcam), 1:10,000 rabbit anti-hsp70 (K-20) antibody (Santa Cruz Biotechnology), and 1:250 rabbit anti-SERINC5 antibody (Abcam). The HRP-conjugated secondary antibodies were 1:2,000 rabbit anti-goat (Invitrogen), 1:2,000 goat anti-human (Invitrogen), and 1:10,000 anti-rabbit (Sigma-Aldrich).
## Expression and processing of HIV-1 Env variants
HEK 293T cells were transfected transiently either with (i) pSVIIIenv plasmids encoding HIV-1 Env and Rev along with a plasmid encoding HIV-1 Tat at an 8:1 ratio or (ii) pcDNA3.1 plasmids encoding Env. In some experiments, the psPAX2 plasmid encoding HIV-1 packaging proteins and a luciferase-expressing HIV-1 vector were transfected with the Env-expressing plasmids (93,95). Forty-eight hours later, the cell lysates were prepared in 1× PBS/0.5% NP-40/protease inhibitors. Extracellular vesicles (EVs) were prepared from the medium of Env-expressing cells; the cell supernatant was cleared (600 × g for 10 min), followed by centrifugation at 14,000 × g for 1 h at 4°C. In experiments in which pseudotyped VLPs or provirus-produced virions were studied, HEK 293T cell supernatants were cleared (600 × g for 10 min), filtered through a 0.45 µm membrane, and centrifuged at 14,000 × g for 1 h at 4°C.
Env or VLP expression from A549 cell lines was induced by incubation in either 1 or 2 µg/mL doxycycline for 48 h. Then, cell lysates, extracellular vesicles, and VLPs were prepared as described above.
## Immunoprecipitation of Env from the surface of cells and VLPs
HEK 293T cells expressing Env, Rev, and Tat or doxycycline-induced A549-Env cells were washed with 1× PBS. The cells were then incubated with 10 µg/mL anti-Env antibody or soluble CD4 (sCD4-Ig) for 1.5 h at room temperature. After three washes in 1× PBS, the cells were lysed (1× PBS/0.5% NP-40/protease inhibitors (Roche)) for 5 min on ice. The lysates were cleared by centrifugation at 13,200 × g for 10 min at 4°C, and the clarified supernatants were rotated during incubation with Protein A-Sepharose beads for 1 h at room temperature. The beads were pelleted (1,000 rpm for 1 min) and washed twice with wash buffer (1× PBS and 0.5% NP-40) and a third time with 1× PBS without NP-40. The beads were resuspended in 1× PBS/LDS/DTT and analyzed by western blotting. To prepare the Input sample, an equal number of Env-expressing cells were lysed; cell lysates were cleared and analyzed by western blotting, as above.
To evaluate Env antigenicity on the surface of VLPs, aliquots of virus particles (pelleted and resuspended in 1× PBS) were incubated with a panel of antibodies at 10 µg/mL concentration for 1.5 h at room temperature. One mL of chilled 1× PBS was then added, and samples were centrifuged at 14,000 × g for 1 h at 4°C. The pellets were lysed in 100 µL chilled 1× PBS/0.5% NP-40/protease inhibitor cocktail. Lysates were rotated during incubation with Protein A-agarose beads for 1 h at 4°C and washed two times with chilled 1× PBS/0.1% NP-40 and a third time with 1× PBS without NP-40. The beads were resuspended in 1× PBS/LDS/DTT and used for western blotting. To prepare the Input sample, an equal volume of virus suspension was mixed with 1 mL chilled 1× PBS and centrifuged at 14,000 × g for 1 h at 4°C; the pellet was resuspended in 1× PBS/LDS/DTT and analyzed by western blotting.
## Deglycosylation of Env
Cells, VLPs, and virions were lysed in 1× PBS/0.5% NP-40. The viral lysate was then boiled in denaturing buffer (New England BioLabs) for 10 min and treated with PNGase F or Endo Hf enzymes (New England BioLabs) for 1.5 h at 37°C according to the manufactur er's protocol. The treated proteins were analyzed by reducing SDS-PAGE and western blotting.
## Analysis of Env and SERINC5 by western blotting
Clarified cell lysates, extracellular vesicles, and pelleted virus or virus-like particles were analyzed by western blotting using a nitrocellulose membrane and wet transfer (350 A, 75 min, Bio-Rad). Western blots were developed with 1:2,000 goat anti-gp120 polyclonal antibody (Invitrogen), 1:2,000 4E10 anti-gp41 antibody, 1:3,000 rabbit anti-Gag (p55/p24/ p17) antibody (Abcam), and 1:10,000 rabbit anti-hsp70 (K-20) antibody (Santa Cruz Biotechnology). The HRP-conjugated secondary antibodies were 1:2,000 rabbit antigoat (Invitrogen), 1:2,000 goat anti-human (Invitrogen), and 1:10,000 goat anti-rabbit (Sigma-Aldrich). The intensity of protein bands on nonsaturated western blots was quantified using ImageJ software.
To western blot SERINC5, HEK293T cells were transfected with the SERINC5 plasmid, the ΔSERINC5 plasmid, or proviral constructs expressing the HIV-1 AD8 and HIV-1 NL4-3 Env variants along with the SERINC5 or ΔSERINC5 plasmids. The cell lysates were prepared as described above, or the virions were pelleted and lysed. Western blots were devel oped using 1:250 rabbit polyclonal anti-SERINC5 antibody (Abcam), which detected an approximately 47 kDa SERINC5 band.
## Virus infectivity and cytopathic effect
HEK 293T cells were transfected with pNL4-3 plasmids containing infectious HIV-1 proviruses with the Envs of interest. Forty-eight hours after transfection, cell superna tants were collected and subjected to low-speed centrifugation (600 × g for 10 min at room temperature) to remove cell debris. Different volumes of the clarified cell supernatants were incubated with TZM-bl cells for 48 h in a 37°C/5% CO 2 incubator, after which luciferase activity in the cells was measured. Cell viability was measured using the CellTiter 96 Non-Radioactive Cell Proliferation Assay [(3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT)] kit from Promega. Briefly, different volumes of the clarified cell supernatants were incubated with TZM-bl cells for 72 h in a 37°C/5% CO 2 incubator. Fifteen µL of dye solution (from the kit) was added to each well, followed by incubation of the plate at 37°C for 4 h in a humidified CO 2 incubator. One hundred microliters of solubilization/stop solution (from the kit) was then added to each well. Absorbance at 570 nm was measured using a 96-well plate reader.
## Production of recombinant pseudoviruses expressing luciferase
As described previously (93,95), HEK 293T cells were transfected with pSVIIIenv plasmids expressing Env variants, the psPAX2 Gag-Pol packaging construct, and the firefly luciferase-expressing HIV-1 vector at a 1:1:3 µg DNA ratio using Effectene transfection reagent (Qiagen). Recombinant, luciferase-expressing viruses capable of a single round of replication were released into the cell medium and were harvested 48 h later. The virus-containing supernatants were clarified by low-speed centrifugation (600 × g for 10 min) and used for single-round infections.
To measure virus inhibition, the antibodies or the compounds (T20, BNM-III-170, BMS-806, amphotericin B methyl ester, and methyl-β-cyclodextrin) to be tested were incubated with pseudoviruses for 1 h at 37°C. The mixture was then added to Cf2ThCD4/ CCR5 target cells expressing CD4 and CCR5. Forty-eight hours later, the target cells were lysed, and the luciferase activity was measured. To test the effect of cholesterol oxidase on virus inhibition, the pseudoviruses were incubated with the enzyme for 5 h at 37°C, and then, the mixture was added to the target cells. To study the role of cholesterol in the mechanism of amphotericin B methyl ester virus inhibition, the pseudotyped viruses were incubated with 3 mM MBCD or control buffer for 1 h at 37°C. The pseudotyped viruses were then pelleted, resuspended, and incubated with amphotericin B methyl ester at the indicated concentrations for 1 h at 37°C. The viruses were then added to Cf2Th-CD4/CCR5 cells. After 2 days of culture, the cells were lysed, and luciferase activity was measured. The basal infectivity of the pseudotyped viruses treated with 3 mM MBCD was low, but still 15 times the assay background and adequate to assess amphotericin sensitivity.
## Cell-cell fusion assay
For the alpha-complementation assay measuring cell-cell fusion, COS-1 effector cells were seeded in black-and-white 96-well plates and then cotransfected with plasmids expressing α-gal, Env variants, and Tat at a 1:1:0.125 ratio, using Lipofectamine 3000 transfection reagent (Thermo Fisher Scientific) following the manufacturer's protocol. At the same time, Cf2Th-CD4/CCR5 target cells in 6-well plates were cotransfected with a plasmid expressing ω-gal using Lipofectamine 3000 transfection reagent. Forty-eight hours after transfection, target cells were detached and resuspended in DMEM medium. The medium was aspirated from the effector cells, and target cell suspensions in 50 µL volumes were added to the effector cells (one target-cell well provides sufficient cells for 50 effectorcell wells). Plates were spun at 500 × g for 3 min and then incubated at 37°C in 5% CO 2 for 2 h. The medium was removed, and the cells were lysed in Tropix lysis buffer (Thermo Fisher Scientific). The β-galactosidase activity in the cell lysates was measured using the Galacto-Star Reaction Buffer Diluent with Galacto-Star Substrate (Thermo Fisher Scientific), according to the manufacturer's instructions.
## Statistics
The concentrations of HIV-1 entry inhibitors that inhibit 50% of infection (IC 50 values) were determined by fitting the data in fiveparameter dose-response curves using GraphPad Prism 9. Unpaired t-tests were performed, and P values were calculated using GraphPad Prism 9. Correlation was evaluated by a Spearman rank correlation test.
## References
1. Wyatt, Sodroski (1998) "The HIV-1 envelope glycoproteins: fusogens, antigens, and immunogens" *Science*
2. Chen (2019) "Molecular mechanism of HIV-1 entry" *Trends Microbiol*
3. Sodroski, Goh, Rosen et al. (1986) "Role of the HTLV-III/LAV envelope in syncytium formation and cytopathicity" *Nature*
4. Lifson, Reyes, Mcgrath et al. (1986) "AIDS retrovirus induced cytopathology: giant cell formation and involve ment of CD4 antigen" *Science*
5. Labonte, Patel, Hofmann et al. (2000) "Importance of membrane fusion mediated by human immunodeficiency virus envelope glycoproteins for lysis of primary CD4-positive T cells" *J Virol*
6. Willey, Bonifacino, Potts et al. (1988) "Biosynthesis, cleavage, and degradation of the human immunodefi ciency virus 1 envelope glycoprotein gp160" *Proc Natl Acad Sci*
7. Earl, Moss, Doms (1991) "Folding, interaction with GRP78-BiP, assembly, and transport of the human immunodeficiency virus type 1 envelope protein" *J Virol*
8. Dewar, Vasudevachari, Natarajan et al. (1989) "Biosynthesis and processing of human immunodeficiency virus type 1 envelope glycoproteins: effects of monensin on glycosylation and transport" *J Virol*
9. Merkle, Helland, Welles et al. (1991) "gp160 of HIV-I synthesized by persistently infected Molt-3 cells is terminally glycosylated: evidence that cleavage of gp160 occurs subsequent to oligosaccharide processing" *Arch Biochem Biophys*
10. Stein, Engleman (1990) "Endoproteolytic cleavage occurs in a cis or medial compartment of the Golgi complex" *J Biol Chem*
11. Zhang, Nguyen, Ding et al. (2021) "Dual pathways of human immunodeficiency virus type 1 envelope glycoprotein trafficking modulate the selective exclusion of uncleaved oligomers from virions" *J Virol*
12. Checkley, Luttge, Freed (2011) "HIV-1 envelope glycoprotein biosynthesis, trafficking, and incorporation" *J Mol Biol*
13. Munro, Gorman, Ma et al. (2014) "Conformational dynamics of single HIV-1 envelope trimers on the surface of native virions" *Science*
14. Haim, Salas, Mcgee et al. (2013) "Modeling virus-and antibodyspecific factors to predict human immunodeficiency virus neutralization efficiency" *Cell Host Microbe*
15. Herschhorn, Ma, Gu et al. (2025) *Full-Length Text Journal of Virology*
16. Sodroski (2016) "Release of gp120 restraints leads to an entrycompetent intermediate state of the HIV-1 envelope glycoproteins" *mBio*
17. Guttman, Cupo, Julien et al. (2015) "Antibody potency relates to the ability to recognize the closed, pre-fusion form of HIV Env" *Nat Commun*
18. Castillo-Menendez, Witt, Espy et al. (2018) "Comparison of uncleaved and mature human immunodeficiency virus membrane envelope glycoprotein trimers" *J Virol*
19. Lu, Ma, Reichard et al. (2020) "Shedding-resistant HIV-1 envelope glycoproteins adopt downstream conformations that remain responsive to conformation-preferring ligands" *J Virol*
20. Pancera, Wyatt (2005) "Selective recognition of oligomeric HIV-1 primary isolate envelope glycoproteins by potently neutralizing ligands requires efficient precursor cleavage" *Virology (Auckl)*
21. Chakrabarti, Pancera, Phogat et al. (2011) "HIV type 1 Env precursor cleavage state affects recognition by both neutralizing and nonneutral izing gp41 antibodies" *AIDS Res Hum Retroviruses*
22. Zou, Zhang, Gaffney et al. "Smith AB 3rd, Sodroski J. 2020. Longacting BMS-378806 analogues stabilize the State-1 conformation of the human immunodeficiency virus type 1 envelope glycoproteins" *J Virol*
23. Haim, Salas, Sodroski (2013) "Proteolytic processing of the human immunodeficiency virus envelope glycoprotein precursor decreases conformational flexibility" *J Virol*
24. Postler, Desrosiers (2013) "The tale of the long tail: the cytoplasmic domain of HIV-1 gp41" *J Virol*
25. Anokhin, Spearman (2022) "Viral and host factors regulating HIV-1 envelope protein trafficking and particle incorporation" *Viruses*
26. Murphy, Samal, Vlach et al. (2017) "Solution structure and membrane interaction of the cytoplasmic tail of HIV-1 gp41 protein" *Structure*
27. Piai, Fu, Sharp et al. (2021) "NMR model of the entire membrane-interacting region of the HIV-1 fusion protein and its perturbation of membrane morphology" *J Am Chem Soc*
28. Yang, Spies, Compans (1995) "The human and simian immunodeficiency virus envelope glycoprotein transmembrane subunits are palmitoylated" *Proc Natl Acad Sci*
29. Bhattacharya, Peters, Clapham (2004) "Human immunodeficiency virus type 1 envelope glycoproteins that lack cytoplasmic domain cysteines: impact on association with membrane lipid rafts and incorporation onto budding virus particles" *J Virol*
30. Rousso, Mixon, Chen et al. (2000) "Palmitoylation of the HIV-1 envelope glycoprotein is critical for viral infectivity" *Proc Natl Acad Sci*
31. Chan, Lin, Chen (2005) "Wild-type-like viral replication potential of human immunodeficiency virus type 1 envelope mutants lacking palmitoylation signals" *J Virol*
32. Yang, Ai, Huang et al. (2010) "The cytoplasmic domain of human immunodeficiency virus type 1 transmembrane protein gp41 harbors lipid raft association determinants" *J Virol*
33. Berlioz-Torrent, Shacklett, Erdtmann et al. (1999) "Interactions of the cytoplasmic domains of human and simian retroviral transmembrane proteins with components of the clathrin adaptor complexes modulate intracellular and cell surface expression of envelope glycoproteins" *J Virol*
34. Ohno, Aguilar, Fournier et al. (1997) "Interaction of endocytic signals from the HIV-1 envelope glycoprotein complex with members of the adaptor medium chain family" *Virology (Auckl)*
35. Wyss, Berlioz-Torrent, Boge et al. (2001) "The highly conserved C-terminal dileucine motif in the cytosolic domain of the human immunodeficiency virus type 1 envelope glycoprotein is critical for its association with the AP-1 clathrin adaptor [correction of adapter]" *J Virol*
36. Boge, Wyss, Bonifacino et al. (1998) "A membrane-proximal tyrosine-based signal mediates internalization of the HIV-1 envelope glycoprotein via interaction with the AP-2 clathrin adaptor" *J Biol Chem*
37. Byland, Vance, Hoxie et al. (2007) "A conserved dileucine motif mediates clathrin and AP-2-dependent endocytosis of the HIV-1 envelope protein" *Mol Biol Cell*
38. Rowell, Stanhope, Siliciano (1995) "Endocytosis of endoge nously synthesized HIV-1 envelope protein. Mechanism and role in processing for association with class II MHC" *J Immunol*
39. Wilk, Pfeiffer, Bosch (1992) "Retained in vitro infectivity and cytopathogenicity of HIV-1 despite truncation of the C-terminal tail of the env gene product" *Virology (Auckl)*
40. Murakami, Freed (2000) "The long cytoplasmic tail of gp41 is required in a cell type-dependent manner for HIV-1 envelope glycoprotein incorporation into virions" *Proc Natl Acad Sci*
41. Akari, Fukumori, Adachi (2000) "Cell-dependent requirement of human immunodeficiency virus type 1 gp41 cytoplasmic tail for Env incorporation into virions" *J Virol*
42. Dubay, Roberts, Hahn et al. (1992) "Truncation of the human immunodeficiency virus type 1 transmembrane glycoprotein cytoplasmic domain blocks virus infectivity" *J Virol*
43. Yu, Yuan, Mclane et al. (1993) "Mutations in the cytoplasmic domain of human immunodeficiency virus type 1 transmembrane protein impair the incorporation of Env proteins into mature virions" *J Virol*
44. Gabuzda, Lever, Terwilliger et al. (1992) "Effects of deletions in the cytoplasmic domain on biological functions of human immunodeficiency virus type 1 envelope glycoproteins" *J Virol*
45. Staubus, Alfadhli, Barklis et al. (2019) "Replication of HIV-1 envelope protein cytoplasmic domain variants in permissive and restrictive cells" *Virology (Auckl)*
46. Pezeshkian, Groves, Van Engelenburg (2019) "Single-molecule imaging of HIV-1 envelope glycoprotein dynamics and Gag lattice association exposes determinants responsible for virus incorporation" *Proc Natl Acad Sci*
47. Fernandez, Hoffman, Pezeshkian et al. (2020) "Elucidating the basis for permissivity of the MT-4 T-cell line to replication of an HIV-1 mutant lacking the gp41 cytoplasmic tail" *J Virol*
48. Qi, Williams, Chu et al. (2013) "Rab11-FIP1C and Rab14 direct plasma membrane sorting and particle incorporation of the HIV-1 envelope glycoprotein complex" *PLoS Pathog*
49. Qi, Chu, Chen et al. (2015) "A tyrosine-based motif in the HIV-1 envelope glycoprotein tail mediates cell-type-and Rab11-FIP1C-dependent incorporation into virions" *Proc Natl Acad Sci*
50. Kirschman, Qi, Ding et al. (2018) "HIV-1 envelope Full-Length Text Journal of Virology December"
51. "glycoprotein trafficking through the endosomal recycling compart ment is required for particle incorporation" *J Virol*
52. Fernandez-De Céspedes, Hoffman, Carter et al. (2022) "Rab11-FIP1C is dispensable for HIV-1 replication in primary CD4 + T cells, but its role is cell type dependent in immortalized human T-cell lines" *J Virol*
53. Freed, Martin (1996) "Domains of the human immunodeficiency virus type 1 matrix and gp41 cytoplasmic tail required for envelope incorporation into virions" *J Virol*
54. Murakami, Freed (2000) "Genetic evidence for an interaction between human immunodeficiency virus type 1 matrix and alpha-helix 2 of the gp41 cytoplasmic tail" *J Virol*
55. Freed, Martin (1984) "Virion incorporation of envelope glycoproteins with long but not short cytoplasmic tails is blocked by specific, single amino acid substitutions in the human immunodefi ciency virus type 1 matrix" *J Virol*
56. Mammano, Kondo, Sodroski et al. (1995) "Rescue of human immunodeficiency virus type 1 matrix protein mutants by envelope glycoproteins with short cytoplasmic domains" *J Virol*
57. Chen, Kovacs, Peng et al. (2015) "Effect of the cytoplasmic domain on antigenic characteristics of HIV-1 envelope glycoprotein" *Science*
58. Piai, Fu, Cai et al. (2020) "Structural basis of transmembrane coupling of the HIV-1 envelope glycoprotein" *Nat Commun*
59. Ahmed, Parthasarathy, Newhall et al. (2023) "Enhancing anti-viral neutralization response to immunization with HIV-1 envelope glycoprotein immunogens" *NPJ Vaccines*
60. Haider, Snetkov, Jolly (2021) "HIV envelope tail truncation confers resistance to SERINC5 restriction" *Proc Natl Acad Sci*
61. Waheed, Ablan, Mankowski et al. (2006) "Inhibition of HIV-1 replication by amphotericin B methyl ester: selection for resistant variants" *J Biol Chem*
62. Waheed, Ablan, Roser et al. (2007) "HIV-1 escape from the entry-inhibiting effects of a cholesterol-binding compound via cleavage of gp41 by the viral protease" *Proc Natl Acad Sci*
63. Waheed, Ablan, Soheilian et al. (2008) "Inhibition of human immunodeficiency virus type 1 assembly and release by the cholesterol-binding compound amphotericin B methyl ester: evidence for Vpu dependence" *J Virol*
64. Waheed, Ablan, Sowder et al. (2010) "Effect of mutations in the human immunodeficiency virus type 1 protease on cleavage of the gp41 cytoplasmic tail" *J Virol*
65. Abrahamyan, Mkrtchyan, Binley et al. (2005) "The cytoplasmic tail slows the folding of human immunodefi ciency virus type 1 Env from a late prebundle configuration into the sixhelix bundle" *J Virol*
66. Kalia, Sarkar, Gupta et al. (2003) "Rational site-directed mutations of the LLP-1 and LLP-2 lentivirus lytic peptide domains in the intracytoplasmic tail of human immunodeficiency virus type 1 gp41 indicate common functions in cell-cell fusion but distinct roles in virion envelope incorporation" *J Virol*
67. Wyss, Dimitrov, Baribaud et al. (2005) "Regulation of human immunodeficiency virus type 1 envelope glycoprotein fusion by a membrane-interactive domain in the gp41 cytoplasmic tail" *J Virol*
68. Korber, Foley, Kuiken et al. (1998) "Numbering positions in HIV relative to HXB2"
69. Walker, Huber, Doores et al. (2011) "Broad neutraliza tion coverage of HIV by multiple highly potent antibodies" *Nature*
70. Blattner, Lee, Sliepen et al. (2014) "Structural delineation of a quaternary, cleavage-dependent epitope at the gp41gp120 interface on intact HIV-1 Env trimers" *Immunity*
71. Maley, Trimble, Tarentino et al. (1989) "Characteriza tion of glycoproteins and their associated oligosaccharides through the use of endoglycosidases" *Anal Biochem*
72. Nguyen, Wang, Anang et al. (2023) "Characterization of the human immunodeficiency virus (HIV-1) envelope glycoprotein conformational states on infectious virus particles" *J Virol*
73. Zhang, Anang, Nguyen et al. (2024) "Membrane HIV-1 envelope glycoproteins stabilized more strongly in a pretriggered conformation than natural virus Envs"
74. Ding, Nguyen, Li et al. (2024) "Inducible cell lines producing replication-defective human immunodeficiency virus particles containing envelope glycoproteins stabilized in a pretriggered conformation" *J Virol*
75. Lea, Faulds (1996) *Ritonavir. Drugs (Abingdon Engl)*
76. Ghorpade, Nukuna, Che et al. (1998) "Human immunodeficiency virus neurotropism: an analysis of viral replication and cytopathicity for divergent strains in monocytes and microglia" *J Virol*
77. Sullivan, Sun, Li et al. (1995) "Replicative function and neutralization sensitivity of envelope glycoproteins from primary and T-cell line-passaged human immunodeficiency virus type 1 isolates" *J Virol*
78. Verplank, Bouamr, Lagrassa et al. (2001) "Tsg101, a homologue of ubiquitin-conjugating (E2) enzymes, binds the L domain in HIV type 1 Pr55 Gag" *Proc Natl Acad Sci*
79. Stuchell, Garrus, Müller et al. (2004) "The human endosomal sorting complex required for transport (ESCRT-I) and its role in HIV-1 budding" *J Biol Chem*
80. Martin-Serrano, Perez-Caballero, Bieniasz (2004) "Contextdependent effects of L domains and ubiquitination on viral budding" *J Virol*
81. Jennings, Bracey, Hong et al. (2024) "The HIV-1 capsid serves as a nanoscale reaction vessel for reverse transcription" *PLoS Pathog*
82. Ohtani, Irie, Uekama et al. (1989) "Differential effects of α-, β-and γ-cyclodextrins on human erythrocytes" *Eur J Biochem*
83. Klein, Gimpl, Fahrenholz (1995) "Alteration of the myometrial plasma membrane cholesterol content with. Beta.-cyclodextrin modulates the binding affinity of the oxytocin receptor" *Biochemistry*
84. Kilsdonk, Yancey, Stoudt et al. (1995) "Cellular cholesterol efflux mediated by Full-Length Text Journal of Virology December"
85. *J Biol Chem*
86. Vrielink, Ghisla (2009) "Cholesterol oxidase: biochemistry and structural features" *FEBS J*
87. Matsuoka, Murata (2002) "Cholesterol markedly reduces ion permeability induced by membrane-bound amphotericin B" *Biochim Biophys Acta*
88. Cotero, Rebolledo-Antúnez, Ortega-Blake (1998) "On the role of sterol in the formation of the amphotericin B channel" *Biochim Biophys Acta*
89. Dufourc, Smith, Jarrell (1984) "Amphotericin and model membranes. The effect of amphotericin B on cholesterol-containing systems as viewed by 2H-NMR" *Biochim Biophys Acta*
90. Matsumori, Yamaji, Matsuoka et al. (2002) "Amphotericin B covalent dimers forming sterol-dependent ionpermeable membrane channels" *J Am Chem Soc*
91. Chattopadhyay, Jafurulla (2011) "A novel mechanism for an old drug: amphotericin B in the treatment of visceral leishmaniasis" *Biochem Biophys Res Commun*
92. Alfadhli, Romanaggi, Barklis et al. (2023) "Analysis of HIV-1 envelope cytoplasmic tail effects on viral replication" *Virology (Auckl)*
93. Haim, Strack, Kassa et al. (2011) "Contribution of intrinsic reactivity of the HIV-1 envelope glycoproteins to CD4-independent infection and global inhibitor sensitivity" *PLoS Pathog*
94. Ht, Anang, Zhao et al. (2022) "Functional and highly cross-linkable HIV-1 envelope glycoproteins enriched in a pretriggered conformation" *J Virol*
95. Wang, Esnault, Zhao et al. (2022) "Global increases in human immunodeficiency virus neutralization sensitivity due to alterations in the membrane-proximal external region of the envelope glycoprotein can be minimized by distant State 1-stabilizing changes" *J Virol*
96. Anang, Richard, Bourassa et al. (2022) "Characterization of human immunodeficiency virus (HIV-1) envelope glycoprotein variants selected for resistance to a CD4-mimetic compound" *J Virol*
97. Zhang, Wang, Nguyen et al. (2023) "Alterations in gp120 glycans or the gp41 fusion peptideproximal region modulate the stability of the human immunodefi ciency virus (HIV-1) envelope glycoprotein pretriggered conformation" *J Virol*
98. Anang, Zhang, Fritschi et al. (2023) "V3 tip determinants of susceptibility to inhibition by CD4-mimetic compounds in natural clade A human immunodeficiency virus (HIV-1) envelope glycoproteins" *J Virol*
99. Rosa, Chande, Ziglio et al. (2015) "HIV-1 Nef promotes infection by excluding SERINC5 from virion incorporation" *Nature*
100. Usami, Wu, Göttlinger (2015) "SERINC3 and SERINC5 restrict HIV-1 infectivity and are counteracted by Nef"
101. Lu, Pan, Rong et al. (2011) "The IFITM proteins inhibit HIV-1 infection" *J Virol*
102. Walker, Phogat, Wagner et al. (2009) "Broad and potent neutralizing antibodies from an African donor reveal a new HIV-1 vaccine target" *Science*
103. Sok, Van Gils, Pauthner et al. (2014) "Recombinant HIV envelope trimer selects for quaternary-dependent antibodies targeting the trimer apex" *Proc Natl Acad Sci*
104. Chen, Kwon, Zhou et al. (2009) "Structural basis of immune evasion at the site of CD4 attachment on HIV-1 gp120"
105. Liu, Bartesaghi, Borgnia et al. (2008) "Molecular architecture of native HIV-1 gp120 trimers" *Nature*
106. Zhou, Lynch, Chen et al. (2015) "Structural repertoire of HIV-1neutralizing antibodies targeting the CD4 supersite in 14 donors" *Cell*
107. Ozorowski, Pallesen, De Val et al. (2017) "Open and closed structures reveal allostery and pliability in the HIV-1 envelope spike" *Nature*
108. Wang, Zhang, Go et al. (2023) "Asymmetric conformations of cleaved HIV-1 envelope glycoprotein trimers in styrene-maleic acid lipid nanoparticles" *Commun Biol*
109. Lee, Lu, Zhang et al. (2024) "HIV-1-envelope trimer transitions from prefusion-closed to CD4-boundopen conformations through an occluded-intermediate state" *Comput Struct Biotechnol J*
110. Hoffmann, Yang, Huey-Tubman et al. (2023) "ESCRT recruitment to SARS-CoV-2 spike induces virus-like particles that improve mRNA vaccines" *Cell*
111. Clements, Mj, Stephens et al. (1991) "The V3 loops of the HIV-1 and HTV-2 surface glycoproteins contain proteolytic cleavage sites: a possible function in viral fusion?" *AIDS Res Hum Retroviruses*
112. Ebenbichler, Westervelt, Carrillo et al. (1993) "Structure-function relationships of the HIV-1 envelope V3 loop tropism determinant: evidence for two distinct conformations" *AIDS*
113. Gu, Westervelt, Ratner (1993) "Role of HIV-1 envelope V3 loop cleavage in cell tropism" *AIDS Res Hum Retroviruses*
114. Johnson, Lin, Padmanabhan et al. (1994) "Conformational rearrangements required of the V3 loop of HIV-1 gp120 for proteolytic cleavage and infection" *FEBS Lett*
115. Zhang, Wang, Wang et al. (2021) "Asymmet ric structures and conformational plasticity of the uncleaved full-length human immunodeficiency virus envelope glycoprotein trimer" *J Virol*
116. Heijnen, Schiel, Fijnheer et al. (1999) "Activated platelets release two types of membrane vesicles: microvesicles by surface shedding and exosomes derived from exocytosis of multivesic ular bodies and α-granules" *Blood*
117. Raposo, Stoorvogel (2013) "Extracellular vesicles: exosomes, microvesicles, and friends" *J Cell Biol*
118. Kourembanas (2015) "Exosomes: vehicles of intercellular signaling, biomarkers, and vectors of cell therapy" *Annu Rev Physiol*
119. (2025) *Full-Length Text Journal of Virology*
120. You, Fu, Zou (2021) "Mechanism and potential of extracellular vesicles derived from mesenchymal stem cells for the treatment of infectious diseases" *Front Microbiol*
121. Bourinbaiar (1994) "The ratio of defective HIV-1 particles to replication-competent infectious virions" *Acta Virol*
122. Layne, Merges, Dembo et al. (1992) "Factors underlying spontaneous inactivation and susceptibility to neutralization of human immunodeficiency virus" *Virology (Auckl)*
123. Parrish, Gao, Li et al. (2013) "Phenotypic properties of transmitted founder HIV-1" *Proc Natl Acad Sci*
124. Yang, Kurteva, Ren et al. (2005) "Stoichiometry of envelope glycoprotein trimers in the entry of human immunodefi ciency virus type 1" *J Virol*
125. Klasse (2007) "Modeling how many envelope glycoprotein trimers per virion participate in human immunodeficiency virus infectivity and its neutralization by antibody" *Virology (Auckl)*
126. Magnus, Rusert, Bonhoeffer et al. (2009) "Estimating the stoichiometry of human immunodeficiency virus entry" *J Virol*
127. Zarr, Siliciano (2015) "Stoichiometric parameters of HIV-1 entry" *Virology (Auckl)*
128. Brandenberg, Magnus, Rusert et al. (2015) "Different infectivity of HIV-1 strains is linked to number of envelope trimers required for entry" *PLoS Pathog*
129. Katte, Xu, Han et al. (2025) "Inter-protomer opening cooperativity of envelope trimers positively correlates with HIV-1 entry stoichiometry"
130. Zhang, Anang, Wang et al. (2024) "Stoichiometry of HIV-1 envelope glycoprotein protomers with changes that stabilize or destabilize the pretriggered conformation" *bioRxiv*
131. Clavel, Charneau (1994) "Fusion from without directed by human immunodeficiency virus particles" *J Virol*
132. Kondo, Marin, Kim et al. (2015) "Distinct requirements for HIV-cell fusion and HIV-mediated cell-cell fusion" *J Biol Chem*
133. Nieto-Garai, Arboleya, Otaegi et al. (2021) "Cholesterol in the viral membrane is a molecular switch governing HIV-1 Env clustering" *Adv Sci (Weinh)*
134. Zhang, Dornadula, Alur et al. (1996) "Amphipathic domains in the C terminus of the transmembrane protein (gp41) permeabilize HIV-1 virions: a molecular mechanism underlying natural endogenous reverse transcription" *Proc Natl Acad Sci*
135. Adachi, Gendelman, Koenig et al. (1986) "Production of acquired immunodeficiency syndromeassociated retrovirus in human and nonhuman cells transfected with an infectious molecular clone" *J Virol*
136. Demirov, Orenstein, Freed (2002) "The late domain of human immunodeficiency virus type 1 p6 promotes virus release in a cell typedependent manner" *J Virol*
137. Zhang, Anang, Zhang et al. (2024) "Conformations of membrane human immunodeficiency virus (HIV-1) envelope glycoproteins solubilized in Amphipol A18 lipidnanodiscs" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12835417&blobtype=pdf | # Implications of immune responses to DENV, JEV, and ZIKV Infections for cross-reactivity and considerations for vaccine evaluation
Thi Thanh, Ngan Nguyen, Mya Myat, Ngwe Tun, Thi Bich, Hau Vu, Thi Thu, Thuy Nguyen, Vu Mai, Phuong Hoang, Le Khanh, Hang Nguyen, Trong Phan, Anh Dang, Muhareva Raekiansyah, Ryosuke Suzuki, Jean Balingit, Yuki Takamatsu, Corazon Buerano, Takeshi Urano, Thi Quynh, Mai Le, Futoshi Hasebe, Kouichi Morita
## Abstract
Dengue virus (DENV), Japanese encephalitis virus (JEV), and Zika virus (ZIKV) are mosquito-borne flaviviruses of global concern, causing illnesses ranging from mild fever to severe neurological or hemorrhagic disease. Antibodies against these viruses exhibit complex cross-reactivity, influencing both neutralization and antibodydependent enhancement (ADE). DENV comprises four serotypes (65-70% sequence identity) with multiple genotypes. Secondary heterologous infections increase the risk of dengue with or without warning signs and severe dengue, largely mediated by pre-existing antibodies from prior infection or vaccination. We assessed total antibody levels, neutralizing antibody (NAb) titers, and ADE activity of serum samples from Vietnamese patients with confirmed DENV infection and evaluated potential cross-reactivities with JEV and ZIKV. In primary infections, NAb titers were generally low (<1:10), except for JEV. Secondary infections showed the highest NAb titers against DENV-4, followed by DENV-2, DENV-3, and the lowest against DENV-1. In ADE assays, DENV-2 exhibited the highest enhancement, followed by DENV-1 and DENV-3, whereas DENV-4 showed minimal ADE. ADE was also detected for JEV and ZIKV, demonstrating that cross-reactive antibodies can facilitate infection across flaviviruses. These findings underscore the impact of pre-existing cross-reactive antibodies in endemic regions and highlight the importance of evaluating vaccine-induced cross-reactive immunity.
