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=PMC12690436&blobtype=pdf | # Improving Efficiency in Dutch Cervical Screening by Genotyping: An Analysis of Real World Program Data
Ellen Olthof, | Adriaan, J Van Den Brule, Willem Melchers, Johannes Berkhof, Inge De Kok
## Abstract
High-risk human papillomavirus (hrHPV)-genotype specific risk stratification may improve cervical screening efficiency. This study evaluates the risks of cervical intraepithelial neoplasia (CIN), cancer and unnecessary referrals by hrHPV-genotype in cytology-positive (ASCUS+) women, using data from the Dutch population-based cervical screening program. Data from hrHPV+/ASCUS+ women screened between January 2017 and March 2018 were analyzed using the Dutch Screening and Pathology databases. Risks for CIN2+/3+, cancer, and unnecessary referral (i.e., without CIN2+) were evaluated by hrHPVgenotype (HPV16, HPV18, hrHPV-other (i.e., non-16/18 hrHPV), or mixed HPV16/18) using logistic regression, adjusted for age, laboratory (as proxy for region), sampling method (self-vs. clinician sampling), and stratified by age (< 50/≥ 50 years). HPV16+ women had 3.7 (CI: 3.42-3.95) and 4.6 (CI: 4.24-4.99) times higher risks of CIN2+ and CIN3+, respectively, compared to hrHPV-other. HPV18+ women had 1.6 (CI: 1.43-1.79) and 1.9 (CI: 1.68-2.18) times higher risks. The cervical cancer risk was tenfold higher for both HPV16 (OR: 9.85, CI: 6.50-14.95) and HPV18 (OR: 10.27, CI: 6.33-16.68). Women with HPV16 had 70% and HPV18 40% lower risks of unnecessary referral, compared to hrHPV-other. All risk differences between HPV16 or HPV18 and hrHPV-other were statistically significant in both age groups (< 50 and ≥ 50 years). Given the significantly higher risk of CIN2+/3+ and cancer associated with HPV16 and HPV18 and the reduced likelihood of unnecessary referrals compared to hrHPV other, these findings support the use of genotype-based colposcopy referrals in cervical screening to enhance screening efficiency.
## 1 | Background
The primary cause of cervical neoplasia is infection with the highrisk human papillomavirus (hrHPV), with types HPV16 and HPV18 responsible for nearly 70% of cervical cancer cases worldwide [1,2]. Since the introduction of primary HPV-based screening in the Netherlands in 2017, the number of unnecessary referrals has increased due to the high test sensitivity, but relatively low specificity of HPV testing [3]. Cytology is used as a triage test to improve specificity and reduce unnecessary referrals, but the number of unnecessary referrals remains high. Previous studies have demonstrated that incorporating HPV-genotype specific risk stratification into screening algorithms can improve colposcopy referral decisions and reduce overtreatment [4,5]. Therefore, HPV-genotyping could enhance screening accuracy by better predicting the development of precancerous lesions, ultimately improving the efficiency of screening programs. HPV16 infections, in particular, carry a significant higher risk of developing cervical lesions compared to other hrHPV types [6]. Although robust risk estimates from large-scale studies are available and many countries have adopted HPV-based primary screening, only few have implemented it nationwide together with population-based registries that allow evaluation in real-life screening contexts.
In the Dutch screening program, women are invited every 5 years between ages 30 and 60, where a cervical sample is obtained by a general practitioner or by self-sampling. Women positive for HPV are followed up by cytology triage. HrHPV positive women with ASCUS+ (i.e., atypical squamous cells of undetermined significance or worse) are referred to colposcopy, while those with no cytological intraepithelial lesions (i.e., NILM) are invited for repeat cytology after 6 months.
Until mid-2022, the Dutch cervical cancer screening program only used the outcome of the generic hrHPV test, despite the tests capability for genotype differentiation. Since July 2022, HPV-genotyping has been incorporated into the program, resulting in HPV-genotype specific colposcopy referral recommendations. Previously, all hrHPV-positive, women with ASCUS+ were directly referred to colposcopy. However, since July 2022, only women who test positive for HPV16 or HPV18 with ASCUS+ are referred to colposcopy. Women who test positive for other HPV-types (i.e., "hrHPV-other") are referred to colposcopy only in case of high-grade squamous intraepithelial lesion (i.e., HSIL) at baseline or ASCUS+ at repeat cytology. This strategy may contribute to reducing the number of unnecessary referrals while maintaining the same level of cervical cancer prevention [7]. The true effect in the population is, however, not known yet.
Next to reducing the number of unnecessary referrals, HPVgenotyping can also increase the effectiveness of the screening program. Mainly after the menopause, lesions are often deeper localized in the cervical canal and therefore more difficult to detect by cytology [8]. HPV16/18 positivity is associated with an increased risk of CIN in postmenopausal women with falsenegative or unreliable cytology results. Therefore, assessing the impact of HPV-genotyping is especially interesting for postmenopausal women.
The aim of our study was to assess the impact of adding HPV genotyping to the Dutch primary HPV-based screening program on its overall efficiency. We did so, by analyzing the risk of cervical intraepithelial neoplasia (CIN) and cancer, and unnecessary referrals by HPV-genotype in women with ASCUS or higher-grade cytological abnormalities, using observed population-based data. Additionally, we stratified the analyses by age (< 50/≥ 50 years) as a proxy for menopausal status.
## 2 | Methods
## 2.1 | Data Selection
A prospective cross-sectional cohort analysis was conducted using data from the Dutch screening database (ScreenIT) and the pathology database (Palga). In the Netherlands, all women aged 30-60 years are routinely invited for primary hrHPV-based cervical screening every 5 years. The invitation letter offers two options: women can either visit their general practitioner for a clinician-collected sample or request a self-sampling kit if they prefer not to undergo sampling at the GP. If a self-sample test is hrHPV-positive, women are required to visit their GP for cytology triage, as cytology cannot be performed on self-collected samples.
For the present study, we included all eligible women invited for the Dutch cervical screening program between January 1, 2017 and March 1, 2018 who had an hrHPV-positive sample with cytology classified as ASCUS+. HrHPV-positive samples were selected from ScreenIT and linked to genotyping data and histological data from Palga. All samples were analyzed using the Cobas 4800 HPV test (Roche Molecular Systems, Pleasanton, CA, USA). It is important to note that this cohort does not include vaccinated women. The follow-up for histological outcomes was until November 12, 2019.
## 2.2 | Data Indicators and Definitions
The following HPV-genotypes were identified: HPV16, HPV18, hrHPV-other (i.e., hrHPV-types other than HPV16 and 18) and mixed HPV16/18 (i.e., an infection with both HPV16 and HPV18). "HPV16" include infection with HPV16, with or without an "hrHPV-other" infection. "HPV18" include infection with HPV18, with or without an "hrHPV-other" infection. "Mixed HPV16/18" include infections with both HPV16 and HPV18, with or without "hrHPV-other." Main outcomes include the risk of CIN Grade 2 and higher (i.e., CIN2+) and Grade 3 and higher (i.e., CIN3+), cervical cancer, and the risk of an unnecessary referral. Unnecessary referral was defined as cases with no diagnosis of CIN2+. Covariates include age (30-34; 35-39; 40-44; 45-49; 50-54; 55-59; 60-64 years), screening laboratory (1-5), and sampling method (clinician collected or self-sampling). Screening laboratory was defined from 1 to 5 because there are five screening laboratories in the Netherlands who are responsible for analyzing the cervical samples. The screening laboratories represent different regions in the country.
## 2.3 | Data Analysis
The risk by HPV-genotype on CIN2+/3+, cervical cancer, and unnecessary referral in HPV+ and cytology-positive (ASCUS+) women was evaluated using logistic regression analysis. The analysis was adjusted for age (30-34; 35-39; 40-44; 45-49; 50-54; 55-59; 60-64 years), screening laboratory (1-5), and sampling method (clinician collected or self-sampling). Next, the risk by HPV-genotype on CIN2+/3+, cervical cancer, and unnecessary referral in HPV+ and ASCUS+ women was stratified by menopausal status (< 50/≥ 50 years). The analyses were conducted in IBM SPSS Statistics for Windows Version 25.0 (Armonk, NY: IBM Corp). A p value of < 0.05 was considered as statistically significant. compared in our analysis varied significantly in age distribution, sampling method used, and screening laboratory. The proportion of women aged 30-34 was highest (46.0%) among those with mixed HPV16/18 and lowest (26.8%) among those with hrHPV-other infections. Conversely, the proportion of women aged 60 and older was highest (4.5%) in those with hrHPV-other infections and lowest (2.5%) in those with mixed HPV16/18. Additionally, the use of self-sampling was most common in women with HPV16 infections (5.3%) and least common in women with HPV18 infections (3.3%).
## 3.1 | CIN2+/3+ and Cancer Risk by Genotype in ASCUS+ Women
The results of the univariate regression analysis are shown in Table S1. After adjustment for age, screening laboratory, and sampling method, a 3.7 times higher risk of CIN2+ (OR: 3.67, CI: 3.42, 3.95) and 4.6 times higher risk of CIN3+ (OR: 4.60, CI: 4.24, 4.99) was observed for HPV16, compared to women with hrHPV-other (Table 2). For women with HPV18, the risk of CIN2+ and 3+ was 1.6 (CI: 1.43, 1.79) and 1.9 (CI: 1.68, 2.18) times higher compared to women positive for hrHPV-other. For cervical cancer, the risk was 10 times higher for HPV16 (OR: 9.85, CI: 6.50, 14.95) and HPV18 (OR: 10.27, CI: 6.33, 16.68).
## 3.2 | CIN2+/3+ and Cancer Risk by Menopausal Status
A similar pattern has been observed for CIN2+ and 3+ risk by genotype in women before and after the menopause, although risk differences were smaller after menopause. For detection of CIN3+, before menopause, ORs between HPV-genotype groups ranged from 2.03 (95% CI: 1.75-2.34) for HPV18 to 4.82 (95% CI: 4.40-5.26) for HPV16; after menopause ORs ranged from 1.47 (95% CI: 1.06-2.03) to 3.77 (95% CI: 3.12-4.56; Figure 1 and Table S2). The differences in cervical cancer risk by genotype followed a similar pattern within each age group, although differences were larger after menopause. Before menopause, ORs compared with hrHPV-other types ranged from 9.43 (95% CI: 5.95-14.92) for HPV16 to 11.67 (95% CI: 5.58-24.39) for mixed HPV16/18 infections. After menopause, the corresponding ORs ranged from 11.04 (95% CI: 4.14-29.47) for HPV16 to 21.12 (95% CI: 4.91-90.94) for mixed HPV16/18 infections.
## 3.3 | The Risk of Unnecessary Referral
Women positive for HPV16 had a 70% lower risk (HPV16: OR: 0.27, CI: 0.25, 0.29; mixed HPV16/18: OR: 0.29, CI: 0.24, 0.35) of unnecessary referral compared to women with hrHPV-other (Figure 2 and Table S3). For HPV18-positive women, the risk was 40% (OR: 0.62, CI: 0.56, 0.70) lower compared to hrHPVother. After the menopause, the risk difference of unnecessary referral increased with 8% for HPV16 and 18% for HPV18 compared to HPV-positive ASCUS+ women before the menopause.
## 4 | Discussion
This study is the first to analyze CIN and cancer risks, as well as the risk of unnecessary referrals by HPV-genotype, among The regression analysis was adjusted for age category, screening laboratory, and sampling method. Significant results are indicated in bold (p < 0.005). All CIN2+ and 3+ risk differences between HPV-genotypes were significant, except between HPV16 and mixed HPV16/18 (p < 0.05). For cervical cancer, risk differences between genotypes HPV16, HPV18, and mixed HPV16/18 were not significant.
## FIGURE 1 |
The CIN2+ and 3+ and cervical cancer risk difference per genotype with ASCUS+ compared to hrHPV-other, stratified by menopausal status. These are results from the regression analysis, adjusted for age category, screening laboratory, and sampling method. All CIN2+ and 3+ risk differences between HPV-genotypes (within a menopausal status group) were significant, except between HPV16 and mixed HPV16/18 (p < 0.05). For cervical cancer, only risk differences between each HPV-genotype and hrHPV-other were significant. There were no significant cervical cancer risk differences between the genotypes HPV16, HPV18, and mixed HPV16/18. OR = odds ratio.
## 2 |
The risk of unnecessary referral (i.e., referrals with no diagnosis of CIN2+) per genotype with ASCUS+ compared to hrHPV-other, stratified by menopausal status (i.e., < 50/≥ 50 years). These are results from the regression analysis, adjusted for age category, screening laboratory, and sampling method. All differences between HPV-genotypes on the risk of an unnecessary referral stratified for menopausal status were significant, except between HPV16 and mixed HPV16/18 (p < 0.05). OR = odds ratio.
While our study found a higher relative risk for CIN3+ in HPV16-and HPV18-positive women compared to hrHPV-other, the absolute risks reported in the Norwegian study were lower. They reported absolute CIN3+ risks of 57.8%, 40.2% and 31.4% in HPV16, HPV18, and hrHPV-other ASCUS+ positive women, which correspond to relative risks of 1.8 for HPV16-and of 1.3 for HPV18-positive women compared to hrHPV-other [5]. However, both studies demonstrate a consistent trend of increased risks associated with HPV16 and HPV18 infections, underscoring their significant role in the development of CIN3+. Similar to our study, a Chinese study showed a relative risk of CIN2+ of 3.8 (95% CI: 3.8, 6.8) in HPV16-positive women compared to women positive for other genotypes [9]. The Portland-Kaiser cohort study found a 2.7 (95% CI: 1.0, 7.3) times higher risk of CIN3+ for HPV16-positive women compared to other oncogenic HPV types [10]. A subsequent analysis within this cohort reported a 10-year cumulative CIN3+ risk of 20.1% among HPV16-positive women and of 15.4% among HPV18positive women aged 30 years and older [11]. For hrHPV-other genotypes, the 10-year cumulative CIN3+ risk was only 1.8%. A Korean study showed a relative risk of CIN3+ of 3.9 for HPV16 and of 5.2 for HPV18 compared to other HPV-types in ASCUS women [12]. In addition, a meta-analysis showed no differences in HPV-distribution for normal cytology, ASCUS, LSIL, and CIN1, but they observed a steep increase in the HPV16 positivity rate from CIN1 through CIN2 to CIN3 and cervical cancer [6]. Although these studies also observed higher risks for HPV16 and HPV18 compared to hrHPV-other types, risks estimates are difficult to compare with the current study as none of the studies were performed in a national screening program, different outcome measures were used, not all studies were performed in an ASCUS+ population and differences in prevalent HPV types between populations may exist.
The difference in CIN2/3+ risk by genotype was slightly larger in HPV+/ASCUS+ women aged < 50 and slightly smaller in women aged ≥ 50, compared to all HPV-positive women. This difference by age can be explained by several factors. First, the risk of high-grade lesions is lower at older ages, due to early detection through screening [3,13]. Second, the sensitivity of cytology is decreased after menopause (mainly for lower-grade lesions) [14][15][16]. The postmenopausal ASCUS+ women can therefore be a selected higher risk population (compared to premenopausal ASCUS+ women), in which HPV genotype is less predictive for CIN or cervical cancer risk. The risk of cervical cancer was higher in postmenopausal women compared to all HPV+/ASCUS+ women. However, no risk differences were observed between the HPV16 and HPV18 genotypes, despite expectations that HPV16 carries a higher risk of cervical cancer. This discrepancy may be attributed to the presence of adenocarcinomas (which are more often missed by cytology), which are often associated with HPV18. Additionally, as this study represents the first round of HPV-based screening with a test that is more sensitive than cytology, it is important to note that the results may not yet reflect the long-term or steady-state prevalence of cervical cancer in the screening population.
Previously, a Dutch modeling study showed a 32% decrease in unnecessary referrals for referring HPV16 and HPV18 women with ASCUS+ directly to colposcopy and only referring hrHPV-other women with HSIL to colposcopy at baseline and with ASCUS+ at repeat cytology after 6 months, compared to referring all HPV +/ASCUS+ women [7]. In the current observational study, we observed that, compared to women infected with "hrHPV-other," HPV16 had a 70% lower risk of unnecessary referral and in case of HPV18 this risk was 40% lower. Although these two studies are difficult to compare, both indicate that changing the referral recommendations for colposcopy based on genotype as recently has been implemented in our screening program seems valid.