## 1. Introduction
Dengue virus (DENV) is a mosquito-borne virus that causes dengue fever, a major public health concern in tropical and subtropical regions worldwide (World Health Organization 2012; Holmes and Twiddy, 2003). It belongs to the Flavivirus genus of the Flaviviridae family and is primarily transmitted by Aedes aegypti (Souza-Neto et al., 2019), with Aedes albopictus mosquitoes also contributing in the areas where it is widely distributed (Holmes and Twiddy, 2003). It has four closely related serotypes: DENV-1, DENV-2, DENV-3, and DENV-4 (Guzman and Kouri, 2003). Infection with one serotype confers lifelong immunity to that serotype but not to the others, and secondary infection with a different serotype can increase the risk of severe disease through a phenomenon called antibody-dependent enhancement (ADE) (Holmes and Twiddy, 2003,5;Halsey et al., 2012). Dengue disease presents a spectrum of clinical manifestations, ranging from dengue without warning signs, a self-limiting febrile illness to dengue with warning signs and in some cases progressing to severe dengue, which can involve severe plasma leakage, bleeding, or organ impairment and may be fatal (DENGUE GUIDELINES FOR DIAGNOSIS). Common symptoms include high fever, severe headache, retro-orbital pain, myalgia, rash, and, in severe cases, bleeding and plasma leakage (World Health Organization 2012). There is no specific antiviral treatment for dengue, clinical management therefore relies on supportive care. Vaccination and vector control remain key strategies for prevention (Bhatt et al., 2013). Although vaccines such as Dengvaxia have been licensed, their use is limited by the complexity of immune responses and the requirement for prior DENV exposure to avoid an increased risk of severe disease (Thomas and Yoon, 2019). DENV, Zika virus (ZIKV), and Japanese encephalitis virus (JEV) are closely related flaviviruses that pose significant global health threats (Kuno et al., 1998). Due to their genetic and structural similarities, infections with these viruses often elicit cross-reactive immune responses, particularly antibodies recognizing conserved epitopes across different flaviviruses (Maeki et al., 2019). While cross-reactive neutralizing antibodies (NAbs) may provide partial protection, especially during primary infection, they can also result in complex immunological outcomes. One of the most significant concerns is ADE, in which pre-existing non-neutralizing or sub-neutralizing antibodies from a prior flavivirus infection facilitate enhanced viral entry into Fc receptor-bearing cells, thereby increasing viral replication and disease severity (Vaughn et al.;Islam et al., 2022;Phuong et al., 2004). This phenomenon is well-documented in secondary DENV infections and has also been observed in heterologous infections involving ZIKV and JEV in both in vitro and in vivo studies (Salem et al., 2024;Shukla et al., 2020;Dejnirattisai et al., 2016). Vaccination programs further complicate the immunological interplay among flaviviruses. For example, live-attenuated JEV and DENV vaccines can elicit cross-reactive antibodies, potentially influencing immune responses to subsequent natural infections with heterologous flaviviruses (Li et al., 2016). Understanding the quality, magnitude, and durability of cross-reactive NAbs and their potential to mediate either protection or enhancement-is therefore essential for developing safe and effective vaccines, and for predicting disease outcomes in flavivirus-exposed populations. Several studies have reported cross-neutralizing and cross-reactive antibody responses among individuals exposed to different flaviviruses, emphasizing the need to better understand their role in disease modulation (Maeki et al., 2019;Salem et al., 2024).
Vietnam, located in Southeast Asia, is a hotspot for mosquito-borne flaviviruses, particularly DENV, ZIKV, and JEV. These viruses are endemic or emerging threats, posing a persistent burden on the country's public health system. DENV is the most widespread flavivirus in Vietnam, with all four DENV serotypes co-circulating (Lee et al., 2017). Seasonal outbreaks occur annually, particularly during the rainy season, and the number and severity of cases have increased in recent years. During 2017-2019, Dengue virus serotype 1 (DENV-1) was the most prevalent cause of dengue in Vietnam (Vân, 2023). ZIKV, while less common, emerged with confirmed cases during the global outbreak in 2016 and has continued to be detected sporadically. JEV, historically a major cause of viral encephalitis in children, has declined significantly due to the national immunization program using live-attenuated vaccines; though, sporadic cases still occur, particularly in rural areas. Due to antigenic similarities among these viruses, cross-reactive antibodies are frequently detected in the Vietnamese population. These antibodies may either neutralize or enhance subsequent flavivirus infections, depending on their specificity and concentration. This raises significant concerns about ADE, especially in secondary DENV infections or in individuals with prior JEV vaccination or ZIKV, JEV exposure. In this context, characterizing humoral immune responses-focusing on both neutralizing and enhancing antibodies-is essential for improving clinical management and guiding vaccine development. It is also important to assess how existing vaccination strategies, such as the JEV vaccine, influence susceptibility and disease outcomes following DENV or ZIKV infection in Vietnam's flavivirus-endemic setting. To our knowledge, this is the first study in Vietnam to investigate cross-neutralization and antibody-dependent enhancement (ADE) in DENV, JEV, and ZIKV. Accordingly, our study aimed to elucidate the role of cross-reactive humoral immunity due to these flaviviruses using well-characterized DENV serum samples collected from Vietnamese patients.
## 2. Materials and methods
## 2.1. Sample collection, serotyping and classification of dengue infection
A total of 68 archived serum samples from patients with confirmed DENV-infection were used in this study. These samples obtained within 0-7 days after symptom onset were collected during the 2017 and 2019 dengue outbreaks in Vietnam and were previously serotyped for DENV at the National Institute of Hygiene and Epidemiology, Vietnam based on the institute's existing protocol on real time RT-PCR. The clinical classification (dengue without warning signs [DWOWS], dengue with warning signs [DWWS], or severe dengue [DS]) was determined according to the World Health Organization guidelines in 2009 (DENGUE GUIDELINES FOR DIAGNOSIS), based on the initial status at the time of hospitalization.
## 2.2. Enzyme-linked immunosorbent assays (ELISAs)
We measured IgG antibodies against DENV, JEV and ZIKV using our in house ELISA kit based on the previous publiscations (Inoue et al., 2010;Tun et al., 2021;Tun et al., 2020;Ngwe Tun et al., 2013). Briefly, 96-well plates were coated with purified viral antigens specific to DENV, JEV, or ZIKV diluted in ELISA coating buffer (0.05 M carbonate-bicarbonate, pH 9.6, without sodium azide) and incubated at 37 • C for 1 h or at 4 • C overnight. Plates were blocked with 100 µL of BlockAce (UK-1 B80, Yukijirushi, Sapporo, Japan) per well, except for blank wells, and incubated at room temperature (RT) for 1 h to prevent non-specific binding. Test samples and negative control were diluted at 1:1000. Positive control was diluted serially with blocking buffer. Test and control samples were added in designated wells and incubated for 1 h h at 37 • C. After washing with phosphate-buffered saline containing 0.1 % Tween 20 (PBST), horseradish peroxidase (HRP)-conjugated anti-human IgG secondary antibodies were added and incubated for 1 h h. Plates were washed, and substrate was added for color development. The reaction was stopped with 1 N sulfuric acid, and optical density (OD) was read at 492 nm using a microplate reader (Synergy H1, Bio-Tek, Winooski, VT, USA). Antibody titers were determined based on standard curve. Each sample was tested in duplicate, and appropriate negative and positive controls were included. Dengue virus-specific IgM antibodies were detected using an IgM-capture ELISA (Tun et al., 2020;Ngwe Tun et al., 2023). Briefly, 96-well microplates (Maxisorp, Nalge Nunc International, Roskilde, Denmark) were coated with 100 µL of goat anti-human IgM (5.5 µg/100 µL; Cappel ICN Pharmaceuticals, Aurora, OH) diluted in ELISA coating buffer (0.05 M carbonate-bicarbonate, pH 9.6) and incubated at 37 • C for 1 h or at 4 • C overnight. Plates were blocked with 100 µL of BlockAce (UK-1 B80, Yukijirushi, Sapporo, Japan) per well, except for blank wells, and incubated at RT for 1 h h. After three washes with PBS-T, test samples and positive/negative controls were diluted 1:100 in PBS-T, and 100 µL was added in duplicate wells. Plates were incubated at 37 • C for 1 h h, then washed. Tetravalent DENV antigen was added at 100 µL/well and incubated at 37 • C for 1 h h. Following washing, 100 µL/well of HRP-conjugated anti-flavivirus mouse monoclonal antibody (12D11/7E8) was added at 1:1500 and incubated at 37 • C for 1 h h, followed by washing. Color development was achieved by adding 100 µL per well of o-phenylenediamine dihydrochloride (OPD; 5 mg in 10 mL of 0.05 M citrate phosphate buffer, pH 5.0). Plates were kept in the dark at RT for 30-60 min. The reaction was stopped with 100 µL of 1 N sulfuric acid per well, and optical density (OD) was measured at 492 nm using a microplate reader (Synergy H1, BioTek, Winooski, VT, USA). Samples were considered positive when the OD ratio of the positive control (or test sample) to the negative control was ≥ 2.0. All individuals were classified as having primary or secondary DENV infection based on IgG and IgM results, following World Health Organization (WHO) criteria and laboratory guidelines. Using predefined serology thresholds, infections were categorized as follows: patients with primary DENV infection had DENV-specific IgG titers below 3000, whereas those with secondary infection had IgG titers above 29,000. Furthermore, a DENV IgG titer ≥3000 combined with a positive IgM result, NS1 antigen detection, or PCR confirmation was also classified as secondary (Ngwe Tun et al., 2025).
## 2.3. Cell line and culture conditions
Baby hamster kidney 21(BHK-21) cells were maintained in Eagle's minimum essential medium (EMEM; Gibco, Gaithersburg, MD, USA) supplemented with 10 % heat-inactivated fetal bovine serum (FBS; Gibco). BHK-21 cells stably expressing human Fc gamma receptor IIA (FcyRIIA) were maintained in EMEM supplemented with 10 % heatinactivated FBS, and 0.5 mg/mL neomycin (G418, PAA laboratories) to maintain FcyR expression (Moi et al., 2010). Human embryonic kidney 293T (HEK293T) cells were maintained in high glucose Dulbecco's modified Eagle's medium with L-glutamine (DMEM, Wako) supplemented with 10 % heat-inactivated FBS, 1X MEM non-essential amino acids (Gibco) and 100 U/ml penicillin-streptomycin. All cells were incubated at 37 • C in a humidified atmosphere with 5 % CO 2 .
## 2.4. Plasmid propagation
We used the sub-genomic replicon plasmid of the YFV 17D vaccine strain (X03700), containing the nano-luciferase gene (pCMV-YFV-nlucrep), and expression plasmids encoding the 100 amino-acid mature capsid of YFV 17D (pCAG-YF-C), as well as the precursor membrane (prM) and envelope (E) proteins of various DENV strains, JEV, and ZIKV, as described previously (Yamanaka et al., 2012;Balingit et al., 2025). Strains included DENV-1 D1/Hu/Saitama/NIID100/2014 (Genotype I, LC011945, Japan), DENV-2 36_DENV2 (Cosmopolitan, MN083232, Sri Lanka), DENV-3 Genotype II (CNR_15418), DENV-4 India G11337 (JF262783), ZIKV-Asian I (PRVABC59), and JEV-Genotype I (Mie/41/2002). The DENV-1 and DENV-2, ZIKV, JEV prM/E expression plasmids were constructed using complementary DNA synthesized from viral RNA extracted from infected cell cultures, whereas the DENV-3 and DENV-4 prM/E expression plasmids were generated from synthetic DNA fragments ordered through GENEWIZ. Target plasmids were propagated in Escherichia coli NEB 5-alpha chemically competent cells (C3040H). Briefly, 50-100 ng of plasmid DNA was added to 25 µL of competent cells and incubated on ice for 30 min, followed by heat shock at 42 • C for 30-45 seconds and immediate return to ice for 5 min. After heat shock, 950 µL of pre-warmed SOC medium was added, and the mixture was incubated at 37 • C with shaking at 250 rpm for 1 h. For NEB 10-beta transformations, NEB 10-beta Outgrowth Medium (NEB #B9035) was used. Transformed cells were plated on Luria-Bertani (LB) agar supplemented with the appropriate antibiotic and incubated overnight at 37 • C. Individual colonies were cultured in LB medium prepared as above without agar and supplemented with the same antibiotic at 37 • C with shaking at 200 rpm for 12-16 h. Plasmid DNA was extracted using a QIAGEN Plasmid Midi Kit according to the manufacturer's instructions. DNA concentration and purity were assessed by spectrophotometry (A260/A280) and agarose gel electrophoresis, and purified plasmids were stored at -20 • C until use.
## 2.5. Production of SRIPs
Briefly, HEK293T cells were co-transfected with three plasmids: 2.5 µg of replicon plasmid, 1.25 µg of capsid expression plasmid, and 1.25 µg of prME-expression plasmid, using Polyethylenimine HCl Max (MW 40,000; Polysciences) in Opti-MEM (Gibco), following the procedure described previously (Matsuda et al., 2018). After 5-6 h, the culture medium was replaced with fresh medium. Supernatants containing SRIPs were harvested 2-3 days post-transfection, clarified by centrifugation (2000-3000 rpm for 10 min), filtered through a 0.45 µm syringe filter, aliquoted, and stored at -80 • C. SRIP production using the YFV 17D vaccine strain replicon plasmid (X03700), containing the nano-luciferase gene (pCMV-YFV-nluc-rep), and expression plasmids encoding the 100-amino-acid mature capsid of YFV 17D (pCAG-YF-C), as well as the precursor membrane (prM) and envelope (E) proteins of various DENV strains, JEV, and ZIKV, was performed (Yamanaka et al., 2012;Balingit et al., 2025).
## 2.6. Titration of SRIPs
Titration of SRIPs to determine the working dilution was performed as previously described (Balingit et al., 2024). SRIPs were serially diluted two-fold across a 96-well plate, mixed 1:1 with EMEM supplemented with 2 % FBS, and incubated at 37 • C for 1 h h. Each plate included a cell control (no SRIP), and each SRIP sample was assayed in duplicate. The SRIP dilutions were inoculated onto cell monolayers and incubated at 37 • C in 5 % CO₂ for 5 h. Fresh medium was then added, and the plates were further incubated at 37 • C in 5 % CO₂ for 2 days. Luciferase activity in SRIP-infected cells was measured at 2 days post-inoculation using a microplate reader (Synergy H1, BioTek, Winooski, VT, USA). A criterion of ≥100-fold difference between the SRIP control and the cell control in all two replicates (e.g., 10⁵ relative luminescence units [RLU] for the SRIP control vs. 10³ RLU for the cell control) was set. The maximum dilution of SRIP meeting this criterion was selected as the working dilution for subsequent neutralization and ADE assays.
## 2.7. NAb and ADE assays
Non-expressing-FcγR-BHK and FcγR-expressing-BHK-cells were used for NAb assays, whereas FcγR-expressing-BHK-cells were used for the ADE assays. Cells were seeded in 96-well plates at a density of 2 × 10⁴ cells/well and cultured overnight. On the following day, serum samples were heat-treated at 56 • C for 30 min. Subsequently, they were serially diluted fourfold with EMEM supplemented with 2 % FBS, mixed with SRIPS at a 1:1 ratio, then incubated at 37 • C for one hour. The SRIPserum mixtures were then added to the cells and incubated for 5 h at 37 • C. Each plate included a SRIP control (no serum) and cell control (no SRIP, no serum). Unbound virus-antibody complexes of the mixture were removed and the cells were cultured in MEM containing 2 % FBS for 48 h virus. Viral replication was quantified using the Nano-Glo Luciferase Assay System (Promega, USA) to detect luciferase activity (indicator of viral replication) in the supernatant. Briefly, equal volumes of the collected supernatant and Nano-Glo reagent were mixed according to the manufacturer's instructions, and the luminescence was measured using a microplate reader (Synergy H1, BioTek, Winooski, VT, USA). The luminescence signal was proportional to the amount of viral replication, which reflected the level of infection. The NAb titer was determined as the serum dilution that reduced ≥50 % luciferase signal (NT 50 ) following the formula below:
The fold enhancement described below was determined after this formula:
Fold enhancement = mean RLUtmean RLUc mean RLUsmean RLUc RLU is the relative luminescence unit measured corresponding to the luciferase activity in SRIP-infected cells. The RLU of each of test serum sample, SRIP control and cell control is represented by RLU t , RLU s , RLU c , respectively.
A DENV-immune serum and DENV-naive serum served as the positive and negative controls, respectively. Fold enhancement was defined as an increase in luciferase signal that was greater than the mean ± 3 standard deviations of the negative control serum sample for each DENV serotype, Zika and JEV. The peak enhancement titer (PET) was determined as previously described (Balingit et al., 2024). Specifically, the fold enhancement values were plotted on the y-axis, while the log-reciprocal serum dilutions were plotted on the x axis. The ADE curve was then fitted using a Gaussian distribution. The data point corresponding to the peak of the Gaussian curve (amplitude) was used to identify the log-reciprocal serum dilution, which was designated the PET viral infection.
$$% luciferase reduction = ( 1 - mean RLUt -mean RLUc mean RLUs -mean RLUc ) x100$$
## 2.8. Data analysis
All figures and statistical analyses were performed using GraphPad Prism 10 (GraphPad Software Inc., United States) and Stata/IC 16.0. Data were expressed as mean ± standard deviation (SD) or as median ± interquartile range (IQR), depending on the distribution. A p-value < 0.05 was considered statistically significant.
## 3. Results
## 3.1. Patient characteristics and clinical features
The 68 positive DENV serum samples collected during the 2017 and 2019 DENV outbreak in Vietnam were obtained from a patient cohort consisting of 39 males and 29 females, with a mean age of 30.7 ± 1.86 years (95 % CI: 27.0-34.5). The majority of patients (n = 55) presented with dengue fever without warning signs, while a smaller subset (n = 13) exhibited warning signs based on clinical assessment (Table 1). The time from symptom onset ranged from 0 to 7 days, with 53 patients (78 %) presenting within 0-2 days and 15 patients (22.0 %) between days 3-7. As shown in Table 1, primary infections predominated (48/68, 70.59 %), while secondary infections accounted for 20/68 (29.41 %) of the patients. The most frequent infecting serotype was DENV-1 (47/68, 69.1 %) followed by DENV-2 (18/68, 26.5 %), DENV-4 (2/68, 2.9 %) and the coinfection of DENV-1 and -2 (1/68, 1.5 %). Comparing DENV-1 and DENV-2 in terms of the serotype causing a higher percentage of DWWSaccording to infection status, DENV-2 caused a higher percentage of DWWS in both primary (28.58 % [4/14] for DENV-2 vs. 16.13 % [5/ 31] for DENV-1) and secondary infections (25.0 % [1/4] for DENV-2 vs. 18.75 % [3/16] for DENV-1). Additionally, all primary DENV-4 infections (2/2) and the single primary co-infection with DENV-1 and DENV-2 (1/1) presented with DWOWS.
## 3.2. Comparison between IgG titers to DENV, JEV and, ZIKV
Among the 68 DENV-positive serum samples, the highest IgG titers measured by ELISA were observed against JEV, followed by DENV and ZIKV, in both primary and secondary dengue infections. In patients with primary DENV infection, the mean DENV IgG titer was 751.4 ± 106.9 (95 % CI: 536.4-966.4). JEV IgG titers were markedly higher, with a mean of 6399.9 ± 1527.4 (95 % CI: 3327.3-9472.6), whereas ZIKV IgG titers were comparatively lower, averaging 323.3 ± 57.8 (95 % CI: 207.0-439.6). In patients with secondary DENV infection, the mean IgG titers were 32,872.2 ± 6790.8 against DENV (95 % CI: 18,658.9-47,085.4), 65,508.9 ± 13,508.9 against JEV (95 % CI: 37,234.5-93,783.3), and 3674.9 ± 1091.7 against ZIKV (95 % CI: 1390.0-5959.8) (Fig. 1a,b). These findings indicate robust crossreactive IgG responses against JEV, intermediate responses to DENV, and comparatively low ZIKV reactivity in this cohort. No significant differences in IgG titers against DENV, JEV, or ZIKV were observed between patients with different clinical manifestations (Fig. 1c). However, a trend toward higher IgG titers was noted in patients presenting with warning signs, suggesting a potential link to disease severity (Fig. 1c). To further characterize IgG responses over the course of illness, IgG titers were compared between patients sampled early (0-2 days after onset) and later (3-7 days after onset). As shown in Fig. 1d, no significant differences in DENV or ZIKV IgG titers were observed between these groups (p > 0.05). In contrast, JEV-specific IgG titers were significantly higher in patients sampled at 3-7 days post-onset compared with those sampled at 0-2 days (p < 0.0001).
## 3.3. Characteristics of neutralizing antibodies (NAbs) against DENV, JEV, and ZIKV
NAb titers were determined in parallel by using parental BHK cells (FcγRIIA-negative) and BHK cells stably expressing FcγRIIA. NAb titers were higher in FcγRIIA-negative BHK cells compared with FcγRIIA-positive BHK cells (Fig. 2a,b). The results on FcγRIIA-negative BHK showed that the serum samples from primary DENV infection had generally low (below the 1:10 serum dilution threshold) NAb titers against DENV and ZIKV, whereas the NAb titer against JEV was the highest, with a geometric mean titer (GMT) of 17.2 ± 26.5 (Fig. 2c). In contrast, serum samples from secondary infection had significantly higher NAb titers against all four DENV serotypes (particularly DENV-4), JEV and ZIKV but with the titers against JEV and ZIKV lower than those of the four DENV serotypes (Fig. 2d). The higher response observed for DENV-4 should be interpreted with caution because of its limited representation in the cohort and the relatively small number of secondary infection samples. The markedly elevated NAb50 titers to DENV-1, DENV-2, DENV-3, and DENV-4 had GMTs of 88. 00 ± 194.60, 593.50 ± 2278.77, 202.00 ± 587.75 and 2295.50 ± 9118.80, respectively. NAb50 titers against JEV and ZIKV had GMTs of 29.50 ± 47.29 and 35.50 ± 142.37, respectively.
## 3.4. Characterization of antibody-dependent enhancement (ADE) in sera from DENV-infected patients against DENV, JEV, and ZIKV
For ADE assays, Fold Enhancement (FE) represents the relative increase in viral infection in the presence of antibodies compared to a noantibody control, reflecting the degree of ADE. By using the serum samples from primary DENV infections, the highest viral infection was observed at the lowest serum dilution (1:10) (Fig. 3a-f). With serum samples from secondary DENV infections, strong enhancement was observed for DENV-1 (FE 17.95), DENV-2 (FE 28.21) and DENV-3 (FE 11.54), while DENV-4 was not enhanced (Fig. 3a-d, Table 2). Fold enhancement was also detected for ZIKV (FE 13.15) (Fig. 3f, Table 2).
The FE for JEV due to serum samples from both primary and secondary infections were 6.77 and 7.70 respectively (Fig. 3e; Table 2). Stratification by patient status of the serum sources indicated that enhancement levels for DENV-1, DENV-2 and ZIKV were higher in patients with warning signs compare to those without, whereas enhancement for DENV-3, DENV-4 and JEV was similar between groups (Fig. 3g-m (Fig. 3a-d, Table 2), reflecting the serum dilution that produces maximal enhancement. For JEV, PET was observed due to samples from both primary (PET 261.22) and secondary infections (PET 666.81) (Fig. 3f; Table 2). ADE profiles over time showed that with primary infection samples, ADE was detected only for DENV-1 during the later phase (days 3-7) and for JEV during both early (days 0-2) and later (days 3-7) phases of illness; other DENV serotypes and ZIKV did not show ADE (Fig. 4a-f). With secondary infection samples, ADE was observed for all DENV serotypes, as well as for ZIKV and JEV, generally peaking in the later phase, except for DENV-1, which exhibited higher enhancement during the early phase (days 0-2) (Fig. 4g-m). A one-way ANOVA was performed for each virus to assess differences in mean IgG titers between groups. Data are presented as mean ± 95 % confidence interval. A statistically significant difference was observed in JEV-specific IgG titers between onset groups (p < 0.0001), whereas no significant differences were detected for DENV or ZIKV.
## Fig. 2. Comparison of neutralizing antibody titers (NAb₅₀).
NAb₅₀ titers were compared among DENV serotypes (DENV-1 to DENV-4), JEV, and ZIKV under FcγRII-negative (a) and FcγRII-positive BHK cell conditions (b). NAb₅₀ levels in patients with primary (c) and secondary infections (d) are shown. Data are presented as mean NAb titers ± standard error on a logarithmic scale. Statistical analysis was performed using the Kruskal-Wallis test, followed by multiple comparisons to identify significant differences between virus groups. Significance levels are indicated as: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.
## Fig. 3. Assessment of antibody-dependent enhancement (ADE).
ADE of the four DENV genotypes, ZIKV, and JEV was evaluated by using serum samples from primary and secondary infections (panels a-f) and between dengue cases without and with warning signs (panels g-m). The Y-axis represents fold enhancement, and the X-axis represents serial dilutions (log₁₀ scale). Data were stratified by infection status: primary (blue curve) and secondary (red curve), as well as by clinical manifestation: dengue without warning signs (blue) and with warning signs (chartreuse). Enhancement curves for each group were fitted using non-linear regression with a Gaussian distribution model. The peak enhancement titer (PET) was defined as the antibody concentration at which maximum enhancement occurred. Data are shown as mean ± SD (NAbs) or median (range) (FE and PET, ADE assays). "Interrupted (No ADE)" denotes absence of antibody-dependent enhancement.
## Fig. 4. Assessment of antibody-dependent enhancement (ADE).
ADE of the four DENV genotypes, ZIKV, and JEV was evaluated by using serum samples from primary (panels a-f) and secondary (panels g-m) infections. The Y-axis represents fold enhancement, and the X-axis represents serial dilutions (log₁₀ scale). Data were stratified by day of illness onset: 0-2 days (blue curve) and 3-7 days (red curve). Enhancement curves for each group were fitted using non-linear regression with a Gaussian distribution model. The peak enhancement titer (PET), defined as the antibody concentration at which maximum enhancement occurred, was estimated from the fitted curves.
3.5. Characterization of ADE against DENV, ZIKV, and JEV in sera from serotype-specific DENV infections By using serum samples from patients infected with DENV-1, ADE was strongest across all tested DENV genotypes, as well as for ZIKV and JEV. The highest ADE was observed against DENV-2, with a peak enhancement titer (PET) of 152.75 and a fold enhancement (FE) of 17.56, followed by ZIKV (PET 91.83,FE 10.76). Significant enhancement was also detected against DENV-1 (PET 189.67, FE 8.06), JEV (FE 7.32),. In contrast, serum samples of patients infected with DENV-2 provided strong ADE predominantly toward DENV-2 itself, with notable enhancement for DENV-1 and JEV. The PET assay revealed significantly elevated fold enhancement for these viruses, whereas enhancement of other DENV serotypes and ZIKV was comparatively lower (Fig. 5a andb). These findings indicate distinct ADE profiles depending on the infecting DENV serotype.
## 4. Discussion
Our study highlights the complex and dynamic nature of antibody responses in DENV infections, particularly the interplay between NAb and ADE. Among DENV-infected patients, the highest IgG titers were observed against JEV, followed by DENV and ZIKV, in both primary and secondary infections. This pattern likely reflects prior JEV vaccination or natural exposure, consistent with the widespread implementation of JEV immunization programs in endemic regions (Maeki et al., 2019;Li et al., 2016;Guzman et al., 2013;Yun and Lee, 2014;Saito et al., 2016). The elevated JEV-specific IgG titers underscore the well-documented serological cross-reactivity among flaviviruses, which can complicate diagnostic interpretation and surveillance efforts (Beltramello et al., 2010;Saito et al., 2016;Sato et al., 2015). Interestingly, total IgG levels did not differ significantly between patients with and without warning signs, suggesting that overall IgG including cross-reactive antibodies may not reliably predict clinical severity (Guzman et al., 2013;Dejnirattisai et al., 2010). However, JEV-specific IgG levels varied significantly between early and later phases of illness, while DENV-and ZIKV-specific titers remained stable. This likely reflects an anamnestic response mediated by memory B cells primed through previous JEV exposure, potentially reactivated during acute DENV infection (Saito et al., 2016). These findings emphasize the need for assays that distinguish functional antibody responses, such as neutralization and enhancement, to better understand disease pathogenesis and guide for vaccine development (Beltramello et al., 2010;Screaton et al., 2015).
In line with these observations, our neutralization assays showed that NAb titers were lower in Fcγ receptor-expressing BHK cells compared to standard BHK cells, highlighting the role of Fcγ receptormediated pathways in facilitating viral entry even in the presence of neutralizing antibodies, a mechanism consistent with ADE (Dejnirattisai et al., 2010;Martina et al., 2009). During primary DENV infection, low neutralizing titers against DENV and ZIKV coupled with relatively high titers against JEV suggest pre-existing immunity to JEV, likely due to childhood vaccination or natural exposure (Durbin and Whitehead, 2011). In Vietnam, adults may retain long-lasting JEV antibodies from earlier immunization programs or subclinical infections. A more detailed assessment of prior vaccination (including number of doses and timing) and infection history would provide more conclusive evidence linking JEV-induced cross-reactive antibodies to subsequent immune responses against DENV or ZIKV (Saito et al., 2016;Wang et al., 2022). Pre-existing JEV immunity in adults may modulate both the magnitude and specificity of neutralizing antibody responses, thereby potentially influencing ADE and clinical outcomes. These findings highlight the importance of accounting for prior flavivirus exposure when interpreting antibody profiles in adult cohorts from regions with co-circulating flaviviruses. In this study, detailed vaccination histories were not available for the enrolled patients, including information on prior JEV vaccination, number of doses received, and timing of immunization. As a result, we were unable to directly distinguish between vaccine-induced and naturally acquired JEV immunity, or to formally assess the impact of prior vaccination on subsequent antibody responses to DENV and ZIKV. Future studies incorporating comprehensive immunization and infection histories will be important to more precisely define the role of pre-existing flavivirus immunity in shaping neutralizing and enhancing antibody responses and clinical outcomes. Conversely, secondary infections were characterized by significantly elevated NAb titers against all four DENV serotypes, as well as increased NAb titers against JEV and ZIKV. This reflects a broad recall response driven by cross-reactive memory B cells (Screaton et al., 2015). Secondary dengue virus infection generates higher cross-reactive memory B cells toward Zika virus than primary infection (Andrade et al., 2020;Hattakam et al., 2021). It should be noted that the serotype distribution in this cohort was skewed, with a predominance of DENV-1 and DENV-2 and very limited representation of DENV-4. In addition, the proportion of secondary infection samples was relatively small compared with primary infections. Therefore, although higher NAb titers and lower ADE were observed in association with DENV-4 in this dataset, these findings should be interpreted cautiously and cannot be generalized as serotype-specific characteristics. Larger and more balanced cohorts will be required to determine whether these observations reflect true biological differences or are influenced by sampling bias. In our ADE assay, sera from primary DENV infections showed the highest fold enhancement at the lowest dilution (1:10), with declining enhancement at higher dilutions. This differs from secondary infections, where peak enhancement occurs at intermediate dilutions (Balingit et al., 2024). The low-dilution peak in primary infections likely reflects the presence of high concentrations of cross-reactive antibodies with varying specificities and weak neutralizing activity, rather than a classical ADE pattern. In secondary infections, ADE was evident for all four DENV serotypes, as well as ZIKV and JEV, generally peaking between days 3-7, reflecting the expansion of cross-reactive memory antibodies. Notably, DENV-1 exhibited an atypical pattern, showing higher ADE at days 0-2, suggesting the presence of pre-existing serotype-specific enhancing antibodies, possibly from prior DENV-1 exposure (Martina et al., 2009;Anasir et al., 2020).
Our analysis indicates a potential correlation between PET and disease severity, with higher PET values observed more frequently in patients with dengue with warning signs than in those without warning signs. This observation is consistent with the hypothesis that higher levels of enhancing antibodies may contribute to more severe clinical manifestations through ADE (Katzelnick et al.). However, PET alone is unlikely to fully predict disease outcome. Several additional factors are likely to influence both PET and clinical severity. Secondary dengue infection was associated with a significantly increased risk of severe dengue compared with primary infection (Shih et al., 2024). According to the WHO, variability in the incubation period and disease course is associated with differences in disease severity. Based on these observations, we hypothesize that time since infection or sampling may also be associated with PET, since antibody levels fluctuate over the course of infection and PET values may differ depending on the sampling time point. Host genetic and immunological factors, such as variation in Fcγ receptor expression, HLA type, and innate immune responses, can modulate the effects of enhancing antibodies (Ghosh et al., 2025) .Viral factors, including DENV serotype and genotype, may interact differently with host antibodies, thereby influencing both PET and disease severity (Dhole et al., 2024). Finally, co-morbidities and age, including underlying health conditions, nutritional status, and age-related immune changes, can influence immune responses and susceptibility to severe disease (Ghosh et al., 2025;Dhole et al., 2024;Bhatt et al., 2021). Taken together, while PET may serve as an indicator of ADE risk and potential disease severity, a comprehensive assessment that integrates these additional host and viral factors is necessary to accurately interpret its clinical significance. Future studies with larger cohorts and longitudinal sampling will be critical to disentangle these relationships. These temporal and serotype-specific differences highlight the complex dynamics of antibody-mediated enhancement in dengue pathogenesis. We used DENV-1 genotype I strain in ADE assays, which matched the infecting genotype in most patients, likely contributed to the robust ADE signals observed, particularly in secondary infections. This genotype match enhances the biological relevance of the assay by better reflecting in vivo virus-antibody interactions. The limited ADE detected in primary infections may be due to early-stage, low-affinity antibodies not yet capable of mediating significant enhancement, whereas in secondary infections, pre-existing cross-reactive antibodies effectively bind the homologous assay strain to promote ADE (Dejnirattisai et al., 2010). Moreover, ADE was observed for multiple flaviviruses in secondary dengue infection, supporting the notion that cross-reactive non-neutralizing antibodies facilitate viral entry via Fcγ receptors. ADE for DENV-1 and DENV-2 was predominantly seen in patients with warning signs, suggesting a potential contribution to disease severity (Guzman et al., 2013;St. John and Rathore, 2019). In contrast, ADE for JEV was detected at similar levels regardless of clinical severity, indicating that while cross-reactive anti-JEV antibodies mediate enhancement, they do not necessarily correlate with disease outcomes in this context (Saito et al., 2016;Sato et al., 2015). Together, these findings underscore the complex interplay between neutralizing and enhancing antibodies in flavivirus infections and the nuanced relationship between ADE and clinical severity. They highlight the importance of considering viral genotype-specific interactions and the timing of antibody responses when interpreting ADE and NAb assay results. A deeper understanding of these factors is critical for improving dengue disease management and informing effective vaccine strategies (Beltramello et al., 2010;Screaton et al., 2015).
## References
1. Anasir, Ramanathan, Poh (2020) "Structure-based design of antivirals against envelope glycoprotein of dengue virus" *Viruses*
2. Andrade, Narvekar, Montoya (2020) "Primary and secondary dengue virus infections elicit similar memory b-cell responses, but breadth to other serotypes and cross-reactivity to zika virus is higher in secondary dengue" *J. Infect. Dis*
3. Balingit, Abe, Suzuki (2025) "Cross-genotype immunogenicity and antibody-dependent enhancement of KD-382 dengue vaccine in flavivirus-naïve adults" *NPJ. Vaccines*
4. Balingit, Denis, Suzuki (2024) "Impact of pre-existing cross-reactive antibodies on cyclic dengue outbreaks in the hyperendemic region of Bali, Indonesia" *Virus. Res*
5. Beltramello, Williams, Simmons (2010) "The human immune response to dengue virus is dominated by highly cross-reactive antibodies endowed with neutralizing and enhancing activity" *Cell Host. Microbe*
6. Bhatt, Sabeena, Varma et al. (2021) "Current understanding of the pathogenesis of dengue virus infection" *Curr. Microbiol. Springer*
7. Bhatt, Gething, Brady (2013) "The global distribution and burden of dengue" *Nature*
8. Dejnirattisai, Jumnainsong, Onsirisakul (1979) "Cross-reacting antibodies enhance dengue virus infection in humans" *Science*
9. Dejnirattisai, Supasa, Wongwiwat (2016) "Dengue virus sero-crossreactivity drives antibody-dependent enhancement of infection with zika virus" *Nat. Immunol*
10. "Dengue Guidelines For Diagnosis, Treatment, prevention and control treatment, prevention and control treatment, prevention and control"
11. Dhole, Zaidi, Nariya et al. (2024) "Host immune response to Dengue virus infection: friend or foe?" *Immuno. Multidiscip. Digit. Public Inst*
12. Durbin, Whitehead (2011) "Next-generation dengue vaccines: novel strategies currently under development" *Viruses*
13. Ghosh, Mondal, Sadhukhan et al. (2025) "Dengue virus and the host immune system: a battle of immune modulation, response and evasion" *Pathog. Multidiscip. Digit. Public Inst*
14. Guzman, Alvarez, Halstead (2013) "Secondary infection as a risk factor for dengue hemorrhagic fever/dengue shock syndrome: an historical perspective and role of antibody-dependent enhancement of infection" *Arch. Virol*
15. Guzman, Kouri (2003) "Dengue and dengue hemorrhagic fever in the Americas: lessons and challenges" *J. Clin. Virol. Elsevier*
16. Halsey, Marks, Gotuzzo (2012) "Correlation of serotype-specific dengue virus infection with clinical manifestations" *PLoS. Negl. Trop. Dis*
17. Hattakam, Elong Ngono, Mccauley et al. (2021) "Repeated exposure to dengue virus elicits robust cross neutralizing antibodies against Zika virus in residents of Northeastern Thailand" *Sci. Rep*
18. Holmes, Twiddy (2003) "The origin, emergence and evolutionary genetics of dengue virus" *Infect. Genet. Evol*
19. Inoue, Alonzo, Kurosawa (2010) "Evaluation of a Dengue IgG Indirect Enzyme-Linked Immunosorbent Assay and a Japanese Encephalitis IgG Indirect Enzyme-Linked Immunosorbent Assay for Diagnosis of Secondary Dengue Virus Infection"
20. Islam, Hasan, Kalam (2022) "Clinical profile, severity spectrum, and hospital outcome of dengue patients in a tertiary care hospital in Dhaka city" *Cureus*
21. Katzelnick, Gresh, Halloran "Antibody-Dependent Enhancement of Severe Dengue Disease in Humans"
22. Kuno, Chang, Tsuchiya et al. (1998) *Phylogeny of the Genus Flavivirus. Vol*
23. Lee, Nguyen-Viet, Nam (2017) "Seasonal patterns of dengue fever and associated climate factors in 4 provinces in Vietnam from 1994 to 2013" *BMC. Infect. Dis*
24. Li, Gao, Fan (2016) "Cross-protection induced by Japanese encephalitis vaccines against different genotypes of Dengue viruses in mice" *Sci. Rep*
25. Maeki, Tajima, Ikeda (2019) "Analysis of cross-reactivity between flaviviruses with sera of patients with Japanese encephalitis showed the importance of neutralization tests for the diagnosis of Japanese encephalitis" *J. Infect. ChemOther*
26. Martina, Koraka, Osterhaus (2009) "Dengue virus pathogenesis: an integrated view" *Clin. Microbiol. Rev*
27. Matsuda, Yamanaka, Yato (2018) "High-throughput neutralization assay for multiple flaviviruses based on single-round infectious particles using dengue virus type 1 reporter replicon" *Sci. Rep*
28. Moi, Lim, Takasaki et al. (2010) "Involvement of the fcγ receptor IIA cytoplasmic domain in antibody-dependent enhancement of dengue virus infection" *J. Gen. Virol*
29. Ngwe Tun, John, Nabeshima (2019) "Genomic analysis and molecular profiling of dengue viruses in Sabah, Malaysia during"
30. Ngwe Tun, Raini, Fernando (2023) "Epidemiological evidence of acute transmission of Zika virus infection in dengue suspected"
31. *J. Infect. Public Health*
32. Ngwe Tun, Thant, Inoue (2013) "Serological characterization of dengue virus infections observed among dengue hemorrhagic fever/dengue shock syndrome cases in upper Myanmar" *J. Med. Virol*
33. Phuong, Nhan, Kneen (2004) "Clinical diagnosis and assessment of severity of confirmed dengue infections in Vietnamese children: is the World Health Organization classification system helpful?" *Am. J. Trop. Med. Hyg*
34. Saito, Moi, Takeshita (2016) "Japanese encephalitis vaccine-facilitated dengue virus infection-enhancement antibody in adults" *BMC. Infect. Dis*
35. Saito, Moi, Takeshita (2016) "Japanese encephalitis vaccine-facilitated dengue virus infection-enhancement antibody in adults" *BMC. Infect. Dis*
36. Salem, Galula, Wu (2024) "Antibodies from dengue patients with prior exposure to Japanese encephalitis virus are broadly neutralizing against Zika virus" *Commun. Biol*
37. Sato, Hamada, Kashiwagi (2015) "Dengue hemorrhagic fever in a Japanese traveler with pre-existing Japanese encephalitis virus antibody" *Trop. Med. Health*
38. Screaton, Mongkolsapaya, Yacoub et al. (2015) "New insights into the immunopathology and control of dengue virus infection" *Nat. Rev. Immunol. Nat. Public Group*
39. Shih, Wang, Wang et al. (2024) "Risk of severe dengue during secondary infection: a population-based cohort study in Taiwan" *J. Microbiol. Immunol. Infect*
40. Shukla, Beesetti, Brown (2020) "Dengue and Zika virus infections are enhanced by live attenuated dengue vaccine but not by recombinant DSV4 vaccine candidate in mouse models" *EBioMedicine*
41. Souza-Neto, Powell, Bonizzoni (2019) "Aedes aegypti vector competence studies: a review" *Infect. Genet. Evol. Elsevier BV*
42. St, John, Rathore (2019) "Adaptive immune responses to primary and secondary dengue virus infections" *Nat. Rev. Immunol. Nat. Public Group*
43. Thomas, Yoon (2019) "A review of Dengvaxia®: development to deployment" *Hum. Vaccin. ImmunOther*
44. Tun, Kyaw, Nabeshima (2021) "Detection of genotype-1 of dengue virus serotype 3 for the first time and complete genome analysis of dengue viruses during the 2018 epidemic in Mandalay, Upper Myanmar" *PLoS. One*
45. Tun, Moriuchi, Toizumi (2020) "Congenital zika virus infection in a birth cohort in Vietnam, 2017-2018" *Am. J. Trop. Med. Hyg*
46. Vân (2017) "Đặc điểm dịch tễ các týp huyết thanh của vi rút dengue gây bệnh sốt xuâ ất huyết Denge ở một số quận/huyện Hà Nội" *J. Clin. Med. Phamarcy*
47. Vaughn, Green, Kalayanarooj "Dengue Viremia Titer"
48. Wang, Zhang, Zhang (2022) "Japanese encephalitis vaccine generates cross-reactive memory T cell responses to Zika virus in humans" *J. Trop. Med*
49. (2012) "Global Strategy for Dengue Prevention and Control, 2012-2020"
50. Yamanaka, Matsuda, Okabayashi (2012) "Seroprevalence of flavivirus neutralizing antibodies in Thailand by high-throughput neutralization assay: endemic circulation of Zika virus before"
51. Yun, Lee (2014) "Japanese encephalitis the virus and vacci" *Hum. Vaccin. ImmunOther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645983&blobtype=pdf | # Perturbed pediatric circulating metabolome in mild and severe dengue disease
Paul Soma, Rebekah Gullberg, Barbara Graham, M Nurul Islam, Guillermina Kuan, Angel Balmaseda, Carol Blair, Barry Beaty, John Belisle, Eva Harris, Rushika Perera
## Abstract
Four billion people are at risk of infection with dengue viruses (DENV), and this burden is rapidly increasing due to geographic expansion of the mosquito vector. Infection with any of the four serotypes of DENV can result in a self-limiting but debilitating febrile illness (DF), and some infections progress to severe disease with hemorrhagic manifestations and shock (dengue hemorrhagic fever/dengue shock syndrome [DHF/DSS]). DENV infection drives the metabolic state of host cells for viral benefit and induces a host-immune response with metabolic implications that link to disease. Here, a dynamic metabolic response to DENV infection and disease was measured in 535 pediatric patients from Nicaragua using liquid chromatography-tan dem mass spectrometry. Metabolomic analyses revealed profound disruptions of critical biochemical pathways and metabolites within the circulating metabolome, especially in those with more severe manifestations of dengue disease. A biomarker panel of 28 metabolites was utilized to classify DF versus DHF/DSS with high sensitivity and specificity, equating to a balanced accuracy of 96.88%. Identified metabolites belonged to biochemical pathways of omega-3 and omega-6 fatty acids, sphingolipids, dipeptides, purines, and tryptophan metabolism. Dipeptides emerged as the most critical mole cules for severe disease classification. Additionally, a previously reported trend between serotonin and platelets in DHF patients was expanded upon here, revealing a major depletion of serotonin, but not platelets, in DSS patients. In this study, the perturbed metabolome was used for disease state classification and exploration of the biochemis try of severe dengue disease pathology.
IMPORTANCEThe international burden of dengue is intensifying, as the number of reported cases in only the first 5 months of 2025 exceeded that of the previous annual high in 2023. The occurrence of deadly severe manifestations of dengue disease will escalate as the total cases rise, and pediatric patients are at greater risk of developing the rapidly progressing severe dengue diseases than adults. Suboptimal vaccines, lack of clinically approved therapeutics, and no methodologies for prognosis of severe disease exacerbate the difficulty of preventative and supportive care. Because human metabo lism is rapidly altered due to infection, perturbations in patients' circulating metabolome can be attributed to dengue disease and correlated to severity. This study contributes metabolic biomarkers of dengue disease in pediatric patients from Nicaragua, indicat ing that metabolic biomarkers are conserved across patients of different ages and geographic and genetic backgrounds. With validation across many cohorts, there is potential to improve diagnostics.
D engue viruses (DENV) are mosquito-borne flaviviruses that place 3.97 billion people at risk of infection each year, with up to 390 million infections estimated annually, rendering them the most prevalent arboviruses worldwide (1)(2)(3). In the first 5 months of 2024, there were 7.6 million reported dengue cases, including 16,000 cases of severe disease and 3,000 deaths-quickly surpassing the previous annual high of 4.6 million dengue cases in 2023 (4). There are four DENV serotypes (DENV1-4). While infection with one serotype can cross-protect from infection with a heterologous serotype in the short term, it can lead to antibody-dependent enhancement of disease in the longer term (5). DENV is the etiologic agent of dengue fever (DF), which is an incapacitating but self-limited disease; however, some cases progress to dengue hemorrhagic fever (DHF) or the potentially fatal dengue shock syndrome (DSS). Peak viremia and fever/symptom onset occur several days after the human host is bitten by an infected mosquito. The acute phase lasts 7 days and includes the "critical phase, " which occurs 4 to 7 days after fever onset (6). Most patients present with DF and defervesce during the critical phase. In a minority of cases, severe manifestations present during the critical phase, with the hallmark vascular leakage that can lead to shock and lethal outcomes. Various proteins, peptides, and metabolites in adult patients have been previously proposed as biomark ers for severe dengue disease (7)(8)(9)(10)(11)(12)(13)(14). Metabolite biomarkers hold notable advantages, and the study of metabolism presents an opportunity to help address challenges in biomarker discovery and in triaging rapidly progressing diseases such as dengue.
Metabolites are the small molecule intermediates and products of biochemical reactions in host cells. Metabolic reactions are driven by upstream gene expression (transcription and translation) and enzyme activity, as well as by the current cellular environment, including active metabolic signaling and bystander effects (15)(16)(17). The metabolome is known as an effector of phenotype because it is dynamic and responds quickly to external stimuli (18). Therefore, the metabolome may describe the current physiological state of the system under study with greater temporal resolution than upstream biomolecules. Perturbations in cellular metabolism of host tissues due to viral infection and the host immune response are reflected in the host's circulating metabolite profile. Accordingly, measuring the circulating metabolome of DENV-infected patients can provide information about system-wide metabolic shifts upon DENV infection and onset of disease.
Infection with DENV prompts perturbations in host cellular metabolism to facilitate various stages of the viral lifecycle and induces a host immune response that is also associated with metabolic alterations (11-13, 19, 20). Metabolomic measurements that lead to mapping the dysregulated pathways in DENV infection and disease can improve understanding of pathogenic processes; define enzyme drug targets that, when inhibited, interfere with the viral lifecycle; or identify biomarkers of severe dengue disease that aid in triage. In this study, the DENV infection-induced metabolic perturba tions in pediatric patients were measured using liquid chromatography-tandem mass spectrometry (LC-MS/MS). The measured perturbed metabolome was used for disease state classification, metabolic pathway analysis, and the exploration of the biochemistry of severe dengue disease pathology.
## RESULTS
## Clinical samples
This study included 535 retrospective serum and plasma samples from individual pediatric patients enrolled in two well-established studies in Managua, Nicaragua. Selected patient demographics are summarized in Fig. 1a through e. Statistical assessment of patient demographics and distribution of sample types between disease states (clinical diagnosis) is included in Table S1. Of the 535 suspected dengue cases, 251 were laboratory-confirmed as dengue-positive, and 284 were non-dengue (ND) febrile illnesses. Based on the 1997 World Health Organization disease severity criteria (21), the dengue cases were classified as DF (185), DHF (44), and DSS (22). The study included pediatric patients of ages 1 to 15 years (Fig. 1a) and similar numbers of females (n = 257, 48%) and males (n = 278, 52%) (Fig. 1b). All samples used for this study were collected on days 1 to 7 of illness (Fig. 1c). Dengue patients were positive for DENV serotypes 1, 2, or 3, or had an undetermined serotype (Fig. 1d). Immune status (primary or secondary infection) was reported (Fig. 1e).
## Metabolic pathway analysis enables insight into dengue disease pathology
The circulating metabolome of each patient was measured via untargeted LC-MS/MS, and the final feature list after data preprocessing contained 3,809 features (Fig. 2). For hypothesis generation and to glean biological information from this metabolomics data set, all 3,809 molecular features (RT, m/z, p-value, and t-score) were fed into the mummichog algorithm to be tentatively identified and mapped to known meta bolic pathways. Pathway analysis using all molecular features proposed 94 tentatively enriched metabolic pathways, each pathway having its own magnitude and direction of dysregulation and statistical significance between three pairwise comparisons: DF vs DHF, DF vs DSS, and DHF vs DSS. Of the 94 pathways, 86 were common among the three comparisons, 3 were unique to DF vs DHF, and 5 were unique to DHF and DSS. Though 86 pathways were identified to be altered in all disease states, the extent of perturbation between comparisons was different. Median t-scores and Fisher's exact p-values for all pathways and comparisons are shown in Table S2.
Twenty-five perturbed pathways with known biological relevance to dengue and other pathological phenotypes are shown in Fig. 3. The 25 pathways are related to the metabolism of tryptophan, lipids (e.g., fatty acids, phospholipids, eicosanoids, and sphingolipids), bile acids, amino acids, purines, sugars, and other cellular energy-related small molecules. The TCA cycle, which drives cellular energy generation, was upregulated in DSS, and pathways related to amino acids and purines were downregulated. Dysregulation of lipids, notably upregulation of fatty acid (FA) synthesis during DENV infection, aligns with our previous in vitro and in vivo studies (10,20). Metabolism of bile acids and retinoic acid was upregulated in DSS, which may relate to liver damage observed in DHF/DSS (22,23). Regarding sugars, hyaluronan metabolism, sialic acid metabolism, and heparan sulfate degradation were all downregulated in DSS, which may relate to the endothelial glycocalyx and its role in viral entry (24,25) or its breakdown in the loss of vascular integrity in severe dengue disease (26)(27)(28)(29). The results of pathway analysis were used to inform metabolic pathway and metabolite targets for molecular structure identification and biomarker panel development.
## Circulating metabolites are perturbed in dengue disease pathology
The workflow to identify molecular features of significant differential abundance across disease states is shown in Fig. 2. A total of 1,512 features revealed significant (adjusted Pvalue < 0.05) differential abundance for at least one of three pairwise disease state comparisons: DF vs DSS, DHF vs DSS, or DF vs DHF. Volcano plots were used to visualize the differential abundance of features between disease states (Fig. S1), and histograms were used to visualize the frequency of feature differential abundance (Fig. S2) (Figure S2). In the pairwise comparison of DF vs DSS, 1,117 and 160 features were identified as less abundant (log 2 FC < -1) and more abundant (log 2 FC > 1), respectively, in DSS, and 235 features with significant p-values presented with -1 < log2 FC <1 and were consid ered unperturbed. Comparing DHF vs DSS resulted in 961 and 320 features that were less or more abundant, respectively, in DSS, with 231 features considered unperturbed. Comparing DF vs DHF resulted in 364 features that were less abundant in DHF, 31 features that were more abundant in DHF, and 1,117 that met the defined criteria for "unperturbed. "
The 1,512 significant features possess information that can differentiate disease state but are not clinically useful without being identified. Therefore, to develop a tool with potential clinical utility, a subset of molecules was identified. These molecules were chosen based on pathway analysis, stark differential abundance between disease states, known relevance to DENV infection and disease, and previously published literature. Twenty-eight metabolites were included in the biomarker panel, all of which were identified using analytical standards and/or LCMS/MS and had an adjusted P-value < 0.05 for at least one pairwise dengue disease state comparison. Twelve and 16 metabolites were identified at confidence levels 1 and 2, respectively (30). A total of 48 relevant metabolites were identified at confidence levels 1 through 3, with level 1 representing the highest confidence and level 3 indicating tentative molecular formula prediction based on accurate mass (m/z). Serum metabolites were identified at confidence level 1 by comparing LC-MS/MS metrics to pure, synthetic reference standards, and at confidence level 2 by comparing experimental collision-induced dissociation data to known dissociation pathways and MS/MS spectral libraries. Identification of metabolites at each confidence level is further described in the methods and previous literature (30). The LC-MS/MS data used to validate metabolite identities at confidence level 1 are provided in Fig. S3a through n. Further details and relevant pathways for all metabolites identified in this study are summarized in Table S3 and S4 summarizes the log2FC and adjusted pvalues from a moderated t-test for the pairwise disease state comparisons of these metabolites.
Circulating creatinine level was used as a quality control in this study. Creatinine is well known to positively correlate with increasing age (31), and this trend was recapitula ted in the current data set. Creatinine was identified in the current data set, and a linear regression on creatinine abundance as a function of patient age produced a positive slope, independent of disease state and sex (Fig. S4). Other metabolites identified did not strongly correlate with age.
## Identified metabolites accurately classify severe dengue disease
We evaluated the capacity of the biomarker panel for disease state classification. Using 28 identified metabolites, an ensemble of classification models was trained and tested on their ability to differentiate DF from DHF/DSS or ND from DF. A random forest (rf ) model performed exceedingly well in differentiating DF from DHF/DSS, resulting in a balanced accuracy of 96.88% (Fig. 4a). When challenged with the classification of the test set, rf misclassified only one DHF/DSS sample as DF (Fig. 4b). The variable importance plot for DF vs DHF/DSS classification is displayed in Fig. 4c, indicating the impact of each metabolite on the quality of disease state classification. The high classification accuracy of the rf model was an important outcome, but underprediction of severe disease was not satisfactory. To investigate if the day of illness was a confounding variable for classification of DF vs DHF/DSS, we performed a subset analysis for samples collected on days 3 to 6 of illness. This subset classification analysis correctly classified all samples (100% balanced accuracy), and the single DHF/DSS sample was no longer underpredicted. Further, when plasma samples were excluded from classification, all samples were correctly classified besides one DHF/DSS, and even the days 3 to 6 subset analysis did not properly classify that single sample. For the classification of ND vs DF, performance was poorer, resulting in a balanced accuracy of 69.04% (Fig. 2).
## Dipeptides contribute greatly to classification success
Various dipeptides and amino acids were detected in patient samples (Fig. S7). The dipeptides L-prolyl-proline, valyl-leucine, and leucyl-alanine were major drivers of disease state classification and were depleted in all DSS patients and in some DHF patients, in whom a bimodal distribution of abundance was observed. The amino acid proline was significantly decreased in DHF and DSS patients compared to DF. Phenyla lanyl-phenylalanine circulating levels also significantly decreased with dengue disease severity, but the amino acid phenylalanine was unperturbed across disease states.
## Tryptophan metabolism is reduced in dengue disease
Tryptophan metabolism has been implicated in many disease conditions, including viral infection. In this study, tryptophan and seven of its metabolites were identified in patient samples. Metabolite abundance in each patient sample was stratified by disease state and visualized using boxplots, and tryptophan metabolic pathways were illustrated (Fig. 5). Additionally, the differential abundance of each metabolite is described by log 2 FC and adjusted p-value (Table S4).
Tryptophan abundance was significantly lower in DF patients compared to ND. Regarding dengue disease severity, tryptophan abundance progressively decreased in DHF and DSS patients. Because tryptophan binds to the protein albumin in circulation, sample-matched clinical measures of albumin were correlated with tryptophan levels in a subset of 144 patients. Tryptophan and albumin levels had a positive Spearman's ρ correlation coefficient of 0.45 (P-value < 0.0001) (Fig. 6a). Within each individual disease state, positive correlations were observed but with variable p-values, likely due to differences in sample size. Further, albumin levels in DSS (P-value < 0.0003) and DHF (not significant) were lower on average than in ND or DF patients (Fig. 6b), suggesting that disease state may be the main driver of correlation.
Kynurenine, a metabolite from one of the two host-driven tryptophan metabolic pathways, was significantly elevated in DF and DHF compared to ND. Kynurenine was not significantly elevated in DSS compared to ND, and there were no observed significant differences in kynurenine between dengue disease states. The first metabolic product in the serotonin pathway, 5-hydroxytryptophan, was significantly elevated in DF and DHF, but not significantly different in DSS, compared to ND. Accordingly, 5-hydroxytryptophan was significantly lower in DSS compared to DF.
Serotonin, the downstream metabolite of 5-hydroxytryptophan, was significantly decreased in all dengue disease states. In DF and DHF patients, serotonin abundance was moderately decreased relative to ND. Strikingly, serotonin abundance was massively decreased in DSS patients. Because serotonin is known to affect platelet function and activity, for a subset of 144 patients, sample matched clinical measures of platelets were correlated to serotonin levels. Serotonin and platelet levels had a positive Spearman correlation coefficient of 0.52 (P-value < 0.0001) (Fig. 6c). For DSS patients only, a low correlation coefficient of 0.09 was observed, suggesting differences between disease states may be the main driver of correlation. Further, platelet levels were significantly decreased in all dengue disease states compared to ND, but platelet levels were not significantly different between DHF and DSS patients (Fig. 6d).
A trend in serotonin abundance was observed when stratified by day of illness, and a similar trend was observed for platelets for a subset of patients (Fig. S5). In a cross-sectional analysis of DF patients, serotonin abundance began "normal" or relatively high on day 1 of illness, then decreased until days 5 or 6, where the rate of decrease (slope) levels off. The same decreasing trend was observed in DHF. In DSS patients, serotonin levels were massively decreased on all days (days 3 to 6). Three metabolic products from the indole metabolic pathway of tryptophan via the gut microbiota were measured: indoleacetic acid (IAA), indolelactic acid (ILA), and indolepropionic acid (IPA) (Fig. 5). The abundances of IAA and IPA were significantly decreased in DHF/DSS, and the abundance of IAcrA was significantly decreased in all dengue disease states (Table S4). ILA abundance was not observed to differ significantly among disease states.
## Fatty acid metabolism is elevated in severe dengue disease
The abundances of omega-3 (n-3) and omega-6 (n-6) fatty acids (FA) (Fig. 7; Table S4) and their downstream bioactive lipids were assessed (Fig. S6 and Table S4). Linoleic acid (18:2 n-6), a dietary FA, and three downstream FAs within the desaturation and elongation pathway were identified at confidence level 1 by comparison to synthetic standards. Three bioactive eicosanoids within different pathways of the n-6 arachidonic acid cascade were tentatively identified based on molecular formula prediction by accurate mass (confidence level 3). Regarding the n-3 FAs, α-linolenic acid and docosahexaenoic
## Bioactive sphingolipids are perturbed in dengue disease
Four bioactive sphingolipids were observed in the patient samples: sphingosine (d18:1), sphingosine-1-phosphate, sphinganine (d18:0), and sphinganine-1-phosphate (Fig. S6). Elevated levels of sphingosine (d18:1) and sphinganine (d18:0) were measured in DSS patients compared to all other disease conditions. Decreasing levels of sphingosine-1phosphate and sphinganine-1-phosphate were measured in DHF and DSS patients in a disease severity-dependent manner.
## Other metabolites of interest
Related to purine metabolism, hypoxanthine and inosine were significantly lower in abundance in severe disease (Fig. S6). Amino acids, dipeptides, carnitines, and others are visualized in Fig. S7. Glycerophospholipids and glycerolipids are visualized in Fig. S8. All identified metabolites of interest are described in Table S2 andS3.
## DISCUSSION
## Circulating metabolic profile and biosignature development
To characterize the metabolic changes during dengue disease, circulating metabolites were measured using untargeted LC-MS/MS. Pathway analysis enabled the generation of hypotheses and metabolic pathways of interest and was supplemented by differential abundance analysis of individual features and metabolites. A disparity in the number of unperturbed features was observed between disease state comparisons. More unpertur bed features for DF vs DHF than DF vs DSS indicated that more metabolic changes were measured in DSS. Regarding the direction of differential abundance, most features were less abundant in more severe forms of dengue disease, especially DSS. The direction of change in disease states may have implications as correlates of protection (higher in DF, lower in DSS) or as correlates of severe disease pathogenesis (lower in DF, higher in DSS). Thus, interesting features were structurally identified at confidence levels 1 or 2 for use as biomarkers and to enable inferences about their biological roles.
Multiple metabolomics studies involving DENV-infected patients have been conducted, all of which have contributed potential biomarkers from diverse sample backgrounds (10)(11)(12)(13)(14). Previous findings in adult patients corroborate many of the findings presented here in pediatric patients. Cui et al. observed reduction of dipeptides leucyl-alanine, phenylalanine, and serinyl-cysteine in adult DHF patients compared to DF (11)(12)(13). These studies also reported decreased serotonin and tryptophan metabolites in adults with DHF. Additionally, there is some agreement between the adult and pediatric data on the direction of lysophospholipid perturbation in DHF.
This current pediatric study contributes novel distinguishing biomarkers of pediatric severe dengue disease by leveraging a large dataset generated from 535 samples, including 22 DSS patients. Twenty-eight of the identified metabolites were used as a biomarker panel to classify disease state using the pediatric patient samples.
Additionally, we observed trends in this pediatric metabolomics data set that matched well-established trends in the medical field. For example, it is well known that creatinine levels increase as a function of age in children and adults (31), independent of sex. In our data set, the positive correlation between creatinine and patient age recapitulated these well-established trends (Fig. S4). Creatinine levels in serum or urine are commonly used as a measure of renal health. Kidneys filter creatinine out of the blood, such that an increase in serum creatinine levels may be a sign of liver dysfunction.
## Disease state classification
Classification success for DF vs DHF/DSS implies that phenotype differences were reflected in the patients' circulating metabolome. Only one DHF/DSS sample was misclassified as DF in the full sample set analysis, and all samples were correctly classified by a subset analysis for days 3 to 6 of illness. Exclusion of plasma samples did not adversely affect classification performance. Furthermore, poor classification performance for ND vs DF was likely due to less phenotype difference between febrile DF and other febrile illnesses (ND).
Classification using the 28-metabolite biomarker panel revealed potential for prediction of severe dengue disease, as well as supported the hypothesis that the host metabolome contains dengue disease phenotype information. The biomarker panel presented here should be further validated in diverse sample sets representing different geographic locations, genetic backgrounds, and age ranges.
## Dipeptides
In this study, four identified dipeptides were reduced in DHF/DSS patients and were critical for disease state classification. Dipeptides have been used to treat human papillomavirus-associated disease (32), and a naturally occurring anti-HIV-1 dipeptide was discovered in elite controllers of HIV-1 (33,34). Furthermore, dipeptidyl peptidase-4 (DPP4) inactivates bioactive peptides via cleavage of X-proline or X-alanine dipeptides from their N-terminus (35). DPP4 plays a role in immunity, cancer, and diabetes, but also antagonizes vasoconstriction and platelet aggregation via cleavage of NPY1-36, a neuropeptide that can be found in endothelial cells (36). It is intriguing that two of the highly perturbed dipeptides in DHF/DSS patients (prolyl-proline and leucyl-alanine) contain peptide motifs that are cleavable by DPP4. To our knowledge, this is the first evidence that exhibits association between dipeptides and dengue disease, and this association should be studied further.
## Tryptophan metabolism
Tryptophan can be metabolized into bioactive molecules that function in inflammation and aging, gut-brain homeostasis, immune regulation, cardiovascular diseases, and endothelial dysfunction (37)(38)(39)(40)(41). Tryptophan metabolism is also implicated in infectious diseases, including DENV infection and disease (10)(11)(12). Importantly, dysregulation of tryptophan metabolism observed within our data set is consistent with, and significantly expands upon, the knowledge gained from previous findings (12).
Hydroxylation of tryptophan via tryptophan hydroxylase (TPH) leads to the forma tion of bioactive serotonin. Serotonin and its precursor, 5-hydroxytryptophan, were identified in this study. 5-Hydroxytryptophan is the direct precursor of serotonin. The elevated 5-hydroxytryptophan levels in DF and DHF compared to ND patients could be a protective metabolic response, in which host metabolism shifts to replenish serotonin levels. In DSS patients, 5-hydroxytryptophan levels are not significantly different from ND patients and are significantly lower than DF patients. The lack of elevated 5-hydroxytryp tophan levels in DSS may indicate that patient metabolism did not effectively shift to supplement depleted serotonin levels. It is also important to note that 5-hydroxytrypto phan did not show the same decrease in DSS that was observed for serotonin. Therefore, as hypothesized in a previous study (12), normal 5-hydroxytryptophan levels in DSS patients indicate that serotonin synthesis is not the major perturbation; rather, serotonin release or uptake is highly perturbed.
Serotonin abundance was highest in ND patients, progressively lower in DF and DHF, and drastically depleted in DSS patients. Platelets trended similarly to serotonin, except that they were not significantly depleted in DSS compared to DHF patients. Additionally, serotonin in DF and DHF patients trended with day of illness (Fig. S5), resembling previously reported platelet count behavior during DF progression, where platelets hit a low on day 6 and increased to normal by day 10 (42). A similar platelet trend for days 1 to 6 was observed in this study. Thus, it is hypothesized that serotonin abundance may also recover by day 10 in DF patients. These findings contribute to the hypothesized connection between platelets and serotonin but suggest diverse roles for serotonin, indicating that mechanisms other than thrombocytopenia may contribute to the depletion of circulating serotonin in DSS.
In the mediation of shock and vascular endothelial dysfunction, serotonin may either be protective or pathogenic. In a pathogenic role, serotonin has been shown to induce local vasodilation when released at the vascular endothelium, affecting endothelial function and contributing to vascular leakage and shock (43,44). Systemic shock can be induced by circulating antibody-antigen immune complexes via the platelet Fcγ receptor IIA (44). Within this immune complex-induced mechanism, serotonin is pathogenic and is required for vasodilation. Serotonin can also be derived from mast cells and induce significant platelet activation, leading to thrombocytopenia via platelet aggregation and increased splenic uptake (45). However, platelet activation is also necessary for hemostasis and vascular wall maintenance. Thus, serotonin may also play a protective role.