Because we used nationwide data from an organized primary HPV-based screening setting, we were able to adjust for multiple important confounders and could compute a robust estimation of the CIN2+/3+, cervical cancer, and unnecessary referral-risk by HPV genotype. Furthermore, we measured both the benefits and the harms, while most studies only focus on the benefits (i.e., CIN and cervical cancer risk by HPV type). Still, our study also has some limitations. We were only able to investigate the CIN-and cancer-risk in HPV-positive ASCUS+ women, as those women are referred to the gynecologist according to the screening program. It would be interesting to investigate the CIN-and cancer-risk in all HPV-positive women to quantify the number of lesions and cancer cases that might be missed using HPV-genotyping with cytology triage. Furthermore, the findings of our study might be different in other countries, as there is some geographic variation in most prevalent HPV types across countries [17]. For example, HPV16 accounts for 48% of the cervical cancer cases in Africa compared to 66% in Europe. So, this may require different screening strategies as opposed to our country where the impact of HPV16 on the population is higher. In addition, our study is performed in an unvaccinated population. As of 2023, vaccinated cohorts (vaccinated against HPV16 and HPV18) have entered the screening program. The risk of cervical lesions will be reduced in vaccinated women compared to unvaccinated women, and hence, fewer screening rounds and longer screening intervals might be preferred [18]. Also, the rate of false-positive cytology results (i.e., cases in which hrHPV-positive women are referred to colposcopy but no CIN2+ lesion is detected) is expected to be higher in vaccinated cohorts due to their lower underlying cancer risk, which advocates adapted screening strategies. A Dutch modeling study has shown that the number of CIN2+ diagnosed decreases in these vaccinated cohorts and recommends longer screening intervals for low-risk women (i.e., women vaccinated against HPV16/18) to maintain the program's efficiency [19]. In our study, we showed a four times higher CIN2+ risk of HPV16. The impact of our study might be lower in vaccinated cohorts as lower rates of HPV16 are expected. However, stratification by genotype for colposcopy referral is still recommended given the large risk differences.
In conclusion, our findings support the use of hrHPVgenotyping to guide colposcopy referral decisions in cervical screening and highlight the potential for enhanced screening efficiency across all age groups.
## References
1. Smith, Lindsay, Hoots (2007) "Human Papillomavirus Type Distribution in Invasive Cervical Cancer and High-Grade Cervical Lesions: A Meta-Analysis Update" *International Journal of Cancer*
2. Bosch, Burchell, Schiffman (2008) "Epidemiology and Natural History of Human Papillomavirus Infections and Type-Specific Implications in Cervical Neoplasia" *Vaccine*
3. Aitken, Van Agt, Siebers (2019) "Introduction of Primary Screening Using High-Risk HPV DNA Detection in the Dutch Cervical Cancer Screening Programme: A Population-Based Cohort Study" *BMC Medicine*
4. Xue, Gao, Zheng (2021) "Use of Extended HR-HPV Genotyping in Improving the Triage Strategy of 2019 ASCCP Recommendations in Women With Positive HR-HPV Diagnosis and Simultaneous LSIL Cytology Results" *Journal of Cancer*
5. Hashim, Engesaeter, Baadstrand Skare (2020) "Real-World Data on Cervical Cancer Risk Stratification by Cytology and HPV Genotype to Inform the Management of HPV-Positive Women in Routine Cervical Screening" *British Journal of Cancer*
6. Guan, Howell-Jones, Li (2012) "Human Papillomavirus Types in 115,789 HPV-Positive Women: A Meta-Analysis From Cervical Infection to Cancer" *International Journal of Cancer*
7. Kaljouw, Jansen, Aitken et al. (2021) "Reducing Unnecessary Referrals for Colposcopy in hrHPV-Positive Women Within the Dutch Cervical Cancer Screening Programme: A Modelling Study" *Gynecologic Oncology*
8. Campaner, Fernandes (2024) "Discussion on Cervical Cytology in Postmenopausal Women" *Minerva Obstetrics and Gynecology*
9. Zhou, Chen, Strickland et al. (2021) "Prevalence of Genotype-Specific Human Papillomavirus in Cytology Specimens and Cervical Biopsies, and Its Implication in Cervical Cancer Risk Stratification: A Retrospective Study of 10647 Cases" *Journal of Cancer*
10. Schiffman, Glass, Wentzensen (2011) "A Long-Term Prospective Study of Type-Specific Human Papillomavirus Infection and Risk of Cervical Neoplasia Among 20,000 Women in the Portland Kaiser Cohort Study" *Biomarkers & Prevention*
11. Khan, Castle, Lorincz (2005) "The Elevated 10-Year Risk of Cervical Precancer and Cancer in Women With Human Papillomavirus (HPV) Type 16 or 18 and the Possible Utility of Type-Specific HPV Testing in Clinical Practice" *Journal of the National Cancer Institute*
12. Cho, Park, Woo et al. (2024) "Evaluation of Clinical Usefulness of HPV-16 and HPV-18 Genotyping for Cervical Cancer Screening" *Journal of Gynecologic Oncology*
13. Rijkaart, Berkhof, Rozendaal (2012) "Human Papillomavirus Testing for the Detection of High-Grade Cervical Intraepithelial Neoplasia and Cancer: Final Results of the POBASCAM Randomised Controlled Trial" *Lancet Oncology*
14. Gustafsson, Sparen, Gustafsson (1995) "Low Efficiency of Cytologic Screening for Cancer In Situ of the Cervix in Older Women" *International Journal of Cancer*
15. Gyllensten, Lindell, Gustafsson et al. (2010) "HPV Test Shows Low Sensitivity of Pap Screen in Older Women" *Lancet Oncology*
16. Gyllensten, Gustavsson, Lindell et al. (2012) "Primary High-Risk HPV Screening for Cervical Cancer in Post-Menopausal Women" *Gynecologic Oncology*
17. De Sanjose, Quint, Alemany (2010) "Human Papillomavirus Genotype Attribution in Invasive Cervical Cancer: A Retrospective Cross-Sectional Worldwide Study" *Lancet Oncology*
18. El-Zein, Richardson, Franco (2016) "Cervical Cancer Screening of HPV Vaccinated Populations: Cytology, Molecular Testing, Both or None" *Journal of Clinical Virology*
19. Kaljouw, Jansen, Aitken et al. (2022) "Shift in Harms and Benefits of Cervical Cancer Screening in the Era of HPV Screening and Vaccination: A Modelling Study" *BJOG* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737563&blobtype=pdf | # Correction: Sachse et al. From Entry to Outbreak in a High School Setting: Clinical and Wastewater Surveillance of a Rare SARS-CoV-2 Variant. Viruses 2025, 17, 477
Sven Sachse, Ivana Kraiselburd, Olympia Anastasiou, Carina Elsner, Sarah Goretzki, Stefan Goer, Michael Koldehoff, Alexander Thomas, Jens Schoth, Sebastian Voigt, Rudolf Ross, Ulf Dittmer, Folker Meyer, Ricarda Schmithausen
## Missing Citation
In the original publication [1], eleven references were not cited. Among these, references 15-22 are newly added, while references 9, 14, and 23 (which corresponds to the original reference 25) are pre-existing entries that were not previously cited. Due to an oversight by the authors, some numerals appearing in the Materials and Methods as well as the Results sections should have been referenced, but were not marked in square brackets or included in the reference list.
The following references should be included: Section 2.3.2: "The sequence editing, generation, and translation of the consensus sequence into the corresponding amino acid sequence were performed using Geneious Pro 5.1.7. A multiple-sequence alignment was performed by Clustal Omega version 2.1 using the SARS-CoV-2 isolate Wuhan-HU-1 (NC_045512.2) as a reference. The visualization of the Clustal Omega alignment was carried out with a Highlighter analysis using the Los Alamos National Laboratory pathogen database. Initial manual lineage prediction was performed based on mutations identified in a multiple-sequence alignment using out-break.info [14]. Whole-genome sequencing (see below) later confirmed the initial lineage assignment."
"A phylogenetic tree was inferred using the Maximum Likelihood method and the Tamura-Nei nucleotide substitution model based on the receptor binding domain of the viral spike gene (nucleotides 963 to 1737). The tree was drawn to scale, with branch lengths measured in the number of substitutions per site. Statistical robustness was tested using the bootstrap approach with 1000 replicates. Analysis was conducted in MEGA,version 11 [15]."
Section 2.3.3: "Reads with an average Phred quality below 20 and a length below 30 base pairs were excluded, enabling downstream analysis. Subsequently, data analysis was performed with the UnCoVar bioinformatic pipeline for reconstructing whole viral genomes [16]. UnCoVar performed a series of QC steps, initially attempted de novo assembly, and then resorted to co-assembly for recalcitrant samples; it subsequently used pangolin [17] and Kallisto [18] matching to GISAID [19] to obtain lineage calls. Additionally, Freebayes [20], Delly [21],and Varlociraptor [22] were utilized for variant calling." Section 2.3.4: "During this study, wastewater samples were collected within the metropolitan area Ruhr and processed for SARS-CoV-2 detection, as described [23]. The sampling area comprises both the high school and maximum care hospital described in this work."
"Viral variants were identified with a modified version of UnCoVar [16], based on the detection of variant-specific mutations without the attempt at genome assembly." Section 3.1: "This characteristic pattern of amino acid exchanges had already been reported for variants from the former SARS-CoV-2 B.1.640 lineage, which was renamed to B. 1.640.1 [9]. Analyses of full-lengths sequences obtained by next-generation sequencing confirmed and extended the findings from [9]. Additionally, the formation of a so-called monophyletic group in a tree based on partial SARS-CoV-2 spike gene RBD sequences, which reconstructs the evolutionary history of the viral isolates (Figure 2), shows that the children and their relatives were infected in a single-source outbreak by a yet rare SARS-CoV-2 B.1.640.1 variant, which had possibly originated in the Republic of Congo [9]."
With the above corrections, the reference citation numbers 15-32 have been changed to 15-40, respectively. The authors state that the scientific conclusions are unaffected. This correction was approved by the Academic Editor. The original publication has also been updated.
## References
1. Sachse, Kraiselburd, Anastasiou et al. (2025) "From Entry to Outbreak in a High School Setting: Clinical and Wastewater Surveillance of a Rare SARS-CoV-2 Variant" *Viruses*
2. "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=PMC12548463&blobtype=pdf | # LPS/TLR4-activated M1-polarized macrophage-derived exosomes enhance IBV vaccine efficacy in chickens
Jun Zhou, Shikai Cai, Hongbin Huang, Fan Yang, Kexin Pan, Zhaoyang Sun, Yun Fan, Feng Wen, Limei Qin, Yun Zhang
## Abstract
Infectious bronchitis virus (IBV) imposes substantial economic losses on poultry production due to extensive serotypic diversity and limited cross-protection conferred by conventional vaccines. This study evaluated exosomes derived from M1-polarized chicken macrophages (HD11 M1 -exo) as a novel adjuvant for IBV vac cination. HD11 M1 -exo, isolated from lipopolysaccharide (LPS)-activated HD11 macro phages via ultracentrifugation, demonstrated significant immunomodulatory properties across multiple experimental systems. In vitro analyses demonstrated that HD11 M1 -exo enhances macrophage phagocytosis and promotes cellular immune activation via the LPS/TLR4 signaling pathway. In ovo analyses showed that HD11 M1 -exo pretreatment upregulates tracheal expression of IL-1β, IL-2, IL-4, IFN-γ, TNF-α, and TLR4 at different time points, thereby enhancing viral resistance and reducing pathological damage. In chickens, HD11 M1 -exo administration elevated CD80/CD86 and TGF-β4 expression in respiratory tissues and increased secretory immunoglobulin A (IgA) levels in lacrimal fluid. When co-administered with the H120 vaccine, HD11 M1 -exo significantly improved both humoral immunity (elevated serum IgY and mucosal IgA) and cellular responses (increased CD80/CD86 expression), outperforming commercial adjuvants in efficacy. Following the viral challenge, HD11 M1 -exo + H120-immunized chickens exhibited significantly reduced viral loads and attenuated histopathological lesions compared to controls. These results collectively suggest that exosome-based formulations may serve as promising adjuvants for enhancing the immunogenicity and protective efficacy of poultry vaccines. IMPORTANCE Infectious bronchitis virus (IBV) causes significant global economic losses in the poultry industry despite extensive vaccination programs. Current vaccines often fail to elicit sufficient mucosal and cellular immunity, which are critical for protection against the virus. Although commercial adjuvants have been employed to enhance vaccine efficacy, many exhibit limitations in eliciting comprehensive immune responses. In this study, we comprehensively evaluated HD11 M1 -exo as a novel adjuvant for IBV vaccines across in vitro, in ovo, and in vivo models for the first time. Our results dem onstrate that HD11 M1 -exo enhances macrophage function via lipopolysaccharide (LPS)/ TLR4 signaling, upregulates key cytokines and immune markers in embryonic tissues, and significantly boosts cellular, humoral, and mucosal immunity when co-administered with live-attenuated IBV vaccines, outperforming commercial adjuvants. Importantly, this adjuvant strategy significantly enhanced protective efficacy in challenged chickens. This study provides a foundation for developing exosome-based adjuvants that could advance poultry vaccination strategies against IBV and other avian respiratory patho gens.
A vian infectious bronchitis (IB), an acute and highly contagious disease caused by infectious bronchitis virus (IBV), is characterized by respiratory tract lesions, nephritis, and reproductive tract abnormalities resulting in reduced egg production in layers, manifesting as a complex syndrome with substantial economic consequences for the global poultry industry (1). IBV is an enveloped, pleomorphic virus belonging to the genus Gammacoronavirus within the family Coronaviridae, possessing a linear, single-stranded, positive-sense RNA genome of approximately 27.5-28 kb in length (2). The remarkable genetic plasticity of IBV facilitates ongoing evolutionary adaptation through both point mutations in the spike glycoprotein and homologous recombination events, resulting in the persistent emergence of novel genetic variants and epidemiolog ically distinct serotypes in the field. Moreover, the serotype-specific immunity elicited by a single IBV serotype confers limited heterologous protection against antigenically diverse field strains (3). This immunological constraint poses substantial challenges for the development of broadly protective vaccines, complicating effective control strategies against IBV with its considerable antigenic and epidemiological heterogeneity.
Exosomes are small cell-derived extracellular vesicles originating from multivesicular bodies, ubiquitous in biological fluids, including plasma, lymph, saliva, semen, urine, and cerebrospinal fluid (4,5). These membranous nanovesicles typically measure 40-160 nm in diameter with a characteristic buoyant density of 1.13-1.19 g/mL under ultracentri fugation conditions. Transmission electron microscopy demonstrates their distinctive cup-shaped or spherical morphology with lipid bilayer boundaries (6). Exosomes carry a multifaceted molecular payload from progenitor cells, including functional proteins, bioactive lipids, and regulatory miRNAs, which are selectively packaged and can be transferred to recipient cells through membrane fusion or endocytic uptake mechanisms. These biologically active nanovesicles serve as critical mediators in numerous physio logical and pathological processes by orchestrating intercellular signaling networks, regulating immune system activation and suppression, directing cellular motility and tissue infiltration, facilitating neovascularization, and modulating pathogenic processes in various disease conditions (7,8). Macrophages, serving as ubiquitous sentinel cells throughout tissues, constitute essential components of tissue homeostasis and host defense mechanisms against pathogen invasion (9). Tissue-resident macrophages exhibit remarkable plasticity and polarize into functionally distinct phenotypes in response to microenvironmental cues, conventionally categorized as classically activated (M1) macrophages driving pro-inflammatory responses and alternatively activated (M2) macrophages associated with immunoregulation and tissue repair (10). Experimental studies have established that macrophage polarization is triggered by specific molec ular stimuli: lipopolysaccharide (LPS) induces M1 phenotype development, whereas interleukin-4 (IL-4) directs M2 phenotype differentiation (11). During immune defense responses, M1 macrophages demonstrate bifunctional effector mechanisms. These cells exert direct antimicrobial activity through the generation of reactive oxygen species that compromise microbial integrity and through pathogen phagocytosis. Concurrently, they mediate indirect immunoregulatory functions via cytokine secretion and associated signaling pathways (12).
Exosomes secreted by M1 macrophages transport bioactive molecules from progenitor cells, mediating immunostimulatory activities. Hong et al. showed that dsRNA virus-infected chicken macrophages activated via TLR3 ligands coordinate innate immune responses in naïve macrophages and T lymphocytes through NF-κB signaling cascade modulation (13). Furthermore, Hong et al. systematically characterized the functional capacity of macrophage-derived exosomes in avian immune responses at the cellular level. Their findings documented that LPS-stimulated exosomes possess minimal immunogenicity, increased cellular uptake, and immune-modulating character istics, promoting immune activation and cytokine secretion through MyD88-depend ent NF-κB signaling, thus establishing them as potent immunostimulatory agents (14). Srinivasan et al. elucidated that exosomes secreted by poly(I:C)-activated macrophages drive M1 phenotype differentiation in murine lymph node-resident macrophages and trigger NF-κB signaling cascade activation (15). Tang et al. documented that exo somes secreted by LPS-activated human monocytes upregulate CCL2, ICAM-1, and IL-6 expression in endothelial cells mediated via the NF-κB signaling cascade (16). Collec tively, exosomes derived from LPS-polarized M1 macrophages demonstrate minimal intrinsic immunogenicity, enhanced cellular internalization, and potent immunomodu latory properties. Through activation of specific intracellular signaling cascades, these vesicles augment immune responses and cytokine secretion. These biological nanopar ticles represent promising vaccine adjuvant candidates with target specificity, preci sion delivery capability, and immunostimulatory efficacy. In the present investigation, LPS stimulation was utilized to induce M1 phenotype differentiation in avian HD11 macrophages, with subsequent exosome purification from culture supernatants through differential ultracentrifugation. We developed a comprehensive evaluation framework integrating in vitro cellular, embryonated egg, and in vivo avian models to systematically evaluate the immunostimulatory properties and mechanistic basis of M1-polarized avian macrophage-derived exosomes (HD11 M1 -exo) and quantify their adjuvant synergy with attenuated IB vaccine, benchmarked against commercial immunopotentiators.