The reason for circulating serotonin depletion in dengue disease is still unclear. One potential serotonin sink in dengue disease could be the liver. Liver damage is a clinical presentation of severe dengue disease, indicated by increased levels of liver transaminases and bilirubin in circulation. Multiple studies involving DENV-infected adults and children reported that greater than 90% of patients had elevated circulating liver aminotransferases (46). Through altered expression of serotonin receptor subtypes in the liver (e.g., 5-HT 2A and 5-HT 2B serotonin receptors), serotonin mediates hepatocyte proliferation and restoration of hepatic mass following injury (47)(48)(49)(50). Therefore, the uptake and usage of serotonin by hepatocytes to counteract DENV-induced liver damage may, in part, account for decreased circulating serotonin levels.
Decreased production of albumin by the liver (e.g., liver damage) or increased escape of circulating albumin from the vascular space (increased vascular permeabil ity) can lead to hypoalbuminemia (51). Two studies involving DENV-infected children reported that 60-80% of children present with hypoalbuminemia (46). In blood, most tryptophan (90%) is bound to albumin, and 10% of tryptophan is unbound (free) and available for tissue uptake. Therefore, decreased tryptophan levels detected in dengue patients may be related to decreased albumin, which was supported by the positive correlation observed between the two measures. Also, rapid equilibration between albumin-bound and free tryptophan in blood, paired with sustained tissue uptake, results in depleted tryptophan blood levels (52). Tissue uptake and subsequent usage of tryptophan through one of its four metabolic pathways, without sufficient replenishment of tryptophan, could be another reason for decreased circulating tryptophan in dengue disease.
IPA has been reported to have immunomodulatory, anti-inflammatory, and antioxi dant effects, as well as various protective functions in mammals (39,53). IPA abundance has a positive correlation with factors that promote cardiovascular health, and IPA levels decrease in scenarios where cardiovascular health is diminished (54). Decreased IPA levels in DHF and DSS patients could be related to diminished cardiovascular health. Myocarditis has been reported but is not common in dengue disease.
## Fatty acid metabolism
Lipid metabolism is dysregulated in all disease states in our data set; specifically, FA metabolism and biosynthesis. FAs and other cellular lipids are important structural, signaling, and energy-yielding molecules for viral entry, replication, assembly, and release. Previous work has shown that FA biosynthesis is actively dysregulated during viral infection to benefit assembly and function of viral replication factories (20,(55)(56)(57). FA oxidation pathways and the citric acid cycle (TCA) were observed to be altered across disease states. DENV utilizes FA beta-oxidation to fuel the mitochondrial TCA cycle, which in turn provides adenosine-5′-triphosphate (ATP) as energy for the viral lifecycle. Cellular lipids can also mediate the inflammatory response associated with disease. Increased linolenic acid (n-3) and linoleic acid (n-6, precursor to arachidonic acid) metabolism leads to formation of the pro-and anti-inflammatory and bioactive lipid mediators, such as eicosanoids (leukotrienes, prostaglandins, thromboxanes) (58)(59)(60).
Linoleic acid (18:2 n-6) is an essential n-6 FA that is metabolized to arachidonic acid or other very-long-chain FAs. Enhanced levels of linoleic acid and metabolites serve to increase pools of arachidonic acid and its downstream bioactive lipids. Three eicosanoids were tentatively identified in this study and were elevated in DSS patients (Fig. S6): leukotriene A 4 (LTA 4 ), prostaglandin E 2 (PGE 2 ), and hepoxilin A 3 (HxA 3 ). These three molecules may have implications for increased vascular permeability observed in severe dengue disease.
Leukotrienes are produced by the enzyme 5-lipoxygenase (5-LOX). Initially, 5-LOX generates 5-hydroperoxyeicosatatraenoic acid, which is unstable and rapidly converted to 5-hydroxyeicosatetraenoic acid or to LTA 4 , which is then converted to LTB 4 or LTC 4 . LTC 4 serves to increase vascular permeability and plasma leakage (61,62), both of which are symptoms observed in DHF and DSS.
Prostaglandins are produced via arachidonic acid through cyclooxygenase (COX) isoenzymes. PGE 2 is significantly increased during inflammation, where it causes increased microvascular permeability (59,63). Another study demonstrated involvement of COX-2 in DENV replication in cell culture (64), potentially implying a virally induced mechanism for increased PGE 2 . HxA 3 is a non-canonical eicosanoid that is produced when arachidonic acid is converted via the 12S-LOX enzyme to 12SHpETE, the HxA 3 precursor. HxA 3 increases vascular permeability in rat skin, induces neutrophil chemo taxis, and stimulates release of arachidonic acid and diacylglycerol (65,66). Elevated circulating arachidonic acid levels could, in part, be related to HxA 3 activity.
## Sphingolipids
Sphingosine is the common backbone of the diverse sphingolipid molecules, which can have signaling and structural roles. Sphingosine, or the closely related sphinganine, can be phosphorylated to generate a potent signaling molecule. Based on ELISA measure ments of adult dengue patient sera from Colombo, Sri Lanka, sphingosine-1-phosphate decreased in a severity-dependent manner when compared to healthy controls (67). This current mass spectrometry-based study of pediatric patients from Nicaragua recapitula ted the disease severity-dependent decrease in sphingosine-1-phosphate and revealed that this trend extends to DSS patients. Like serotonin, sphingosine-1-phosphate is stored in and released by platelets (68). Thus, these trends may be associated with thrombocytopenia.
Sphingosine-1-phosphate and its various receptors (S1PR1 through S1PR5) have been implicated in both protection and disruption of the endothelial barrier. Upon binding of sphingosine-1-phosphate, S1PR1 promotes regulation of endothelial cell function via downstream signaling. In contrast, S1PR2 induces vascular permeability via the Rho-ROCK-PTEN signaling cascade that facilitates phosphorylation and loss of VE-Cadherin at adherens junctions (69)(70)(71). Modak et al. hypothesized that low serum sphingosine-1phosphate and DENV-induced upregulation of the high-affinity S1PR2 in endothelial cells result in preferential activation of disruptive signaling pathways leading to vascular permeability (71).
Sphinganine-1-phosphate was progressively decreased in DF, DHF, and DSS patients in this study (Fig. S6). Sphinganine-1-phosphate lacks a double bond and is thus closely related to sphingosine-1-phosphate and can bind S1P receptors. Sphinganine-1phosphate was reportedly depleted in mice after hepatic ischemia-reperfusion, and exogenous replenishment of sphinganine-1-phosphate protected the mice against liver and kidney injury, improving endothelial integrity and vascular function (72). The protective effect of sphinganine-1-phosphate on endothelial cells was found to be related to S1PR1 (73). Thus, depleted sphinganine-1-phosphate in DENV-infected patients may be related to vascular integrity.
## Inosine and hypoxanthine
Decreased levels of inosine and its downstream product hypoxanthine (Fig. S6). In dengue disease, especially severe disease, could be related to higher levels of adenosine (inosine precursor) and lower levels of xanthine and uric acid (hypoxanthine catabolites). Adenosine is a vasodilator and platelet aggregation inhibitor and thus could contrib ute to leak (74)(75)(76)(77). Circulating adenosine was reported to inhibit polymorphonuclear leukocyte function, resulting in decreased synthesis of specialized pro-resolving lipid mediators during coagulation that drive resolution of inflammation (78). Additionally, circulating adenosine deaminase abundance and activity (the enzyme that converts adenosine to inosine) are reportedly altered in various diseases (79,80). Adenosine was not detected in this data set, but its potential importance warrants further investigation.
## Study limitations
Biomarkers with potential for triaging severe disease should be perturbed and measura ble within the first three days of the acute phase of dengue disease. This study included samples that were collected from day 0 to 6 of illness. Thus, the perturbed metabolites described here were associated with DHF/DSS but were not used to predict progression to DHF/DSS. The next step towards identifying biomarkers for early triage will require analyzing the biomarkers presented here in samples from patients on day 3 of illness or earlier who later progress to severe disease. Additionally, further expansion of these studies to other geographical regions, genetic backgrounds, age ranges, and longitudi nal sample collections will help to fortify such biomarker panels.
This study did not include healthy controls. Instead, we used the ND control and various dengue disease severities to assess the circulating metabolic changes that were specific to DF febrile illness and metabolic changes between DF and DHF/DSS.
This study included majority serum and some plasma samples, but disease state classification was equally successful with or without inclusion of plasma samples in the training and test sets. Serum sample preparation involves platelet coagulation, which should be considered when interpreting the relationship between platelets and serum metabolites. Additionally, coagulation can induce production of arachidonic acid-derived lipid mediators.
Like all metabolomics studies, this study had an inherent measurement bias based on molecular polarity. The extraction protocol and LC-MS/MS analyses were optimized for mid-polar molecules, so additional workflows would be necessary if comprehensive analyses of highly polar metabolites, sugars, or nonpolar lipids are desired.
## Conclusion
Here, we demonstrate that the pediatric circulating metabolome is dynamically altered in response to DENV infection and reveals signatures associated with disease severity. Dengue disease state was accurately classified, and the biochemistry of severe dengue disease pathogenesis was explored. Unique to this study was the identification of the metabolic biosignatures of DSS, and biomarkers of DHF were identified that were either novel or that recapitulated those reported in studies from other geographical regions, supporting the hypothesis that the metabolome can provide information that is independent of age, and geographical and genetic backgrounds. These studies could be critical in assembling biomarkers for early triaging of severe dengue disease.
## MATERIALS AND METHODS
## Chemicals and reagents
LC-MS grade water and methanol were purchased from Honeywell (Charlotte, NC). LCMS grade acetonitrile was purchased from Fischer Scientific (Hampton, NH). Analytical grade 5-hydroxy-L-tryptophan, arachidonic acid, arachidonic acid-d 8 , creatinine, creatinined 3 , dihomo-gamma-linolenic acid, dihomo-gamma-linolenic acid-d 6 , gamma-linolenic acid, indole-3-acetic acid, indole-3-lactic acid, indole-3-propionic acid, linoleic acid, linoleic acid-d 4 , L-kynurenine, L-tryptophan, serotonin hydrochloride, and serotonin-d 4 hydrochloride were purchased from Cayman Chemical (Ann Arbor, MI). Analytical grade carnitine, hypoxanthine, and inosine were purchased from Sigma Aldrich (St. Louis, MO).
## Study design
For this study, 535 samples from 535 individuals were retrospectively obtained from two different well-established studies in Managua, Nicaragua (81,82). A set of 122 well-char acterized samples collected between 2012 and 2013 was sourced from the Pediatric Dengue Cohort Study (PDCS), following over 3,800 children between the ages of 2 and 17 years old since 2004 (Fig. 1). Cohort patients were enrolled as healthy volunteers, followed for all medical episodes, and monitored for suspected arboviral diseases. Those who met the case definition of dengue, or undifferentiated febrile illnesses, were worked up for laboratory confirmation using molecular biological, virological, and/or serologi cal methods. Another 413 samples were obtained from patients aged 1 to 14 years old (Fig. 1) in the Pediatric Dengue Hospital-based Study (PDHS), who presented at the Hospital Infantil Manuel de Jesús Rivera, the National Pediatric Reference Hospital in Nicaragua, between 2005 and 2015, with a fever or history of fever for <7 days and one or more of the following signs and symptoms: headache, arthralgia, myalgia, retro-orbital pain, positive tourniquet test, petechiae, or signs of bleeding. Cases were laboratory-confirmed for DENV infection by detection of DENV RNA by RT-PCR, isolation of DENV, seroconversion of DENV-specific IgM antibody titers observed by MAC-ELISA in paired acute-and convalescent-phase samples, and/or seroconversion or a ≥4-fold increase in anti-DENV antibody titer measured using inhibition ELISA (iELISA) in paired acute and convalescent samples (83)(84)(85). Immune status was determined using iELISA in early convalescent samples (14 or more days post-onset of symptoms); a titer <2,560 was considered primary infection, and ≥2,560 was considered secondary infection. Cases were classified by disease severity (DF, DHF, or DSS) using computerized algorithms based on the 1997 WHO schema (21). Samples that were negative for DENV infection were classified as ND. Metadata recorded included patient sex, age, disease outcome (disease state), infection history, DENV serotype, day of illness (number of days of fever at time of enrollment and sample collection), date of sample collection, and detailed clinical data across disease evolution.
These samples were procured as part of the normal dengue diagnosis mission of the laboratory and not as part of an experimental protocol. Parents or legal guardians of participants provided written informed consent; participants 6 years of age and older provided assent; and participants aged 12 years and older in the Hospital-based Dengue Study provided written assent.
## Sample preparation and metabolite extraction
Twenty microliters of each serum sample were aliquoted into individual microcentrifuge tubes, and an additional 20 µL of each serum sample were combined to generate the pooled quality control (QC). For metabolite extraction, patient serum and pooled QC samples were randomized. Five microliters of L-tryptophan-d 5 (80 ng/mL) heavy-iso tope=labeled internal standard were added to 20 µL of patient serum (or pooled QC aliquot) in a microcentrifuge tube. To precipitate proteins, 100 µL of cold methanol was added to the serum, and samples were incubated for 12 h at -80°C. Samples were then centrifuged at 4°C for 15 min at 18,000 × g to pellet proteins. Supernatant was then transferred to a new microfuge tube and dried under nitrogen. Samples were reconsti tuted in 25 µL of methanol/water (50/50), allowed to stand at room temperature for 15 min, vortexed for 20 sec, and centrifuged to pellet insoluble debris. Sample super natants were then transferred to autosampler vials fitted with low-volume inserts and immediately submitted for LC-MS analysis. Serum samples were prepared and analyzed in six randomized batches. A pooled quality control sample was run after every five experimental samples, and a solvent blank was run after every 10 samples.
## Untargeted liquid chromatography-mass spectrometry
Sample order for injection was randomized. Each sample was injected (7.5 µL) with an Agilent 1290 HPLC system where metabolites were separated on an XBridge BEH C18 column (2.5 µm particle size, 2.1 × 100 mm, Waters Millford, MA, USA). The total mobile phase flow rate was 0.250 mL/min, made up of water + 0.1% formic acid (mobile phase A) and 95/5 acetonitrile/water + 0.1% formic acid (mobile phase B). For metabolite separation, the mobile phase composition began at 5% B and held until 0.5 min. From 0.5 to 14 min, the mobile phase composition was adjusted in a linear fashion to 98% B. From 14.5 to 15 min, the mobile phase composition was returned to starting gradient conditions of 5% B. From 15 to 19.5 min, the starting gradient conditions were held to equilibrate the LC column for the subsequent sample injection. The LC column outlet was coupled to the electrospray ionization (ESI) source of an Agilent 6224 time-of-flight mass spectrometry system being operated in positive ionization mode. The ESI emitter was electrically grounded, and the MS-inlet capillary was held at -4,000 V to generate and transmit positive ions from the metabolites eluting from the LC column. The ESI nebulizer nitrogen gas was set to 45 psi, and the heated counter flow of nitrogen ("dry gas" used to aid in droplet desolvation) was flowed at 10 L per minute and held at 310°C. Beyond the MS-inlet capillary, the fragmentor voltage was set to 120 V to aid in ion desolvation and transmission, and the skimmer voltage was set to 50 V. The peak-to-peak voltage of the ion transfer octopole was set to 750 V. The time-of-flight mass analyzer was set to scan between m/z 70 to 1,700, collecting full-scan (MS1) spectra at a rate of 1.66 spectra/sec.
## Metabolite identification via liquid chromatography-tandem mass spectrom etry
To confirm metabolite identity, RT, m/z, and the collision-induced dissociation (CID) product ion spectra were collected for each metabolite. For level 1 identification, RT, m/z, and CID product ion spectrum from a pure synthetic standard were matched to that of the molecule originating from serum. For level 2 identification, RT, m/z, and CID product ion spectra for the molecule in serum were compared to literature, spectral databases, or known dissociation patterns (86)(87)(88). For level 3 identification, the accurate mass (m/z) of a feature measured by the time-of-flight (TOF) mass analyzer was tentatively assigned a molecular formula (within 20 ppm mass error to the exact mass) and structure. Confirmation of metabolite identities was performed on an Agilent 1290 HPLC system coupled to an Agilent 6546 quadrupole time-of-flight (QTOF) mass spectrometry system. LC conditions, ionization polarity, and ion transfer optics in the MS were identical to conditions used for the untargeted LC-MS experiment. To acquire CID product ion spectra for metabolites of interest, the quadrupole mass analyzer was set to transmit the precursor m/z of interest with an isolation width of 1.3 Da. Precursor ions were transmitted through the quadrupole mass analyzer and accelerated into the collision cell filled with N 2 gas (24 psi). Within the collision cell, precursor ions underwent energetic collisions with N 2 gas molecules. Product ion spectra were collected at collision energies of 10 and 40 arbitrary units. After the collision cell, precursor and product ions were pulsed into the TOF for mass analysis and subsequent detection.
## Data processing
## Molecular feature extraction
The Agilent data files were converted from .d to .mzML format using ProteoWizard MS Convert version 3.0.6478 64 bit. Peak picking, retention time correction, chromatogram alignment, and gap filling were performed using XCMS software version 1.46 in R version 3.2.2 (89). The R package IPO was used to optimize XCMS parameters on the dataset (90), and CAMERA was used for deisotoping (91).
## Molecular feature description
Within each data file, the abundance and tentative identity of many potential metabo lites (molecular features) are embedded. Two LC-MS metrics that define a molecular feature are the RT of its chromatographic peak and the m/z of the ion that correlates to the chromatographic peak. Enabled by the accurate mass (m/z) measurement capability of the time-of-flight mass analyzer, each feature was tentatively identified by searching its measured accurate mass (within ± 20 parts-per-million mass error) against the Human Metabolome Database (HMDB), LipidMaps, Metlin, and Kegg databases using CEU mass mediator (92). Additionally, the area under the chromatographic peak of a molecular feature represents its abundance in each serum sample.
## Preprocessing
All data preprocessing steps were conducted in R version 3.4.2 (93). Features were first filtered to remove any that failed to meet the following criteria: (1) present in at least 20% of all samples across all analysis batches; (2) present in at least 75% of all pooled QC samples; and (3) present in at least 70% of samples from at least one disease group (ND, DF, DHF, DSS).
Normalization was conducted stepwise. First, within each batch separately, features that were not present in at least 80% of pooled QC samples in that batch were removed. Features were then normalized using a Tobit regression (left-censored at the minimum value for each batch) fitted to the pooled QC samples, implemented in the R package "AER" version 1.2.5 (94,95). Batches were then combined, retaining only those features present in all batches, and normalized feature-wise across all batches by the ratio of pooled QC mean batch abundance to pooled QC overall mean abundance.
The year of sample collection influenced the presence or abundance of some features, suggesting that length of storage and/or methods of collection and handling may be influencing the results. Therefore, features were removed if found to be present in ≥50% of all samples (irrespective of group) only in samples collected between 2005 and 2009, or only in samples collected between 2011 and 2015. However, if the feature appeared to be specific to a sample infected with a single serotype (i.e., found in >50% of samples of that serotype and <50% of samples from other serotype), then the feature was retained. Finally, features that had a coefficient of variation >30% in pooled QC samples after combining and normalizing batches were deemed unreliable and removed. Abundances were log 2 -transformed. Missing values were imputed using a random forest algorithm implemented in the R package missForest version 1.4 (96). Abundance variance was calculated for each feature across all samples, and features in the lowest quartile were excluded from analysis (97).
## Metabolic pathway analysis
Differential abundance and statistics (t-score and p-value) were calculated feature-wise via limma for each pairwise comparison of dengue disease severity. Pathway analysis was performed on results from each pairwise disease state comparison. Feature m/z, RT, p-value, and t-score were used as inputs for the mummichog metabolic pathway analysis algorithm (98). For mummichog version 2.0 analysis, primary ion types were forced, the metabolite p-value cutoff was 0.05, a minimum of three metabolites were required to flag a pathway, and the pathway library was Homo sapiens (MFN). A 20 ppm mass error tolerance was set for tentative metabolite annotation.
## Statistical analysis
Univariate analysis of features was implemented in the R package limma version 3.32.10, in which linear models with empirical Bayes statistics were applied feature-wise to generate pairwise comparison of feature abundances across disease states (99)(100)(101). The models provided estimated log2 FCs, moderated t-statistics, and false discovery rate adjusted p-values for each feature. Significant features were defined by log 2 FC ≥ 1 and P-value < 0.05 after adjustment for false discovery rate via the Benjamini-Hochberg method.
To develop classification models, the 535 samples were randomly divided within disease states into a training set (75% of samples within each disease state) and a test set (remaining 25% of samples), where disease states were ND, DF and severe disease (DHF/DSS). DHF and DSS were combined into the severe disease category to alleviate the small sample size for each disease state. Training sets were used to fit models. Test sets were used only to test predictions from fitted models.
Using the training sample set, an ensemble of classification models was constructed for two separate classification problems: ND vs DF and DF vs DHF/DSS (102)(103)(104). The R "caret" package version 6.0.93 was used to build several classification models (105), including a generalized linear model, random forest, adaBoost.M1, support vector machines, linear and quadratic discriminate analysis, and k-nearest neighbors. Parame ters were chosen by five-fold cross-validation. Peak areas were scaled and centered. The generalized linear model, random forest, and AdaBoost.M1 gave perfect classification on the training set and were therefore chosen for evaluation with the test set.
## References
1. Brady, Gething, Bhatt et al. (1760) "Refining the global spatial limits of dengue virus transmission by evidence-based consensus" *PLoS Negl Trop Dis*
2. Bhatt, Gething, Brady et al. (2013) "The global distribution and burden of dengue" *Nature*
3. Cattarino, Rodriguez-Barraquer, Imai et al. (2020) "Mapping global variation in dengue transmission intensity" *Sci Transl Med*
4. Who (2024) "Dengue -Global Situation"
5. Reich, Shrestha, King et al. (2013) "Interactions between serotypes of dengue highlight epidemiological impact of crossimmunity" *J R Soc Interface*
6. Halstead (2007) *Dengue. The Lancet*
7. Jusof, Lim, Aziz et al. (2022) "The cytokines CXCL10 and CCL2 and the kynurenine metabolite anthranilic acid accurately predict patients at risk of developing dengue with warning signs" *J Infect Dis*
8. Ghetia, Bhatt, Mukhopadhyay (2022) "Association of dengue virus non-structural-1 protein with disease severity: a brief review" *Trans R Soc Trop Med Hyg*
9. Yuan, Chen, Zhong et al. (2022) "Risk and predictive factors for severe dengue infection: a systematic review and meta-analysis" *PLoS One*
10. Voge, Perera, Mahapatra et al. (2016) "Metabolomics-based discovery of small molecule biomarkers in serum associated with dengue virus infections and disease outcomes" *PLoS Negl Trop Dis*
11. Cui, Lee, Kumar et al. (2013) "Serum metabolome and lipidome changes in adult patients with primary dengue infection" *PLoS Negl Trop Dis*
13. Cui, Lee, Thein et al. (2016) "Serum metabolomics reveals serotonin as a predictor of severe dengue in the early phase of dengue fever" *PLoS Negl Trop Dis*
14. Cui, Pang, Lee et al. (2018) "Serum metabolome changes in adult patients with severe dengue in the critical and recovery phases of dengue infection" *PLoS Negl Trop Dis*
15. Villamor, Villar, Lozano-Parra et al. (2018) "Serum fatty acids and progression from dengue fever to dengue haemorrhagic fever/dengue shock syndrome" *Br J Nutr*
16. Carthew (2021) "Gene regulation and cellular metabolism: an essential partnership" *Trends Genet*
17. Baker, Rutter (2023) "Metabolites as signalling molecules" *Nat Rev Mol Cell Biol*
18. Bi, Parysek, Warnick et al. (1993) "In vitro evidence that metabolic cooperation is responsible for the bystander effect observed with HSV tk retroviral gene therapy" *Hum Gene Ther*
19. Cloviss (2022) "Innate metabolic responses against viral infections" *Nat Metab*
20. Mackenzie, Khromykh, Parton (2007) "Cholesterol manipula tion by West Nile virus perturbs the cellular immune response" *Cell Host Microbe*
21. Heaton, Perera, Berger et al. (2010) "Dengue virus nonstructural protein 3 redistributes fatty acid synthase to sites of viral replication and increases cellular fatty acid synthesis" *Proc Natl Acad Sci*
22. (1997) "Dengue haemorrhagic fever: diagnosis, treatment, prevention, and control. World Health Organiza tion"
23. Shirakami, Lee, Clugston et al. (2012) "Hepatic metabolism of retinoids and disease associations" *Biochimica et Biophysica Acta (BBA) -Molecular and Cell Biology of Lipids*
24. Perez, Briz (2009) "Bile-acid-induced cell injury and protection" *WJG*
25. Chen, Maguire, Hileman et al. (1997) "Dengue virus infectivity depends on envelope protein binding to target cell heparan sulfate" *Nat Med*
26. Ling, Li, Khan et al. (2022) "Is heparan sulfate a target for inhibition of RNA virus infection?" *Am J Physiol Cell Physiol*
27. Tang, Alonso, Ng et al. (2017) "Increased serum hyaluronic acid and heparan sulfate in dengue fever: association with plasma leakage and disease severity" *Sci Rep*
28. Lennon, Singleton (2011) "Hyaluronan regulation of vascular integrity" *Am J Cardiovasc Dis*
29. Queisser, Mellema, Petrey (2021) "Hyaluronan and its receptors as regulatory molecules of the endothelial interface" *J Histochem Cytochem*
30. Espinosa, Beatty, Puerta-Guardo et al. (2019) "Increased serum sialic acid is associated with morbidity and mortality in a murine model of dengue disease" *J Gen Virol*
31. Schrimpe-Rutledge, Codreanu, Sherrod et al. (2016) "Untargeted metabolomics strategies-challenges and emerging directions" *J Am Soc Mass Spectrom*
32. Uemura, Honda, Matsuyama et al. (2011) "Age, gender, and body length effects on reference serum creatinine levels determined by an enzymatic method in Japanese children: a multicenter study" *Clin Exp Nephrol*
33. Makatsariya, Bitsadze, Khizroeva et al. (2019) "Efficacy and safety of glucosaminylmuramyl dipeptide in treatment of human papillomavirus-associated diseases: a systematic review" *Akušerstvo, ginekologiâ i reprodukciâ*
34. Ceña-Diez, Narayanan, Ray et al. (2023) "Naturally occurring dipeptide from elite controllers with dual anti-HIV-1 mechanism" *Int J Antimicrob Agents*
35. Giammarino, Sönnerborg, Ceña-Diez (2024) "Broad synergistic antiviral efficacy between a novel elite controller-derived dipeptide and antiretrovirals against drug-resistant HIV-1" *Front Cell Infect Microbiol*
36. Mentlein (1999) "Dipeptidyl-peptidase IV (CD26)-role in the inactiva tion of regulatory peptides" *Regul Pept*
37. Li, Demuth, Zukowska (2006) "Dipeptidyl peptidase IV: a molecular switch of vascular actions of neuropeptide Y" *Adv Exp Med Biol*
38. O'mahony, Clarke, Borre et al. (2015) "Serotonin, tryptophan metabolism and the brain-gut-microbiome axis" *Behav Brain Res*
39. Fiore, Murray (2021) "Tryptophan and indole metabolism in immune regulation" *Curr Opin Immunol*
40. Guijas, Horton, Hoang et al. (2022) "Microbial metabolite 3-indolepropionic acid mediates immunosuppression" *Metabolites*
41. Song, Ramprasath, Wang et al. (2017) "Abnormal kynurenine pathway of tryptophan catabolism in cardiovascular diseases" *Cell Mol Life Sci*
42. Wang, Zhang, Ding et al. (2014) "Activation of NAD(P)H oxidase by tryptophan-derived 3-hydroxykynur enine accelerates endothelial apoptosis and dysfunction in vivo" *Circ Res*
43. Yasuda, Saito, Suzuki et al. (2021) "Unique characteristics of new complete blood count parameters, the immature platelet fraction and the immature platelet fraction Count, in dengue patients" *PLoS One*
44. (2025) *Full-Length Text Journal of Virology*
45. Vanhoutte (2016) "Regenerated endothelium and its senescent response to aggregating platelets" *Circ J*
46. Cloutier, Allaeys, Marcoux et al. (2018) "Platelets release pathogenic serotonin and return to circulation after immune complexmediated sequestration" *Proc Natl Acad Sci*
47. Masri, Mantri, Rathore et al. (2019) "Peripheral serotonin causes dengue virus-induced thrombocytopenia through 5HT 2 receptors" *Blood*
48. Samanta (2015) "Dengue and its effects on liver" *World J Clin Cases*
49. Ruddell, Mann, Ramm (2008) "The function of serotonin within the liver" *J Hepatol*
50. Clavien (2008) "Liver regeneration: a spotlight on the novel role of platelets and serotonin" *Swiss Med Wkly*
51. Lesurtel, Clavien (2012) "Serotonin: a key molecule in acute and chronic liver injury" *Clin Res Hepatol Gastroenterol*
52. Papadimas, Tzirogiannis, Mykoniatis et al. (2012) "The emerging role of serotonin in liver regeneration" *Swiss Med Wkly*
53. Soeters, Wolfe, Shenkin (2019) "Hypoalbuminemia: pathogene sis and clinical significance" *JPEN J Parenter Enteral Nutr*
54. Badawy (2019) "Tryptophan metabolism: a versatile area providing multiple targets for pharmacological intervention" *Egypt J Basic Clin Pharmacol*
55. Negatu, Gengenbacher, Dartois et al. (2020) "Indole propionic acid, an unusual antibiotic produced by the gut microbiota, with antiinflammatory and antioxidant properties" *Front Microbiol*
56. Konopelski, Mogilnicka (2022) "Biological effects of indole-3propionic acid, a gut microbiota-derived metabolite, and its precursor tryptophan in mammals' health and disease" *Int J Mol Sci*
57. Perera, Riley, Isaac et al. (2012) "Dengue virus infection perturbs lipid homeostasis in infected mosquito cells" *PLoS Pathog*
58. Chotiwan, Andre, Sanchez-Vargas et al. (2018) "Dynamic remodeling of lipids coincides with dengue virus replication in the midgut of Aedes aegypti mosquitoes" *PLoS Pathog*
59. Gullberg, Steel, Pujari et al. (2018) "Stearoly-CoA desaturase 1 differentiates early and advanced dengue virus infections and determines virus particle infectivity" *PLoS Pathog*
60. Dennis, Norris (2015) "Eicosanoid storm in infection and inflammation" *Nat Rev Immunol*
61. Ricciotti, Fitzgerald (2011) "Prostaglandins and inflammation"
62. Wang, Wu, Chen et al. (2021) "Metabolism pathways of arachidonic acids: mechanisms and potential therapeutic targets"
63. Dahlén, Björk, Hedqvist et al. (1981) "Leukotrienes promote plasma leakage and leukocyte adhesion in postcapillary venules: in vivo effects with relevance to the acute inflammatory response" *Proc Natl Acad Sci*
64. Björk, Hedqvist, Arfors (1982) "Increase in vascular permeability induced by leukotriene B4 and the role of polymorphonuclear leukocytes" *Inflammation*
65. Funk (2001) "Prostaglandins and leukotrienes: advances in eicosanoid biology" *Science*
66. Lin, Tseng, Wu et al. (2017) "Cyclooxygenase-2 facilitates dengue virus replication and serves as a potential target for developing antiviral agents" *Sci Rep*
67. Laneuville, Corey, Couture (1991) "Hepoxilin A3 increases vascular permeability in the rat skin" *Eicosanoids*
68. Cr (2015) "Pathophysiology of the hepoxilins" *Biochim Biophys Acta*
69. Gomes, Fernando, Fernando et al. (2014) "Sphingosine 1-Phosphate in acute dengue infection" *PLoS One*
70. Tolksdorf, Moritz, Wolf et al. (2022) "Platelet-derived S1P and its relevance for the communication with immune cells in multiple human diseases" *Int J Mol Sci*
71. Sanchez, Skoura, Wu et al. (2007) "Induction of vascular permeability by the sphingosine-1-phosphate receptor-2 (S1P2R) and its downstream effectors ROCK and PTEN" *Arterioscler Thromb Vasc Biol*
72. Anupriya, Singh, Hulyalkar et al. (2018) "Sphingolipid signaling modulates trans-endothelial cell permeability in dengue virus infected HMEC-1 cells" *Prostaglandins Other Lipid Mediat*
73. Modak, Mishra, Awasthi et al. (2024) "Fingolimod (FTY720), an FDA-approved sphingosine 1-phosphate (S1P) receptor agonist, restores endothelial hyperpermeability in cellular and animal models of dengue virus serotype 2 infection" *IUBMB Life*
74. Park, Kim, Chen et al. (2010) "Sphinganine-1phosphate attenuates both hepatic and renal injury induced by hepatic ischemia and reperfusion in mice" *Shock*
75. Sang Won, Kim, Chen et al. (2010) "Sphinganine-1-phosphate protects kidney and liver after hepatic ischemia and reperfusion in mice through S1P1 receptor activation" *Laboratory Investigation*
76. Morgan, Mccormack, Griffiths et al. (1991) "Adenosine as a vasodilator in primary pulmonary hyperten sion" *Circulation*
77. Hansen, Schnermann (2003) "Vasoconstrictor and vasodilator effects of adenosine in the kidney" *Am J Physiol Renal Physiol*
78. Johnston-Cox, Ravid (2011) "Adenosine and blood platelets" *Purinergic Signal*
79. Fuentes, Pereira, Mezzano et al. (2014) "Inhibition of platelet activation and thrombus formation by adenosine and inosine" *PLoS One*
80. Norris, Serhan (2018) "Metabololipidomic profiling of functional immunoresolvent clusters and eicosanoids in mammalian tissues" *Biochem Biophys Res Commun*
81. Beyazit, Koklu, Tas et al. (2012) "Serum adenosine deaminase activity as a predictor of disease severity in ulcerative colitis" *J Crohns Colitis*
82. Xuan, Tian, Zhang et al. (2019) "Serum adenosine deaminase activity and coronary artery disease: a retrospective case-control study based on 9929 participants" *Ther Adv Chronic Dis*
83. (2025) *Full-Length Text Journal of Virology*
84. Kuan, Gordon, Avilés et al. (2009) "The Nicaraguan pediatric dengue cohort study: study design, methods, use of information technology, and extension to other infectious diseases" *Am J Epidemiol*
85. Narvaez, Montenegro, Juarez et al. (2025) "Dengue severity by serotype and immune status in 19 years of pediatric clinical studies in Nicaragua" *PLoS Negl Trop Dis*
86. Katzelnick, Gresh, Halloran et al. (2017) "Antibody-dependent enhancement of severe dengue disease in humans" *Science*
87. Gutiérrez, Gresh, Pérez et al. (2013) "Evaluation of the diagnostic utility of the traditional and revised WHO dengue case definitions" *PLoS Negl Trop Dis*
88. Narvaez, Gutierrez, Pérez et al. (1397) "Evaluation of the traditional and revised WHO classifications of Dengue disease severity" *PLoS Negl Trop Dis*
89. "2024. mzCloud -Advanced Mass Spectral Database"
90. Pauling, Hermansson, Hartler et al. (2017) "Proposal for a common nomenclature for fragment ions in mass spectra of lipids" *PLoS One*
91. Uwpr (2007) "Peptide Fragmentation"
92. Smith, Want, 'maille et al. (2006) "XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification" *Anal Chem*
93. Libiseller, Dvorzak, Kleb et al. (2015) "IPO: a tool for automated optimization of XCMS parameters" *BMC Bioinformatics*
94. Kuhl, Tautenhahn, Böttcher et al. (2012) "CAMERA: an integrated strategy for compound spectra extraction and annotation of liquid chromatography/mass spectrometry data sets" *Anal Chem*
95. Gil-De-La-Fuente, Godzien, Saugar et al. (2019) "CEU mass mediator 3.0: a metabolite annotation tool" *J Proteome Res*
96. (2017) "R: A Language and Environment for Statistical Computing"
97. Tobin (1958) "Estimation of relationships for limited dependent variables" *Econometrica*
98. Wehrens, Hageman, Van Eeuwijk et al. (2016) "Improved batch correction in untargeted MS-based metabolomics" *Metabolomics (Los Angel)*
99. Stekhoven, Bühlmann (2012) "MissForest-non-parametric missing value imputation for mixed-type data" *Bioinformatics*
100. Hackstadt, Hess (2009) "Filtering for increased power for microarray data analysis" *BMC Bioinformatics*
101. Li, Park, Duraisingham et al. (2013) "Predicting network activity from high throughput metabolomics" *PLoS Comput Biol*
102. Smyth (2004) "Linear models and empirical bayes methods for assessing differential expression in microarray experiments" *Stat Appl Genet Mol Biol*
103. Phipson, Lee, Majewski et al. (2016) "Robust hyperparameter estimation protects against hypervariable genes and improves power to detect differential expression" *Ann Appl Stat*
104. Ritchie, Phipson, Wu et al. (2015) "Limma powers differential expression analyses for RNA-sequencing and microarray studies" *Nucleic Acids Res*
105. Breiman (2001) "Random Forests" *Mach Learn*
106. Freund, Schapire (1996) "Experiments with a new boosting algorithm"
107. Breiman (1998) "Arcing classifiers" *Ann Stat*
108. Kuhn, Wing, Weston et al. (2025) "Can Candan TH, R Core Team. 2022. Caret: classification and regression training Full-Length Text Journal of Virology November" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12656901&blobtype=pdf | # Pre-Clade IIb Mpox Virus Exposure in Ghana: A Retrospective Serological Analysis
David Schwartz, Robert Colebunders, Patrick Katoto, Christopher Dorcoo, Grace Gyamfi, Franziska Kaiser, Elvis Lomotey, Jeffrey Sumboh, Robert Fischer, Claude Kwe, Vincent Munster, Joseph Bonney, Irene Donkor
## Abstract
Monkeypox virus (MPXV), a zoonotic Orthopox virus endemic to West and Central Africa, causes mpox disease. Although Ghana had no confirmed human cases before 2022, the 2003 U.S. mpox outbreak was traced to rodents exported from Ghana, suggesting potential undetected exposure in the local population. This study assessed mpox exposure prior to the emergence of Clade IIb in humans. We tested 457 serum samples collected across 14 regions of Ghana using a commercial anti-MPXV IgG ELISA. These samples comprised 365 archived sera from 2021 SARS-CoV-2 surveillance and 92 sera from suspected mpox cases during the 2022 outbreak. Multivariable logistic regression was performed to examine associations between MPXV seropositivity and demographic factors, including age, sex, region, urban/rural status and inferred smallpox vaccination status. Overall MPXV seroprevalence was 6.6%. Participants from the Western Region had significantly increased odds of seropositivity (aOR = 6.70, 95% CI: 1.75-25.62, p = 0.005), whereas those from Greater Accra had decreased odds (aOR = 0.28, 95% CI: 0.09-0.90, p = 0.033). The findings suggest localized MPXV circulation or repeated zoonotic spillover may have occurred undetected, challenging the prevailing assumption that Ghana was unaffected by human mpox prior to 2022, underscoring the importance of strengthened surveillance and preparedness in Ghana.
## 1. Introduction
Mpox, caused by the Monkeypox virus (MPXV), is an emerging zoonotic disease of increasing global health concern. MPXV is an enveloped double-stranded DNA virus of the Orthopoxvirus genus [1]. Two distinct clades of Monkeypox virus (MPXV) are recognized: Clade I, historically circulating in Central Africa, and Clade II, originating in West Africa, with further subdivisions into lineages Ia/Ib and IIa/IIb, respectively [2,3]. While outbreaks were once sporadic and largely linked to zoonotic spillover, the past two decades have seen a steady rise in incidence and geographic expansion [4].
In May 2022, Clade IIb MPXV triggered an unprecedented multinational outbreak that rapidly spread across non-endemic regions, prompting the World Health Organization (WHO) to declare mpox a Public Health Emergency of International Concern (PHEIC) [5]. More recently, Clade Ib has fueled a resurgence of cases in the Democratic Republic of the Congo (DRC) and neighboring countries, including Uganda, Burundi, Rwanda, Kenya and Zambia [6]. with evidence of exportation beyond Africa-Sweden, India, Germany and the United States of America [7,8]. As of 26 January 2025, 21 countries have reported 19,837 confirmed cases, including 70 deaths [9]. Case Fatality Ratios (CFR) have been reported to differ substantially between clades, ranging from <1% in many clade II outbreaks to >10% in Clade I outbreaks, particularly in Central Africa [10][11][12].
Ghana represents a paradox in the epidemiology of mpox. Although the country had no confirmed human cases before 2022, it played a central role in the 2003 U.S. outbreak when rodents exported from Ghana seeded transmission in prairie dogs [13,14].
In 2022, Ghana reported its first human cases of mpox, interestingly in a male traveler returning from the United States of America with mild symptoms [15,16]. By the last quarter of the year, Ghana Health Service recorded a cumulative total of 87 suspected cases, of which 76 were laboratory-confirmed, with four associated deaths [15,17]. This has raised urgent questions about whether undetected transmission might have occurred locally before the recognition of Clade IIb spread. Given Ghana's ecological diversity, bushmeat consumption, and wildlife trade links, silent MPXV circulation cannot be excluded.
To address this gap, we conducted a retrospective serological analysis of archived sera collected across Ghana prior to the 2022 outbreak to investigate evidence of undetected MPXV exposure. In addition, we analyzed clinical samples obtained during the 2022 outbreak to provide a contemporary comparator. Together, these datasets allowed us to evaluate both pre-outbreak exposure and outbreak-associated serological responses, thereby offering new insights into the hidden epidemiology of mpox in Ghana and informing regional preparedness in West Africa.
## 2. Methods
## 2.1. Study Design
This study employed a retrospective cross-sectional study approach using samples collected during a SARS-CoV-2 surveillance study conducted across 16 regions of Ghana. A total of 5939 serum samples were collected between February and December 2021 for this study [18]. Samples for this study were selected based on reported suspected and confirmed cases during the outbreak, availability of sufficient sample volume and completeness of demographic metadata. Clinical samples from suspected cases during the 2022 national mpox outbreak were screened for MPXV antibodies. All sample-related data were deidentified to ensure confidentiality.
## 2.2. Study Population and Sample Size
This study analyzed a total of 457 serum samples representing 14 regions of Ghana. A subset of 365 archived sera collected during SARS-CoV-2 surveillance activities in 2021 [18] and 92 clinical sera obtained from individuals during the 2022 national outbreak [19]; however, data on clinical symptoms associated with the clinical samples collected during the 2022 national outbreak were unavailable. Samples were collected by venipuncture into Serum Separating Tubes (SSTs) by trained healthcare personnel following standard biosafety procedures. Immediately after collection, whole blood was processed to obtain serum, which was aliquoted into cryovials, stored at -20 • C at sentinel sites, and later transported on a cold chain to the Noguchi Memorial Institute for Medical Research (NMIMR), where they were preserved at -80 • C until analysis. To ensure broad geographic representation, archived samples were selected across all regions with emphasis on highly populated areas and regions with suspected or confirmed mpox cases, particularly Greater Accra and Ashanti, which are Ghana's largest urban centers and major hubs of domestic and international travel (Figure 1).
## 2.3. Sample Processing
All samples were processed and analyzed at the Noguchi Memorial Institute for Medical Research and the Rocky Mountain Laboratories, NIAID, USA.
## 2.4. Detection of IgG Antibodies Against MPVX
All sera were screened for anti-MPXV IgG antibodies using a commercial ELISA kit (RayBiotech, Peachtree Corners, Peachtree Corners, GA, USA) targeting the MPXV E8L protein (RayBiotech, Peachtree Corners, GA, USA), following the manufacturer's instructions. Serum samples were diluted 1:200 using the assay diluent provided in the RayBiotech ELISA kit (Catalog No. ELV-MPXVE8L), in accordance with the manufacturer's instructions for serum/plasma samples. The diluted samples were then added to the pre-coated 96-well microplate and processed following the recommended protocol. Each 96-well plate included negative control sera from unexposed individuals (individuals with no known history of exposure to mpox or vaccination). The cut-off for seropositivity was defined as the mean optical density of the negative controls plus ten (10×) standard deviations. This ELISA has a sensitivity of 0.11 pg/mL and is designed with high specificity as antibody pair detects Monkeypox Virus (MPXV) Envelope protein E8L with no reported cross-reactivity to other orthopoxviruses [20].
## 2.5. Data Curation and Analysis
Sample metadata, including GPS location, demographic variables from surveillance records and clinical status, were collated using Microsoft excel. All records were reviewed for completeness, standardized, and de-identified prior to analysis to ensure confidentiality. Smallpox vaccination status was defined based on the existing criteria [21]. Briefly, individuals born before the official smallpox eradication declaration in 1980 (≥45 years) were classified as vaccinated. Those born between 1975 and 1985 (40-50 years) were considered to have an uncertain vaccination status, while individuals born after 1985 (≤39 years) were classified as unvaccinated. The finalized dataset was imported into Stata version 17 MP4 (StataCorp, College Station, TX, USA) for statistical analyses. Seroprevalence was estimated with corresponding 95% confidence intervals (CIs). Associations between MPXV seropositivity and demographic variables (age group, sex), geographic region, and smallpox vaccination status were assessed using multivariable logistic regression. Results were expressed as odds ratios (ORs) with 95% CIs, and statistical significance was defined as p < 0.05.
## 3. Results
A total of 457 serum samples were analyzed, comprising both archived sera and clinical specimens collected during the 2022 outbreak. Just under two-thirds of participants resided in urban areas (62.6%), while 17.3% were from rural settings. The age distribution was skewed toward younger individuals: more than half were ≤25 years, with nearly one quarter under 16 years. Females constituted a slight majority (55.4%). Regional representation was greatest from Greater Accra (30.0%) and Ashanti (26.0%), reflecting their high population density and role as major travel hubs. Vaccination status, inferred from birth year, indicated that two-thirds of participants were classified as unvaccinated against smallpox (67.8%), while only one in five were considered vaccinated (22.1%) (Table 1).
## Factors Associated with MPXV Seropositivity in the Tested Samples
In multivariable models, neither sex, age, nor smallpox vaccination status was significantly associated with MPXV seropositivity. Geographic factors, however, showed strong associations. Participants from the Western Region had significantly higher odds of seropositivity (aOR = 6.70, 95% CI: 1.75-25.62; p = 0.005), whereas those from Greater Accra had significantly lower odds (aOR = 0.28, 95% CI: 0.09-0.90; p = 0.033). Strikingly, clinical samples with undisclosed location data also showed elevated odds of seropositivity (aOR = 5.65, 95% CI: 1.52-20.94; p = 0.010) (Table 2).
## 4. Discussion
This retrospective serological study suggests nationwide evidence of prior mpox virus (MPXV) exposure in Ghana before the detection of Clade IIb in humans in 2022. We observed an overall seroprevalence of 6.6%, with signals of elevated exposure in the Western Region and among younger, largely unvaccinated individuals. These findings suggest that localized MPXV circulation or repeated zoonotic spillover may have occurred undetected, challenging the prevailing assumption that Ghana was unaffected by human mpox prior to 2022 [12]. Moreover, this interpretation lends further support to earlier evidence implicating Ghana as the source of the reservoirs linked to the 2003 outbreak in the United States [14].
The geographic divergency warrants particular attention as well as careful interpretation. While samples from the Western Region showed significantly increased odds of seropositivity compared to the Ashanti Region, Greater Accra showed reduced odds of seropositivity; these patterns may reflect both genuine epidemiological differences and/or potential biases arising from the sample constitution. Ecological interfaces, wildlife trade, and bushmeat consumption practices in the Western Region plausibly elevate zoonotic spillover, all of which modulate human mpox exposure risk [22,23], but the uneven distribution of archived samples across regions, coupled with purposive inclusion of areas considered high-risk, may have contributed to the apparent differences; hence, it should be regarded as exploratory and interpreted within the context of the study's sampling limitations. Nonetheless, these findings align with ecological hypotheses of localized exposure and are consistent with current epidemiological patterns, wherein 71% of recent mpox cases were reported from the Western Region [24,25].
The Western Region is rich in tropical forest ecosystems with dense rodent and primate populations, potential reservoirs for orthopoxvirus [26,27]. On the other hand, Greater Accra, while a hub for international travel, has lower direct zoonotic exposure. Such patterns suggest the importance of integrating ecological and behavioral factors into mpox risk assessments across West Africa [28].
Our findings revealed that seropositivity was highest among individuals under 35 years, the cohort not covered by routine smallpox vaccination. This aligns with the broader African and global experience, where waning Orthopoxvirus immunity after the cessation of smallpox vaccination has widened the pool of susceptible hosts [29,30]. Although our analysis did not demonstrate a statistically significant protective effect of inferred smallpox vaccination status, likely due to limited sample size, the pattern is consistent with reports of increased mpox burden in younger, unvaccinated populations during the 2022 multinational outbreak [31].
Our findings also intersect with Ghana's paradoxical role in mpox epidemiology. Despite the absence of confirmed human cases until 2022, Ghana was central to the 2003 U.S. outbreak via exported rodents [14]. The present serological evidence suggests that humans within Ghana may have experienced unrecognized exposures long before international recognition, highlighting the limitations of passive surveillance systems in detecting lowlevel or atypical infections. This echoes broader concerns that mpox epidemiology in Africa is incompletely understood, with the true burden underestimated by reliance on syndromic case detection, i.e., identifying cases based solely on clinical symptoms [10,32].
Several limitations merit consideration. First, serological assays for MPXV may crossreact with other orthopoxviruses, potentially inflating prevalence estimates. However, the use of the E8L-targeting ELISA provides enhanced specificity. Second, the modest sample size, particularly within certain regions, limited statistical power to detect finerscale associations. Third, the lack of confirmatory assays like seroneutralization, PCR and/or viral sequencing constrains definitive attribution of exposure to MPXV versus related viruses. Finally, the absence of detailed information on the stage of illness at sample collection may have influenced antibody detection and interpretation of serological patterns. Nonetheless, the consistency of our findings with known ecological and epidemiological patterns lends credibility to the observed patterns. Moreover, future studies should address these limitations by integrating molecular and serological approaches, including IgM/IgG detection, PCR and neutralization assays; expanding sample size and geographic coverage; and collecting detailed clinical, ecological, behavioral and exposure data to enable more comprehensive interpretation of mpox transmission dynamics.
In conclusion, this study suggests that Ghanaians have likely been exposed to MPXV prior to the recognition of Clade IIb transmission, with implications for surveillance, preparedness, and risk communication. Strengthened One Health-based surveillance that integrates human, animal, and ecological data is essential to detect early zoonotic transmission events. Regional and global health security strategies must also recognize that the epidemiology of mpox in West Africa extends beyond recognized outbreaks. Future research should prioritize longitudinal serological studies, reservoir host investigations, and genomic surveillance to better define the landscape of MPXV circulation and mitigate the risk of future widespread outbreaks.
## References
1. Arita, Henderson (1968) "Smallpox and monkeypox in non-human primates" *Bull. World Health Organ*
2. Ulaeto, Agafonov, Burchfield et al. (2023) "New nomenclature for mpox (monkeypox) and monkeypox virus clades" *Lancet Infect. Dis*
3. Wawina-Bokalanga, Akil-Bandali, Kinganda-Lusamaki et al. (2024) "Co-circulation of monkeypox virus subclades Ia and Ib in Kinshasa Province, Democratic Republic of the Congo"
4. Olawade, Wada, Fidelis et al. (2024) "Strengthening Africa's response to Mpox (monkeypox): Insights from historical outbreaks and the present global spread" *Sci. One Health*
5. Miura, Van Ewijk, Backer et al. (2022) "Estimated incubation period for monkeypox cases confirmed in the Netherlands"
6. Branda, Ceccarelli, Maruotti et al. (2024) "Global spread of mpox Clade I: Implications for travel and public health" *Travel Med. Infect. Dis*
7. Lancet, Mpox "The need for a coordinated international response" *Lancet*
8. (2022) *Mpox (Monkeypox) Outbreak: Global Trends. Available online*
9. Beer, Bhargavi Rao (2019) "A systematic review of the epidemiology of human monkeypox outbreaks and implications for outbreak strategy" *PLoS Neglected Trop. Dis*
10. Bunge, Hoet, Chen et al. "The changing epidemiology of human monkeypox-A potential threat? A systematic review"
11. Africa (2024)
12. Ligon (2004) "Monkeypox: A review of the history and emergence in the Western hemisphere" *Semin. Pediatr. Infect. Dis*
13. Reed, Melski, Graham et al. (2004) "The Detection of Monkeypox in Humans in the Western Hemisphere" *N. Engl. J. Med*
14. Adadi, Mensah, Abdul-Razak (2023) "The outbreak of monkeypox (MPX) in Ghana" *J. Med. Virol*
15. (2025) "Greater Accra Has More Than 50% of Monkeypox Cases-Citinewsroom-Comprehensive News in Ghana"
16. Mpox (monkeypox)-Africa, Available Online (2025)
17. Owusu Donkor, Mensah, Dwomoh et al. (2023) "Modeling SARS-CoV-2 antibody seroprevalence and its determinants in Ghana: A nationally representative cross-sectional survey" *PLoS Glob. Public Health*
18. (2025) "RayBio ® Monkeypox Virus E8L Protein Human IgG ELISA Kit RayBio® Monkeypox Virus E8L Protein Human IgG ELISA Kit. Available online: www.RayBiotech.com"
19. Leendertz, Stern, Theophil et al. (2017) "A cross-sectional serosurvey of anti-orthopoxvirus antibodies in central and Western Africa" *Viruses*
20. Sklenovská, Van Ranst (2018) "Emergence of Monkeypox as the Most Important Orthopoxvirus Infection in Humans" *Front. Public Health*
21. Durski, Mccollum, Nakazawa et al. (1970) *Morb. Mortal. Wkly. Rep*
22. (2025) "Mpox Crisis: 71% of Cases Concentrated in Western Region-MyJoyOnline"
23. (2025) "Mpox Outbreak: Western Regional Health Directorate Springs Into Action" *Available online*
24. Curaudeau, Besombes, Nakouné et al. (2023) "Identifying the Most Probable Mammal Reservoir Hosts for Monkeypox Virus Based on Ecological Niche Comparisons" *Viruses*
25. Rimoin, Mulembakani, Johnston et al. (2010) "Major increase in human monkeypox incidence 30 years after smallpox vaccination campaigns cease in the Democratic Republic of Congo" *Proc. Natl. Acad. Sci*
26. Yinka-Ogunleye, Aruna, Dalhat et al. (2019) "Outbreak of human monkeypox in Nigeria in 2017-18: A clinical and epidemiological report" *Lancet Infect. Dis*
27. Karem, Reynolds, Hughes et al. (2007) "Monkeypox-induced immunity and failure of childhood smallpox vaccination to provide complete protection" *Clin. Vaccine Immunol*
28. Reynolds, Damon (2012) "Outbreaks of human monkeypox after cessation of smallpox vaccination" *Trends Microbiol*
29. Thornhill, Barkati, Walmsley et al. (2022) "Monkeypox Virus Infection in Humans across 16 Countries"
30. Adetifa, Muyembe, Bausch et al. (2023) "Mpox neglect and the smallpox niche: A problem for Africa, a problem for the world" *Lancet*
31. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645968&blobtype=pdf | # Nephropathogenic infectious bronchitis virus induces epithelialmesenchymal transition of renal tubular epithelial cells through the TGF-β/p-P38 pathway causing uric acid excretion disorder in chickens
Yunfeng Chen, Yan Shi, Cheng Huang, Haoyu Huang, Yizhou Zeng, Gaofeng Cai, Zhanhong Zheng, Ping Liu, Xiaona Gao, Xiaoquan Guo
## Abstract
Nephropathogenic infectious bronchitis virus (NIBV) infection usually causes kidney enlargement and urate deposition in chickens, leading to sudden death. However, the mechanism by which NIBV causes urate deposition in the kidneys has not yet been elucidated. The coordinated operation of uric acid transporters is crucial for the kidneys to maintain uric acid homeostasis, and existing studies have shown that the occurrence of cell epithelial-mesenchymal transition (EMT) can affect the expression of uric acid transporters. Thus, this study aimed to explore the effect of NIBV on urate transporters and elucidate the mechanism of EMT in NIBV-induced urate deposition in chicken kidneys in vivo and in vitro. The results revealed that NIBV infection led to an abnormal increase in uric acid levels in chicks, affected the expression of uric acid transport proteins, induced EMT in renal tubular epithelial cells, and activated the TGF-β/p-p38 pathway. The changes in uric acid concentration induced by NIBV were related to the uric acid excretion protein ABCG2, whose expression is negatively regulated by EMT. The occurrence of NIBV-induced EMT coincided with the time point at which NIBV activated the TGF-β/p-p38 pathway. After siRNA knockdown of p38 MAPK, EMT did not occur, and ABCG2 expression returned to normal. In summary, the mechanism by which NIBV causes abnormal elevation of uric acid levels in chickens is through the induction of EMT in renal tubular epithelial cells via the TGF-β/p-p38 pathway, which strongly inhibits the expression of ABCG2, thereby causing uric acid excretion disorders in chickens.IMPORTANCE NIBV infection results in a reduction in uric acid transporter expression in the kidneys of chickens. ABCG2 plays a pivotal role in the excretion of uric acid in chickens. The mechanism by which NIBV causes an abnormal increase in uric acid levels in chickens involves the induction of renal tubular epithelial cell EMT through the TGF-β/P-p38 pathway and the subsequent strong inhibition of ABCG2 expression, causing uric acid excretion disorders in chickens.
NIBV occurs mainly in Asian and Middle Eastern countries and poses a major threat to the chicken industry (7). Our previous research demonstrated that NIBV infection results in kidney enlargement, urate deposition, and the occurrence of white loose stools in poultry. These conditions not only increase mortality rates among affected poultry but also impose significant economic burdens on the poultry breeding sector (8,9). Urate deposition is typically linked to elevated levels of uric acid (UA), as prolonged high concentrations of uric acid lead to the precipitation of urate in the bloodstream and its subsequent accumulation in the internal organs of poultry (8). Therefore, investigating the mechanisms through which NIBV induces an abnormal increase in UA levels in chicks is highly important.
UA is generated through purine metabolism. Under normal conditions, the produc tion and excretion of uric acid in poultry maintain a dynamic equilibrium (9). When the body consumes an excessive amount of purine-rich foods or when the kidneys encounter disorders in uric acid excretion, this balance is disrupted (8). The kidney accounts for up to two-thirds of the excretion of urate and constitutes an essential organ for regulating uric acid homeostasis (10). The regulation of uric acid homeostasis by the kidney depends on uric acid transporters located in renal tubular epithelial cells. Uric acid transporters are categorized into uric acid reabsorption proteins and uric acid excretion proteins, which play roles in reabsorbing and secreting uric acid, respectively (11,12). Currently, uric acid transporters have been extensively investigated; renal uric acid transporters mainly belong to the organic anion transporter (OAT) family, the glucose transporter glucose transporter (Slc2a9) family, the ATP-binding cassette transporter superfamily (ABCs), the sodium phosphate cotransporter (NTP) family, and others (13). ABCG2 is an ATP-dependent urate export pump that plays a crucial role in the transport of uric acid in cells, significantly contributing to uric acid excretion within the body (14). However, whether uric acid transporters are involved in NIBV-induced gout in chickens has not yet been reported.
Epithelial-mesenchymal transition (EMT) is a biological process in which epithelial cells transform into cells with a mesenchymal phenotype, and this process constitutes the early basis of fibrosis (15). During this process, alterations can occur in both the phenotypes of epithelial cells and gene expression profiles, manifested by diminished cell adhesion, enhanced migratory ability, and changes in cell morphology (16). During EMT, the expression of the epithelial cell marker E-cadherin is downregulated (17), and the expression of the fibroblast marker fibronectin (FN) is significantly decreased (18). Although multiple signals can regulate EMT, transforming growth factor-β (TGF-β) typically plays a predominant role as an inducer of EMT (19)(20)(21). Additionally, studies have demonstrated that TGF-β-induced EMT potently inhibits ABCG2 expression and that the removal of TGF-β can restore the cell phenotype to that of an epithelial type and reinstate ABCG2 expression (22). Hence, TGF-β-induced EMT is associated with the uric acid excretion-related protein ABCG2. Furthermore, studies have confirmed that the p38 MAPK signaling pathway is closely linked to MET (23,24). p38 MAPK serves as a downstream gene of TGF-β and is capable of being activated by the latter. The regulation of signal transduction and activation within this pathway is governed by the expression level of TGF-β. When TGF-β binds to receptors on cells, it can phosphorylate MKK4, and P-MKK4 can phosphorylate p38 to p-p38, which subsequently enhances the promoters of genes related to EMT (15,25).
Research has indicated that coronaviruses are closely associated with EMT and that infection by coronaviruses can lead to an increase in fibrin levels within the organism, resulting in tissue fibrosis (26,27). For example, the new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus that emerged in 2019 can induce pulmonary fibrosis (PF) in infected individuals, and its mechanism includes promoting the expres sion of TGF-β (28). In addition, Reyhaneh Niayesh-Mehr et al. elucidated EMT-related processes related to the origin of PF in COVID-19 through the latest investigation of cellular and molecular mechanisms (29). Coincidentally, NIBV is classified into the family Coronaviridae and the genus Gamma coronavirus. Therefore, we hypothesized that EMT may occur in the kidney tissue of chicks infected with NIBV.
This study aims to elucidate the mechanism through which NIBV induces abnormal elevations in serum uric acid levels in the body by exploring the impact and mechanism of NIBV infection on kidney uric acid transporters in chicks. Our research elucidated the mechanism through which NIBV infection leads to urate deposition in chicks, thereby providing a theoretical foundation for the prevention and treatment of avian gout.
## MATERIALS AND METHODS
## Animal feeding and sample collection
A total of 300 one-day-old healthy Hy-Line brown chickens were raised in the laboratory of the College of Animal Science and Technology at Jiangxi Agricultural University with free access to drinking water and food. At 28 days of age, the chicks were randomly divided into a control group (Con) consisting of 100 chicks and a model group (NIBV) of 200 chicks. The model group was inoculated with the SX9 virus (10^5 ELD 50 /0.2 mL), while the control group was inoculated with physiological saline. To prevent cross-infec tion, contact between the two groups of chickens was avoided. Kidney tissues were collected at 1, 3, 5, 7, 9, 11, 13, 15, 18, 21, and 28 days post-infection (dpi) and stored at -80°C until they were used in subsequent experiments. The isolated IBV SX9 strain was deposited at the College of Animal Science and Technology of Jiangxi Agricultural University, and its specific sequence was identified as MN707951.1 in the NCBI database.
## Chicken renal tubular epithelial cell isolation, culture, and treatment
In accordance with previous methods (30), primary renal tubular epithelial cells were extracted from 1-to 7-day-old chicks for in vitro experiments. The isolated chicken primary renal tubular epithelial cell suspension was cultured in low-sugar DMEM (Solarbio, Beijing, China) supplemented with 10% FBS (Excel, Shanghai, China), adjusted to a seeding density of 1 × 10^6 cells/mL, and cultured in a cell incubator at 37°C with 5% CO 2 . After 24 h of incubation, the original culture medium was discarded, and the cells were washed once with PBS (Sevier, Wuhan, China). The cells were subsequently cultured in DMEM supplemented with 5% FBS until they reached confluence (70-80%) within the culture dish.
When the cell density reached approximately 70-80%, the cells were processed. The control group (Con, C) was treated with pure DMEM (Solarbio, Beijing, China) for 2 h, and the NIBV infection group (NIBV, N) was infected with 1 MOL of virus based on TCID50. During this two-hour treatment period, the cell culture plate was gently shaken every half hour. Afterward, the medium was switched to DMEM containing 2% FBS and 1% double antibiotics for continued culture.
The specific small interfering RNA (siRNA) targeting p38 (sip38) used in this study was purchased from HANBIO (Shanghai, China). In accordance with the manufactur er's instructions, siRNA was transfected using RNA Fit Transfection Reagent (HANBIO, HH20250220WY-SI02) at a final concentration of 20 nM. CY3 siRNA NC (sense strand: U UCUCCGAACGUGUCACGUTT; antisense strand: ACGUGACACGUUCGGAGAATT); siRNA NC (sense strand: UUCUCCGAACGUGUCACGUTT; antisense strand: ACGUGACACGUUCGGAG AATT); Si-p38MAPK3 (sense strand: CGAUGAAGUAAUCAGCUUUGUTT; antisense strand: A CAAAGCUGAUUACUUCAUCGTT).
## Cell viability assay
Cell viability was determined using CCK-8 detection reagent according to the manufac turer's instructions (YEASEN, Shanghai, China). Renal tubular epithelial cells were spread in a 96-well plate. After adhesion and growth, the cells were cultured with different concentrations of inhibitors (0, 2, 4, 6, 8, 10, 12, and 14 µg/mL) for 24 h. After the cells were washed once with PBS, 110 µL of CCK-8 reagent was added to each well, and the cells were incubated at 37°C for 2 h in the dark. Afterward, the absorbance value (OD) of each well was measured at 450 nm using a microplate reader, and the cell viability was calculated as the OD. Each concentration was established in eight replicate wells.
## Measurement of serum uric acid content
The collected serum was used to measure the UA concentration in the serum using a fully automatic biochemical detection analyzer (Hitachi Group, 3100, Japan).
## Total RNA extraction and real-time RT-PCR
RNA was extracted from kidney tissue and cell samples using the TRIzol method (Vazyme, Nanjing, China). The RNA concentration was measured and adjusted to 1000 ng before it was reverse transcribed into cDNA. The reverse transcription system and PCR amplification were performed following the instructions of the reverse transcription and qPCR kit, and the amplification procedure was 94 predenaturation at 94°C for 5 min and denaturation at 94°C for 50 s, annealing at 56°C for 30 s, and extension at 72°C for 1 min for 38 cycles. The fluorescence quantitative PCR system was prepared according to the instructions of the fluorescence quantitative kit (TransGen, Beijing, China). The primers were designed on NCBI and synthesized by Shanghai Qingke Biotechnology Co., Ltd. The gene sequences are shown in Table 1.
## Protein extraction and western blot
Kidney tissue and cell samples were digested and lysed using a preprepared lysis solution of RIPA: phosphatase inhibitor: protease inhibitor = 100:1:1 (Solarbio, Beijing, China). After lysis, the supernatant was centrifuged, and the concentration of the supernatant was determined with a BCA kit (Solarbio, Beijing, China), which was set to 5 mg/mL (tissue) or 1.5 mg/mL (cell). After loading buffer was added to the sample, it was placed in a water bath at 100°C for 10 min to denature the protein. In this way, the protein sample was prepared before storage at -80°C for later use.
The WB process involved electrophoresis at 80-120 V, wet transfer at 300 mA, and blocking with Fast Blocking Solution (Epizyme, Shanghai, China) for 20 min. The membranes were washed with PBST three times for 10 min, incubated with primary antibody at 4°C overnight, and washed again with PBST three times for 10 min. The sections were incubated with a goat secondary antibody for 40 min and finally developed using ECL supersensitive luminescent solution (Abbkine, Wuhan, China). The protein bands were quantified using Image J, and the resulting protein bands were recorded and saved for statistical analysis. The primary antibodies used for WB were ABCG2, FN, MKK4, P-MKK4, TGF-β (Wanleibio, Shenyang, China), ABCC4 (ABclonal, Wuhan, China), SLC2A9, E-cadherin, p-p38, p38 (Abmart, Shanghai, China), and the secondary antibody (Proteintech Group, Wuhan, China).
## Determination of viral load
The pEASY-T3-N-positive plasmid was constructed according to the laboratory's previous method (31), the number of virus copies in the positive plasmid was calculated, and a standard curve was established by equal dilution. The absolute fluorescence quantitative PCR method was used to measure the viral load in the kidneys of the chicks at each time point.
## Immunofluorescence staining
During sampling at 1, 5, 11, and 18 dpi, portions of the kidney tissue were excised and preserved in 4% paraformaldehyde for subsequent embedding and the preparation of wax blocks. The cell slides were fixed with 4% paraformaldehyde following NIBV treatment for 6, 12, 18, and 24 h. After dewaxing, antigen retrieval, and serum blocking, the wax blocks were incubated overnight at 4°C with the appropriate primary antibody. The corresponding secondary antibody was subsequently added, and the samples were incubated at 37°C for 1 h. Finally, after the nuclei were stained with DAPI, anti-fade mounting medium was added to the tissue, and a coverslip was placed on top. The prepared fluorescent slides were stored at 4°C in the dark for later observation via fluorescence microscopy.
## Cell scratch test
When the cell density reached 70% to 80%, viral infection treatment was performed, and uniform scratches were made on the bottom of the culture dish. Observation and imaging under the microscope were carried out at 6, 12, 18, and 24 h post-scratching, with three replicate wells for each group, and the experiment was repeated three times.
## Statistics
In this experiment, Image J was used to conduct grayscale analysis of protein bands and quantify the scratch area. All the data were statistically analyzed using GraphPad Prism 9.0 software. The independent sample t-test was used to compare the differences between the two groups. One-way analysis of variance (ANOVA) was used to compare the differences between multiple groups. The data are displayed as the means ± SD. The data were graphed using GraphPad Prism 9.0 software. The threshold for statistical significance was P < 0.05 (*), P < 0.01 (**), and P < 0.001 (***); if P < 0.05, the difference was considered significant.