## MATERIALS AND METHODS
## Cells, chicken embryos, and animals
The HD11 cell line was maintained by the Guangdong Provincial Key Laboratory of Animal Molecular Design and Precise Breeding, Foshan University. One-day-old chicks and fertilized chicken embryos were purchased from Nanhai Poultry Breeding Co., Ltd., Foshan City, Guangdong Province.
## Viruses and reagents
The IBV attenuated vaccine (H120 strain) and the IBV-HX strain (GenBank accession number: OP846990) are preserved by the Guangdong Provincial Key Laboratory of Animal Molecular Design and Precision Breeding at Foshan University. LPS was purchased from Shanghai Biyuntian Biotechnology Co., Ltd. (Shanghai, China); Chicken Immunoglobulin A (IgA) enzyme-linked immunosorbent assay (ELISA) kit was obtained from Quanzhou Ruixin Biotechnology Co., Ltd. (Fujian, China); aluminum hydroxide adjuvant was acquired from Beijing Boao Long Biotechnology Co., Ltd. (Beijing, China); Commercial avian adjuvant (IMS 1313 VG N ST, referred to as 1313) was sourced from SEPPIC Special Chemicals Co., Ltd. (Shanghai, China); Chicken Infectious Bronchitis Antibody Test Kit (Indirect ELISA) was purchased from Shenzhen Zhenrui Biotechnology Co., Ltd. (Shenzhen, China).
## Exosome isolation and identification
To obtain exosomes from HD11 M1 -exo, when the density of HD11 cells reaches 80% and they are in good growth condition, wash the cells three times with phosphate-buf fered saline (PBS) to remove the original culture medium, then replace it with RPMI 1640 complete medium containing 1% exosome-depleted serum. Randomly select one bottle of cells as a blank control group, and add LPS to the remaining culture bottles to a final concentration of 1 µg/mL. Incubate all culture bottles at 37°C with 5% CO 2 for 24 h. For the isolation of HD11 M1 -derived exosomes, avian macrophage HD11 cells were maintained until reaching 80% confluence and exhibiting log-phase growth. Cultures were rinsed thrice with sterile PBS (pH 7.4) to eliminate residual growth medium, followed by supplementation with RPMI 1640 medium containing 1% exosome-depleted FBS. Cell cultures were designated as either non-stimulated controls or experimental groups that received LPS stimulation (1 µg/mL). All cultures were subsequently incubated at 37°C in a humidified atmosphere containing 5% CO 2 for 24 h. After incubation, collect the HD11 cells from both the control group and the treated group, as well as the supernatant from the treated group cells. Use quantitative PCR (qPCR) to identify typical markers of M1 macrophages (iNOS, IL-6, IL-12, IL-23, CD80, CD86) to confirm macrophage polarization to the M1 type. Then, use ultracentrifuga tion to isolate exosomes from the cell supernatant, and conduct transmission electron microscopy, particle size analysis, and western blot experiments to characterize the extracted exosomes.
## Cellular experiment design
Avian macrophage HD11 cells were seeded in 96-well microplates (1 × 10⁵ cells/well) and maintained until reaching 80% confluence under standard culture conditions. Cells were subsequently assigned to either a negative control (NC) or exosome-treated (Exo) group, with triplicate wells per group. The Exo group received 20 µg/mL HD11 M1 -derived exosomes in serum-free medium, while the NC group was administered an equivalent volume of sterile PBS (pH 7.4). Following 24 h incubation at 37°C in a humidified 5% CO 2 atmosphere, culture supernatants were aspirated. For the neutral red uptake assay, a 0.1% (wt/vol) neutral red solution was freshly prepared by dissolving neutral red powder in sterile DPBS and filtering through a 0.22 µm membrane. Add 100 µL to each well to achieve a final concentration of neutral red at 1 µg/mL in each well. Incubate the culture plates in a 37°C incubator for an additional 1 h. Subsequently, wash the wells three times with PBS and add 100 µL of cell lysis buffer (a mixture of ice acetic acid and anhydrous ethanol in a 1:1 ratio) to each well. Gently shake for 15 s and let stand at room temperature for 2 h to allow complete cell lysis. Finally, measure the absorbance of each well at a wavelength of 540 nm.
HD11 cells in good growth condition were seeded into a 6-well culture plate. The experiment was divided into an NC group and an Exo group, with three independent repeats for each group. When the cell confluence reached approximately 80%, the Exo group was stimulated with 50 µg/mL of HD11 M1 -exo for 48 h, while the NC group was treated with an equal volume of PBS. The transcriptional levels of the LPS/TLR4 signaling pathway-related genes CD14, TLR4, TRAM, TRAF3, IRF3, and CD80 were detected by qPCR. Expression levels of TLR4, IRF3, and CD80, key proteins in the LPS/TLR4 signaling pathway, were detected using western blotting analysis.
## Chicken embryo experimental design
Chicken embryos at 18 days of age were divided into four groups (see Fig. 3): NC group, Exo group, IBV group, and Exo+IBV group. The Exo and Exo+IBV groups received an injection of HD11 M1 -exo (50 µg per embryo) into the allantoic cavity, while the NC and IBV groups received an equal volume of PBS. The injection holes were sealed with wax. The embryos were returned to the incubator for an additional 24 h. Subsequently, the IBV and Exo+IBV groups were inoculated with IBV strain (0.2 mL EID50 per embryo) into the allantoic membrane, while the NC and Exo groups received an equal volume of PBS. The embryos were then placed back into the chicken embryo incubator for further cultivation. At 24, 48, and 72 h after inoculation, the embryos were dissected, and the clinical symptoms and pathological changes in each group were observed. The tracheas of the embryos were collected, and qPCR was used to detect changes in the levels of tracheal cell factors.
## Animal experiment design
During this experiment, the immunization method was via eye drops and nasal drops. The dosage of HD11 M1 -exo was 50 µg per chicken, the dosage of aluminum hydroxide adjuvant was 31.4 µg per chicken, the dosage of the 1313 adjuvant was 0.05 mL per chicken, and the dosage of the H120 IB attenuated vaccine was 0.2 mL EID50.
To evaluate the immunostimulatory potential of HD11 M1 -exo in vivo, day-old chicks (n = 50) were randomly allocated into two experimental groups (n = 25 per group) as illustrated in Fig. 5. Chicken received oculonasal administration of either HD11 M1 -exo (Exo group) or physiological saline (NC group) at days 1, 3, 5, and 14 post-hatch.
The clinical status of the chicken was monitored daily throughout the experimental period. At 7, 14, 21, 28, and 35 days post-vaccination (dpv), five chickens were randomly selected from each group at each time point for collection of lachrymal fluid, followed by euthanasia and subsequent tissue sampling. To evaluate the adjuvant effect of HD11 M1exo on immune responses to the infectious bronchitis (IB) live attenuated vaccine, day-old chicks (n = 100) were randomly allocated into four groups (n = 25 per group) as detailed in Fig. 8. H120 group: vaccinated with H120 strain alone; HD11+Exo group: vaccinated with H120 strain combined with HD11 M1 -exo; H120+AL group: vaccina ted with H120 strain adjuvanted with aluminum hydroxide; and H120+1313 group: vaccinated with H120 strain adjuvanted with Montanide IMS 1313. Initial vaccination was performed at day 1 post-hatch via the oculonasal route. At 3 and 5 dpv, chickens received a second and third dose of their respective adjuvants (HD11 M1 -exo, saline, aluminum hydroxide, or Montanide IMS 1313) via the same route. A booster vaccination following the identical prime-vaccination protocol was administered at 14 dpv. Clinical status was monitored daily throughout the study period. At 7, 14, 21, 28, and 35 dpv, lachrymal fluid samples were collected from five randomly selected chickens per group, after which the chickens were euthanized for tissue collection.
To assess the protective efficacy of HD11 M1 -exo as an adjuvant during IBV challenge, day-old chicks were assigned to six experimental groups, and the NC group was housed separately under identical environmental conditions. Primary vaccination was adminis tered at day 1 post-hatch with the following regimens (Fig. 12): the H120 group was inoculated only with the H120 vaccine; the HD11+exo group was inoculated with both the H120 vaccine and HD11 M1 -exo; the H120+AL group received the H120 vaccine with aluminum hydroxide adjuvant; the H120+1313 group received the H120 vaccine with the 1313 adjuvant; the NC group and IBV group were inoculated with an equal volume of physiological saline. At 14 dpv, all groups of chicks received a booster immunization following the same protocol as at 1 dpv. Seven days after the booster immunization, the IBV challenge was conducted; except for the NC group, each chick in the remaining groups was inoculated with 0.1 mL of EID50 IBV virus solution, while the NC group received 0.1 mL of PBS. Throughout the experiment, the clinical manifestations of the chicks in each group were observed daily, and samples were collected on the 5th day post-infection (designated as 5 dpi).
## Real-time quantitative PCR
For quantitative reverse transcription PCR (RT-qPCR) analysis, expression levels of LPS/ TLR4 signaling pathway genes (CD14, TLR4, TRAM, TRAF3, IRF3, and CD80) were quantified in vitro using cultured cells. In the embryonated egg model, TNF-α mRNA expression was assessed in embryonic tracheal tissues at predetermined time points. In the in vivo study, mRNA expression of LPS/TLR4 pathway components was measured in tracheal tissues at 21 dpv. Additionally, expression of mucosal immune factor (TGF-β4) and co-stimulatory molecules (CD80 and CD86) was analyzed in tracheal and Harder ian gland samples at various sampling points. Viral RNA load in tracheal tissues was determined at 5 dpi. Primer sequences used for gene amplification are listed in Table 1.
## Tear fluid IgA antibody testing
Lachrymal fluid was collected from five chickens per group at 7, 14, 21, 28, and 35 dpv. Briefly, lacrimation was induced by applying approximately 1 mg of molecular biology-grade sodium chloride crystals to the bilateral ocular surfaces. Tears were collected within 30-45 s post-stimulation using a calibrated micropipette. Samples were immediately transferred to sterile microcentrifuge tubes containing protease inhibitor cocktail (1:100 vol/vol) and stored at -20°C until analysis. Tear IgA concentrations were determined using a commercial chicken-specific IgA ELISA kit according to the manufac turer's instructions.
## Serum antibody level testing
At 7, 14, 21, 28, and 35 dpv, blood samples were collected from chickens in each experimental group via wing vein puncture. Samples were allowed to clot at room temperature for 2 h and subsequently centrifuged at 5,000 × g for 15 min to obtain serum fractions. Serum antibody titers against avian IBV were determined using a commercially available indirect ELISA kit (Company, Location) per manufacturer's specifications.
## Histopathological observations
Collect tracheal tissues from different groups of chickens in the immunological detoxification protection experiment for histopathological analysis. The tissues are fixed in 4% formalin solution. After dehydration with ethanol, the paraffin-embedded tracheal tissues are sectioned to a thickness of 5 microns and stained with hematoxylin and eosin. Following staining, the sections are dehydrated with absolute ethanol and washed with xylene for transparency. Finally, transparent sections are mounted with neutral resin and examined under a microscope to evaluate the lesions in each group. The observations are photographed and recorded.
## Statistical analysis
The experimental data are presented as the means ± standard deviation. Statistical analyses were performed using GraphPad Prism (9.5.1), with comparisons made using a one-way analysis of variance or multiple comparisons. Significance levels are indicated as follows: different letters indicate that the differences are significant at the P < 0.05 level. Specifically, treatment A (a) does not differ significantly from treatment B (ab), but it does differ significantly from treatment C (b); treatment B (ab) also does not differ significantly from treatment C (b). Data visualization was accomplished using GraphPad Prism (9.5.1) software.
## RESULTS
## Isolation and characterization of exosomes derived from M1-type macro phages
To determine whether macrophages polarize to the M1 type after stimulation with LPS, qPCR was used to detect typical markers of M1 type in macrophages from the stimulated group (LPS group) and the control group (NC group) after 24 h of stimulation. The results indicated that, compared to the NC group, the expression levels of iNOS, IL-6, IL-12, IL-23, CD80, and CD86 in the LPS group were significantly elevated (P < 0.05) (Fig. 1A), demonstrating that macrophages had indeed polarized to the M1 type following LPS stimulation (HD11 M1 -exo).
The HD11 M1 -exo were separated using ultrahigh-speed centrifugation, and their characterization was identified through transmission electron microscopy, particle size analysis, and western blot. The experimental results indicate that HD11 M1 -exo has an intact structure and exhibits cup-shaped or spherical forms, with particle sizes primarily distributed between 30 and 200 nm, in accordance with the characteristics of exosomes (Fig. 1B andC). Western blot analysis detected positive signals for the typical exosomal markers HSP70 (70 kDa) and ALIX (105 kDa) in HD11 M1 -exo (Fig. 1D). These results suggest that successful separation of HD11 M1 -exo has been achieved.
## HD11 M1 -exo enhances the immune activity of macrophages through activating the LPS/TLR4 signaling pathway
After 24 h of stimulation with HD11 M1 -exo, compared to the NC group, macrophages in the exosome group showed a significant increase in their phagocytic capacity (P < 0.05) (Fig. 2A). Additionally, quantitative analysis revealed that 48 h after HD11M1exo stimulation, expression levels of core LPS/TLR4 signaling pathway components-CD14, TLR4, TRAM, TRAF3, IRF3, and CD80-were significantly upregulated in the exosome-treated group compared to controls (P < 0.05) (Fig. 2B). Correspondingly, western blotting analysis confirmed significant upregulation of key regulatory proteins within the LPS/TLR4 pathway, including TLR4, IRF3, and CD80 (P < 0.05) (Fig. 2C). The proposed molecular mechanism underlying these observations is graphically summar ized in Fig. 2D. These experimental results indicate that HD11 M1 -exo can activate the LPS/TLR4 signaling pathway in macrophages, enhancing their phagocytic ability and thus boosting their immune activity.
## HD11 M1 -exo pretreatment enhanced antiviral resistance to IBV infection in embryonated chicken eggs
The experimental design of HD11 M1 -exo resisting IBV infection in chicken embryos is shown in Fig. 3.
After stimulating the chicken embryos with HD11 M1 -exo for 24 h, IBV was introduced. At 24, 48, and 72 h post-infection, the IBV group showed 2, 1, and 1 chicken embryos, respectively, exhibiting symptoms of IBV infection (marked by blue arrows), while the NC group, Exo group, and Exo+IBV group showed no signs of lesions (Fig. 4A). Cytokine expression in tracheal tissues of chicken embryos was evaluated by qPCR at 24, 48, and 72 h following IBV infection with or without HD11M1-exo pretreatment. At 24 h postinfection, TNF-α and TLR4 expression levels were significantly elevated in the exosometreated group (Exo) compared to the negative control group (NC) (P < 0.05). Similarly, the exosome-pretreated and subsequently infected group (Exo+IBV) exhibited significantly higher expression of these genes compared to the IBV-only group (P < 0.05) (Fig. 4B, a, andb). At 48 h, IL-2, IL-4, and TNF-α were significantly upregulated in the Exo group compared to NC controls (P < 0.05), with the Exo+IBV group also demonstrating significantly enhanced expression of these cytokines relative to the IBV group (P < 0.05) (Fig. 4B, c, d, ande). By 72 h post-infection, the Exo group showed significant upregula tion of IL-1β, IFN-γ, and TNF-α compared to the NC group (P < 0.05), while the Exo+IBV group maintained significantly higher expression of IFN-γ and TNF-α compared to the IBV group (P < 0.05) (Fig. 4B, f, g, and h).
## HD11 M1 -exo potentiates cellular immune responses in chickens
The experimental design for the immune activation induced by HD11 M1 -exo in chickens is shown in Fig. 5.
Quantitative PCR analysis was performed to assess the mRNA expression of costimulatory molecules CD80 and CD86 in tracheal and Harderian gland tissues from HD11 M1 -Exos-treated chickens at 7, 14, 21, 28, and 35 dpv. Expression of both CD80 and CD86 was significantly upregulated (P < 0.05) in both tissues from the HD11 M1 -Exos group compared to the negative control group at all examined time points (Fig. 6A through D). This consistent upregulation throughout the 35-day observation period suggests sustained immunomodulatory effects of HD11 M1 -Exos on co-stimulatory molecule expression.
To elucidate the molecular mechanisms underlying the immunostimulatory effects of HD11 M1 -exo in chickens, the expression of genes involved in the LPS/TLR4 signaling pathway was evaluated in tracheal tissues from immunized chickens using quantitative PCR. Analysis revealed significant upregulation (P < 0.05) of key components of the LPS/ TLR4 signaling cascade, including CD14, TLR4, TRAM, TRAF3, IRF3, and CD80, in the Exo group compared to the NC group (Fig. 6E).