## RESULTS
## Determination of pathological modeling results and time period in vivo
Analysis of serum uric acid concentration revealed that uric acid levels in the model group (NIBV) began to increase at 5 dpi compared with control group (Con). This increase reached its peak at 11 dpi, followed by a gradual decline, and eventually returned to baseline levels by 18 dpi (Fig. 1A). The uric acid concentration distribution diagram revealed that the UA concentration in the Con group was less than 400 µmol/L at all time points, whereas at 9, 11, and 13 dpi, 50% of the chicks in the NIBV group had a UA concentration exceeding 400 µmol/L, and at 11 dpi, it was even greater than 2000 µmol/L (Fig. 1B). Concurrently, the renal viral load also reached its maximum at 11 dpi before gradually decreasing thereafter (Fig. 1C). Renal anatomy at 11 dpi revealed that the kidneys in the NIBV group were significantly enlarged and had urate deposition (Fig. 1D).
## Expression results of genes and proteins related to uric acid transport
Based on the previously measured viral load and serum uric acid concentration results, we selected four time periods-1, 5, 11, and 18 dpi-for subsequent experiments. To explore the influence of NIBV infection on kidney uric acid transport-related proteins in chicks, we detected the molecular expression of the reabsorption-related protein Slc2a9 and the uric acid excretion proteins ABCG2 and ABCC4 through qPCR and WB (Fig. 2A through F). Compared with that in the Con group, the expression of these transporters in the NIBV group decreased significantly at 11 dpi. Interestingly, at 18 dpi, compared with that in the Con group, the expression of ABCC4 in the NIBV group still tended to significantly decrease, whereas the expression of ABCG2 tended to significantly increase.
## Localization and expression of uric acid excreting protein in the kidney
To elucidate why the expression trends of ABCG2 and ABCC4 in the NIBV group were completely opposite at 18 dpi, we selected the kidney tissue of healthy chicks for immunofluorescence costaining of ABCC4 and ABCG2. The results indicated that ABCG2 was expressed mainly in the medullary area of chicken kidneys, whereas ABCC4 was expressed in the cortical area. Additionally, ABCG2 (green fluorescence) was expressed mainly in the kidney on the luminal side of the tubule, and ABCC4 (red fluorescence) was expressed mainly in the basement membrane and apical mold of renal tubular epithelial cells (Fig. 3). In general, those expressed on the basement membrane are mainly responsible for the reabsorption of uric acid, whereas those expressed on the apical membrane are responsible for the excretion of uric acid (32,33). Combined with the previous data, the uric acid level returned to normal at 18 dpi, and simultaneously, the expression of ABCG2 also returned to normal, while the expression of ABCC4 was irreversibly impaired. We speculate that ABCG2 is a major factor in the excretion of uric acid in poultry.
## Expression of genes related to mesenchymal transformation and pathwayrelated genes and proteins
Studies have demonstrated that the expression of ABCG2 is inhibited by TGF-β-induced EMT (22). To explore whether ABCG2, the main contributor to uric acid excretion, is affected by EMT, we detected the expression of E-cadherin and FN. The results indicated that compared with that in the Con group, the expression of FN significantly increased and that of E-cadherin significantly decreased in the NIBV group at 11 dpi (Fig. 4A through E), suggesting that EMT occurred in the kidney at this time. During the period when EMT occurs, the expression of ABCG2 is strongly inhibited. Moreover, studies have shown that p38-MAPK signaling is among the pathways that mediate EM (23,24). We explored whether NIBV-induced EMT is mediated through the p38-MAPK pathway. Compared with those in the control group, the levels of TGF-β in the NIBV group significantly increased at 11 dpi (Fig. 4G), and the phosphorylation levels of MKK4 and p38 MAPK significantly increased (Fig. 4H andI), indicating that the TGF-β/p-p38 pathway was significantly activated at 11 dpi and that its activation time coincided with the occurrence of EMT.
## Transcriptional level of extracellular uric acid transport proteins and the impact of NIBV infection on renal tubular epithelial cells
To validate the in vivo results, we established an in vitro model of renal tubular epithelial cells. Based on previous studies, 6 h after the cells were subjected to NIBV treatment, the virus started to enter the cells (34). Hence, cell samples were collected at 6, 12, 18, 24, 30, 36, 48, and 60 h, and qPCR was performed to measure the transcription levels of uric acid transport-related proteins. Compared with that in the control group, the expression of uric acid excretion proteins in the NIBV group significantly increased at 12 h but then significantly decreased after 24 h (Fig. 5A andB). Uric acid reabsorption significantly decreased compared with that in the control group after 12 h (Fig. 5C). Based on these outcomes, we selected four time periods-6, 12, 18, and 24 h-for subsequent experi ments.
Additionally, by detecting the NIBV-N protein, we demonstrated that NIBV success fully infected primary chicken renal tubular epithelial cells. Immunofluorescence staining revealed that the Con group had almost no red fluorescent spots, whereas the NIBV group exhibited numerous red fluorescent spots (Fig. 6A). WB results indicated that the NIBV-N protein could be detected in renal tubular epithelial cells at 6, 12, 18, and 24 h post-infection (Fig. 6B). These findings suggest that NIBV caused infection in primary chicken renal tubular epithelial cells.
## Expression of uric acid excretion protein in vitro
To validate the gene expression levels, we measured the protein levels of uric acid transporters during four time periods. The results demonstrated that the protein expression was in line with the gene expression (Fig. 7A through D). Compared with that in the control group, the fluorescence intensity in the NIBV group increased at 12 h but decreased at 24 h (Fig. 7E).
## Mesenchymal transformation-and pathway-related gene and protein expression results in vitro
Similarly, we detected the expression levels of EMT and TGF-β/p-38 pathway-related proteins in vitro. Compared with that in the control group, the expression of the renal tubular epithelial cell marker E-cadherin significantly decreased at 18 and 24 h, and the expression of the fibroblast marker FN significantly increased in the NIBV group (Fig. 8A through E). Additionally, the results of the scratch assay revealed that the wound healing rate in the NIBV group was faster than that in the control group, especially in the NIBV group, where the scratches were almost completely covered by cells at 24 h, indicating enhanced cell migration ability after NIBV treatment (Fig. 8F). These results suggest that EMT occurred in cells after 18 h of NIBV treatment in vitro. The qPCR and WB results demonstrated that, compared with those in the control group, the levels of TGF-β in the NIBV group significantly increased at 18 and 24 h, and the phosphorylation levels of MKK4 and p38 MAPK significantly increased (Fig. 8G through J), indicating that the TGFβ/p-38 pathway was activated at 18 and 24 h and that its activation time coincided with the occurrence time node of EMT. These findings are consistent with the in vivo results, suggesting that the p38 MAPK Pathway is likely involved in the initiation and progression of EMT.
## Effect of p38 MAPK knockdown on renal tubular epithelial cells
To confirm that the occurrence of EMT is induced via the p38 MAPK pathway, we knocked it down using siRNA. Cy3 fluorescence labeling revealed that the plasmids were successfully transfected into the cells (Fig. 9A). We subsequently assessed the knockout efficiency of p38 MAPK (Fig. 9B andC). In addition, compared with that in the Con group, the expression of E-cadherin in the NIBV group significantly decreased, while the expression of FN significantly increased, suggesting that EMT occurred in renal tubular epithelial cells. Compared with that in the NIBV group, the expression of mesenchymal transition proteins in the p38 MAPK knockdown group was reversed, indicating that blocking the p38 MAPK pathway could prevent NIBV-induced EMT in renal tubular epithelial cells (Fig. 10A through E). The expression of uric acid transport-related proteins was subsequently detected in the four groups. Compared with that in the NIBV group, the expression of the uric acid excretion protein ABCG2 in the Si-p38 MAPK group was significantly increased, while the expression of ABCC4 was not reversed, which also verified that ABCG2 was the main contributor to uric acid excretion in the late stage of NIBV infection (Fig. 10F through J). The siRNA knockdown results were consistent with the outcomes of p38 MAPK inhibitor (SB203580) treatment (Fig. S2 andS3). In summary, the results of the siRNA assay indicated that NIBV induced EMT by activating the p38 MAPK pathway, thereby inhibiting the expression of ABCG2.
## DISCUSSION
In poultry infected with NIBV, there is often an abnormal surge in uric acid levels within their bodies, resulting in hyperuricemia. These conditions subsequently cause loose white stools, slowed growth, and even sudden death, causing losses to the poultry breeding industry (35,36). The primary target organ of NIBV infection is the kidney, and the kidney serves as the main pathway through which poultry excrete uric acid (10,30). Hence, we hypothesized that the damage inflicted on the kidney by NIBV inevitably affects the excretion of uric acid. Therefore, our research focuses on the influence of NIBV on uric acid excretion. The kidneys regulate the homeostatic balance of uric acid in the body by relying on the coordinated operation of uric acid transporters located in renal tubular epithelial cells (11). Uric acid is excreted from the kidneys through reabsorption and subsequent excretion. Research by Minghui Wang et al. indicated that ABCG2, ABCC4, and Slc2a9 play crucial roles in uric acid transport in chickens (37)(38)(39). The peak value at 11 dpi in the viral load measurement results is in line with the changing trend of serum uric acid levels in chicks, which is also in accordance with the results presented by previous studies (31,40). The expression of uric acid reabsorption proteins does not increase but rather decreases when uric acid levels are at their highest, revealing that the cause of elevated uric acid is not uric acid reabsorption proteins. This leads us to focus our subsequent research on uric acid excretion proteins. Prestin K et al. reported that transient HNF4α overexpression enhanced the activity of the hepatocyte nuclear factor (HNF) 4α binding site in the Slc2a9 isoform 1 promoter, whereas mutations in the binding site reduced activation (41). These findings indicate that Slc2a9 expression is indeed reduced under certain circumstances. The mechanism by which NIBV causes a reduction in Slc2a9 expression has not been explored in depth here because our focus is not on reabsorption proteins. We will further explore this topic if necessary.
Interestingly, in the later stages of NIBV modeling, the expression of the uric acid excretion-related protein ABCG2 was restored and even significantly increased, indicat ing that the ability of the kidney to excrete uric acid at this time depended mainly on ABCG2. Previous studies on chickens revealed that renal urate transport is compromised in cases of decreased renal function and that extrarenal ABCG2 seems to play a compen satory role (42), and our results align with theirs, suggesting that ABCG2 compensates for reduced ABCC4 expression. The increased compensatory expression is sufficient to restore chicks with abnormally elevated uric acid levels back to normal. However, an intriguing question is why the trends of ABCG2 and ABCC4 in the later stages of NIBV modeling are completely opposite. The expression and localization of both ABCG2 and ABCC4 within chicken kidneys are particularly critical for understanding this phenom enon. Notably, proteins typically located at the apical membrane (luminal side) of the renal tubules serve as excretory proteins that are responsible for transporting substances from renal tubular cells into the tubular lumen. Conversely, transporters located at the basolateral membrane function as reabsorptive proteins, tasked with reclaiming certain substances from the urine back into the circulation (32,33). Our double immunofluores cence staining results indicate that the expression and localization of ABCG2 and ABCC4 are not entirely identical. Additionally, studies by Yuting Wu et al. have shown that impaired uric acid excretion in chickens is associated with ABCG2 (BCRP) (43). We hypothesize that, in the late stage of NIBV infection, ABCG2 may serve as the primary transporter involved in uric acid excretion. Based on this premise, we will guide our subsequent research efforts to focus on ABCG2. Previous studies have indicated that cytokines and growth factors can significantly influence the expression of the ABCG2 gene (22,44). Yin et al. reported a reduction in ABCG2 gene expression when the ABCG2-positive cell population was isolated from MCF-7 cells and subjected to TGF-β treatment. The decrease in ABCG2 expression is strongly suppressed during TGF-β-directed EMT and is restored when cells return to an epithelial phenotype (22). The trends observed in ABCG2 expression were consistent with these findings. Consequently, we subsequently focused on exploring the relation ship between EMT and ABCG2 expression. The in vivo results revealed that the expression of ABCG2 decreased at the time when EMT occurred, and in the later stages of NIBV modeling, ABCG2 expression was also restored when the epithelial phenotype was restored. This reversible repair may benefit from the reversibility of the EMT. Research has consistently demonstrated that the reverse process of EMT, referred to as MET, occurs frequently during developmental processes such as heart development, kidney morpho genesis, and somite formation, as well as in cancer (45,46). Studies have shown that p38-MAPK is one of the pathways that mediates cellular EMT. p38 can be phosphorylated by TGF-β to form active p-p38 (23,25). We previously reported that the suppression of ABCG2 expression is influenced by the EMT induced by TGF-β. Thus, we believe that p38 MAPK is the "connector" between TGF-β and NIBVinduced EMT. To investigate this hypothesis, we detected the gene and protein levels of the p38 MAPK pathway at various time points in vivo and in vitro. The p38 MAPK pathway is indeed activated during EMT. Furthermore, in the stage of epithelial phenotype recovery following NIBV in vivo modeling, the levels of phosphorylated p38 (p-p38) were comparable to those observed in the normal group. Furthermore, to further validate that NIBV-induced EMT in renal tubular epithelial cells is mediated by p38 MAPK, we knocked down p38 MAPK expression using small interfering RNA. After p38 MAPK knockdown, NIBV failed to induce EMT in renal tubular epithelial cells, and the expression of ABCG2 was minimally affected. These findings indicate that NIBV-induced EMT in renal tubular epithelial cells is mediated through the p38-MAPK signaling pathway.
Our research provides a foundation for the potential of interventions to regulate uric acid levels in poultry, with the goal of preventing and alleviating chicken gout associated with NIBV infection. Additionally, this study provides valuable insights for the develop ment of therapeutic drugs targeting hyperuricemia and avian gout in chickens. Uric acid metabolism in poultry and mammals is not identical, and a homology analysis of the ABCG2 gene also revealed significant differences between avian species and humans (Fig. S4). Our research contributes to the understanding of uric acid metabolism mecha nisms in poultry.
## Conclusion
NIBV infection typically causes an abnormal elevation in uric acid levels in chickens. The mechanism by which NIBV causes an abnormal increase in uric acid levels in chickens involves the induction of renal tubular epithelial cell EMT through the TGF-β/P-P38 pathway and the subsequent strong inhibition of ABCG2 expression, causing uric acid excretion disorders in chickens. This study elucidates the mechanism through which NIBV induces an increase in uric acid levels in chicks, laying the foundation for the prevention and alleviation of NIBV infection-induced gout in chickens.
## References
1. Whitton, Bogueva, Marinova et al. (2021) "Are we approach ing peak meat consumption? Analysis of meat consumption from 2000 to 2019 in 35 countries and its relationship to gross domestic product" *Animals (Basel)*
2. Liu, Zhang, Yang et al. (2022) "Chicken intestinal microbiota modulation of resistance to nephropathogenic infectious bronchitis virus infection through IFN-I" *Microbiome*
3. Chen, Shi, Feng et al. (2024) "A novel highly virulent nephropathogenic QX-like infectious bronchitis virus originating from recombination of GI-13 and GI-19 genotype strains in China" *Poult Sci*
4. Sharma, Rahman, Sharma et al. (2024) "Immuno-pathogenesis study of local infectious bronchitis virus G1-1 lineage variant showed altered tissue tropism in experimental broiler chickens" *Vet Res Commun*
5. Matthijs, Van Eck, Landman et al. (2003) "Ability of Massachusetts-type infectious bronchitis virus to increase colibacillosis susceptibility incommercial broilers: a comparison between vaccine and virulent field virus" *Avian Pathol*
6. Zhang, Xu, Cao (2021) "Host antiviral responses against avian infectious bronchitis virus (IBV): focus on innate immunity" *Viruses*
7. Bande, Arshad, Omar et al. (2017) "Global distributions and strain diversity of avian infectious bronchitis virus: a review" *Anim Health Res Rev*
8. Afanador, Roberts (1994) "Effect of nephropathogenic infectious bronchitis viruses on renal function in young male broiler chickens" *Br Poult Sci*
9. Mandal, Mount (2015) "The molecular physiology of uric acid homeostasis" *Annu Rev Physiol*
10. Wu, Xu, Lv (2020) "Goose astrovirus infection affects uric acid production and excretion in goslings" *Poult Sci*
11. Dalbeth, Gosling, Gaffo et al. (2021) *Gout. Lancet*
12. Adomako, Moe (2023) "Uric acid transport, transporters, and their pharmacological targeting" *Acta Physiol (Oxf )*
13. Du, Zong, Li et al. (2024) "Hyperuricemia and its related diseases: mechanisms and advances in therapy" *Signal Transduct Target Ther*
14. Eckenstaler, Benndorf (2021) "The role of ABCG2 in the pathogenesis of primary hyperuricemia and gout-an update" *Int J Mol Sci*
15. Zhang, Zhang, Wang et al. (2023) "What role does PDL1 play in EMT changes in tumors and fibrosis?" *Front Immunol*
16. Nieto, Huang, Jackson et al. (2016) *Cell*
17. Serrano-Gomez, Maziveyi, Alahari (2016) "Regulation of epithelial-mesenchymal transition through epigenetic and post-translational modifications" *Mol Cancer*
18. Nowak, Bednarek (2021) "Aspects of the epigenetic regulation of EMT related to cancer metastasis" *Cells*
19. David, Massagué (2018) "Contextual determinants of TGFβ action in development, immunity and cancer" *Nat Rev Mol Cell Biol*
20. Pickup, Novitskiy, Moses (2013) "The roles of TGFβ in the tumour microenvironment" *Nat Rev Cancer*
21. Lee, Massagué (2022) "TGF-β in developmental and fibrogenic EMTs" *Semin Cancer Biol*
22. Yin, Castagnino, Assoian (2008) "ABCG2 expression and side population abundance regulated by a transforming growth factor betadirected epithelial-mesenchymal transition" *Cancer Res*
23. Munkonda, Akbari, Landry et al. (2018) "Podocytederived microparticles promote proximal tubule fibrotic signaling via p38 MAPK and CD36" *J Extracell Vesicles*
24. Avila-Carrasco, Majano, Sánchez-Toméro et al. (2019) "Natural plants compounds as modulators of epithelial-to-mesenchymal transition" *Front Pharmacol*
25. Hickson, Huo, Griend et al. (2006) "The p38 kinases MKK4 and MKK6 suppress metastatic colonization in human ovarian carcinoma" *Cancer Res*
26. Majumder, Minko (2021) "Targeted nanotherapeutics for respiratory diseases: cancer, fibrosis, and coronavirus" *Adv Ther (Weinh)*
27. Uhler, Shivashankar (2021) "Mechanogenomic coupling of lung tissue stiffness, EMT and coronavirus pathogenicity" *Curr Opin Solid State Mater Sci*
28. George, Wells, Jenkins (2020) "Pulmonary fibrosis and COVID-19: the potential role for antifibrotic therapy" *Lancet Respir Med*
29. Niayesh-Mehr, Kalantar, Bontempi et al. (2024) "The role of epithelialmesenchymal transition in pulmonary fibrosis: lessons from idiopathic pulmonary fibrosis and COVID-19" *Cell Commun Signal*
30. Qi, Li, Ding et al. (2024) "Wogonin inhibits apoptosis and necroptosis induced by nephropathogenic infectious bronchitis virus in chicken renal tubular epithelial cells" *Int J Mol Sci*
31. Chen, Huang, Shi et al. (2022) "Investigation of the crosstalk between GRP78/ PERK/ATF-4 signaling pathway and renal apoptosis induced by nephropathogenic infectious bronchitis virus infection" *J Virol*
32. Park, Jo, Lee (2020) "Renal effects of uric acid: hyperuricemia and hypouricemia" *Korean J Intern Med*
33. (2025) *Full-Length Text Journal of Virology*
34. Wen, Arakawa, Tamai (2024) "Uric acid in health and disease: from physiological functions to pathogenic mechanisms" *Pharmacol Ther*
35. Li, Qi, Chen et al. (2025) "RIPK3 activation of CaMKII triggers mitochondrial apoptosis in NIBV-infected renal tubular epithelial cells" *Vet Microbiol*
36. Li, Huang, Li et al. (2022) "Nephropathogenic infectious bronchitis virus mediates kidney injury in chickens via the TLR7/NF-κB signaling axis" *Front Cell Infect Microbiol*
37. Chen, Feng, Huang et al. (2024) "Preparation of polyclonal antibodies to chicken P62 protein and its application in nephropathogenic infectious bronchitis virusinfected chickens" *Int J Biol Macromol*
38. Wang, Wu, Jiao et al. (2022) "Enterocyte synthesizes and secrets uric acid as antioxidant to protect against oxidative stress via the involvement of Nrf pathway" *Free Radic Biol Med*
39. Jiang, Cao, Su et al. (2022) "Exercise serum regulates uric acid transporters in normal rat kidney cells" *Sci Rep*
40. Nakanishi, Ohya, Shimada et al. (2013) "Functional cooperation of URAT1 (SLC22A12) and URATv1 (SLC2A9) in renal reabsorption of urate" *Nephrol Dial Transplant*
41. Xu, Shi, Liu et al. (2020) "16S rRNA gene sequencing reveals an altered composition of the gut microbiota in chickens infected with a nephropathogenic infectious bronchitis virus" *Sci Rep*
42. Prestin, Wolf, Feldtmann et al. (2014) "Transcriptional regulation of urate transportosome member SLC2A9 by nuclear receptor HNF4α" *Am J Physiol Renal Physiol*
43. Ding, Li, Peng et al. (2019) "Uric acid transporters BCRP and MRP4 involved in chickens uric acid excretion" *BMC Vet Res*
44. Wu, Liu, Liu et al. (2022) "Bromoacetic acid causes oxidative stress and uric acid metabolism dysfunction via disturbing mitochondrial function and Nrf2 pathway in chicken kidney" *Environ Toxicol*
45. Robey, To, Polgar et al. (2008) "ABCG2: a perspective" *Adv Drug Deliv Rev*
46. Polyak, Weinberg (2009) "Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits" *Nat Rev Cancer*
47. Huang, Hong, Wei (2022) "The molecular mechanisms and therapeutic strategies of EMT in tumor progression and metastasis" *J Hematol Oncol*
48. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12569597&blobtype=pdf | # Field Evaluation of Mobile Molecular Differential Tests in DRC and Nigeria
Martin Faye, Sheila Makiala-Mandana, Moussa Diagne, Oumar Faye, Susanne Boehlken-Fascher, Jonas Kissenkoetter, Jean-Jaques Muyembe-Tamfum, Steve Ahuka-Mundeke, Placide Mbala-Kingebeni, Patient Okitale-Talunda, Gracia Mujinga, Marc-Antoine De, La Vega, Gary Kobinger, Oyinloye Elijah, Adeleye Bakarey, Raifu Kolawole, Olusegun Ademowo, Cheikh Diagne, Amadou Sall, Ahmed Abd, El Wahed, Ousmane Faye, Manfred Weidmann, Medizinische Hochschule, Brandenburg Theodor
## Abstract
Background. Accurate and timely differential diagnoses are a challenge for health care, particularly in infrastructure-poor settings. Methods. To investigate fevers of unknown origin in Africa, a mobile suitcase laboratory was deployed to DRC and Southwest Nigeria to support the control of the 2018-2020 Ebola virus disease outbreak in North-Kivu and Ituri provinces (DRC) and to provide a point-of-need solution for malaria confirmation during the dry season, respectively.Results. In DRC, the samples were tested for Ebola virus and the differentials Plasmodium falciparum, Salmonella enterica, yellow fever virus, Dengue virus, and chikungunya virus. In Southwest Nigeria, the samples were not tested for Ebola virus but were tested for the same differentials and additionally for Rickettsia spp., Leptospira, and Streptococcus pneumoniae. Plasmodium falciparum was detected in 23% (n = 192) and 47% (n = 88) of cases, respectively, and Salmonella enterica was detected in only 1 case in each cohort.Conclusions. The etiological agents circulating in febrile patients in Sub-Saharan Africa and the true incidence of neglected tropical diseases are still underestimated.
The 2018-2020 Ebola virus disease (EVD) outbreak in the east of the Democratic Republic of Congo (DRC) was associated with >3200 confirmed cases, including 56% female and 28% children under 18 years old [1]. Confirmation of positive cases relied on testing using a specific reverse transcription polymerase chain reaction (RT-PCR) assays targeting the viral nucleoprotein (NP) and glycoprotein (GP), Xpert Ebola Assay (GeneXpert Instrument Systems; Cepheid, Sunnyvale, CA, USA).
Malaria remains the primary cause of acute undifferentiated fever (or fever of unknown origin) in children across West Africa [2]. As the burden of this illness declines on the continent [3,4] and parasitological testing increases [5], a growing number of patients who would have been treated for malaria are now being diagnosed with different conditions [6]. Managing these patients appropriately still presents challenges for health care providers [7].
Despite the satisfactory performance and availability of malaria rapid diagnostic kits (mRDTs), there is evidence of antimalarial treatment of patients with negative mRDTs undermining the benefits of diagnostic screening [8]. In some contexts, the decrease in antimalarial consumption after introducing mRDTs was accompanied by increased use of antibiotics [8,9], raising concerns about potential emergence of antibiotic resistance.
A mobile suitcase laboratory for Ebola virus (EBOV) point-of-care detection at Ebola treatment centers was successfully implemented in Guinea during the large EVD outbreak in West Africa in 2014-2015. It was shown that isothermal amplification (recombinase polymerase amplification [RPA]) could be efficiently used to test suspected EVD cases, and local teams were trained in and successfully deployed this rapid method [10]. In the EVD outbreak in Guinea, up to 90% of patients fitting the case definition for EVD tested negative for EBOV, but these individuals were rarely diagnosed for differentials. Malaria and other diseases went underdiagnosed, which most likely led to more deaths from these infectious agents. Another important aspect is that the fear of patients of having Ebola while they are suffering from other diseases must be overcome by providing correct and accurate diagnosis of the type of pathogens involved in good time.
Accurate and timely differential diagnosis is a challenge for health care, particularly in the tropical regions of Sub-Saharan Africa. Syndromic differential diagnosis is very limited in point-of-care concepts. We deployed a point-of-care mobile suitcase lab to 2 African countries to demonstrate its efficacy in support of differential detection of infectious agents.
To investigate fevers of unknown origin in Africa, our mobile suitcase laboratory was deployed to East DRC and Southwest Nigeria to support control of the 2018-2020 EVD outbreak in North-Kivu and Ituri provinces (DRC) and to provide a point-of-need solution for malaria confirmation during the dry season, respectively.
## METHODS
## Study Design
A prospective cross-sectional study was designed to identify the etiology of febrile cases fitting the World Health Organization (WHO) EVD case definition [11] who were negative for EBOV in East DRC. Differentials tested were the main "suspected epidemic febrile" differentials, P. falciparum and Salmonella enterica, as well as epidemic arboviruses (Orthoflavivirus flavi [yellow fever virus {YFV}], Orthoflavivirus denguei [Dengue virus 1-4 {DENV}], and Alphavirus chikungunya [Chikungunya virus {CHIKV}]).
In Southwest Nigeria, the main study site was a primary health care center in Abanla, located in the outskirts of Ibadan, Southwest Nigeria. The study was done in the dry season, which is known for a relatively low prevalence of malaria. Patients presenting with onset of fever within the last 24 hours were recruited into the study upon consent. By default for every sample, a blood smear was stained with Giemsa and screened by conventional microscopy for detection of Plasmodium using the WHO protocol [12]. The differential RPA panel described above was extended to include Streptococcus pneumonia as Nigeria in particular suffers from the second highest burden globally [13,14]. Also included were Rickettsia spp. and Leptospira as high incidences have been reported from tropical regions including Nigeria for both pathogens [15,16].
Calibration and validation of all assays were performed at the University of Leipzig and the Institut Pasteur de Dakar before deployment, in comparison with quantitative polymerase chain reaction (qPCR), to ensure accurate and reliable results. In the field, indeterminate results were addressed by retesting samples.
## Ethical Approval
In collaboration with staff from the Ebola Treatment Unit (ETU) in the Katwa district (North-Kivu province, East DRC) all EVD-negative cases were included (national ethical approval assigned number: ESP/CE/098/2019). Teams at the mobile suitcase laboratory worked closely with the clinical staff of the health care services facility in Southwest Nigeria, where all febrile patients were invited to participate in the study (national ethical approval UI/UCH ethics committee assigned number: UI/EC/17/0005). Use of a patient consent form procedure was a condition for both approvals.
## RPA Assays
A recombinase polymerase assay (RPA) for the detection of Plasmodium falciparum was adapted for fluorescent RPA detection (Table 1) [17]. The RPA kit TwistGlow Salmonella kit (Twist Dx, Maidenhead, UK) was used for detection of Salmonella enterica. The remaining RPAs were all previously described: Streptococcus pneumonia [18], Rickettsia spp. [19], Leptospira [20], YFV [21], DENV [22], and CHIKV [23]. Additionally, an RPA for detection of the human β-actin gene was developed and used as a control to test for sampling and extraction quality and as a generic RPA-positive control (GPC) (Table 1). No template controls (NTCs) and strict procedures such as dedicated lab coats, proper use of personal protective equipment, and meticulous decontamination of the bench and the device were implemented to avoid contamination.
All of the assays were implemented in the existing mobile suitcase laboratory concept [24] to allow for testing of the major confounding epidemic differentials in an EVD outbreak and a malaria-endemic context. The amplification concept is sequential, with all extracted samples first being tested for the human ß-actin gene followed by parallel RPA amplification using the 8-strip RAA kits from the Jiangsu Qitian Gene Biotechnology company (China), with each tube targeting 1 of the individual differentials [25].
## Nucleic Acid Extraction
In DRC, RNA was extracted from blood samples using the QIAamp Viral RNA mini kit (QIAGEN, Hilden, Germany) according to the manfuacturer`s instructions. In DRC and Nigeria, total nucleic acid extraction was done from serum samples using the SpeedExtract kit (Qiagen, Hilden, Germany) according to the manufacturer`s instructions. Briefly, 200 µL of SL-Buffer, 30 µL of bead suspension, and a 20-µL serum sample were mixed in a 2-mL tube and vortexed. The mix was incubated at 95°C in a small heat block and repetitively vortexed and reinserted every 2 minutes during a 10-minute incubation period. Thereafter, the tube was briefly spun down and placed into a magnetic stand for 1 minute. The supernatant was pipetted into a new tube, and 5 µL of supernatant was used in the RPA reaction. Five microliters of eluate each was tested in the respective differential RPA assays. The RPAs were performed in an 8-strip format with dried primer and probe mixes for each reaction as previously described [26].
## Data Presentation and Statistical Analysis
Participants were categorized according to age, following the same pattern of the population's statistics and demography guidelines (<15, 16-30, 31-45, 46-60, and >60 years).
Positive results were classified according to the category of the pathogen (protozoa, bacteria, virus). Most analyses consist of simple proportions of positive samples (ie, samples that indicated the presence of a pathogen by total samples analyzed).
## RESULTS
## Differential Testing Results in DRC
In DRC, a total of 220 samples were tested by the differential RPA panel from patients with a median age of 20 years (age range, 0-98 years) and a sex ratio of 1.02. Reference testing by GeneXpert targeting 2 EBOV genes scored 119/220 EBOZV-negative and 1/220 EBOZV-positive. With the differential RPA panel, 192 (87%) were positive by GPC; failure of a GPC signal in 28 samples indicated bad sampling efficiency and low quality of human sample material. Forty-six of 192 samples scored positive for P. falciparum (23%), and 1 was positive for Salmonella sp. (0.5%). One hundred forty-five samples were negative for all parameters (75%). In addition, the EBOZV-positive samples tested negative for all parameters.
To confirm the RPA results, the 145 remaining samples were reference-tested by qRT-PCR for P. falciparum (LightMix Modular Plasmodiium genus [530], Roche, Germany), CHIKV [27], and DENV [28] at Institut National de Recherche Biomédicale (INRB), to which the samples had been transferred. All tested negative for DENV and CHIKV, thus confirming the negative RPA results. Surprisingly, all other infectious agents tested negative by PCR, leaving 41 samples P. falciparum positive by RPA but negative by PCR.
All RPA positives were also positive in the IPC RPA, indicating that the RPA reaction worked correctly. It is quite possible that there is a major sensitivity issue with the malaria qPCR used, which appears not to be sensitive enough for P. falciparum; alternatively, the oligonucleotides used in qRT-PCR need to be cross-checked against indigenous P. falciparum target gene sequences.
## Differential Testing Results in Nigeria
In Southwest Nigeria, 88 blood samples were tested from patients (age range, 1-63 years), of whom 41 (47%) tested positive using the P. falciparum RPA assay. Only 1 patient sample (1.1%) was positive using the Salmonella RPA assay. None of the other pathogens were detected. The positivity rate for P. falciparum found in Nigeria was significantly higher than that obtained in DRC (P < .0001). All RPA-positive cases were confirmed by smear test microscopy.
## DISCUSSION
A mobile suitcase laboratory for point-of-care differential testing of probable etiologies of non-EVD illness was successfully implemented near the ETU in the Katwa district (North-Kivu province) during the 2018-2020 EVD outbreak in DRC and in a malaria-endemic context in Nigeria for rapid confirmation of etiologies in febrile patients. The climate of the Kivu region in DRC is described as humid-tropical tempered by altitude. A growing population including a large displaced population due to the ongoing civil war is seeing a constantly rising malaria incidence, which was calculated at 15 501/100 000 in 2015 [29]. Southwest Nigeria has a tropical climate with significant rainfall and a short dry season. Malaria is endemic, with stable ongoing transmission and a prevalence of >94% recently recorded in a study of 300 patients visiting hospitals [30]. Given these conditions, the main differential pathogen detected at both study sites was P. falciparum in 23% and 47% of febrile EVD-negative cases and of febrile suspected malaria cases, respectively. Not surprisingly, a roughly 2-fold higher positivity rate was detected in the cohort of suspected malaria cases as opposed to the cohort of EBOV-negative cases in DRC.
Isothermal amplification RPA could be efficiently used to perform differential testing of EVD-negative cases and suspected malaria cases. P. falciparum was detected in 23% and 47% of cases, respectively. This demonstrates once more that molecular detection and in particular isothermal amplification methods can be a useful tool for malaria diagnostics [31]. A recent meta-analysis of 29 studies on the use of PCR and LAMP vs RDTs and microscopy for malaria diagnostics in Ethiopia demonstrated this conclusively [32].
The discrepancy between the commercial PCR kit results and the RPA results could not be followed up as we had no access to the oligonucleotide sequences of the commercial kit. However, once again it was made clear that for all molecular assays target erosion needs to be constantly monitored. This is now a postmarket surveillance requirement of the In vitro Diagnostic Regulation (IVDR) regulations in Europe. A better PCR reference test needs to be identified in order to determine potential RPA false-positive rates.