These findings demonstrate that HD11 M1 -exo treatment significantly upregulates expression of co-stimulatory molecules CD80 and CD86 in tracheal and Harderian gland tissues, activates the LPS/TLR4 signaling pathway, and potentiates cellular immune responses in chickens.
## HD11 M1 -exo augments mucosal immunity in the respiratory tract of chickens
Using qPCR and a chicken IgA kit (ELISA), we detected the levels of the mucosal immunerelated gene TGF-β4 in the trachea and IgA in the tears of experimental chickens at 7, 14, 21, 28, and 35 dpv. The experimental results showed that, compared to the NC group, the transcription level of the mucosal immune-related gene TGF-β4 in the trachea of the Exo group was significantly elevated at each time point (P < 0.05) (Fig. 7A). Additionally, the level of chicken immunoglobulin IgA in the tears of the Exo group was significantly elevated at each time point compared to the NC group (P < 0.05) (Fig. 7B). These results suggest that stimulation with HD11 M1 -exo can enhance the transcription level of the mucosal immune-related gene TGF-β4 and the level of chicken immunoglobulin IgA, thereby improving the mucosal immunity in chickens.
## HD11 M1 -exo enhances vaccine H120-induced cellular immune responses in chickens
The experimental design for immunological activation induced by the combined use of IB vaccines in chickens is shown in Fig. 8.
In this experiment, qPCR was used to detect the transcription levels of the costimulatory factors CD80 and CD86 in the trachea and Harderian glands of each group of experimental chickens at 7, 14, 21, 28, and 35 dpv. The experimental results indicated that, compared to the use of the vaccine alone, the combination of HD11 M1 -exo with the vaccine significantly increased the transcription levels of CD80 and CD86 in the trachea and Harderian glands of immunized chickens. Furthermore, the effects were more pronounced at 7 and 28 dpv compared to the H120+AL group and the H120+1313 group Expression of co-stimulatory molecules CD80 and CD86 in tracheal and Harderian gland tissues was assessed by quantitative PCR at 7, 14, 21, 28, and 35 dpv across all experimental groups. The HD11 M1 -exo adjuvanted vaccine significantly upregulated CD80 and CD86 mRNA levels in both tissues compared to vaccination with H120 alone. This enhancement was particularly evident at 7 and 28 dpv, where the H120+exo group had significantly higher expression than both the H120+AL and H120+1313 groups (P < 0.05) (Fig. 9A through D). Collectively, these data demonstrate that HD11 M1 -exo, when combined with the attenuated infectious bronchitis vaccine H120, potentiates the expression of key co-stimulatory molecules in mucosal-associated lymphoid tissues, suggesting enhanced vaccine-induced cellular immunity compared to conventional adjuvant formulations.
## HD11 M1 -exo significantly elevates vaccine H120-induced humoral immunity in chickens
Serum IgY antibody levels against infectious bronchitis virus were quantified using an indirect ELISA at 7, 14, 21, 28, and 35 dpv. Administration of HD11 M1 -exo in combination with the attenuated infectious bronchitis vaccine H120 significantly enhanced vaccineinduced antibody responses throughout the observation period. This enhancement was particularly pronounced at 14 dpv, when the HD11+exo group exhibited significantly higher antibody titers compared to both the H120+AL and H120+1313 groups (P < 0.05) (Fig. 10). These findings demonstrate that HD11 M1 -exo effectively potentiates the humoral immune response elicited by the H120 vaccine. 11A and B). These results demonstrate that HD11 M1 -exo markedly enhances vaccineinduced mucosal immune responses at respiratory surfaces.
## HD11 M1 -exo enhances the protective efficacy of the H120 vaccine against IBV challenge
The experimental design for the combined use of HD11 M1 -exo and IB vaccine to resist IBV infection in chickens is shown in Fig. 12. Viral load in tracheal tissues at 5 dpi was assessed by quantitative PCR, measuring relative expression of the IBV nucleocapsid (N) gene. Analysis revealed significantly reduced viral burdens in tracheal tissues from chickens in both the HD11+exo and H120+Al groups compared to the H120 group (P < 0.05) (Fig. 13A). In contrast, viral loads in the H120+1313 group did not differ significantly from those in the H120 group. These findings demonstrate that HD11 M1 -exo adjuvantation of the H120 vaccine effectively reduces viral replication in the respiratory tract following challenge with infectious bronchitis virus.
At 5 dpi, the body weight of each group of chickens was measured and subjected to statistical analysis. The results showed that the body weights of the NC group, H120 group, HD11+exo group, H120+AL group, and H120+1313 group were significantly higher than that of the IBV group (P < 0.05) (Fig. 13B). Additionally, the body weights of the HD11+exo group, H120+AL group, and H120+1313 group were all significantly higher than that of the H120 group (P < 0.05). Body weights were recorded and analyzed at 5 dpi. All vaccinated groups (H120, HD11+exo, H120+AL, and H120+1313) and the NC group exhibited significantly higher body weights compared to the unvaccinated, challenged IBV group (P < 0.05) (Fig. 13B). Moreover, chickens in the adjuvanted vaccine groups (HD11+exo, H120+AL, and H120+1313) maintained significantly higher body weights than those receiving the H120 vaccine alone (P < 0.05). Necropsy examinations performed at 5 dpi revealed normal tracheal mucosa without lesions in the NC group, whereas numerous petechial hemorrhages were observed in tracheal tissues from the IBV group. The H120 group displayed moderate hemorrhagic lesions, while all adjuvan ted groups (HD11+exo, H120+AL, and H120+1313) showed reduced tracheal pathology (Fig. 13C). These findings confirm the successful establishment of the challenge model and demonstrate that HD11 M1 -exo, when used as an adjuvant with the H120 vaccine, mitigates IBV-induced pathology and weight loss in infected chickens. The pathological examination of the tracheal tissue samples indicates that the NC group of experimental chickens exhibited intact histological structure with no pathologi cal changes. The IBV group displayed a complete loss of cilia, with the cell nuclei exhibiting deep staining and significant infiltration of lymphocytes. The H120 vaccine group showed a more pronounced loss of cilia, with similarly deep-stained nuclei, but a comparatively lower number of lymphocyte infiltrates. The HD11+exo group showed only a small amount of cilia loss, deep-stained nuclei, but no significant lymphocyte infiltration was observed. Both the H120+AL group and the H120+1313 group demon strated slight cilia loss, deep-stained nuclei, accompanied by some sloughing of epithelial cells (Fig. 13D). In summary, compared to the H120 group, the combination of HD11 M1 -exo with the H120 vaccine can mitigate the pathological damage caused by IBV infection in the trachea of chickens.
## DISCUSSION
Immune cell-derived exosomes are currently the most extensively studied exosomal subtype. Several studies have demonstrated that exosomes originating from immune cells possess immunostimulatory properties (15,16) and can be utilized for immune activation and disease prevention. M1 macrophages not only present antigens to naive T cells but also secrete IL-12, thereby promoting the differentiation of T helper type 1 (Th1) immune responses. These cells play a crucial role in host defense against microbial pathogens and represent key components of the immune system (17). In this study, the chicken macrophage cell line HD11 was stimulated with LPS, and qPCR was employed to assess the expression of typical M1 macrophage markers, including iNOS, IL-6, IL-12, IL-23, CD80, and CD86, which serve as established indicators for M1 phenotypic characterization (18,19). This study demonstrated macrophage polarization to the M1 phenotype and successfully isolated exosomes from M1 chicken macrophages (HD11 M1exo) using ultrahigh-speed centrifugation, with characterization confirming typical morphological and molecular profiles. Although previous research has established that exosomes derived from LPS-stimulated macrophages exhibit immunostimulatory effects at the cellular level (14), their immunomodulatory potential at the organismal level remains unreported.
Phagocytosis represents a fundamental function of macrophages, and enhanced phagocytic capacity serves as a key indicator of macrophage activation status (20). Our experimental results demonstrate that following 24 h stimulation with HD11 M1 -exo, macrophages display significantly enhanced capacity to phagocytose neutral red, indicating that HD11 M1 -exo stimulates innate immune responses in macrophages. The Toll-like receptor (TLR) signaling pathway serves as a pivotal regulator of immune responses, operating through two distinct pathways: the MyD88-dependent pathway and the TRIF-dependent pathway (21). Established research has demonstrated that LPS stimulation of chicken macrophage-derived exosomes can induce immune responses and cytokine production through the MyD88/NF-κB signaling pathway (14). However, the signaling mechanisms of M1 macrophage-derived exosomes via the TRIF pathway Full-Length Text confer enhanced resistance against IBV infection. Bal Krishan Sharma et al. experimen tally demonstrated the inhibitory effects of TLR agonists on IBV replication in chicken embryos, elucidating the underlying mechanism by which this protection is mediated through modulation of proinflammatory cytokine and antiviral gene expression (22). TNF-α serves as a pleiotropic cytokine that mediates inflammatory responses, antitumor immunity, and immune homeostasis across diverse physiological processes (23,24). Similarly, cytokines including IL-1β, IL-2, IL-4, and IFN-γ function as critical mediators in immunomodulation and are essential components of innate immunity. Our investiga tion demonstrated that HD11M1-exo pretreatment significantly upregulated IL-1β, IL-2, IL-4, TLR4, TNF-α, and IFN-γ expression in chicken embryonic tracheal tissues at 24, 48, and 72 h post-IBV challenge. This enhanced immune response effectively attenu ated IBV-induced pathological lesions. Collectively, these findings provide compelling evidence that HD11M1-exo administration elicits robust immunostimulatory effects in avian embryonic tissues, suggesting its potential application as an immunomodulatory agent.
Furthermore, in vivo trials were conducted to evaluate HD11 M1 -exo immunomodula tory properties, which revealed that HD11 M1 -exo administration significantly upregula ted CD80 and CD86 transcription in tracheal and Harderian gland tissues of vaccinated chickens, thereby enhancing cell-mediated immune responses in avian tissues. As a critical component of the host defense network, the mucosal immune system serves a pivotal function in avian immunity. Chickens employ their distinctive mucosal immune apparatus to mount effective immune responses against pathogen invasion, thus providing protection against infectious agents (25). IgA, as the predominant antibody subtype in the mucosal immune system, is widely distributed on the mucosal surfaces of the gastrointestinal tract, respiratory tract, vagina, and in bodily fluids, such as tears and saliva, forming a crucial defense line for the body's mucosal immune protection (26). For poultry, localized immunoglobulins in tears, such as IgA, are produced by lymphocytes in the Harderian gland, providing local protection to the upper respiratory tract tissues (27). Additionally, TGF-β4 can induce B cells to produce the non-inflammatory immunoglobulin subclasses IgG4 and IgA, playing a crucial role in IgA generation (28). In our investigation, IgA concentrations in the lacrimal fluid and TGF-β4 transcrip tional activity in tracheal tissues of chickens exposed to HD11 M1 -exo were significantly elevated compared to control subjects. This study represents the first demonstration at the organismal level that avian M1 macrophage-derived exosomes activate both cell-mediated and mucosal immune responses in chickens, indicating that HD11 M1 -exo represents a novel candidate immunostimulant for veterinary applications.
The immunomodulatory effects of M1 macrophage-derived exosomes on vaccineinduced immunity represent a central research objective of this investigation. In this study, HD11 M1 -exo was evaluated as an adjuvant coadministered with the IBV live attenuated vaccine H120. Results demonstrated that HD11 M1 -exo/H120 coadministra tion significantly upregulated CD80 and CD86 transcription in tracheal and Harderian gland tissues at multiple time points post-vaccination. This combination also enhanced IgA concentrations in lacrimal fluid of vaccinated chickens and increased TGF-β4 gene expression in tracheal tissues. Furthermore, H120-induced IgY antibody titers were augmented, with the HD11 M1 -exo group exhibiting significantly higher antibody levels at 14 dpv compared to the adjuvant control group. These findings represent the first experimental evidence that HD11 M1 -exo coadministration with IBV live attenuated vaccine H120 significantly enhances cell-mediated, humoral, and mucosal immune responses to IBV vaccination, thereby potentiating vaccine efficacy. These data indicate that HD11 M1 -exo constitutes a novel vaccine adjuvant candidate for avian respiratory vaccines.
The focus of this study is to investigate whether HD11 M1 -exo can serve as an adjuvant to resist strong pathogenic infections of IBV and whether it can enhance the cross-protective effects of existing vaccines. Therefore, this study conducted animal experiments on the combination of HD11 M1 -exo and vaccines to resist IBV infection. The initial infection site of IBV is typically located in the upper respiratory tract, primarily targeting ciliated cells and mucous-secreting cells (29). Multiple studies have demon strated that during IBV infection, the viral load in the trachea typically peaks at 5 dpi and subsequently declines gradually (30). The IBV HX strain utilized in this study is a highly pathogenic strain isolated from Guangxi Province, China, exhibiting a 5% nucleotide sequence difference in the S1 gene, a threshold associated with insufficient cross-protection from commercial vaccines (31,32). Our preliminary research demon strates that sequence similarity analysis of the IBV HX strain with vaccine strains H120, 4/91, and LDT3-A reveals relatively low similarity of the S1 gene at both nucleotide (58.49%, 58.80%, and 59.65%, respectively) and amino acid (52.33%, 52.09%, and 53.51%, respectively) levels (33). Therefore, the HX strain provides an appropriate model to evaluate whether HD11 M1 -exo can enhance the cross-protective efficacy of the lowvirulence vaccine H120. Our experimental results demonstrate that combining HD11 M1exo with the low-virulence IB vaccine H120 significantly reduces IBV viral load in the trachea of immunized chickens, significantly alleviating pathological damage in the trachea and mitigating clinical signs in infected chickens. This study provides the first evidence that exosomes derived from M1 macrophages can enhance the cross-protec tive efficacy of existing vaccines, effectively protecting against infections by virulent IBV strains. HD11 M1 -exo may serve as a vaccine adjuvant to provide protection for commer cial poultry.
## Conclusion
In summary, this study investigates exosomes derived from LPS-stimulated M1 macrophages, providing the first comprehensive analysis at cellular, embryonated egg, and animal levels (Fig. 14). Cellular-level analysis demonstrated that HD11 M1 -exo activates the LPS/TLR4 signaling pathway in macrophages, enhancing their immunologi cal activity. In chicken embryos, HD11 M1 -exo pretreatment increased the expression of TNF-α, IL-1β, IL-2, IL-4, TLR4, and IFN-γ genes in the trachea, thereby enhancing resistance to IBV infection. At the animal level, HD11 M1 -exo significantly enhances cellular, humoral, and mucosal immunity in immunized chickens. Furthermore, challenge experiments demonstrated that HD11 M1 -exo, when used as a vaccine adjuvant, effectively enhan ces the cross-protective efficacy of existing IB vaccines against highly pathogenic IBV infections.
## References
1. Bande, Arshad, Omar et al. (2017) "Global distributions and strain diversity of avian infectious bronchitis virus: a review" *Anim Health Res Rev*
2. Falchieri, Coward, Reid et al. (2024) "Infectious bronchitis virus: an overview of the "chicken coronavirus" *J Med Microbiol*
3. Zhao, Zhao, Zhang (2023) "Key aspects of coronavirus avian infectious bronchitis virus" *Pathogens*
4. Bakhshandeh, Kamaleddin, Aalishah (2017) "A comprehensive review on exosomes and microvesicles as epigenetic factors" *Curr Stem Cell Res Ther*
5. Lu, Xing, Xun et al. (2018) "Exosome-based small RNA delivery: progress and prospects" *Asian J Pharm Sci*
6. Deb, Gupta, Mazumder (2021) "Exosomes: a new horizon in modern medicine" *Life Sci*
7. Gross, Chaudhary, Bartscherer et al. (2012) "Active Wnt proteins are secreted on exosomes" *Nat Cell Biol*
8. Mathivanan, Ji, Simpson (2010) "Exosomes: extracellular organelles important in intercellular communication" *J Proteomics*
9. Davies, Jenkins, Allen et al. (2013) "Tissue-resident macrophages" *Nat Immunol*
10. Martinez, Gordon (2014) "The M1 and M2 paradigm of macrophage activation: time for reassessment"
11. Sica, Mantovani (2012) "Macrophage plasticity and polarization: in vivo veritas" *J Clin Invest*
12. Patel, Rajasingh, Samanta et al. (2017) "Macrophage polarization in response to epigenetic modifiers during infection and inflammation" *Drug Discov Today*
13. Hong, Lee, Vu et al. (2021) "Immunomodula tory effects of poly(I:C)-stimulated exosomes derived from chicken macrophages" *Poult Sci*
14. Hong, Lee, Vu et al. (2021) "Exosomes of lipopolysaccharide-stimulated chicken macrophages modulate immune response through the MyD88/NF-κB signaling pathway" *Dev Comp Immunol*
15. Srinivasan, Su, Ravishankar et al. (2017) "TLR-exosomes exhibit distinct kinetics and effector function" *Sci Rep*
16. Tang, Sun, Gupta et al. (2016) "Monocyte exosomes induce adhesion molecules and cytokines via activation of NF-κB in endothelial cells"
17. Philips, Ernst (2012) "Tuberculosis pathogenesis and immunity" *Annu Rev Pathol*
18. Yunna, Mengru, Lei et al. (2020) "Macrophage M1/M2 polarization" *Eur J Pharmacol*
19. Atri, Guerfali, Laouini (2018) "Role of human macrophage polarization in inflammation during infectious diseases" *Int J Mol Sci*
20. Murray, Wynn (2011) "Protective and pathogenic functions of macrophage subsets" *Nat Rev Immunol*
21. Duan, Du, Xing et al. (2022) "Toll-like receptor signaling and its role in cell-mediated immunity" *Front Immunol*
22. Sharma, Kakker, Bhadouriya et al. (2020) "Effect of TLR agonist on infections bronchitis virus replication and cytokine expression in embryonated chicken eggs" *Mol Immunol*
23. Croft (2009) "The role of TNF superfamily members in T-cell function and diseases" *Nat Rev Immunol*
24. Aggarwal (2003) "Signalling pathways of the TNF superfamily: a double-edged sword" *Nat Rev Immunol*
25. Nochi, Jansen, Toyomizu et al. (2018) "The well-developed mucosal immune systems of birds and mammals allow for similar approaches of mucosal vaccination in both types of animals" *Front Nutr*
26. Li, Chen (2020) "The effects of secretory IgA in the mucosal immune system" *Biomed Res Int*
27. Baba, Masumoto, Nishida et al. (1988) "Harderian gland dependency of immunoglobulin a production in the lacrimal fluid of chicken" *Immunology*
28. Akdis, Blaser, Akdis (2004) "Genes of tolerance"
29. Raj, Jones (1996) "Protectotypic differentiation of avian infectious bronchitis viruses using an in vitro challenge model" *Vet Microbiol*
30. Okino, Alessi, Montassier et al. (2013) "Humoral and cell-mediated immune responses to different doses of attenuated vaccine against avian infectious bronchitis virus" *Viral Immunol*
31. Marandino, Pereda, Tomás et al. (2015) "Phylody namic analysis of avian infectious bronchitis virus in South America" *J Gen Virol*
32. Villarreal, Sandri, Souza et al. (2010) "Molecular epidemiology of avian infectious bronchitis in Brazil from 2007 to 2008 in breeders, broilers, and layers" *Avian Dis*
33. Yang, Zhou, Huang et al. (2024) "Isolation of a more aggressive GVI-1 genotype strain HX of the avian infectious bronchitis virus" *Poult Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772287&blobtype=pdf | # Isolation and characterization of TayeBlu, a novel bacteriophage of Azotobacter vinelandii
Taiwo Akanbi, Mariana Labat, Tyler Sun, Derek Smith, Sarah Bagby
## Abstract
Soil microbial communities drive global biogeochemical cycles and alter crop yields through nitrogen fixation. As agents of genetic mobility, mortality, and nutrient release, viruses have been shown to influence microbial community structure and activity in numerous marine and aquatic systems. However, their impacts on terrestrial ecosystems are less well understood, in part because few model phage-host systems have been established for soils. To fill this gap, we sought to develop a new model system for viral infection of nitrogen-fixing bacteria derived from agricultural soil. Here, we report the isolation, characterization, and sequencing of the novel bacterio phage TayeBlu, which infects the globally distributed aerobic soil bacterium Azotobacter vinelandii, a facultative diazotroph. TayeBlu was isolated from the rhizosphere of tomato plants in a farm greenhouse. We find that the availability of nitrogen to host cells strongly influences TayeBlu infection physiology at the level of adsorption kinetics, time to lysis, and burst size. Taxonomic and comparative genome analyses reveal that TayeBlu belongs to an understudied family in class Caudoviricetes in which a small core of structural and assembly genes has persisted through adaptive diversification on different bacterial hosts. IMPORTANCE Understanding the forces regulating soil microbial activity is critical for building accurate ecosystem models that can inform land-management strategies for mitigating climate risks and stabilizing global food supply. For agricultural sustainability, it is particularly important to understand the dynamics of nitrogen-fixing soil bacteria like Azotobacter vinelandii, a well-studied and globally distributed species whose activity promotes plant growth and soil fertility. To support detailed investigations of the impact of viruses on diazotroph ecosystem outputs, we isolated and investigated a novel soil virus that infects A. vinelandii. The resulting new phage-host system expands the limited toolkit of model experimental systems for soil viral studies. By enabling investigations of viral impacts on terrestrial nitrogen-fixing bacteria, this work sets the stage for future studies illuminating a critical but poorly understood aspect of soil ecology. Moreover, TayeBlu belongs to a novel viral family, and this study provides the first description of this family's conserved features. KEYWORDS bacteriophages, soil microbiology, diazotrophs S oil microbial communities shape global biogeochemical cycles and local ecosystem function through metabolic processes and ecological interactions that determine the bioavailability of carbon (C) and nitrogen (N) (1, 2), affecting plant productivity and soil function. Within these communities, N-fixing bacteria (diazotrophs) play a crucial role by converting atmospheric N 2 into biologically available forms. Annually, biological N fixation contributes ∼200 million metric tons of bioavailable N to Earth's ecosystems (3,4), including nearly half the total N present in crop fields (5,6). By releasing organic N as a public good, soil diazotrophs facilitate the growth of plants and other microbes (7,
8), which otherwise may scavenge N by degrading recalcitrant soil organic matter (7,8), potentially increasing carbon dioxide emissions (9).
Infection of bacteria by their viruses (bacteriophages, or phages) alters host ecological impacts (10). Phages influence nutrient cycling and control bacterial population dynamics and mortality (11). In addition to direct lysis, phage infection impacts bacterial communities and biogeochemical processes through several distinct mechanisms: phages can (i) reprogram host metabolism by introducing auxiliary metabolic genes (AMGs) (12,13); (ii) accelerate nutrient turnover through release of host cellular contents (14,15); and (iii) transfer genetic material between hosts (16)(17)(18). Phage infection dynamics can be highly dependent on host physiological state, with low-nutrient conditions reducing phage infectivity and burst size (e.g., see references 19 and 20). Viral lysis accounts for substantial bacterial mortality and subsequent nutrient release in well-studied marine systems (14,17,21). While available evidence suggests that soil environments contain substantial phage populations (∼10 3 to 10 9 viral particles per gram of soil, depending on soil type and conditions [22]), our knowledge of phage-host dynamics in soil ecosystems lags considerably behind that in marine systems (23,24).
A major challenge in soil virology is the limited number of established phage-host systems available for experimental investigation, with current models capturing only a handful of the predominant soil bacterial phyla (25,26). This limitation is especially pronounced for N-fixing bacteria, a problematic gap given the critical importance of diazotrophy for soil function. In particular, the observation that N flow to the soil system can differ substantially between diazotrophs (e.g., between plant-associated and free-living species) (27) points to a need for diverse experimental systems to support comparative analysis of the ecosystem impacts of phage infection of different diazotrophic hosts. Recent work has identified phages infecting Klebsiella sp. M5al (28), a facultative diazotroph from a genus that often grows in association with a plant host. Klebsiella spp. fix N anaerobically, with low levels of O 2 rapidly inhibiting nitroge nase, the enzyme responsible for N fixation (29). By contrast, the globally distributed soil bacterium Azotobacter vinelandii is a free-living and obligately aerobic faculta tive diazotroph (30,31). Azotobacters are noteworthy for their specialized metabolic adaptations (e.g., extraordinarily high respiration rates) to protect nitrogenase from oxygen (reviewed in reference 30), offering an opportunity to investigate infection physiology under widely different host metabolic states. With a versatile metabolism, a long history as a soil model system, and modern tools for mechanistic investigation (32,33), A. vinelandii is an ideal host for phage model system development. Despite numerous reports of characterized A. vinelandii phage isolates several decades ago (34)(35)(36), to our knowledge, none of these phages remain in cultivation, and none have been sequenced or characterized by modern methods.
To fill this gap, we sought to isolate a novel phage of A. vinelandii that can be developed as an experimental model system for investigation of phage impacts on free-living soil diazotrophs. Here, we report the isolation, characterization, and sequenc ing of the novel rhizosphere siphovirus TayeBlu infecting A. vinelandii strain OP, and we demonstrate that phage infection is significantly influenced by N availability in the medium. Through comprehensive genomic analysis, we identify this phage as belonging to a novel viral family within an unclassified order of the class Caudoviricetes, with a conserved core of structural and replication genes but low genomic similarity to other known soil phages. This novel phage shows promise as an experimental system for studies on the influence of soil phages on N-fixing bacterial communities in variable soil environments and their impact on global biogeochemical cycles (23).
## MATERIALS AND METHODS
## Host strain, growth media, and phage buffer
All experiments were performed with Azotobacter vinelandii strain OP (hereafter A. vinelandii), which was the generous gift of Xinning Zhang. A. vinelandii was grown in three media spanning high and low C and N availabilities. Peptone-yeast-calcium (PYCa) medium is a rich and N-replete medium consisting of 15 g L -1 peptone, 1 g L -1 yeast extract, 1 g L -1 dextrose, and 4.5 mM CaCl 2 . Dean's Burke (Dean) medium is a minimal, N-free medium consisting of 1 mM phosphate buffer, 20 g L -1 sucrose, 0.
## Isolation and purification of phage TayeBlu
Phage TayeBlu (named from "Taye, " a Nigerian-Yoruba name for "first child to taste the world, " reflecting its status as the first sequenced and genomically characterized A. vinelandii phage isolate, and "Blu, " a local reference reflecting the isolate's Cleveland origin) was isolated from rhizosphere soil collected at the base of a tomato plant in a greenhouse at the Case Western Reserve University Farm, Hunting Valley, OH (41°29′36″N, 81°25′24″W), on 30 August 2023. Using a hand shovel sterilized with 70% ethanol, 50-55 g of dry, dark soil was collected from within 2 cm of the plant base (Fig. S1), where fine roots were visible. Ambient temperature was 18°C. Samples were stored on ice and kept at 4°C until further processing.
Phage isolation used the enrichment method (37), followed by spot test analysis and agarose overlays (38). Briefly, fresh soil samples (∼25 g) were enriched with 500 μL of an overnight A. vinelandii culture and incubated at 30°C for 48 h with shaking at 225 rpm. The mixture was centrifuged at 4,700 × g for 30 min, and the supernatant was passed through a 0.22-µm pore-size polyethersulfone (PES) filter to remove bacterial cells. Filtrates were screened for phage activity using the spot test method (39) with A. vinelandii cultured in the three media described above. Briefly, 500 μL of overnight culture was mixed with 4.5 mL of 0.6% (wt/vol) molten agar and spread evenly on a 1.6% (wt/vol) agar plate. After solidification of the top agar matrix, 10 μL of filtrate was spotted onto the surface and allowed to dry. Plates were incubated at 30°C for 24 h and monitored for plaque formation. The presence of a clear or turbid clearing on the plated bacterial lawn was considered a putative phage plaque.
Putative phage plaques were cored and suspended in 100 μL phage buffer, serially diluted, and then plated using the agarose overlay method. Ten microliters of the target dilution (typically 10 -4 to 10 -6 ) was added to 500 μL overnight culture and 4.5 mL of 0.6% (wt/vol) molten agar and spread evenly on a 1.6% (wt/vol) agar plate. Plates were incubated at 30°C overnight. For three consecutive rounds of purification, a single, isolated plaque was re-suspended in 100 μL of phage buffer, diluted, and plated to obtain a clonal phage population. Purified clonal phage was then amplified, filtered (0.22-μm pore-size PES filters), titered, and stored (4°C for immediate use, -80°C in 6.54% dimethyl sulfoxide or 20% glycerol for freezer stock) (39). Only stocks stored at 4°C were used for further characterization in this study.
## Phage-host adsorption kinetics
To determine the adsorption rate constants (k) of phage TayeBlu and A. vinelandii strain OP in PYCa, Dean, and AC_Dean media, we performed adsorption assays following previously described protocols (40,41). Briefly, in each medium, triplicate cultures of A. vinelandii were grown to mid-exponential phase (∼10 8 colony-forming units [CFU] mL -1 ) as determined by optical density (OD) measurements and OD-to-CFU calibration curves established previously for each medium. An aliquot of 2 × 10 8 cells was removed from each culture and transferred to a five-mL glass culture tube, and the volume was adjusted to 2 mL with fresh medium to achieve a culture density of 1 × 10 8 CFU mL -1 . High-titer TayeBlu lysate (4.3-7.3 × 10 10 PFU mL -1 ) was added to a multiplicity of infection (MOI) of 0.1.
Immediately following infection (t 0 ) and at pre-determined intervals, samples were collected for both total and free phage quantification. For free phage measurements, samples were immediately filtered through 0.22-µm pore-size PES filters to remove bacterial cells and any cell-associated phages. For total phage measurements, parallel samples were left unfiltered. All samples were serially diluted in phage buffer and enumerated at two sequential dilutions using the agarose overlay method. AC_Dean was used as the base medium for plaque assays from AC_Dean adsorption experiments; PYCa was used for plaque assays from experiments in both PYCa and Dean. This use of PYCa ensured reliable plaque formation, as our preliminary experiments demonstrated that PYCa provides consistently good conditions for TayeBlu plaque formation on this host. Plates were incubated at 30°C, and plaques were enumerated after 24, 48, and 72 h to ensure complete development of all viable plaques.
Statistical analyses of the decrease in free phage concentration over time were performed using R (v.4.4.3) (42) with the tidyverse (43), Hmisc (44), errors (45), ggtext (46), BSDA (47), and patchwork (48) packages. Data were filtered to exclude dilutions with zero counts and those with plaques too numerous to count (TNTC). For each measurement, the phage concentration (plaque-forming units [PFU] per milliliter) was calculated and assigned an uncertainty scaled to the square root of the number of plaques observed (truncated at a minimum of 1.2), reflecting Poisson error in counting statistics. Weighted means and standard deviations (SDs) were calculated for each time point, with weights assigned as the inverse of the relative error for each measurement. In cases where the weighted coefficient of variation was greater than 0.8 and at least four measurements (of the six attempted, three biological replicates × two dilutions) produced quantitative (non-zero, non-TNTC) results, we checked for influential outliers using jackknife resampling as follows. For each such measurement, we calculated the absolute values of (i) the difference between the full weighted mean and the jackknife weighted mean, relative to the jackknife weighted mean, to judge the measurement's influence on sample mean; and (ii) the difference between the full weighted SD and the jackknife weighted SD, relative to the full weighted SD, to judge the measurement's contribution to the sample error. We set thresholds for both quantities at 0.5 and considered measurements outliers only if they exceeded both thresholds. For adsorption experiments, no measurements were excluded as outliers.
Following reference 40, the fraction of free phage remaining (P t /P 0 ) was plot ted against time. Adsorption rate constants (k, in units of mL min -1 ) were deter mined by fitting the natural log-transformed data to a first-order kinetic model, k = -ln(P t /P 0 )/Nt, where P t is the concentration of free phage at time t, P 0 is the initial free phage concentration, and N is the bacterial concentration. All error propagation used the errors package. Differences in fitted k values were tested with Welch's modified two-sample t-test (BSDA tsum.test).
## One-step growth curves
To determine the latent period and average burst size of phage TayeBlu and A. vine landii strain OP in PYCa, AC_Dean, and Dean media, we performed one-step growth experiments following previously described protocols (40,49,50), with an MOI of ∼0.1 (PYCa, 0.106; AC_Dean, 0.090; Dean, 0.091) and an 11-min adsorption period. In each one-step assay, triplicate cultures were grown in the test medium to mid-logarithmic phase (1.1-1.7 × 10 8 CFU mL -1 ) as determined by optical density measurements and OD-to-CFU calibration curves established previously for each medium. Aliquots of 1 × 10 8 cells were removed from each culture and transferred to 1.5-mL microcentrifuge tubes, then adjusted to concentrations of ∼1 × 10 8 CFU mL -1 using fresh medium. High-titer TayeBlu lysate was added to achieve the target MOI with a working volume of 1 mL for each medium during the adsorption period.
After the adsorption period, 500 μL of each replicate culture was transferred to fresh 250-mL flasks containing 49.5 mL of fresh medium pre-incubated at 30°C, diluting the phage-host mixture a hundredfold to prevent new infections. Flasks were incubated at 30°C with shaking at 225 rpm. At pre-determined intervals, samples were collected for total phage (10 μL) and free phage (400 μL filtered through 0.22-µm pore-size PES filters). Phage abundance in free and total phage samples was enumerated using plaque assays on PYCa (PYCa and Dean experiments) or AC_Dean (AC_Dean experiment). Plates were incubated at 30°C for 24 h in PYCa medium or 48 h in AC_Dean medium to allow for slower host growth in this medium. Plaque counts were stable after the incubation period. Plates were permitted to develop at room temperature for up to 24 h after incubation for additional re-examination to facilitate accurate counting of the very small plaques obtained in AC_Dean. Analysis of plaque counts, assignment of uncertainties, error propagation, and outlier identification were conducted as described for phage-host adsorption kinetics. Three measurements of total phage counts (one each from PYCa at 90 min, AC_Dean at 28 min, and AC_Dean at 90 min) were excluded as outliers. Burst size was calculated as the ratio of free phage concentration at the plateau after a single infective cycle to the initial number of infected cells, where the number of infected cells was determined by subtracting the initial free phage concentration from the initial total phage concentration. (Plaque counts in total phage measurements are the sum of plaques due to free phage and plaques due to infected cells; the latter are expected to yield one plaque per infected cell.) Differences in burst sizes were tested with Welch's modified two-sample t-test (BSDA tsum.test). Latent period was determined as the time between phage adsorption and the onset of the release of phage progeny (40).