In both cases, only 1 Salmonella-positive case was identified (0.5% and 1.1%). This may be due to the known low pathogen concentration in blood in acute cases [33]. Another study that tested 741 febrile children in Guniea-Bissau with a PCR panel and scored 27/544 positive hits for Salmonella enterica (4%) confirms the low detection rate in our study. The slightly elevated value may be due to the target population being exclusively children <5 years of age and the larger sample size [6].
Arbovirus epidemics can have a significant impact, and in 2023 Burkina Faso experienced a Dengue outbreak with 154 867 suspected cases, 70 433 confirmed by rapid diagnostic tests, and 709 recorded deaths [34].
In DRC and Nigeria, no etiological agent was identified in 75% and 51.9% of samples, respectively. The study in febrile children in Guinea-Bissau yielded 27% negatives, indicating that, as commonly circulating arboviruses were ruled out in all 3 studies, current evidence about possibly circulating etiological agents in febrile patients in Sub-Saharan Africa is still underestimating the true incidence of neglected tropical diseases (NTDs). Although Sub-Saharan Africa recorded the most significant decline in NTD cases over the past 3 decades [35], NTDs continue to exert an immense toll, rivalling major infectious and noncommunicable diseases, particularly in Western Saharan Africa, which has the highest incidence of NTDs globally [35]. Thus, sustained investments in strengthening local health systems [36] will be crucial to achieving the goal of 90% reduction in impact and transmission of NTDs by 2030 [37].
In addition, local teams were trained in and successfully deployed this rapid method (Supplementary Figure 1).
Although malaria rapid diagnostic tests play a crucial role in increasing access to diagnosis in regions where highquality microscopy is not feasible [38], their effectiveness is directly influenced by the density of parasites. Recent research showed a sensitivity >95% when the density is ≥400 parasites/μL and nearing 100% at a density of 4000 parasites/μL, with negative predictive values of 88% during the rainy season and 95% in the dry season [39]. This project trained teams in the DRC and Nigeria and expanded the RPA testing capacity to the differentials for EVD and malaria recommended by the WHO [1,12]. Through this successful cooperation, the DRC and Nigeria teams are now able to provide field diagnostic response capacity for infectious disease outbreaks. The collaboration with teams from West Africa and Central Africa that are already proficient in the use of the laboratory suitcase has helped to build a stronger network of collaborating African outbreak response teams [40].
The presented approach demonstrates the versatility of the mobile suitcase extension to bespoke differential testing.
## References
1. (2019) "Communicable Diseases Disease Control in Humanitarian Emergencies. Ebola Virus Disease Democratic Republic of Congo.World Health Organization External Situation Report"
2. Bhatt, Weiss, Cameron (2015) "The effect of malaria control on Plasmodium falciparum in Africa between 2000 and 2015" *Nature*
3. Weiss, Lucas, Nguyen (2019) "Mapping the global prevalence, incidence, and mortality of Plasmodium falciparum, 2000-17: a spatial and temporal modelling study" *Lancet*
4. Snow, Sartorius, Kyalo (2017) "The prevalence of Plasmodium falciparum in Sub-Saharan Africa since 1900" *Nature*
5. Mandai, Francis, Challe (2024) "High prevalence and risk of malaria among asymptomatic individuals from villages with high prevalence of artemisinin partial resistance in Kyerwa district of Kagera region" *Malar J*
6. Gutierrez, Landa, Sambou (2024) "Aetiology of non-malaria acute febrile illness fever in children in rural Guinea-Bissau: a prospective cross-sectional investigation" *Front Epidemiol*
7. Nabisubi, Kanyerezi, Kebirungi (2025) "Beyond the fever: shotgun metagenomic sequencing of stool unveils pathogenic players in HIV-infected children with non-malarial febrile illness" *BMC Infect Dis*
8. Huth, Addo, Daniel (2021) "Extensive antibiotic and antimalarial prescription rate among children with acute febrile diseases in the lake Victoria Region" *J Trop Pediatr*
9. O'boyle, Bruxvoort, Ansah (2020) "Patients with positive malaria tests not given artemisinin-based combination therapies: a research synthesis describing under-prescription of antimalarial medicines in Africa" *BMC Med*
10. Faye, Soropogui (2015) "Development and deployment of a rapid recombinase polymerase amplification ebola virus detection assay in Guinea in 2015" *Euro Surveill*
11. Nsio, Ardiet, Coulborn (2023) "Differential symptomology of possible and confirmed Ebola virus disease infection in the Democratic Republic of the Congo: a retrospective cohort study" *Lancet Infect Dis*
12. (2022) "World Health Organization. WHO Guidelines for Malaria"
13. Lo, Hawkins, Jibir (2023) "Molecular characterization of Streptococcus pneumoniae causing disease among children in Nigeria during the introduction of PCV10 (GSK)" *Microb Genom*
14. Kalata, Nyazika, Swarthout (2019) "Pneumococcal pneumonia and carriage in Africa before and after introduction of pneumococcal conjugate vaccines, 2000-2019: protocol for systematic review" *BMJ Open*
15. Onyiche, Labruna, Saito (2022) "Unraveling the epidemiological relationship between ticks and rickettsial infection in Africa" *Front Trop Dis*
16. Gizamba, Mugisha (2023) "Leptospirosis in humans and selected animals in Sub-Saharan Africa 2014-2022: a systematic review and meta-analysis" *BMC infectious diseases*
17. Kersting, Rausch, Bier (2014) "von Nickisch-Rosenegk M. Rapid detection of Plasmodium falciparum with isothermal recombinase polymerase amplification and lateral flow analysis" *Malar J*
18. Clancy, Higgins, Forrest (2015) "Development of a rapid recombinase polymerase amplification assay for the detection of Streptococcus pneumoniae in whole blood" *BMC Infect Dis*
19. Kissenkotter, Hansen, Bohlken-Fascher (2018) "Development of a panrickettsial molecular diagnostic test based on recombinase polymerase amplification assay" *Anal Biochem*
20. Uduwawala, Manamperi, Gunaratna (2024) "Detection of pathogenic Leptospira with rapid extraction followed by recombinase polymerase amplification (RPA) and quantitative polymerase chain reaction (qPCR) assay-a comprehensive study from Sri Lanka" *PLoS One*
21. Escadafal, Sall (2014) "Rapid molecular assays for the detection of yellow fever virus in low-resource settings" *PLoS Negl Trop Dis*
22. El Wahed, Patel, Faye (2015) "Recombinase polymerase amplification assay for rapid diagnostics of dengue infection" *PLoS One*
23. Patel, Wahed (2015) "Rapid point-of-need diagnostics for Chikungunya virus using recombinase polymerase amplification assay" *Trop Med Int Health*
24. Wahed, Patel, Maier (2021) "Suitcase lab for rapid detection of SARS-CoV-2 based on recombinase polymerase amplification assay" *Anal Chem*
25. Ceruti, Faye, Diagne (2024) "Rapid detection of Ebolavirus using isothermal recombinase-aided amplification" *J Med Virol*
26. Weidmann, Faye (2018) "Development of mobile laboratory for viral haemorrhagic fever detection in Africa" *J Infect Dis*
27. Laurent, Roux, Grivard (2007) "Development of a sensitive real-time reverse transcriptase PCR assay with an internal control to detect and quantify chikungunya virus" *Clin Chem*
28. Callahan, Wu, Schultz (2001) "Development and evaluation of serotype-and group-specific fluorogenic reverse transcriptase PCR (TaqMan) assays for Dengue virus" *J Clin Microbiol*
29. Bigirinama, Ntaongo, Batumbo (2020) "Environmental and anthropogenic factors associated with increased malaria incidence in South-Kivu Province, Democratic Republic of the Congo" *Trop Med Int Health*
30. Awosolu, Yahaya, Haziqah et al. (2021) "Epidemiology of falciparum malaria among residents of some rural and periurban communities in Ekiti State" *Southwestern Nigeria. Trop Biomed*
31. Varo, Balanza, Mayor et al. (2021) "Diagnosis of clinical malaria in endemic settings" *Expert Rev Anti Infect Ther*
32. Feleke, Alemu, Yemanebirhane (2021) "Performance of rapid diagnostic tests, microscopy, loop-mediated isothermal amplification (LAMP) and PCR for malaria diagnosis in Ethiopia: a systematic review and meta-analysis" *Malar J*
33. Nga, Karkey, Dongol (2010) "The sensitivity of real-time PCR amplification targeting invasive Salmonella serovars in biological specimens" *BMC Infect Dis*
34. Manigart, Ouedraogo, Ouedraogo et al. (2024) "Dengue epidemic in Burkina Faso: how can the response improve?" *Lancet*
35. Lv, Zhang, Li et al. (2019) "Global, regional, national epidemiology and trends of neglected tropical diseases in youths and young adults aged 15-39 years from 1990 to 2019: findings from the Global Burden of Disease Study" *BMC Public Health*
36. Reddy, Gill, Kalkar et al. (2007) "Oral drug therapy for multiple neglected tropical diseases: a systematic review" *JAMA*
37. Vieri, Logora, Rafiq et al. (2021) "The World Health Organization road map for neglected tropical diseases 2021-2030: implications for onchocerciasis elimination programs" *Infect Dis Poverty*
38. Cunningham, Jones, Gatton (2019) "A review of the WHO malaria rapid diagnostic test product testing programme (2008-2018): performance, procurement and policy" *Malar J*
39. Bisoffi, Sirima, Menten (2010) "Accuracy of a rapid diagnostic test on the diagnosis of malaria infection and of malaria-attributable fever during low and high transmission season in Burkina Faso" *Malar J*
40. (2025) "The mobile suitcase lab" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724345&blobtype=pdf | # A SUMO interacting motif in the replication initiator protein of tomato yellow leaf curl virus is required for viral replication
Nicolas Frédéric Gaertner, Francesca Maio, Manuel Arroyo-Mateos, Ana Luna, Blanca Sabarit, Mark Kwaaitaal, Sandra Eltschkner, Marcel Prins, Eduardo Bejarano, Harrold Van Den Burg, Nicolas Gaertner
## Abstract
Circular Rep-encoding single-stranded (CRESS)-DNA viruses comprise a diverse group of viruses that use rolling-circle replication to copy their genomes. They infect organisms in almost all branches of the eukaryotic tree of life. A hallmark of all CRESS-DNA viruses is the presence of a single conserved protein, the replication initiator protein (Rep), that orchestrates viral replication by exploiting the host DNA replication machinery. In the case of the plant-infecting Geminiviridae, this multifunctional protein both recruits the host DNA replication machinery and manipulates post-translational modification including small ubiquitin-like modifier (SUMO) conjugation. In fact, Rep from two different geminiviruses, tomato yellow leaf curl virus (TYLCV; Begomovirus coheni) and tomato golden mosaic virus (TGMV; B. solanum aureimusivi), was earlier shown to interact with the SUMO conjugating enzyme SCE1. Here, we demonstrate that TYLCV C1/Rep protein also interacts with Arabidopsis SUMO1 and identify a SUMO interacting motif (SIM) located in the C-terminal SF3 helicase/ATPase domain. Remarka bly, a functional SIM proved to be important for the interaction of Rep with both SUMO1 and SCE1. The same motif in the Rep ORF was essential for TYLCV viral replication, disease symptom formation, and systemic movement from an infectious clone, and Rep ATPase activity. Together, our findings thus connect the interaction between Rep and the SUMO machinery with TYLCV viral replication.
IMPORTANCEThe identification of a non-canonical SUMO-interacting motif (SIM) within the Rep protein of tomato yellow leaf curl virus (TYLCV) reveals a connection between viral replication and a protein modification, i.e., SUMOylation. Importantly, the motif is conserved between Rep proteins from different geminiviruses. Functionally, the motif was critical for Rep's interaction with components of the SUMO machinery, for viral DNA replication, and for its ATPase activity. In particular, the third position of the motif was important for each of these activities. We thus uncover a hitherto undescribed mechanism on how geminiviruses recruit the SUMO machinery.
ubiquitin or small ubiquitin-like modifier (SUMO) conjugation (SUMOylation) onto host proteins (12)(13)(14). It is well established that many viruses manipulate host SUMOyla tion to suppress antiviral responses or to enhance virulence (15). Yet, how and why Rep recruits and directs SUMOylation via one or more target proteins remains enigmatic.
Three protein domains have been identified in Rep. Its N-terminal domain, called the HUH (His-hydrophobic residue-His) endonuclease domain, nicks and joins the singlestranded viral genome at the origin of replication using an active-site tyrosine that forms a 5′-phosphotyrosine bond with the nicked DNA (16). The central domain is involved in Rep oligomerization (17)(18)(19), whereas the C-terminal domain shares homology with the adenosine triphosphate (ATP)ase helicase superfamily 3 (SF3 helicase) and is composed of a canonical Walker A and B motifs. This domain is presumed to act as a replicative helicase during RCR elongation (19)(20)(21)(22)(23)(24). It was reported earlier that Rep from different geminiviruses is also able to reprogram the host cell cycle, thereby reactivating DNA replication in terminally differentiated plant cells (25,26). In this process, Rep interacts with a multitude of host proteins that are implicated in the cell cycle (notably the DNA replication fork) and in DNA repair. For example, Rep was found to interact with transcriptional regulators such as retinoblastoma-related protein (RBR) (27,28), as well as with proliferating cell nuclear antigen (PCNA), replication factor C (RFC), replication protein A32 (RPA32), and minichromosome maintenance protein 2 (MCM2) (29)(30)(31)(32).
The general notion is that by the process of SUMOylation, the entire protein complexes become decorated with SUMO proteins attached to different sites, rather than that one substrate is modified at a single site by a single SUMO protein (33). In this way, SUMOylation modulates a wide range of nuclear processes, including the cell cycle, DNA replication, DNA damage repair, RNA processing, and gene expression (34). SUMO attachment involves a cascade of enzymatic reactions that starts with (i) precursor maturation, followed by (ii) SUMO activation by the SUMO E1-activating enzyme (SAE1/ SAE2 dimer), and (iii) SUMO transfer to target proteins by the SUMO E2-conjugating enzyme 1 (SCE1) (35)(36)(37)(38)(39).
Rep connects DNA replication to the host's SUMO modification machinery, modifying nuclear processes, such as cell cycle regulation and DNA repair (12,14). Importantly, in the case of two geminiviruses, i.e., tomato yellow leaf curl virus (TYLCV; Begomovi rus coheni) and tomato golden mosaic virus (TGMV; B. solanum aureimusivi), Rep was confirmed to interact with the SCE1 protein (12,14,40). Interestingly, the HUH domain of Rep TGMV was found to interact with SCE1 (12). Furthermore, in the presence of Rep, SUMOylation of PCNA was impaired, both in vitro and in vivo (14). At least in yeast, loss of PCNA SUMOylation is known to cause an increase in homologous recombination (41), a process critical for geminivirus DNA replication (42). Two lysine residues in the HUH endonuclease domain of Rep TGMV proved to be essential for SCE1 binding and viral replication/spread inside the host plant (40). However, the same lysines were not essential for the interaction between Rep TYLCV and SCE1 (43). Instead, mutating these lysine residues plus one additional lysine resulted in cytosolic accumulation of Rep TYLCV (43). Notwithstanding, Rep TYLCV was found to interact with SCE1 in nuclear protein aggregates/foci called nuclear bodies (NBs), which possibly serve as hubs for interactions between viral and host proteins (43).
Here, we report how Rep TYLCV interacts with Arabidopsis SUMO1 by revealing the existence of a hitherto unknown non-canonical SUMO-interacting motif (SIM) in the SF3 helicase domain. Our data imply that this SIM mediates interactions with both SUMO1 and SCE1. Mutation of this motif significantly impairs viral accumulation, indicating its functional importance. Since this SIM is positioned next to the Walker A motif, we further investigated its role in Rep's ATPase activity. Our results demonstrate that the SIM not only promotes the interaction with proteins of the SUMO pathway, but it is also critical for ATPase activity and viral replication.
## RESULTS
## Rep from TYLCV interacts with SUMO1 and SCE1 via a non-canonical SIM in the SF3 helicase domain
Previously, two Rep proteins (from TGMV and TYLCV) were found to interact with SCE1 from Nicotiana benthamiana and Arabidopsis thaliana (12,14,40,43). As proteins involved in SUMOylation often interact with SUMO too in a non-covalent manner, we wondered whether (i) Rep TYLCV (hereafter referred to simply as "Rep, " unless sta ted otherwise) also interacts with SUMO and (ii) whether this interaction involves a previously unknown SIM in Rep. To test this hypothesis, we first used the yeast twohybrid (Y2H) split-ubiquitin system (SUS) (Fig. 1A). This system was chosen over the more commonly used GAL4-based Y2H system, as Arabidopsis SUMO1 (hereafter referred to as SUMO1) exhibited autoactivation when expressed as a bait protein (BD-SUMO) in the GAL4 system. Using the Y2H SUS, we found that Rep interacts with mature SUMO1 (GG), as well as with a conjugation-deficient variant (ΔGG) of SUMO1 that lacks the characteris tic diGly motif at the C-terminus required for SUMOylation. This strongly suggests that a hitherto unknown SIM could be present in Rep. To further explore this possibility, we tested whether Rep could interact with a SUMO1 variant where two residues critical for SIM binding (F32/I34A) were mutated (Fig. 1B) (44). As expected, Rep failed to interact with this SUMO1 F32/I34A variant (Fig. 1A).
Earlier, it was reported that SCE1 can bind SUMO through two distinct mechanisms: (i) covalently via its catalytic pocket and (ii) non-covalently via a second binding site distal to its catalytic pocket (44,45). Mutating this second site impairs SUMO binding. As the Rep-SCE1 interactions appear to be relatively weak, we wondered whether this noncovalent interaction could, in a cooperative manner, promote the Rep-SCE1 interaction.
To test this, we employed the SCE1 variant R14E/R18E/H21D (hereafter SCE1 SUM1 ), which lacks the ability to interact with wild-type (WT) SUMO1 (44). Using the GAL4 Y2H assay, we observed that Rep interacts with WT SCE1, but not with SCE1 SUM1 (Fig. 1C). Thus, at least in yeast, the formation of a Rep-SCE1 complex depends (in part) on the ability of SCE1 to interact with SUMO1. Importantly, we anticipate that the yeast homologs of SUMO and SCE1 may serve as functional substitutes for the two plant proteins in this Y2H assay considering the strong conservation of the SUMO pathway across eukaryotes, which we did not investigate further.
Next, using the bimolecular fluorescence complementation (BiFC) assay, we investiga ted whether the ability of SUMO1 to bind the SIM also controls in planta the ability of Rep to interact with SUMO1. Previously, we had demonstrated that (i) the SUMO1-SCE1 protein complex tends to aggregate in large bodies in the nucleus, termed nuclear bodies (NBs), and that (ii) formation of these NBs depends on the recruitment of active SCE1 proteins to this complex (44). Similar to the SUMO1-SCE1 NBs, we now found that (i) Rep interacts with SUMO1 and SCE1 inside such NBs in this BiFC assay, and (ii) reconstitu tion of the fluorophore halves inside these NBs was suppressed when Rep was coexpressed with SUMO1 ΔGG+F32/I34A (Fig. 1D) or SCE1 SUM1 (Fig. 1E). These data suggest that Rep interacts with SUMO1 and possibly SCE1 via a SIM in Rep.
Classically, SIMs are characterized by a stretch of at least three long-branched aliphatic residues, [VIL]-X-[VIL][VIL] (where "X" denotes any residue), flanked by a stretch of negatively charged residues (i.e., Glu, Asp, or phosphorylated residues) (46)(47)(48)(49)(50). To screen for candidate SIMs in Rep, the tool GPS-SUMO was used (https:// sumo.biocuckoo.cn/https://sumo.biocuckoo.cn/). This yielded one non-canonical SIM motif in the C-terminal SF3 helicase domain consisting of a stretch of three longbranched aliphatic residues, i.e., I215, V216, and I217. This candidate SIM does not overlap with the coding sequence of the two known viral ORFs (C2 and C4) that overlap with Rep (C1) (51,52). Yet, it is present in a β-strand adjacent to the Walker A motif of the Rep SF3 helicase domain (Fig. 1F). This means that the SIM is present in a different protein domain than the HUH endonuclease domain, which was previously reported to be important for the interaction between Rep TGMV and SCE1 (40). To assess whether this consensus motif is generally conserved across geminiviruses or even CRESS-DNA viruses, the degree of conservation of the motif was determined. This pan-genome comparison revealed that the candidate SIM is highly conserved in the Rep protein sequence of both mono-and bipartite begomoviruses (Fig. S1A) as well as CRESS-DNA viruses (Fig. S1B). In contrast, the same motif was detected in only 35% of the SF3 helicase sequences retrieved of more distant DNA viruses (Fig. S1C andD). To evaluate the structural impact of substituting the IVI motif for three alanines (AAA), structure predictions were performed using AlphaFold2 (AF2) (53) and AlphaFold3 (AF3) (54). Both AF2 and AF3 predicted the SF3 helicase domain of Rep with high confidence, based on the predicted local distance difference test (plDDT) scores (Fig. S2). No major structural differences were observed between WT Rep and Rep sim , with a root mean square deviation (RMSD) of 2.57 Å (Fig. S2B andC), suggesting that the IVI-to-AAA substitution does not induce a severe conformational change or negatively impact the SF3 helicase domain.
To test that this motif serves as a bona fide SIM critical for the interaction of Rep with SUMO1 and SCE1, we then used a mutation strategy and tested the interactions of the mutants in the Y2H assay. To this end, each of the three aliphatic residues of the SIM was mutated to alanine, i.e., alone as single residue mutations and in all possible combinations (double and triple mutants). In the case of SUMO1, we found that the Rep single and double mutants I215A, V216A, I215/V216A, and I215/217A were still capable of interacting with SUMO1 ΔGG , while the variants I217A, V216/I217A, and the triple mutant I215/V216/I217A (hereafter Rep sim ) did not (Fig. 2C; Fig. S4A). For SCE1, we noticed that the double mutant I215/V216A and all Rep mutants that included the I217A mutation failed to interact (Fig. 2D; Fig. S4B). However, we cannot exclude that the absence of a protein interaction in the Y2H assay was caused by reduced protein expression of the fusion products in yeast.
To confirm our Y2H findings, we used two in planta assays: the BiFC and splitluciferase protein-protein interaction assays (Fig. 2C through 2F). In the case of the BiFC assays (Fig. 2C andD), we employed the Rep protein from TYLCV isolate "Alb13, " which is 93% similar to Rep from the isolate "Alm"; the latter Rep was used throughout the remainder of this study. Regardless of the isolate used, the results were consistent. For clarity, the respective residue positions of the SIM of the two variants Rep Alb13 and Rep Alm are given in Table S1 (as there is a single residue shift in the SIM position). We found that the WT Rep and the single mutants localized in NBs with SUMO1 (Fig. 2C) and SCE1 (Fig. 2D). However, the Rep sim triple mutant failed to aggregate inside NBs when coexpressed with SUMO1 (Fig. 2C; Fig. S3A) or SCE1 (Fig. 2D; Fig. S3B), which corroborates the Y2H data.
To assess whether the two BiFC protein pairs Rep-SCE1 and Rep-SUMO1 co-localize inside the same NBs, we also performed a multicolor BiFC experiment (Fig. S4). We observed that the reconstituted fluorescence signal of Rep-SCFP C /Venus N -SUMO1 overlapped nearly completely with the fluorescence signal of Rep-SCFP C /SCFP N -SCE1 inside NBs (Fig. S4A, R = 0.936). To further confirm specificity, we also examined whether the reconstituted BiFC signal of Rep-SUMO overlapped with the signal of SCE1 tagged with RFP (RFP-SCE1) inside NBs (Fig. S4B). This was indeed the case (R = 0.919). Thus, these findings indicate that Rep is recruited by SUMO1 and SCE1 inside the same NBs, resulting in the apparent formation of a ternary complex. Moreover, these NBs are likely formed as a result of SUMO conjugation activity, as previously highlighted by Mazur et al. (44). In an earlier study, we used the Rep protein from TYLCV Alb13 (43), an isolate distinct from the model strain TYLCV Alm used throughout the rest of this study. In the following experiments, we used the TYLCV Alm Rep to align with the commonly used model strain.
To independently validate the role of the Rep SIM, we examined its interaction with SUMO1 and SCE1 using a second assay, the split-luciferase assay (55). In line with our hypothesis, we found that expression of Rep with SUMO1 ΔGG or SCE1 resulted in enhanced luciferase activity compared to the negative controls (Fig. 2E andF). In contrast, expression of Rep sim with SUMO1 ΔGG (Fig. 2E) or SCE1 (Fig. 2F) yielded a luciferase signal similar to or below the background signal observed for the negative controls, e.g., Rep + CLuc empty vector, indicative of a compromised interaction. When we tested the interaction of Rep TGMV with SUMO1 ΔGG and SCE1 in this assay, we noticed that the interaction of Rep TGMV is weaker than that of Rep TYLCV with both SUMO1 ΔGG and SCE1 (Fig. S5A andB). When Rep TGMV -SCE1 was co-expressed, a relatively high back ground luminescence signal was observed for the pair Rep TGMV -NLuc/CLuc empty control (Fig. S5B). Nevertheless, we find that an intact SIM is required for Rep TYLCV to interact with SUMO1 ΔGG in both yeast and in planta and that Rep TGMV also interacts with SUMO1 ΔGG in planta in a SIM-dependent manner. However, the split-luciferase assay did not confirm the role of the Rep TGMV SIM for the interaction with SCE1.
## Rep SIM is essential for viral DNA replication activity in N. benthamiana
Next, we assessed whether the Rep SIM was critical for Rep viral DNA replication. To this end, we used a virus-free plant reporter system that mimics viral DNA replication (2IR-GFP N. benthamiana; Fig. 3A) (56). In this system, transient expression of Rep (in absence of any other viral protein) is sufficient to promote RCR of a 2IR-GFP transgene cassette, thereby forming circular ssDNA extrachromosomal molecules (ECMs) that comprise a GFP expression cassette (35S Pro ::GFP) (56). A TYLCV infection in this plant will result in mass production of GFP in cells where the virus replicates. Importantly, this plant reporter line was originally selected for the fact the GFP transgene is partially silenced, and this silencing does not affect the latter expression of GFP from ECMs formed. Nevertheless, the line already displays some basal GFP expression, which is independent of Rep expression.
Using this reporter, we tested whether the different SIM mutations disrupt the ability of a RFP-tagged Rep protein (Rep-RFP) to mediate viral DNA replication upon transient expression in planta (Fig. 3B andC). As negative controls, we expressed (i) RFP alone and (ii) an RFP-tagged replication-deficient variant of Rep that no longer accumulates in the nucleus (Rep nls ; K65/69/98A) (43). First, we assessed whether the Rep SIM variants accumulated in planta and properly localized to the nucleus (43,55). Mutating the Rep SIM did not change the nuclear localization of three Rep-RFP fusion variants tested (Fig. S6A) (43,57). We then observed the DNA replication activity of the Rep SIM variants using the 2IR-GFP reporter (Fig. 3D). All the Rep-RFP fusions showed protein accumula tion, but some exhibited only 50% of the level of the WT Rep protein (Fig. 3E andF) (43). When assessing DNA replication activity, we found that the Rep variants I215/217A, V216/I217A, and I215/V216/I217A each showed no clear DNA replication activity; their GFP signals were comparable to those of the two negative controls, whereas expression of WT Rep and the other Rep SIM variants resulted in increased GFP signals (Fig. 3D, F andG). Consistently, the ECM levels were also lower for these three Rep SIM variants com pared to WT Rep, demonstrating that the cause was reduced DNA replication activity rather than enhanced GFP silencing or translational arrest (Fig. 3H).
To further exclude that low GFP levels correlated positively with low Rep-RFP levels, both the RFP and GFP protein levels were quantified using immunoblotting and normalized to WT Rep-RFP (Fig. 3F). Based on the RFP fluorescence signal, I215/217A, V216/I217A, and I215/V216/I217A each accumulated to protein levels similar to WT Rep (i.e., in the range of 0.90-1.10) (Fig. 3E andF). As previously observed (43), the Rep nls variant K65/69/98A, which accumulates in the cytosol, showed increased protein levels compared to WT Rep, likely due to differences in the protein extraction efficiencies between the cytosol and nucleus (Fig. 3E andF; Fig. S6B). These results indicate that these three Rep SIM variants with reduced SUMO1-and SCE1-binding (I215/217A, V216/ I217A, and I215/V216/I217A) also exhibit impaired DNA replication activity without affecting their expression pattern or protein stability. It thus appears that mutating residue I217 has a strong negative impact on Rep DNA replication activity in combina tion with mutating residue I215 or V216.
## Rep SIM is also required for TYLCV accumulation and viral spread in plants
Having confirmed that an intact SIM is important for Rep DNA replication activity outside the context of the virus, we introduced the SIM mutations in a TYLCV infectious clone to exclude whether other viral ORFs can functionally compensate for the SIM mutationsboth locally and in a systemic infection. To this end, a TYLCV infectious clone was generated as previously reported (58). In parallel, three mutant clones were generated, each containing different alanine substitutions in the Rep SIM coding sequence, while leaving the other known protein-coding viral ORFs intact (59). Based on the residual DNA replication activity of Rep I215/V216A, we hypothesized that the corresponding TYLCV clone would still be replication active, while the viral clones encoding the Rep variants I215/217A and I215/V216/217A (TYLCV sim ) were expected to display strongly compromised replication activity. The TYLCV clones were each delivered into the 2IR-GFP N. benthamiana plants using agroinfiltration, and the GFP levels were quantified in the infiltration zone (Fig. S7A andB). Wild-type TYLCV and TYLCV Rep I215/V216A caused, respectively, strong and weak GFP fluorescence in the infiltration site (Fig. S7B andC). The viral clones encoding Rep I215/217A and Rep sim instead yielded a GFP fluorescence signal comparable to that of the mock control. Quantification of viral DNA levels in these experiments using real-time PCR yielded similar results (Fig. S7D). The clone encoding Rep I215/V216A showed near-WT activity in viral DNA replication. As expected, the TYLCV clones encoding Rep I215/217A and Rep sim showed a clear reduction in DNA replication activity, with viral DNA titers being 1-2 magnitudes lower than those of WT TYLCV.
To assess whether viral replication was affected systemically, we monitored the infection 28 days post-agroinfiltration of 2IR-GFP N. benthamiana. The WT TYLCV clone induced clear disease symptoms (including leaf curling, leaf puckering, and stunted plant growth) indicative of a systemic infection (Fig. 4A). As expected based on the former experiment, the TYLCV clone encoding Rep I215/V216A caused disease symptoms but to a lesser extent. In contrast, the TYLCV infectious clones encoding Rep I215/217A and Rep sim failed to induce disease symptoms, thereby corroborating our hypothesis that the SIM of the Rep TYLCV is critical for viral infectivity. Real-time PCR analysis confirmed this phenotypic data (Fig. 4B), showing that viral DNA titers were low for the TYLCV clones encoding Rep I215/I217A and Rep sim compared to the WT clone in systemically infected tissue. For the TYLCV clone encoding Rep I215/V216A, the viral titers were intermediate, approximately one order of magnitude lower than those observed for WT TYLCV.
The 2IR-GFP N. benthamiana reporter line was previously selected for its low basal GFP fluorescence, indicative of partial silencing of the GFP transgene in the reporter cassette, and for its capacity to exhibit increased fluorescence upon activation of viral DNA replication (56). Increased GFP fluorescence is visible up to 14 days post-infiltration, while its silencing is re-established from 21 days post-infiltration (60). To indirectly assess viral replication, we imaged GFP fluorescence in the 2IR-GFP N. benthamiana plants. Delivery of the WT TYLCV or TYLCV Rep I215/V216A clones led to reduced GFP fluorescence, consistent with active viral replication and subsequent silencing of the 2IR-GFP construct. In contrast, TYLCV Rep I215/217A and TYLCV sim maintained their basal GFP fluorescence, consistent with an impaired replication capacity and failure to promote further GFP silencing (Fig. S7E). To exclude the possibility of a spontaneous reverting mutation that restores the mutated SIM sequence back to the original sequence and thereby enables viral spread, Sanger sequencing was performed on viral DNA extracted from systemic tissue. No clear evidence was seen in the Sanger reads that the mutated SIM region had undergone a reversion mutation in the viral genome in the systemically
## Mutating the SIM impairs Rep ATPase activity
Due to the proximity of the SIM to the Walker A motif, mutating the SIM might also affect Rep ATPase and helicase activity (20) (Fig. 1F). It has been earlier shown that substitut ing K225 in the SF3 helicase domain of Rep TYLCV for an alanine strongly impairs its ATPase activity (20). To investigate this possibility, we expressed and purified several Rep variants and tested their ATPase activity. Specifically, we focused on the I217A mutations, considering the involvement of this residue in both the SUMO1/SCE1 interaction and viral accumulation (Fig. 2 and4). The WT Rep protein purified exhibited clearly more ATPase activity than the three SIM mutants tested, i.e., less than 10% of the activity of WT Rep. Notably, Rep I217A protein sample exhibited markedly reduced ATPase activity, even lower than that of the negative control, Rep K225A, a mutation previously reported to abolish ATPase activity (20) (Fig. 5). Given the residual activity observed for Rep K225A, we hypothesized that the residual ATPase signal in this negative control could originate from co-purifying E. coli chaperones with ATPase activity, such as Heat Shock Protein HSP70 (61), rather than from Rep protein itself (62). Nevertheless, these findings indicate that the I217A mutation already reduced Rep ATPase activity, and this activity was even lower for the Rep double mutations I215/216A and V216/217A. Collectively, these results underscore again the key role of the I217 residue in the Rep SIM, not only for the SUMO1/ SCE1 interactions, but also for viral replication and ATPase activity.
## DISCUSSION
CRESS-DNA viruses all encode a single conserved protein, Rep, which orchestrates viral DNA replication. Previous studies reported that Rep TGMV interacts with the SUMO E2 conjugating enzyme SCE1 via its N-terminal HUH endonuclease domain (12,14,40). Here, we established that the SUMO protein itself is apparently involved in this interac tion between Rep TYLCV and SCE1. Specifically, we provided evidence for the existence of a not earlier studied non-canonical SIM in the SF3 helicase domain of this Rep. This SIM was not only required for Rep TYLCV to interact with SUMO1, but also with SCE1. Using a virus-free reporter system for viral DNA replication activity (2IR-GFP N. benthamiana) (56), we demonstrated that replacing two residues in this SIM for alanines, specifically the third residue along with either the first or second residue in the motif, was sufficient to apparently strongly inhibit RCR by Rep. Likewise, mutating the first and third positions combined was sufficient to prevent viral replication, systemic spread, and disease symptom development of an otherwise functional infectious clone of TYLCV, demon strating that none of the other viral transcripts or translated proteins were able to compensate for the introduced mutations in the Rep ORF. The overall significance of this SIM is underscored by its apparent conservation across different geminivirus Rep proteins, including both monopartite and bipartite begomoviruses (Fig. 1F), suggesting a conserved role in the viral infection cycle. Notably, while the SIM is conserved in geminiviruses and CRESS-DNA viruses that employ RCR, the motif was not strictly conserved in other SF3 helicase domains in unrelated DNA viruses, e.g., cottontail rabbit papillomavirus or bovine parvovirus 1 (Fig. S1). This reduced conservation suggests that the SIM is not essential for ATPase/helicase activity per se, unlike the Walker A motif that was strictly conserved across all the protein sequences analyzed (Fig. S1).