## Morphological characterization with transmission electron microscopy
Virion morphology was examined by transmission electron microscopy (TEM) follow ing a protocol modified from reference 51. Purified high-titer phage lysate (1 mL at 1 × 10 10 PFU mL -1 ) was centrifuged at 20,630 × g for 50 min. The supernatant was carefully removed and replaced with 500 μL of 0.1 M ammonium acetate. The sam ple was centrifuged again at 20,630 × g for 50 min. For visualization, 5 μL of the concentrated phage suspension was applied to a carbon/Formvar-coated copper grid (Electron Microscopy Sciences, FCF300CU50, purchased from Fisher Scientific, catalog no. 50-260-36) and allowed to adsorb for 1 min. Without removing excess liquid, 5 μL of 2% uranyl acetate (wt/vol) was added for negative staining and left for an additional minute. Excess liquid was then wicked away using filter paper, and the grid was air-dried. Grids were examined using a TECNAI SPIRIT T12 TEM operating at 100 kV. Virion head diameter and tail length measurements were made by analyzing transmission electron micrographs in ImageJ (52), using the line feature to draw a line across the heads and along the tails of individual virions. The lengths of these lines were recorded, and the average measurements were calculated.
## DNA extraction and whole-genome nanopore sequencing
High-molecular-weight DNA was extracted following a modified protocol based on reference 53 and the DNeasy PowerSoil Pro (QIAGEN, catalog no. 47014) DNA kit protocol. Briefly, phage lysate (6 mL) was treated with DNase I-RNase A (added at a ratio of 0.5 µL of 10 mg mL -1 stock per mL lysate, for a total of 3 μL nuclease mixture) and incubated at 37°C for 30 min to eliminate contaminating nucleic acids. Phage particles were concentrated by adding 0.5 mL of 30% PEG-8000 per mL of treated lysate, followed by overnight incubation at 4°C and centrifugation at 10,000 × g for 10 min. The resulting pellet was re-suspended in 300 μL of 5 mM MgSO 4 and transferred to PowerBead Pro tubes (QIAGEN, catalog no. 47014) for capsid disruption using an MP Biomedicals FastPrep 24 bead beater (6 m s -1 for 30 s, repeated once). DNA purification continued according to the manufacturer's protocol using the DNeasy PowerSoil Pro Kit. DNA was barcoded with a single barcode using the SQK-NBD114.24 Native Barcoding Kit 24 (v.14), then pooled with other separately barcoded samples for adaptor ligation (SQK-NBD114.24 Native Barcoding Kit 24 v.14) and multiplex sequencing on a MinION Mk1C long-read sequencer (model MIN-101C) with an R10.4.1 flowcell (FLO-MIN114).
## Phage genome assembly and annotation
Phage genomes were assembled using Oxford Nanopore Technology sequencing data. Raw reads were basecalled using Dorado (v.0.7.2) (54) with the super-accuracy model (dna_r10.4.1_e8.2_400bps_sup@v5.0.0) with 400 bps parameters (55). Demultiplexing was performed with the --both-ends argument, followed by adapter trimming using the same software (54). BAM files were converted to FASTQ format using Samtools (v.1.21) (56). Quality assessment of the FASTQ files was conducted using FastQC (v.0.11.9) (57), and the results were aggregated with MultiQC (v.1.9) (58). Quality trimming was implemented with Prowler (59) using the F1 option (Q-score threshold of 20) to retain the longest high-quality fragments. Assembly was performed using Metaflye (v.2.9) (60) with the --meta option, which is optimized for viral genomes, and the resulting graph was visualized with Bandage (61). Because polishing tools like Medaka may reduce assembly quality for modern Dorado super-accuracy (5 Hz SUP) basecalled data sets (62), we did not perform polishing. Genome coverage depth was assessed using coverage_histogram in CoverM (v.0.7.0) (63). The completeness and potential contamina tion of the assembled viral genome was assessed using CheckV with the end_to_end option (64).
Because assembly into circular contigs can be an artifact of the terminal repeats or circular permutation used in packaging many phage genomes, we used Terminus.SE to detect the genomic positions of terminus sequences from sequencing coverage patterns (65). To predict the packaging mechanism, we performed phylogenetic analysis of the large terminase protein, using ClustalW2 (66) for multiple sequence alignment of TayeBlu with 44 reference phages of known packaging mechanisms, followed by neighbor-join ing tree construction with bootstrap analysis (1,000 replicates) using the boot.phylo function in the ape package in R (67). To reflect the biological packaging organization detected, the genome sequence was linearized and re-numbered starting at the site of the predicted 5′ terminus (position 3560 in the circular contig assembly). Gene positions and locus numbering reflect this curation step.
Coding sequences (CDSs) were predicted using multiple tools including GeneMark, GeneMarkS, Glimmer (v.3.02) (68), Prodigal (v.2.6.3) (69), and MetaGeneAnnotator (v.1.0) (70). Predictions were integrated using a weighted scoring system as described in reference 28, with scores calculated based on gene length, overlap, protein identification, and programming potential. Functional annotations were assigned based on significant alignment scores from BLASTP (e-value <10 -5 ) (71), Swiss-Prot (release 2023_01) (72), HHpred (probability >90% and e-value <1) (73), and phold (v.0.2.0) (74), which converts protein sequences into structural tokens for comparison against >1 M phage protein structural models (75)(76)(77)(78)(79). Phage genome visualization was performed using the final manually curated GenBank-formatted TayeBlu sequences and a modified version of the phold python plotting script (plot.py).
Assignments were manually curated to resolve conflicts between predictions with attention to genomic context; across tools, HHpred results were weighted most heavily and phold least, in line with current levels of benchmarking support for these tools. Additional genomic features, including tRNAs, introns, and spanins, were screened for using the structural and functional phage annotation pipelines in Apollo (80). Rho-inde pendent transcription terminators were identified using ARNold (81), which employs both RNAmotif and Erpin algorithms. Predicted terminators were manually curated based on their genomic context and structural features. Putative AMGs were screened using DRAM-v (v.1.3.0) (82) with default parameters, retaining predictions with an M flag and an auxiliary score of ≤3.
## Phage taxonomy classification
Identification of related phages and taxonomic classification followed a multi-tiered approach to ensure accuracy (83). We recruited phage sequences similar to TayeBlu at the nucleotide level by comparing the complete TayeBlu genome to the National Center for Biotechnology Information (NCBI) non-redundant (nr) nucleotide database using BLASTN. The average nucleotide identity (ANI) of the closest relative was estimated by multiplying the genome coverage by the percent identity of the hit. Next, we used ViPTree's web interface (84) and vConTACT3 (v.3.1.3) (85) to recruit additional relatives based on similarity at the protein level to phages represented in the Virus-Host DB and Viral RefSeq (v.230), respectively. This analysis identified 44 viruses with significant protein-level similarity (genomic similarity [S G ] values ≥0.020) from across currently known viral diversity, in addition to the three close relatives originally identified at the nucleotide level. One of the 44 viruses was identified as a much shorter satellite phage and excluded from further analysis. The phylogenetic tree of TayeBlu and the remaining 46 relatives was visualized using the package ggtree (v.3.10.1) in R (v.4.3.3) (42,86). TayeBlu and its eight closest relatives were submitted to VIRIDIC (87) for pairwise genomic distance analysis. VIRIDIC output was visualized using a modified version of the original VIRIDIC R script and the dendextend package (88) for dendrogram presentation.
We used vConTACT3's hierarchical clustering of gene-sharing networks with this set of phage genomes to place TayeBlu among the Caudoviricetes. To refine this prediction, we broadened our comparison set to include all Caudoviricetes virus species exemplars and additional isolates in the latest International Committee on the Taxonomy of Viruses (ICTV) Virus Metadata Resource (VMR) database (https://ictv.global/vmr, accessed 29 May 2025). Using the VMR MSL 40.v1 metadata sheet, we compiled GenBank acces sions for ICTV VMR Caudoviricetes exemplars and additional isolates (5,822, of which 5,821 were available in GenBank), then compared this list against all accessions in the NCBI Viral RefSeq (v.230) database, removing duplicates. This identified 311 Caudovir icetes phages that represent known species in ICTV's database and were not yet captured in Viral RefSeq. We concatenated these 311 sequences with the genomes of TayeBlu and the three NCBI nr relatives to build a non-redundant set for comparison against Viral RefSeq (v.230). We used vConTACT3 with default parameters to perform protein clustering and network-based classification leveraging the ICTV framework. Final taxonomic assignments were determined based on the optimal distance thresholds identified for each taxonomic rank.
## Comparative genome and core gene analysis
TayeBlu, the three best BLAST hits from NCBI nr, and the five other Viral RefSeq phages of novel_family_3 were analyzed for genome comparison and core gene identification. In order to perform these analyses, we downloaded the FASTA files of all eight phages from NCBI and performed genome annotation using pharokka with the -g prodigal-gv option to ensure consistent annotation strings (75,89). The resulting GenBank files, together with our manually curated TayeBlu .gbk file, were analyzed and visualized for gene cluster synteny using clinker genome analysis with the default minimum alignment sequence identity (0.3) (90). Core gene analysis was performed by clustering predicted protein sequences using MMseqs2 with a minimum sequence identity threshold of 30% and a coverage threshold of 80% (-min-seq-id 0.30 -c 0.8). The protein cluster results and GenBank annotations were further analyzed in R (v.4.4.3) with dplyr, tidyr, and stringr (all tidyverse v.2.0.0) for data wrangling, then visualized with ggplot2 (v.3.5.2), scales, and patchwork (42,43,48,91). Clusters with representatives from all nine analyzed phages were considered core genes; those with representatives from eight of the nine were considered near-core.
## Comparison with PIGEON database environmental phages
We downloaded 515,763 metagenomic viral operational taxonomic units (vOTUs) from natural soil and rhizosphere environments from the PIGEONv2.0 database (https:// datadryad.org/stash/dataset/doi:10.25338/B8C934, accessed May 2025 [92]). This data set includes 53,391 vOTUs specifically derived from the tomato rhizosphere, an environment similar to the sample from which TayeBlu was isolated, among 192,008 vOTUs from a range of soil environments. Species-level clustering of TayeBlu with the complete PIGEONv2.0 database was performed using CheckV (v.0.8.1) and its associated scripts, anicalc.py and aniclust.py (64), with a threshold of 95% ANI over 80% genome coverage. To comprehensively assess TayeBlu's relationship to soil viral diversity, we performed protein clustering analysis using vConTACT3 (v.3.1.3) (85) on the concaten ated genome sequences of the nine phages of TayeBlu's novel family and all 192,008 soil vOTUs in PIGEONv2.0. We examined the complete set of vConTACT3 taxonomic assignments to identify the vOTUs that cluster with TayeBlu at each taxonomic level.
## RESULTS AND DISCUSSION
## Isolation and morphological characterization of novel phage TayeBlu
TayeBlu was successfully isolated from soil samples collected from the greenhouse farm enclosure at the CWRU University Farm (Hunting Valley, OH). As an initial investigation of the effects of nitrogen availability on infection, we asked whether plaque morphology differed across media. We consistently observed the clearest plaques on rich medium (PYCa, diameter ∼1.6 mm; Fig. 1A) and tiny plaques on minimal medium amended with ammonium chloride as a nitrogen source (AC_Dean, ∼0.6 mm; Fig. 1B andE). Plaques on minimal medium (Dean) varied, with some plates showing temperate-like plaques (∼7.2 mm, Fig. 1C) and others showing no clear plaques, though parallel inoculation of AC_Dean plates confirmed the presence of active phage in the inocula (Fig. 1F; compare to panel E). Morphological characterization by TEM revealed an icosahedral head with a mean diameter of 67 ± 3 nm (n = 15) and a long, non-contractile tail (mean length 148 ± 6 nm, n = 11; Fig. 1D), features historically associated with siphoviruses (93).
## Infection dynamics
Phage TayeBlu infection dynamics on A. vinelandii strain OP in liquid medium, like its plaque morphologies on solid medium, were significantly influenced by media composition, with effects spanning the entire infection cycle. TayeBlu's adsorption rate constant and total extent of adsorption were both lower by an order of magnitude in minimal medium than in rich medium (Fig. 2A). Ammonium amendment of the minimal medium substantially rescued adsorption, although it remained significantly slower than in rich medium (P = 7.7 × 10 -12 , Welch modified two-sample t-test). Similarly, bursts were both later (first clear evidence of burst, 55 min vs 28 min) and far smaller (burst size 13 ± 15 vs 117 ± 13, P = 1.119 × 10 -5 , Welch modified two-sample t-test) in minimal medium than in rich medium (Fig. 2B); ammonium amendment of minimal medium produced burst dynamics much closer in timing (first clear evidence of burst, 35 min vs 28 min) and scale (burst size 59 ± 13 vs 117 ± 73) to those in rich medium, though the difference in burst size remained significant (P = 0.003717).
The extent of rescue by ammonium amendment (AC_Dean) of minimal medium strongly suggests that TayeBlu infection dynamics in minimal medium are shaped primarily by the requirement for diazotrophy in these conditions. Previous studies have shown that nutrient availability can modulate the expression of outer membrane proteins and lipopolysaccharides that commonly serve as phage receptors (94,95). In addition to any such changes, Azotobacter in well-aerated minimal medium is expected to grow diazotrophically and to deploy nitrogenase-protective mechanisms that are likely to alter phage infection. First, under N-fixing conditions, many Azotobacter strains substantially increase production of the extracellular product alginate, thought to limit diffusion of O 2 from the medium to the cell (96). Although A. vinelandii strain OP is generally considered a "non-gummy, " alginate non-producing strain due to a spontane ous loss of function in its algU gene (97), we have frequently observed that a gummy phenotype spontaneously re-emerges during propagation under N limitation (Fig. S2); if present, an alginate capsule or other extracellular polysaccharide layer could easily alter receptor availability for phage, hampering adsorption (98). Second, Azotobacter greatly increases its respiration rate to draw down intracellular O 2 , creating high C-substrate demand under N limitation (31,99,100). In minimal medium, this high respiration rate could compromise resource availability for phage reproduction, extending the latent period and limiting burst size. The large change to infection dynamics observed here between rich and minimal media demonstrates the potential variability of TayeBlu impacts across soils with varying nutrient availability, highlighting the importance of characterizing soil phage infection dynamics across a range of growth conditions to support quantitative modeling of ecosystem impacts (101).
## Genome analysis
Genome sequencing (median coverage: 2,191) revealed that TayeBlu possesses a double-stranded DNA genome of 59,885 bp with a G + C content of 50.34% (Table 1). The assembly was circular, as expected for completely sequenced dsDNA phage with linear genomes and terminal repeats or circularly permuted genomes. Coverage analysis identified a 5′ terminus on the forward strand, just upstream of the genes encoding the small and large terminase proteins, and no clear 3′ terminus, consistent with the pattern expected for headful packaging (65,102). Phylogenetic analysis of the large terminase protein with those of phages of known packaging mechanism (103,104) confirmed that TayeBlu's terminase clusters with those of known headful packaging phages (Sf6 group, Fig. S3).
Structural and functional annotation using automated tools followed by manual curation in Apollo predicted 100 CDSs with no identifiable tRNA genes. We identified 15-18 putative transcription terminators from the 30 predicted using ARNold (81) (Table 1; Table S1). Start codon usage showed a predominance of ATG, with lesser usage of GTG and TTG. Functional annotation of the genome classified the CDSs into several catego ries (Fig. 3): structural and assembly genes (23 genes, comprising head and packaging proteins, connector proteins, and tail proteins), DNA, RNA, and nucleotide metabolism (16 genes), lysis (3 genes), transcription regulation (3 genes), and other functions (14 genes), while the rest of the genes encode hypothetical proteins (41) with no known function (Fig. 3).
## Genes with potential functional significance
Intriguingly, TayeBlu encodes a putative phosphofructokinase (locus AZP_TayeBlu_0026), previously reported as an AMG in marine and gut viruses (13,105). This enzyme catalyzes the conversion of fructose-6-phosphate to fructose-1,6-bisphosphate, an ATP-consuming early step in the Embden-Meyerhof-Parnas (EMP) glycolytic pathway that is the point at which a substrate glucose molecule is energetically committed to the pathway. Thus, any phage-mediated expression of phosphofructokinase during infection should partition glycolytic flux toward this pathway and away from the competing Entner-Doudoroff (ED) pathway. Fluxomic analysis of A. vinelandii has shown that only a small minority of glycolytic flux in uninfected cells uses EMP, while the bulk is directed through ED despite its lower ATP yield (99). Hypothesizing that the trade-off lowers the N demand for glycolytic enzyme synthesis, Wu et al. argue that A. vinelandii tunes central C metabo lism so that the redox state of the cell supports respiratory protection of nitrogenase (99). Phage-mediated rebalancing of glycolytic flux might thus put nitrogenase at risk. We hypothesize that TayeBlu will express phosphofructokinase during infection under conditions where phage replication is limited more by ATP than by N availability. Under this hypothesis, TayeBlu could not only directly influence carbon use efficiency but also indirectly influence host nitrogen fixation despite the absence of recognized nitrogencycling AMGs.
Beyond central metabolism, several genes annotated in TayeBlu may influence phage-host interactions. Putative carbohydrate-active enzymes include a lyase/tailspike with 63D sialidase homology (locus AZP_TayeBlu_0064) and an alpha-2,3-/2,8-sialyltrans ferase (AZP_TayeBlu_0047), which might interact with A. vinelandii's exopolysaccharides, including alginate. Viral sialidases often facilitate cell entry or exit by modifying surface glycans (106); earlier work has demonstrated alginate lyase activity in an unsequenced A. vinelandii phage (36). Additionally, we identified a putative paratox (Prx) domain protein (AZP_TayeBlu_0029) with sequence similarity to inhibitors of the quorum sensing receptor ComR (107). Quorum sensing regulation plays a critical role in biofilm forma tion, e.g., by altering production of extracellular polymeric substances (EPS) including alginates (reviewed in reference 108); a paratox protein might allow TayeBlu to manip ulate host cell signaling to alter biofilm formation or EPS production. Finally, the genome encodes a protein with homology to recently described ribosome hibernation factors (AZP_TayeBlu_0033) (109), which may influence host translation machinery under nutrient-limited conditions. The roles of these genes in the TayeBlu infection cycle across growth conditions remain to be experimentally validated.
## Taxonomic placement
Following current ICTV guidelines (93), we sought to classify TayeBlu on the basis of its genome (110). BLAST searches of all viruses in the NCBI Nucleotide collection (nr/nt) database against the complete TayeBlu genome revealed three moderately similar phages with non-trivial query coverage: Enterobacter phage Mulvp2 (OR508996.1; Next, we scored protein similarity to phages in the Virus-Host DB (111), identifying 43 phages (including the four family-level relatives identified by vConTACT3, but excluding one short satellite phage) with moderate similarity (ViPTree genomic similarity score ≥0.020) to TayeBlu, for a total of 47 at least moderately similar relatives (Table S2). In this expanded group, all phages whose hosts are known infect hosts within the phylum Pseudomonadota. A ViPTree proteome-based phylogenetic tree for TayeBlu, its three NCBI relatives, and all phages in Virus-Host DB revealed that TayeBlu forms a single, cohesive monophyletic lineage with the seven close relatives identified above and an additional Salmonella phage, SP069 (TayeBlu and moderately similar relatives, Fig. 4; full tree, Fig. S5). This proteome tree demonstrates that a significant number of orthologous genes are shared among TayeBlu and its eight close relatives, consistent with ICTV family-level classification standards.
To refine TayeBlu's taxonomic placement, we sought to compare it against all established Caudoviricetes diversity, as captured by ICTV's Virus Metadata Resource. We repeated the vConTACT3 analysis, clustering the members of the TayeBlu ViPTree family with the ICTV Caudoviricetes species exemplars and additional isolates, as well as the rest of Viral RefSeq (v.230). TayeBlu clustered with 5,378 Caudoviricetes viruses, with 534 of these in novel_order_21. No additional ICTV phages were assigned to TayeBlu's family, novel_family_3 (Fig. S4; Table S3), indicating that this is a previously uncharacterized lineage within Caudoviricetes. Two of the four phages in TayeBlu's novel_subfamily_0 are known only from metagenomic sequence. By contrast, several phages of the 9NA group in sister clade novel_subfamily_1 have been characterized experimentally, and the members of this clade have been shown to be only distantly related to other known Siphoviridae (104). We propose that these two subfamilies constitute a novel phage family.
Analysis of novel_family_3 at the nucleotide level suggests substantial diversification both between and within the two subclades evident in the proteome tree. As expected, VIRIDIC intergenomic similarity analysis showed that the closest relative to TayeBlu was its top BLAST hit, sharing only 65.3% ANI (Fig. 5). Notably, TayeBlu exhibited minimal nucleotide similarity (≤10%) with the ViPTree-and vConTACT3-identified family members in the sister subclade. This observation is consistent with the growing recognition that phage taxonomy should rely primarily on protein conservation patterns, as the extent of diversification among related phage can obscure relationships at the nucleotide level (112,113) and, further, that family-level groupings represent broader genetic diversity among bacteriophage than among eukaryotic viruses (114). With current ICTV inter genomic similarity thresholds of ≥95% for the species level and ≥70% for the genus level, the highest intergenomic similarity observed here (67.6%) falls short of the genus-level threshold. However, proteome-based clustering places the uncharacterized Enterobacter phage Mulvp2_OR50 and Caudoviricetes sp. ctumj2 in the same genus (novel_genus_0) as TayeBlu, a discrepancy that reflects well-documented challenges in viral taxonomy (115). Whether or not continued improvements in phage taxonomy conclusively show these three phages to be congeneric, TayeBlu is the first phage of its genus and species to be characterized physiologically.
## Genome comparison and core gene identification
TayeBlu and its eight identified relatives have genomes of 41-59 kb with 69-100 predicted open reading frames (ORFs) (Fig. 6), a total of 799 ORFs across the family. We applied protein clustering (sequence identity ≥30%, coverage ≥80%) to identify this novel family's core genes, finding 397 unique clusters (Table S4; Fig. S5) of which 11 were present in all 9 examined genomes. These 11 core genes, ∼12% of the total ORFs in this phage family, constitute the family's conserved genetic framework. An additional five genes were shared between TayeBlu and all but one phage in the family, extending the near-core genome to 16 genes. Functional annotation of these conserved genes revealed that they primarily encode structural components (capsid and tail proteins), DNA packaging machinery (terminase and portal proteins), and enzymes involved in DNA replication (DNA helicase, polymerase, and dUTPase), alongside an endonuclease. Genome regions encoding structural tail proteins and DNA packaging machinery also show substantial synteny (Fig. 6).
The predominance of structural and assembly genes in the core genome is consistent with previous observations that these functional categories tend to be the most highly conserved genes across diverse phage lineages (116,117). The critical importance of these functions in the phage life cycle likely places these genes under higher selective pressure than the rest of the genome, as seen previously in tailed bacteriophage large terminase genes (reviewed in reference 103). The limited but syntenic core genome suggests that the essential framework for virion assembly and DNA replication has remained intact during adaptive radiation to the family's ecological niches. Notably absent from the core genome are genes involved in transcriptional regulation and host takeover functions (Fig. S6), suggesting diverse host interaction strategies. The conservation pattern observed in this novel phage family, with structural proteins, DNA packaging components, and replication machinery forming the stable core within a family marked by low overall genome similarity, is consistent with the observation that phages can maintain a stable core genome over extended periods while acquiring variable complements of genes through recombination from the phage pan-genome (118), highlighting the balance between functional conservation and adaptive diversification.
## Clustering with phages isolated from natural soil and rhizosphere
Finally, we investigated TayeBlu's genomic relatedness to vOTUs from environmental sampling, as collected in the PIGEONv2.0 database (92), using clustering based on average nucleotide identity and gene-sharing networks. In the complete database of 515,763 environmental vOTUs, no sequences clustered with TayeBlu at species-level thresholds (ANI ≥95% with ≥80% of sequences aligned). Protein-based clustering of the nine phages in TayeBlu's family with the 192,008 vOTUs from soil environments found 56,968 vOTUs in class Caudoviricetes, but of these only a single environmental vOTU, sampled in a Northern California saltwater wetland (119), clustered with TayeBlu at the family level (see Table S5 at https://zenodo.org/records/16696721). Together, these analyses indicate that TayeBlu represents a novel viral lineage. Soil viromes are known to harbor extensive genetic novelty (120,121), and this vast diversity is still largely unexplored (122). Even across different rhizosphere habitats, phages display exception ally diverse communities depending on plant species, soil type, and geographic location (122)(123)(124), with most identified viral sequences showing little similarity to known phages from any environment (123). This suggests highly specialized adaptation to ecological niches that are subject to plant-soil-microbe interactions. Strikingly, the environmental vOTUs examined here include 53,391 from the tomato rhizosphere, like TayeBlu itself. The lack of phage genomes closely related to TayeBlu even within this set underscores the limited representation of soil viral diversity in current databases and the pressing need for expanded efforts to catalog and study phages from diverse terrestrial ecosys tems.
## Conclusion
The novel phage TayeBlu, isolated on Azotobacter vinelandii strain OP from agricultural rhizosphere soil, is a member of a novel family and represents a distinct genus and species within this viral lineage. Its infection physiology varies dramatically depending on nutrient availability and particularly on the exogenous supply of fixed nitrogen to its facultatively diazotrophic host. We hypothesize that this variation will alter the phage's ecological impacts across different soils and agricultural practices, potentially influencing bacterial community structure and evolution (10,125), N flow within the rhizosphere, and ecosystem outputs from the globally significant process of biological N fixation. Genomic analysis reveals extensive diversification even among the few members of this phage family, highlighting the need for future isolation efforts to illuminate the ecological roles and impacts of phage in soil systems. Derek A. Smith http://orcid.org/0000-0002-3675-9153 Sarah C. Bagby http://orcid.org/0000-0002-1795-3161
## References
1. Banerjee, Van Der Heijden (2023) "Soil microbiomes and one health" *Nat Rev Microbiol*
2. Aqeel, Ran, Hu et al. (2023) "Plant-soil-microbe interactions in maintaining ecosystem stability and coordinated turnover under changing environmental conditions" *Chemosphere*
3. Graham (1992) "Stress tolerance in Rhizobium and Bradyrhizobium, and nodulation under adverse soil conditions" *Can J Microbiol*
4. Peoples, Brockwell, Herridge et al. (2009) "The contributions of nitrogen-fixing crop legumes to the productivity of agricultural systems" *Symbiosis*
5. St, Ormeño-Orrillo, Rogel et al. (2019) "La diversidad de rizobios nativos de México a la luz de la genómica" *Rev Mex Biodivers*
6. Soumare, Diedhiou, Thuita et al. (2020) "Exploiting biological nitrogen fixation: a route towards a sustainable agriculture" *Plants*
7. Li, Jia, Lv et al. (2017) "Nitrogen fertilization decreases the decomposition of soil organic matter and plant residues in planted soils" *Soil Biol Biochem*
8. Birt, Bonnett (2018) "Microbial mechanisms of carbon and nitrogen acquisition in contrasting urban soils" *Eur J Soil Biol*
9. (2026) *Research Article Microbiology Spectrum*
10. Fontaine, Mariotti, Abbadie (2003) "The priming effect of organic matter: a question of microbial competition?" *Soil Biol Biochem*
11. Koskella, Brockhurst (2014) "Bacteria-phage coevolution as a driver of ecological and evolutionary processes in microbial communities" *FEMS Microbiol Rev*
12. Yoshida, Morimoto, Kimura (2019) "Bacteria-virus interactions"
13. Luo, Wang, Li et al. (2022) "Viral community-wide auxiliary metabolic genes differ by lifestyles, habitats, and hosts" *Microbiome*
14. Hurwitz, Hallam, Sullivan (2013) "Metabolic reprogramming by viruses in the sunlit and dark ocean" *Genome Biol*
15. Wilhelm, Suttle (1999) "Viruses and nutrient cycles in the sea: viruses play critical roles in the structure and function of aquatic food webs" *BioScience*
16. Tong, Wang, Yu et al. (2023) "Viral lysing can alleviate microbial nutrient limitations and accumulate recalcitrant dissolved organic matter components in soil" *ISME J*
17. Sullivan, Lindell, Lee et al. (2006) "Prevalence and evolution of core photosystem II genes in marine cyanobacterial viruses and their hosts" *PLoS Biol*
18. Breitbart, Thompson, Suttle et al. (2007) "Exploring the vast diversity of marine viruses" *Oceanography*
19. Gregory, Solonenko, Espinoza et al. (2016) "Genomic differentiation among wild cyanophages despite widespread horizontal gene transfer" *BMC Genomics*
20. Fowler, Cohen (1948) "Chemical studies in host-virus interactions: IV. A method of determining nutritional requirements for bacterial virus multiplication" *J Exp Med*
21. Nilsson, Li, Hoetzinger et al. (2022) "Nutrient driven transcriptional changes during phage infection in an aquatic Gammaproteobacterium" *Environ Microbiol*
22. Suttle (2007) "Marine viruses-major players in the global ecosystem" *Nat Rev Microbiol*
23. Williamson, Fuhrmann, Wommack et al. (2017) "Viruses in soil ecosystems: an unknown quantity within an unexplored territory" *Annu Rev Virol*
24. Braga, Spor, Kot et al. (2020) "Impact of phages on soil bacterial communities and nitrogen availability under different assembly scenarios" *Microbiome*
25. Emerson, Roux, Brum et al. (2018) "Host-linked soil viral ecology along a permafrost thaw gradient" *Nat Microbiol*
26. Paez-Espino, Eloe-Fadrosh, Pavlopoulos et al. (2016) "Uncovering earth's virome" *Nature*
27. Delgado-Baquerizo, Oliverio, Brewer et al. (2018) "A global atlas of the dominant bacteria found in soil" *Science*
28. Rousk, Sorensen, Michelsen (2016) "Nitrogen transfer from four nitrogen-fixer associations to plants and soils" *Ecosystems*
29. Gittrich, Sanderson, Wainaina et al. (2025) "Isolation and characterization of 24 phages infecting the plant growth-promoting rhizobacterium Klebsiella sp" *PLoS One*
30. Pengra, Wilson (1958) "Physiology of nitrogen fixation by Aerobacter aerogenes" *J Bacteriol*
31. Noar, Bruno-Bárcena (2018) "Azotobacter vinelandii: the source of 100 years of discoveries and many more to come" *Microbiology*
32. Setubal, Santos, Goldman et al. (2009) "Genome sequence of Azotobacter vinelandii, an obligate aerobe specialized to support diverse anaerobic metabolic processes" *J Bacteriol*
33. Karmakar, Kumar, Roy et al. (2023) "Comparative genomics reveals the diversity of CRISPR-Cas locus in Azotobacter organisms" *Lett Appl Microbiol*
34. Russell, Garcia, Kaçar (2024) "A CRISPR interference system for engineering biological nitrogen fixation"
35. Duff, Wyss (1961) "Isolation and classification of a new series of Azotobacter bacteriophages" *J Gen Microbiol*
36. Domingo, Escarmis, Warner (1975) "Transcription of bacterio phage deoxyribonucleic acid. Comparison of Escherichia coli and Azotobacter vinelandii σ subunits" *J Biol Chem*
37. Davidson, Lawson, Sutherland (1977) "An alginate lysate from Azotobacter vinelandii phage" *J Gen Microbiol*
38. Van Twest, Kropinski (2009) "Bacteriophage enrichment from water and soil"
39. Poxleitner, Pope, Jacobs-Sera et al. (2018) "HHMI SEA-PHAGES phage discovery guide"
40. Carlson (2004) "Working with bacteriophages: common techniques and methodological approaches"
41. Hyman, St (2009) "Practical methods for determining phage growth parameters"
42. Holmfeldt, Howard-Varona, Solonenko et al. (2014) "Contrasting genomic patterns and infection strategies of two coexisting Bacteroidetes podovirus genera" *Environ Microbiol*
43. (2024) "R: a language and environment for statistical computing"
44. Wickham, Averick, Bryan et al. (2019) "Welcome to the tidyverse" *JOSS*
45. Harrell (2025) "Hmisc: Harrell miscellaneous"
46. Ucar, Pebesma, Azcorra (2018) "Measurement errors in R" *R J*
47. Wilke, Wiernik (2022) "ggtext: improved text rendering support for "ggplot2"
48. Arnholt, Evans (2023) "BSDA: basic statistics and data analysis. R package version 1"
49. Pedersen (2024) "patchwork: the composer of plots. R package version 1"
50. Dang, Howard-Varona, Schwenck et al. (2015) "Variably lytic infection dynamics of large Bacteroidetes podovirus phi38:1 against two Cellulophaga baltica host strains" *Environ Microbiol*
51. Howard-Varona, Roux, Dore et al. (2017) "Regulation of infection efficiency in a globally abundant marine Bacteriodetes virus" *ISME J*
52. Ackermann (2009) "isolation, characterization, and interactions"