The requirement for an intact SIM for Rep TYLCV to mediate the interaction with both SUMO1 and SCE1 was demonstrated using the Y2H assay and confirmed in planta through two independent methods: BiFC and the split-luciferase. It was previously reported that Rep overexpression did not result in a global inhibition of the SUMO conjugate levels in planta (40). Here, we make a related observation that, at least in the BiFC system, the protein pair Rep-SCE1 resides in NBs, similar to the SUMO1-SCE1 protein pair. As the formation of SUMO1-SCE1 NBs strictly depends on the presence of SCE1 SUMO conjugation activity in these bodies (44), the formation of these Rep-SUMO/SCE1 NBs corroborates again that the viral role of Rep is not to suppress global SCE1 enzyme activity (Fig. 2C andD), but more likely that it has a role in stimulating or repressing SUMOylation of one or more specific host proteins, as shown in the latter case for tomato PCNA (14).
Previous studies have reported that Rep TGMV interacts with SCE1 via its N-terminal HUH endonuclease domain, with lysine mutations in this domain disrupting this interaction (12,14,40). Here, we identified a novel SIM present in the SF3 helicase domain of Rep TYLCV and Rep TGMV that is important for both proteins to interact in a non-covalent manner with a conjugation-deficient form (ΔGG) of SUMO1. However, in the case of the Rep TGMV -SCE1 protein pair, a relatively high background signal was observed in our split-luciferase assay (Fig. S4D), which prevents us from drawing firm conclusions on the role of the SIM in the case of the Rep TGMV -SCE1 interaction, while it was needed for the Rep TYLCV -SCE1 interaction. These findings do, nevertheless, support a conserved function for the SIM across geminiviruses, including mono-and bipartite begomoviruses. Our data also corroborate the earlier proposed notion by Maio et al. (43) that the mechanism that stabilizes the SCE1-Rep interaction likely differs between Rep TYLCV and Rep TGMV .
We propose that, in the case of Rep TYLCV , a ternary complex is formed between Rep TYLCV , SCE1, and SUMO1, and this ternary interaction allows what would otherwise be a weak interaction to be detected in our assays. In this model, no single interaction predominates; rather, all three components contribute to the formation and stability of the ternary complex. This cooperative binding may alleviate competition between SUMO1 and SCE1. Notably, the Rep residue I217 appears to have a central role in coordinating this interaction between Rep TYLCV and SUMO1/SCE1. Most other single and double residue mutations in the SIM gave a Rep variant that still interacted to some extent with SUMO1, ΔGG except for Rep I217A, V216/I217A, and the triple mutant I215/ V216/I217A (Fig. 2A). In the case of the interaction between Rep and SCE1, the double mutant I215/V216A, as well as any mutant carrying the I217A substitution, disrupted this interaction (Fig. 2B). The I217A mutation led to a more pronounced reduction in viral DNA accumulation (quantified using both GFP expression levels from ECMs and accumulation of viral DNA) compared to the single I215 or V216 single residue mutations (Fig. 3F). These findings thus add a novel function to the SF3 helicase domain of Rep, that is, SUMO binding.
Finally, we investigated the impact of the position of the SIM adjacent to the Walker A and B motifs within the SF3 helicase domain to determine whether SIM mutations negatively impact Rep ATPase activity. Indeed, mutating the SIM impaired Rep ATPase activity, providing a mechanistic explanation for the loss of RCR activity, as ATP hydrolysis is essential for Rep helicase function (63). In general, Walker A motifs are located on a flexible loop (also known as the "phosphate-binding loop" or in short "P-loop") preceded by a β-strand and followed by an α-helix. Notably, previous studies have shown that a conserved stretch of hydrophobic residues is located on the β-strand upstream of the Walker A (P-loop) motif and identified the presence of hydrophobic residues at distinct positions on the subsequent α-helix (64). Those hydrophobic residues possibly interact with each other and thereby stabilize the arrangement of the two secondary structure elements relative to each other, which might in turn facilitate nucleotide (e.g., ATP) binding by the P-loop. By substituting isoleucines and valines in Rep that likely coincide with the hydrophobic stretch in the β-strand preceding the P-loop (Fig. S1) for alanines, the overall hydrophobic character of the motif was preserved. However, the smaller size of the alanine side chain compared to valine and isoleucine may have an effect on the proximity/angular arrangement of the two secondary structure motifs flanking the P-loop. Structural predictions with AlphaFold3 of Rep sim suggested that the introduction of the alanines did not appear to negatively impact the integrity of the Walker A (P-loop motif ) of the ATPase domain and resulted in comparable plDDT confidence values for this protein domain with respect to WT Rep (Fig. S2). Yet, these latter protein models should be taken with caution as AlphaFold was not designed to capture subtle effects of one or a few amino acid changes (65). To conclude, we find a three-residue motif that contributes to both the SUMO1/SCE1 interaction and viral accumulation. Whether this SIM is also involved in the suppression of PCNA SUMOylation by Rep remains an open question (14). While our findings advance our understanding of the Rep interaction with the SUMO machinery, further studies are needed to elucidate the specific mechanisms by which the SIM may modulate PCNA activity.
## MATERIALS AND METHODS
## General methods and cloning
All molecular techniques were performed using standard methods (66). Escherichia coli strain DH5α was used to clone gene fragments. All primers and plasmids used are described in the Tables S2 andS3, respectively. The different fragments were PCR-ampli fied using Phusion DNA polymerase (Thermo Fisher), and the amplicons obtained were introduced in pENTR207 or pENTR221 (Thermo Fisher) using Gateway BP Clonase II (Thermo Fisher). Transfer of the inserts to destination vectors was done using Gateway LR Clonase II (Thermo Fisher). Point mutations were introduced using QuikChange site-directed mutagenesis. All inserts were confirmed by DNA sequencing. The order of the different protein fusions (i.e., tag-REP or REP-tag) is indicated in the figures. Rep ORF from the TYLCV isolate "Almeria" was used in this study, unless mentioned otherwise.
## Yeast two-hybrid assays
For the GAL4 Y2H assay, all gene fragments were cloned into Gateway-compatible variants of pGADT7 and pGBKT7 (Clontech) (59). Resulting plasmids were introduced into Saccharomyces cerevisiae strain PJ69-4α (67) using a standard lithium acetate/sin gle-stranded DNA/polyethylene glycol 3350 transformation protocol (68). Transformed colonies were selected on yeast minimal medium (MM) lacking the amino acids Leu and Trp. To select for protein-protein interactions (PPIs), three independent transformants were randomly picked. After resuspension in 100 µL sterile double-distilled water, a tenfold serial dilution was spotted on MM agar plates lacking Leu, Trp, and His (-LWH) supplemented with 1 mM 3-amino-1,2,4-triazole (3AT). The plates were then incubated at 30°C for 3 days prior to scoring yeast growth.
The Gateway-ready plasmids for the split-ubiquitin Y2H system, pMET/pCub (57), were transformed into S. cerevisiae JD53. Selection of transformants and positive PPIs was performed as described (69). Briefly, transformed colonies were selected on yeast MM lacking His and Trp (-HW). To select for PPIs, three independent transformants were resuspended, and a tenfold serial dilution was spotted (similar to GAL4 assay) on MM agar plates supplemented with 0.1 mM CuSO 4 and the appropriate amino acids. To increase selection specificity, 1 mg/mL 5-FOA (5-fluoroorotic acid, Sigma) was added to the selection plates. Plates were incubated at 30°C for 4 days prior to scoring.
## Transient expression of proteins in N. benthamiana using agroinfiltration
All binary constructs were transformed in Agrobacterium tumefaciens (Rhizobium radiobacter) strain GV3101 (70) by electroporation (25 µF, 400 Ω, 1.25 kV/mm). Single colonies were grown overnight to an optical density at 600 nm (OD 600 ) of 0.8-1.5 in low salt LB broth (1% [wt/vol] tryptone, 0.5% [wt/vol] yeast extract, 0.25% [wt/vol] NaCl, pH 7.0). Bacterial cells were collected by centrifugation (3,000 × g for 5 min), washed, and resuspended in leaf infiltration medium (1 × MS [Murashige and Skoog] salts, 10 mM MES pH 5.6, 2% [wt/vol] sucrose, 200 µM acetosyringone). A. tumefaciens carrying the pBIN61 binary vector to express the P19 silencing suppressor (referred to as pBIN61:P19) from tomato bushy shunt virus (TBSV, Tombusvirus lycopersici) (71) was added to every experimental sample.
## Bimolecular fluorescence complementation (BiFC) assay
Rep from the TYLCV isolate "Alb13" (Rep Alb13 ), the Arabidopsis SUMO1 (gene ID: AT4G26840), and SCE1 (AT3G57870) coding sequence were cloned in the vectors pDEST-GWSCYCE to express fusion protein with the C-terminal half of S(CFP)3A (residues 156-239; referred to as SCFPC) or pDEST-SCYNEGW (for protein fusions with the N-terminal half of S(CFP)3A residues 1-173; SCFPN) (72). Four-week-old N. benthamiana leaves were syringe-infiltrated with A. tumefaciens suspensions at a final OD 600 of 1.0 when a single construct was delivered. When two cultures were co-infiltrated for BiFC analysis, the cultures were mixed at a ratio of 1:1 to a final OD 600 of 1.0. A. tumefaciens carrying pBIN61:P19 was added to every experimental sample at a final OD 600 of 0.5. Three days post-infiltration, N. benthamiana leaf material was collected to analyze the expression of the infiltrated constructs. SCFP fluorescence was detected using an excitation wavelength of 458 nm (argon laser), primary beam-splitting mirrors 458/514 nm, secondary beam splitter 515 nm, and band filter BP 470-500 nm.
## In planta protein localization
Rep Alb13 was cloned in the plasmid pGWB654, which carries a C-terminal monomeric red fluorescent protein (mRFP) tag (73). The expression and detection were the same procedure as for the BiFC assay. with (i) A. tumefaciens strain carrying pGWB654 plasmid encoding different Rep-mRFP variants ( 77) and (ii) A. tumefaciens carrying the pBIN61:P19 to suppress gene silencing (71). The bacterial cultures were mixed at a ratio of 2:1 to final OD 600 values of 0.8 and 0.4, respectively, prior to infiltration. Four days post-infiltration, whole leaves were imaged for GFP and RFP fluorescence using a Chemidoc MP (BioRad) imager with the presettings "Alexa 488" and "Rhodamine, " respectively. To estimate the relative fluorescence intensity, three leaves of the same age from different plants were agroinfiltrated and imaged (n = 3). Mean fluorescence intensities per infiltrated area were calculated. To normalize fluorescence between leaves, the fluorescence is given as a proportion of the total intensity per leaf. Visualization and ANOVA tests were performed using Prism 9.0v (GraphPad).
## Quantification of extrachromosomal molecules using qPCR
To quantify the level of ECMs, leaf disk samples were snap-frozen in liquid nitrogen and stored at -80°C prior to tissue processing. Frozen tissue was ground to powder using a steel ball in a bead mill (TissueLyser II, Qiagen). Total DNA was extracted from approximately 30 mg of leaf tissue using a routine hexadecyltrimethylammonium bromide (CTAB) method (78). DNA concentrations were estimated using the absorbance at 260 nm on a Nanodrop (Thermo Fisher). In total, 250 ng of total DNA was used as template for a real-time PCR reaction (QuantStudio3, Thermo Fisher) using the Hot FIREPol EvaGreen qPCR kit according to the suppliers' instructions (Solis Biodyne). The ECM signal was normalized using the 25S rRNA (GenBank ID, KP824745.1) as an internal reference. The primers to detect the viral DNA were designed such that they only amplify ECMs and not the 2IR-GFP genomic insert (Table S2). All cycle threshold (Ct) values were corrected for primer efficiencies. All expression data were analyzed following the standard workflow provided by qBASE+ (Biogazelle) using three independent biological replicates (n = 3). CNRQ was calculated using qBASE+. Data visualization and ANOVA tests were performed using Prism 9.0v (GraphPad).
## Protein extraction from plant tissue and immunoblotting
Total protein fraction was extracted using approximately 30 mg of frozen N. benthami ana leaf tissue. To this end, leaf material was snap-frozen and ground using plastic pestles, after which RIPA buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1% [vol/vol] Triton X-100, 0.5% [wt/vol] deoxycholate, 0.1% [wt/vol] SDS, 5 mM DTT, 1 × cOmplete Protease Inhibitor Cocktail [Roche]) was added at a 4:1 (vol/wt) buffer-to-tissue ratio. Samples were thawed on ice and vortexed three times for 10 s. After incubating the homogenates prior on a rotating wheel for 1 h (25 rpm) at 4°C, they were centrifuged at 16,000 × g (4°C), and 80 µL of the supernatant was mixed with 2 × Laemmli buffer (100 mM Tris pH 6.8, 20% [wt/vol] glycerol, 4% [wt/vol] SDS, 100 mM DTT, 0.001% [wt/vol] Bromophenol blue) in 1:1 ratio. Total protein extract was denatured by heating at 96°C for 10 min. Upon protein denaturation, extracts were centrifuged at maximum speed (16,000 × g, ambient temperature) for 5 min, and the supernatant was separated on a 12% SDS-PAGE gel and subsequently transferred onto a PVDF membrane (Immobilon-P, Millipore). Immunode tection of the proteins was performed according to the standard protocols using the antibodies detailed in Table S4. For detecting chemiluminescence (ECL), a home-made solution was used (0.1 M Tris-HCl pH 8.5, 1.25 mM luminol [Sigma-Aldrich] in DMSO, 0.2 mM p-coumaric acid [Sigma-Aldrich] in DMSO, 0.01% [vol/vol] H 2 O 2 ), and the signals were imaged using a Chemidoc MP (Bio-Rad) or a light-sensitive X-ray film (Fuji Super RX). Equal loading of the different protein extracts was confirmed by examining the Rubisco levels using Ponceau S staining of the membranes. The intensity of the Rubisco signal was also used to normalize the relative GFP and mRFP protein levels.
## Quantification of viral infections
Infectious clones are described in detail in Table S3. Agrobacterium carrying the pGreen TYLCV constructs was cultivated 24-30 h to saturation (OD 600 3.5-4) and then pelleted by centrifugation for 10 min (3,000 × g) before being resuspended in fresh low salt LB medium without antibiotics to a final OD 600 of 7.5-8. Then, they were introduced into three-week-old N. benthamiana by injecting the suspension in an abaxial bud. Plants were maintained at 25°C with natural daylight supplemented to a 16-hour photoper iod with sodium lamps (Sylvania Gro-Lux). Three weeks post-inoculation, apical leaves were harvested near the apical shoot to isolate total DNA and quantify the relative viral accumulation of TYLCV coat protein in N. benthamiana by quantitative real-time PCR. To confirm that TYLCV mutants did not revert back to WT Rep protein sequence by spontaneous mutations and selection pressure, the virus samples were isolated and sequenced. One representative sequencing result of Rep (552-720 bp from the start codon) is shown (Fig. S7F). Normalization for viral DNA was performed using the ribosomal RNA 25S as internal reference. Primers used for real-time quantification are detailed in Table S2. Real-time PCR primers were designed to only amplify the circular ized viral DNA.
## Split-luciferase complementation (SLUC) assay
N. benthamiana leaves were co-infiltrated with a mixture of A. tumefaciens harboring different pGWB402 ( 77) constructs (vectors used can be found in Table S3) and A. tumefaciens pBIN61:P19 with a final OD 600 of 0.8 and 0.4 for both strains, respectively. Rep Alb13 variant was used for this SLUC. The infiltrated leaf areas were marked using a paint marker. Three days post-infiltration, agroinfiltrated leaves were brushed twice with D-luciferin buffer (235 mM D-luciferin [Duchefa Biochemie] dissolved in Milli-Q water, 0.02% [wt/vol] Silwet L-77 [Crompton Europe]). After 2 h dark incubation, the chemilu minescence signal was captured using a CCD imaging system (Princeton Instruments) set at -70°C. An exposure time of 15 min with 10 × 10 binning was used for images. Data acquisition was performed using the MetaVue program (X-Rite). Each data point consisted of at least three replicates, and four independent experiments were performed for each assay. As a negative control, unfused half of the luciferases were used. The log10 value of the mean signal (integrated density/area size), corrected for the leaf background signal and expressed as a proportion of total signal of all samples infiltrated in a single leaf, was used for further statistical tests. The data of each independent experiment were pooled for the ANOVA followed by Dunnett's multiple comparison test in Prism v10.0 (GraphPad).
## Rep protein expression and purification
To express and purify recombinant Rep, the E. coli strain BL21-Gold (Agilent Technolo gies) harboring Rep in pGEX (with an N-terminal glutathione S-transferase (GST)-tag and PreScission-cleavage site) was used. Two-liter cultures were started with an overnight pre-culture to an OD 600 of 0.1 and grown at 37°C and 200 rpm until reaching an OD 600 of 0.7. Rep protein expression was induced using 0.5 mM IPTG, and the temperature was lowered to 30°C. After 2 h of induction, the bacterial cells were harvested at 5,000 × g for 15 min at room temperature, and the cell pellet was stored at -20°C prior to purification. The frozen pellet was thawed and resuspended in 30 mL of lysis buffer (50 mM Tris-HCl pH 7.5, 300 mM NaCl, 10% [wt/vol] glycerol, 2 mM MgCl 2 , 0.5 mM CaCl 2 , 2 mM DTT, 1% [wt/vol] Tween-20, 0.5 mg/mL lysozyme, 6 U/mL DNase I, and 1 mM PMSF) and stirred for 1 h at 4°C. Cells were disrupted by passing them three to five times through a French press (Thermo Fisher) and centrifuged at 50,000 × g for 1 h at 4°C. Supernatant was loaded onto a 25 mL Glutathione Sepharose 4 Fast Flow column (Cytiva) and washed with 1.5 column volumes (CV) of wash buffer I (50 mM Tris-HCl pH 7.5, 200 mM NaCl, 5% [wt/vol] Glycerol, 2 mM MgCl2, and 2 mM DTT). Subsequently, the column was washed with 1.5 CV of wash buffer II (50 mM Tris-HCl pH 7.5, 800 mM NaCl, 5% [wt/vol] glycerol, 2 mM MgCl 2 , and 2 mM DTT). The column was equilibrated with 1.5 CV of wash buffer I, and 50 μL (8 mg/mL) of the His-tagged PreScission protease was added for overnight digestion. Protein elution was undertaken with 1.5 CV of elution buffer (50 mM Tris-HCl pH 7.5, 200 mM NaCl, 5% [wt/vol] glycerol, 2 mM MgCl 2 , 2 mM DTT, and 20 mM GSH). The eluted protein (approximate mass 40 kDa) was further concentrated using a spin concentrator (10 kDa, Amicon, Millipore) and rebuffered in 20 mM HEPES pH 7.5, 200 mM NaCl, 2 mM MgCl 2 , 5% (wt/vol) glycerol, and 2 mM DTT. To quantify the concentration of purified protein in the samples, a bicinchoninic acid (BCA) assay was conducted using the BCA kit following the manufacturer's recommendations (Sigma Aldrich).
## ATPase activity assay
The decrease in NADH absorbance, which is proportional to the rate of ATP hydrolysis, was measured continuously at a wavelength of 340 nm for 120 min at 20°C using the Synergy H1MF (BioTek) plate reader. The assay was performed in 25 mM HEPES/NaOH pH 7.5, 200 mM NaCl, 10 mM KCl, 10 mM MgCl 2 , 1 mM DTT, 5% (wt/vol) glycerol, 1.5 mM NADH, 10 U each of pyruvate kinase (PK) and lactate dehydrogenase (LDH), and 3 mM of phosphoenolpyruvate (PEP). The quantity of the sample was adjusted to achieve complete consumption of NADH within the measurement time. For WT Rep, Rep I217A, Rep I215/217A, Rep V216/I217A, and Rep K225A, we used a final concentration of 8 µM, 80 µM, 80 µM, 130 µM, and 50 µM of protein, respectively, to observe ATP hydrolysis within the measurement time. Experimental replicates were measured. The negative control contained all the reaction components except for the Rep protein. ATP hydrolysis was initiated by adding 1 mM ATP to each well.
## Determination of ATP turnover rate
A linear decrease in NADH absorbance over time was selected to calculate the ATP turnover rate for each measurement (as described above), spanning at least 15 min (79), i.e., the slope of the curve. An NADH standard curve ranging from 0.0 mM to 2.0 mM NADH was used to correlate the absorbance values at 340 nm with the respective NADH concentrations. Hence, the slope of the curve (absorbance decrease per time interval) could be transformed into NADH consumption over time (mM/min). Spontane ous oxidation of NADH observed in the negative control, i.e., a mock sample without Rep, was subtracted from the corresponding absorbance values. The slope, intercept, and R-squared values were determined by fitting a linear model using RStudio (R version 4.1.0).
## Accession numbers
DNA clones of TYLCV isolate "Alb13" Rep (Genebank ID: FJ956702.1) were kindly provided by Keygene (Wageningen, Netherlands). Rep from TYLCV isolate "Almeria" (Begomovirus coheni AJ489258.1) was synthesized by GenScript. Clones for the coding sequences of AtSCE1 (At3g57870) and AtSUMO1 (At4g26840) were previously described (71)
## References
1. Kazlauskas, Varsani, Koonin et al. (2019) "Multiple origins of prokaryotic and eukaryotic single-stranded DNA viruses from bacterial and archaeal plasmids" *Nat Commun*
2. Zhao, Rosario, Breitbart et al. "Eukaryotic circular repencoding single-stranded DNA (CRESS DNA) viruses: ubiquitous viruses with small genomes and a diverse host range" *Adv Virus Res*
3. Desingu, Nagarajan (2022) "Genetic diversity and characterization of circular replication (Rep)-encoding single-stranded (CRESS) DNA viruses" *Microbiol Spectr*
4. Zhao, Lavington, Duffy (2021) "Truly ubiquitous CRESS DNA viruses scattered across the eukaryotic tree of life" *J Evol Biol*
5. Rizvi, Choudhury, Tuteja (2015) "Insights into the functional characteristics of geminivirus rolling-circle replication initiator protein and its interaction with host factors affecting viral DNA replication" *Arch Virol*
6. Ruhel, Chakraborty (2019) "Multifunctional roles of geminivirus encoded replication initiator protein. Virusdisease"
7. Wei, Lozano-Durán (2023) "The primase subunits of DNA polymerase α, PRIM1 and PRIM2, are required for the replication of the geminivirus tomato yellow leaf curl virus in the host plant" *MicroPublication Biol. htt*
8. Wu, Wei, Tan et al. (2021) "Plant DNA polymerases α and δ mediate replication of geminiviruses" *Nat Commun*
9. Ghosh, Chakraborty (2022) "Selective REcruitmeNt of plant DNA polymerases by geminivirus" *Trends Genet*
10. Shakir, Mubin, Nahid et al. (2023) "REPercussions: how geminiviruses recruit host factors for replication" *Front Microbiol*
11. Bonnamy, Blanc, Michalakis (2023) "Replication mechanisms of circular ssDNA plant viruses and their potential implication in viral gene expression regulation"
12. Castillo, Kong, Hanley-Bowdoin et al. (2004) "Interaction between a geminivirus replication protein and the plant sumoylation system" *J Virol*
13. (2017) "The replication initiator protein of a geminivirus interacts with host monoubiquitination machinery and stimulates transcription of the viral genome" *PLoS Pathog*
14. Arroyo-Mateos, Sabarit, Maio et al. (2018) "Geminivirus replication protein impairs SUMO conjugation of proliferating cellular nuclear antigen at two acceptor sites" *J Virol*
15. Fan, Li, Zhang et al. (2022) "SUMOylation in viral replication and antiviral defense" *Adv Sci (Weinh)*
16. Chandler, De La Cruz, Dyda et al. (2013) "Breaking and joining single-stranded DNA: the HUH endonu clease superfamily" *Nat Rev Microbiol*
17. Orozco, Miller, Settlage et al. (1997) "Functional domains of a geminivirus replication protein" *J Biol Chem*
18. Missich, Ramirez-Parra, Gutierrez (2000) "Relationship of oligomeri zation to DNA binding of Wheat dwarf virus RepA and Rep proteins" *Virology (Auckl)*
19. Tarasova, Khayat (2023) "Dynamics and conformations of a full-length CRESS-DNA replicase" *Viruses*
20. Gorbalenya, Koonin, Wolf (1990) "A new superfamily of putative NTP-binding domains encoded by genomes of small DNA and RNA viruses" *FEBS Lett*
21. Ilyina, Koonin (1992) "Conserved sequence motifs in the initiator proteins for rolling circle DNA replication encoded by diverse replicons from eubacteria, eucaryotes and archaebacteria" *Nucleic Acids Res*
22. Koonin (1993) "A common set of conserved motifs in a vast variety of putative nucleic acid-dependent ATPases including MCM proteins involved in the initiation of eukaryotic DNA replication" *Nucleic Acids Res*
23. Desbiez, David, Mettouchi et al. (1995) "Rep protein of tomato yellow leaf curl geminivirus has an ATPase activity required for viral DNA replication" *Proc Natl Acad Sci*
24. Tarasova, Dhindwal, Popp et al. (2021) "Mechanism of DNA interaction and translocation by the replicase of a circular Repencoding single-stranded DNA virus"
25. Hanley-Bowdoin, Bejarano, Robertson et al. (2013) "Geminiviruses: masters at redirecting and reprogramming plant processes" *Nat Rev Microbiol*
26. Gupta, Reddy, Bhattacharyya et al. (2021) "Plant responses to geminivirus infection: guardians of the plant immunity" *Virol J*
27. Kong, Orozco, Roe et al. (2000) "A geminivirus replication protein interacts with the retinoblastoma protein through a novel domain to determine symptoms and tissue specificity of infection in plants" *EMBO J*
28. Xie, Suárez-López, Gutiérrez (1995) "Identification and analysis of a retinoblastoma binding motif in the replication protein of a plant DNA virus: requirement for efficient viral DNA replication" *EMBO J*
29. Luque, Sanz-Burgos, Ramirez-Parra et al. (2002) "Interaction of geminivirus Rep protein with replication factor C and its potential role during geminivirus DNA replication" *Virology (Auckl)*
30. Castillo, Collinet, Deret et al. (2003) "Dual interaction of plant PCNA with geminivirus replication accessory protein (Ren) and viral replication protein (Rep)" *Virology (Auckl)*
31. Singh, Islam, Choudhury et al. (2007) "The 32 kDa subunit of replication protein A (RPA) participates in the DNA replication of mung bean yellow mosaic India virus (MYMIV) by interacting with the viral Rep protein" *Nucleic Acids Res*
32. Suyal, Mukherjee, Srivastava et al. (2013) "Arabidopsis thaliana MCM2 plays role(s) in mungbean yellow mosaic India virus Full-Length Text Journal of Virology December"
33. Mymiv "DNA replication" *Arch Virol*
34. Psakhye, Jentsch (2012) "Protein group modification and synergy in the SUMO pathway as exemplified in DNA repair" *Cell*
35. Vertegaal (2022) "Signalling mechanisms and cellular functions of SUMO" *Nat Rev Mol Cell Biol*
36. Yunus, Lima (2006) "Lysine activation and functional analysis of mediated conjugation in the SUMO pathway" *Nat Struct Mol Biol*
37. Geiss-Friedlander, Melchior (2007) "Concepts in sumoylation: a decade on" *Nat Rev Mol Cell Biol*
38. Saracco, Miller, Kurepa et al. (2007) "Genetic analysis of SUMOylation in Arabidopsis: conjugation of SUMO1 and SUMO2 to nuclear proteins is essential" *Plant Physiol*
40. Gareau, Lima (2010) "The SUMO pathway: emerging mechanisms that shape specificity, conjugation and recognition" *Nat Rev Mol Cell Biol*
41. Cappadocia, Lima (2018) "Ubiquitin-like protein conjugation: structures, chemistry, and mechanism" *Chem Rev*
42. Sánchez-Durán, Dallas, Ascencio-Ibañez et al. (2011) "Interaction between geminivirus replication protein and the SUMOconjugating enzyme is required for viral infection" *J Virol*
43. Mohiuddin, Evans, Rahman et al. (2018) "SUMOylation of PCNA by PIAS1 and PIAS4 promotes template switch in the chicken and human B cell lines" *Proc Natl Acad Sci*
44. Lefeuvre, Moriones (2015) "Recombination as a motor of host switches and virus emergence: geminiviruses as case studies" *Curr Opin Virol*
45. Maio, Arroyo-Mateos, Bobay et al. (2019) "A lysine residue essential for geminivirus replication also controls nuclear localization of the tomato yellow leaf curl virus Rep protein" *J Virol*
46. Mazur, Kwaaitaal, Mateos et al. (2019) "The SUMO Conjugation complex self-assembles into nuclear bodies independent of SIZ1 and COP1" *Plant Physiol*
47. Knipscheer, Van Dijk, Olsen et al. (2007) "Noncova lent interaction between Ubc9 and SUMO promotes SUMO chain formation" *EMBO J*
48. Minty, Dumont, Kaghad et al. (2000) "Covalent modification of p73α by SUMO-1: two-hybrid screening with p73 identifies novel SUMO-1-interacting proteins and a SUMO-1 interaction motif" *J Biol Chem*
49. Song, Zhang, Hu et al. (2005) "Small ubiquitin-like modifier (SUMO) recognition of a SUMO binding motif: a reversal of the bound orientation" *J Biol Chem*
50. Hecker, Rabiller, Haglund et al. (2006) "Specification of SUMO1-and SUMO2-interacting Motifs" *Journal of Biological Chemistry*
51. Wang, Dasso (2009) "SUMOylation and deSUMOylation at a glance" *J Cell Sci*
52. Lascorz, Codina-Fabra, Reverter et al. (2022) "SUMO-SIM interactions: from structure to biological functions" *Seminars in Cell & Developmental Biology*
53. Gong, Tan, Zhao et al. (2021) "Geminiviruses encode additional small proteins with specific subcellular localizations and virulence function" *Nat Commun*
54. Zhao, Gong, Ren et al. (2022) "The novel C5 protein from tomato yellow leaf curl virus is a virulence factor and suppressor of gene silencing" *Stress Biol*
55. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
56. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
57. Chen, Zou, Shang et al. (2008) "Firefly luciferase complementation imaging assay for protein-protein interactions in plants" *Plant Physiol*
58. Morilla, Castillo, Preiss et al. (2006) "A versatile transreplication-based system to identify cellular proteins involved in geminivirus replication" *J Virol*
59. Deslandes, Olivier, Peeters et al. (2003) "Physical interaction between RRS1-R, a protein conferring resistance to bacterial wilt, and PopP2, a type III effector targeted to the plant nucleus" *Proc Natl Acad Sci*
60. Klinkenberg, Stanley (1989) "Fate of African cassava mosaic virus coat protein deletion mutants after agroinoculation" *Journal of General Virology*
61. Chien, Bartel, Sternglanz et al. (1991) "The two-hybrid system: a method to identify and clone genes for proteins that interact with a protein of interest" *Proc Natl Acad Sci*
62. Lozano-Durán, Rosas-Díaz, Luna et al. (2011) "Identification of host genes involved in geminivirus infection using a reverse genetics approach" *PLoS One*
63. Li, Li, Zhang et al. (2020) "Dynamic subcellular localization, accumulation, and interactions of proteins from tomato yellow leaf curl china virus and its associated betasatellite" *Front Plant Sci*
64. Gorovits, Czosnek (2017) "The involvement of heat shock proteins in the establishment of tomato yellow leaf curl virus infection" *Front Plant Sci*
65. Clérot, Bernardi (2006) "DNA helicase activity is associated with the replication initiator protein rep of tomato yellow leaf curl geminivirus" *J Virol*
66. Romero, Yang, Lin et al. (2018) "Simple yet functional phosphateloop proteins" *Proc Natl Acad Sci*
67. Pak, Markhieva, Novikova et al. (2023) "Using AlphaFold to predict the impact of single mutations on protein stability and function" *PLoS One*
68. Green, Sambrook (2012) "Molecular cloning: a laboratory manual"
69. James, Halladay, Craig (1996) "Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast" *Genetics*
70. Folters, Rgh (2011) "Yeast protein-protein interaction assays and screens" *Plant Transcr Factors Methods*
72. Reichel, Johnsson (2005) "The split-ubiquitin sensor: measuring interactions and conformational alterations of proteins in vivo" *Methods Enzymol*
73. Koncz, Schell (1986) "The promoter of TL-DNA gene 5 controls the tissue-specific expression of chimaeric genes carried by a novel type of Agrobacterium binary vector" *Molec Gen Genet*
74. Jay, Brioudes, Voinnet (2023) "A contemporary reassessment of the enhanced transient expression system based on the tombusviral silencing suppressor protein P19" *Plant J*
75. Gehl, Waadt, Kudla et al. (2009) "New GATEWAY vectors for high throughput analyses of protein-protein interactions by bimolecular fluorescence complementation" *Mol Plant*
76. Nakamura, Mano, Tanaka et al. (2010) "Gateway binary vectors with the bialaphos resistance gene, bar, as a selection marker for plant transformation" *Biosci Biotechnol Biochem*
77. Schindelin, Arganda-Carreras, Frise et al. (2012) "Fiji: an opensource platform for biological-image analysis" *Nat Methods*
78. Meng, Goddard, Pettersen et al. (2023) "UCSF ChimeraX: Tools for structure building and analysis"
79. Mirdita, Schütze, Moriwaki et al. (2022) "ColabFold: making protein folding accessible to all" *Nat Methods*
80. Nakagawa, Suzuki, Murata et al. (2007) "Improved gateway binary vectors: high-performance vectors for creation of fusion constructs in transgenic analysis of plants" *Biosci Biotechnol Biochem*
81. Aboul-Maaty, Oraby (2019) "Extraction of high-quality genomic DNA from different plant orders applying a modified CTAB-based method" *Bull Natl Res Cent*
82. Radnai, Stremel, Sellers et al. (2019) "A semihigh-throughput adaptation of the NADH-coupled ATPase assay for screening small molecule inhibitors" *JoVE*
83. (2025) *Full-Length Text Journal of Virology* |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.