53. Schroeder, Dobson, Rueden et al. (2021) "The ImageJ ecosystem: open-source software for image visualization, processing, and analysis" *Protein Sci*
54. Boeckman, Liu, Ramsey et al. (2023) "Phage DNA extraction, genome assembly, and genome closure"
55. Nanopore (2023)
56. Wick, Judd, Holt (2019) "Performance of neural network basecalling tools for Oxford Nanopore sequencing" *Genome Biol*
57. Danecek, Bonfield, Liddle et al. (2021) "Twelve years of SAMtools and BCFtools. Gigascience 10:giab008"
58. Brown, Pirrung, Mccue (2017) "FQC Dashboard: integrates FastQC results into a web-based, interactive, and extensible FASTQ quality control tool" *Bioinformatics*
59. Ewels, Magnusson, Lundin et al. (2016) "MultiQC: summarize analysis results for multiple tools and samples in a single report" *Bioinformatics*
60. Lee, Nguyen, Hayes et al. (2021) "Prowler: a novel trimming algorithm Oxford Nanopore sequence data" *Bioinformatics*
61. Kolmogorov, Yuan, Lin et al. (2019) "Assembly of long, errorprone reads using repeat graphs" *Nat Biotechnol*
62. Wick, Schultz, Zobel et al. (2015) "Bandage: interactive visualization of de novo genome assemblies" *Bioinformatics*
63. Wick (2023) "Yet another ONT accuracy test"
64. Aroney, Newell, Nissen et al. (2025) "CoverM: read alignment statistics for metagenomics" *Bioinformatics*
65. Nayfach, Camargo, Schulz et al. (2021) "CheckV assesses the quality and completeness of metagenomeassembled viral genomes" *Nat Biotechnol*
66. Chung, Walter, Yang et al. (2017) "Predicting genome terminus sequences of Bacillus cereus-group bacteriophage using next generation sequencing data" *BMC Genomics*
67. Larkin, Blackshields, Brown et al. (2007) "Clustal W and Clustal X version 2.0"
68. Paradis, Strimmer (2004) "APE: analyses of phylogenetics and evolution in R language" *Bioinformatics*
69. Delcher, Bratke, Powers et al. (2007) "Identifying bacterial genes and endosymbiont DNA with Glimmer" *Bioinformatics*
70. Hyatt, Chen, Locascio et al. (2010) "Prodigal: prokaryotic gene recognition and translation initiation site identification" *BMC Bioinformatics*
71. Noguchi, Taniguchi, Itoh (2008) "MetaGeneAnnotator: detecting species-specific patterns of ribosomal binding site for precise gene prediction in anonymous prokaryotic and phage genomes" *DNA Res*
72. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol*
73. Gasteiger, Jung, Bairoch (2001) "SWISS-PROT: connecting biomolecular knowledge via a protein database" *Curr Issues Mol Biol*
74. Zimmermann, Stephens, Nam et al. (2018) "A completely reimplemented MPI bioinformatics toolkit with a new HHpred server at its core" *J Mol Biol*
75. Bouras (2024) "Phold-phage annotation using protein structures"
76. Bouras, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics*
77. Van Kempen, Kim, Tumescheit et al. (2024) "Fast and accurate protein structure search with Foldseek" *Nat Biotechnol*
78. Heinzinger, Weissenow, Sanchez et al. (2024) "Bilingual language model for protein sequence and structure" *NAR Genom Bioinform*
79. Mirdita, Schütze, Moriwaki et al. (2022) "ColabFold: making protein folding accessible to all" *Nat Methods*
80. Terzian, Ndela, Galiez et al. (2021) "PHROG: families of prokaryotic virus proteins clustered using remote homology" *NAR Genom Bioinform*
81. Ramsey, Rasche, Maughmer et al. (2020) "Galaxy and Apollo as a biologist-friendly interface for high-quality cooperative phage genome annotation" *PLoS Comput Biol*
82. Naville, Ghuillot-Gaudeffroy, Marchais et al. (2011) "ARNold: a web tool for the prediction of Rho-independent transcription terminators" *RNA Biol*
83. Shaffer, Borton, Mcgivern et al. (2020) "DRAM for distilling microbial metabolism to automate the curation of microbiome function" *Nucleic Acids Res*
84. Zhu, Tang, Wang et al. (2024) "A comparative analysis of phage classification methods in light of the recent ICTV taxonomic revisions" *Virology*
85. Nishimura, Yoshida, Kuronishi et al. (2017) "ViPTree: the viral proteomic tree server" *Bioinformatics*
86. Bolduc, Zablocki, Turner et al. (2025) "Scalable and systematic hierarchical virus taxonomy with vConTACT3"
87. Yu, Smith, Zhu et al. (2017) "GGTREE: an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data" *Methods Ecol Evol*
88. Moraru, Varsani, Kropinski (2020) "VIRIDIC-a novel tool to calculate the intergenomic similarities of prokaryote-infecting viruses" *Viruses*
89. Galili (2015) "dendextend: an R package for visualizing, adjusting and comparing trees of hierarchical clustering" *Bioinformatics*
90. Larralde, Camargo (2023) "Pyrodigal-gv: a pyrodigal extension to predict genes in giant viruses and viruses with alternative genetic code"
91. Gilchrist, Chooi "2021. clinker & clustermap.js: automatic generation of gene cluster comparison figures" *Bioinformatics*
92. Wickham (2016) "ggplot2: elegant graphics for data analysis"
93. Horst, Fudyma, Sones et al. (2023) "Dispersal, habitat filtering, and eco-evolutionary dynamics as drivers of local and global wetland viral biogeography" *ISME J*
94. Turner, Shkoporov, Lood et al. (2023) "Abolish ment of morphology-based taxa and change to binomial species names: 2022 taxonomy update of the ICTV bacterial viruses subcom mittee" *Arch Virol*
95. Schwartz (1976) "The adsorption of coliphage lambda to its host: effect of variations in the surface density of receptor and in phagereceptor affinity" *J Mol Biol*
96. Lourenço, Chaffringeon, Lamy-Besnier et al. (2022) "The gut environment regulates bacterial gene expression which modulates susceptibility to bacteriophage infection" *Cell Host Microbe*
97. Sabra, Zeng, Lünsdorf et al. (2000) "Effect of oxygen on formation and structure of Azotobacter vinelandii alginate and its role in protecting nitrogenase" *Appl Environ Microbiol*
98. Martínez-Salazar, Moreno, Nájera et al. (1996) "Characterization of the genes coding for the putative sigma factor AlgU and its regulators MucA, MucB, MucC, and MucD in Azotobacter vinelandii and evaluation of their roles in alginate biosynthesis" *J Bacteriol*
99. Forde, Fitzgerald (1999) "Analysis of exopolysaccharide (EPS) production mediated by the bacteriophage adsorption blocking plasmid, pCI658, isolated from Lactococcus lactis ssp. cremoris HO2" *Int Dairy J*
100. Wu, Herold, Knoshaug et al. (2019) "Fluxomic analysis reveals central carbon metabolism adaptation for diazotroph Azotobacter vinelandii ammonium excretion" *Sci Rep*
101. Inomura, Bragg, Follows (2017) "A quantitative analysis of the direct and indirect costs of nitrogen fixation: a model based on Azotobacter vinelandii" *ISME J*
102. Zimmerman, Howard-Varona, Needham et al. (2020) "Metabolic and biogeochemical consequences of viral infection in aquatic ecosystems" *Nat Rev Microbiol*
103. Casjens, Gilcrease (2009) "Determining DNA packaging strategy by analysis of the termini of the chromosomes in tailed-bacteriophage virions"
104. Casjens (2005) "Comparative genomics and evolution of the tailedbacteriophages" *Curr Opin Microbiol*
105. Zeng, Gilcrease, Hendrix et al. (2019) "DNA packaging and genomics of the Salmonella 9NA-like phages" *J Virol*
106. Anderson, Sullivan, Fernando (2017) "Dietary energy drives the dynamic response of bovine rumen viral communities" *Microbiome*
107. Cohen, Varki (2014) "Modulation of glycan recognition by clustered saccharide patches" *Int Rev Cell Mol Biol*
108. Rutbeek, Rezasoltani, Patel et al. (2021) "Molecular mechanism of quorum sensing inhibition in Streptococcus by the phage protein paratox" *J Biol Chem*
109. Duplantier, Lohou, Sonnet (2021) "Quorum sensing inhibitors to quench P. aeruginosa pathogenicity" *Pharmaceuticals*
110. Rybak, Ekemezie, Sullivan et al. (2024) "A new family of bacterial ribosome hibernation factors" *Nature*
111. Turner, Kropinski, Adriaenssens (2021) "A roadmap for genome-based phage taxonomy" *Viruses*
112. Mihara, Nishimura, Shimizu et al. (2016) "Linking virus genomes with host taxonomy" *Viruses*
113. Hyde, Herring, Hope et al. (2024) "Diversity and conservation of the genome architecture of phages infecting the Alphaproteobacteria" *Microbiol Spectr*
114. Jang, Bolduc, Zablocki et al. (2019) "Taxonomic assignment of uncultivated prokaryotic virus genomes is enabled by gene-sharing networks" *Nat Biotechnol*
115. Simmonds, Aiewsakun (2018) "Virus classification-where do you draw the line?" *Arch Virol*
116. Caetano-Anollés, Claverie, Nasir (2023) "A critical analysis of the current state of virus taxonomy" *Front Microbiol*
117. Mavrich, Hatfull (2017) "Bacteriophage evolution differs by host, lifestyle and genome" *Nat Microbiol*
118. Hatfull, Hendrix (2011) "Bacteriophages and their genomes" *Curr Opin Virol*
119. Bellas, Schroeder, Edwards et al. (2020) "Flexible genes establish widespread bacteriophage pan-genomes in cryoconite hole ecosystems" *Nat Commun*
120. Durham, Sieradzki, Horst et al. (2022) "Substantial differences in soil viral community composition within and among four northern California habitats" *ISME Commun*
121. Pratama, Terpstra, De Oliveria et al. (2020) "The role of rhizosphere bacteriophages in plant health" *Trends Microbiol*
122. Roux, Emerson (2022) "Diversity in the soil virosphere: to infinity and beyond" *Trends Microbiol*
123. Santos-Medellin, Zinke, Horst et al. (2021) "Viromes outperform total metagenomes in revealing the spatiotemporal patterns of agricultural soil viral communities" *ISME J*
124. Buée, Boer, Martin et al. (2009) "The rhizosphere zoo: an overview of plant-associated communities of microorganisms, including phages, bacteria, archaea, and fungi, and of some of their structuring factors" *Plant Soil*
125. Bi, Yu, Du et al. (2021) "Diversity and potential biogeochemical impacts of viruses in bulk and rhizosphere soils" *Environ Microbiol*
126. Weitz, Poisot, Meyer et al. (2013) "Phage-bacteria infection networks" *Trends Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12361357&blobtype=pdf | # Author 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, Scientifc Reports
## Abstract
As a result, in the Data availability section, "Hiroshi Nishiura should be contacted to request the data from this study. " now reads: "Access to the raw epidemiological data used in the analysis can be requested through an application to the |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12792578&blobtype=pdf | Jacob Gerstenberg, Grit Barten-Neiner, Florian Voit, Norbert Suttorp, Christoph Boesecke, Christian Hoffmann, Daiana Stolz, Mathias Pletz, Gernot Rohde, Martin Witzenrath, Marcus Panning, Andreas Essig, Jan Rupp, Olaf Degen, Christoph Stephan, Benjamin Schleenvoigt
## References
1. (2025) "12:15 PM Background. Community-acquired pneumonia (CAP) is a major cause of hospitalization among people living with HIV (PLWH)" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12810925&blobtype=pdf | ## Abstract
Alexandre da Costa Linhares graduated from the Faculty of Medicine of the Federal University of Pará in Belém in 1975 and completed his PhD in Parasite Biology at the Oswaldo Cruz Institute, Oswaldo Cruz Foundation (IOC/Fiocruz), Rio de Janeiro, in 2002. Linhares conducted pioneering studies on rotavirus in Brazil while still a medical student and later as an intern at the Evandro Chagas Institute (IEC) in Belém during the 1970s, where he remained until his retirement in 2019. He served as the head of the Virology section from 1987 to 2019 and as director of the IEC between 1981 and 1987. In 1975, he carried out his initial investigation of rotavirus in children with acute diarrhea treated at the Hospital da Santa Casa de Misericórdia do Pará in Belém.Rotavirus was detected in the feces of four children out of the 25 examined. The agent was identifi ed using electron microscopy at the Bernard Nocht Institute in Hamburg, Germany, and via enzymelinked immunosorbent assay at the IEC. In 1977, Dr. Linhares demonstrated that rotavirus was responsible for an explosive epidemic of acute diarrhea among indigenous individuals of the Tiriyó tribe, located in the northern part of the state of Pará 1 . The epidemic occurred in July and August of that year, aff ecting 157 (70%) of the 224 inhabitants of the main village, with one recorded death in a 1-year-old child.
This was the first epidemic of acute diarrhea caused by rotavirus in indigenous tribes in the Americas. The epidemic aff ected the main village of the Tiriyó tribe, as well as other smaller villages of this indigenous group 2 . In 1978, Dr. Linhares completed an internship at the East Birmingham Hospital in the United Kingdom under the guidance of Professor Thomas Henry Flewett, during which rotavirus samples detected in the state of Pará were analyzed. One of these samples was identifi ed as serotype 1 (Birmingham), the strain responsible for the epidemic in the Tiriyó tribe. In 1984, he completed another study in the UK in Colindale under the guidance of Professor Flewett 3 . These studies were later expanded to include other urban areas and additional viral agents associated with diarrhea, such as the Norwalk agent (currently norovirus), astrovirus, calicivirus, and others 4 . Linhares also made important contributions to clinical studies of rotavirus vaccines, and the trials conducted by the group he led were the only clinical vaccine studies conducted in Brazil 5 . Dr. Francisco Pinheiro, former director of the IEC (1979-1981), was the mentor of Dr. Linhares since his days as a medical student, and his testimony is remarkable: "Linhares had an impressive simplicity and scientifi c Alexandre da Costa Linhares (1952 2025) intuition, and a stoic dedication to studies with children with acute diarrhea." Professor James LeDuc of the University of Texas Medical Branch, who worked with Linhares at the IEC in Belém, stated, "He was a kind and generous person, as well as being a superb clinician and scientist." According to virologist Pedro Vasconcelos, former director of the IEC (2014-2019) and past president of the Brazilian Society of Tropical Medicine, "Linhares had a vocation for medical-scientifi c research, being a gentleman and always very attentive to the students and technicians who collaborated with him." Alexandre Linhares also conducted important studies on acute hemorrhagic conjunctivitis, erythema infectiosum, and HTLV-I in the Amazon 6 . He also served as an editor and contributor to several scientifi c publications and was a member of multiple international editorial boards. His scientifi c legacy includes 225 articles published in journals, three books, and 39 book chapters, in addition to contributing to the training of dozens of researchers. José Paulo Gagliardi Leite, Public Health Researcher and former director of IOC/Fiocruz, and advisor of Alexandre Linhares' doctoral thesis, described their many fruitful years of collaboration as follows: "Linhares was a great friend, a unique person, owing to his professional and personal qualities.
As a person, he was a gentleman, always available to listen and, when necessary, to speak. As a professional, he was always at the forefront of studies on gastroenteric viruses, particularly rotavirus A, one of his great passions. Not only did he describe rotavirus A for the fi rst time in Brazil, but he was also a pioneer in clinical studies with rotavirus vaccines, including Rotarix® 7 , which has been part of the national immunization program since March 2006". He is survived by his wife, Suely; his children Leonardo, Adriana, and Alexandre; nine grandchildren; and fi ve siblings. [1] and Pedro Fernando Vasconcelos
## Francisco de Paula Pinheiro
## References
1. Linhares, Pinheiro, Freitas et al. (1981) "An outbreak of Rotavirus Diarrhoea among a nonimmune isolated South American Indian Community" *Am J Epidemiol*
2. Linhares, Gabbay, Mascarenhas et al. (1996) "Immunigenicity, Safety and Efficacy of tetravalent Rhesus-human, Reassortant Rotavirus vaccine in Belém" *Brazil. Bull World Health Org*
3. Linhares, Gabbay, Mascarenhas et al. (1988) "Epidemiology of Rotavirus subgroups and serotypes in Belém, Brazil: A Three-year study" *Annales Virol. (Institut Pasteur)*
4. Linhares, Gabbay, Costa et al. (2005) "Prevalence and genetic diversity of Astroviruses in children with diarrheain São Luís" *Mem Inst Oswaldo Cruz*
5. Salinas, Perez-Schael, Linhares et al. (2005) "Evaluation of safety, immunogenicity and efficacy of an attenuated Rotavirus vaccine, RIX4414: A randomized, placebo-controlled trial in Latin American infants" *Pediatr Infect Dis J*
6. Freitas, Freitas, Linhares (2004) "Human Herpesvirus-7 as a cause of exanthematous illnesses in Belém, Pará, Brazil" *Rev Inst Med Trop São Paulo*
7. Linhares (2004) "Rotavirus vaccine for prophylaxis against rotavirus gastroenteritis" *Pediatr Infect Dis J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12652929&blobtype=pdf | # COL10A1 Overexpression Promotes Gastric Cancer Aggressiveness Through EMT and Major Oncogenic Pathways
Laura Necula, Denisa Dragu, Lilia Matei, Ioana Pitica, Simona Dima, Coralia Bleotu, Carmen Diaconu, Mihaela Chivu-Economescu
## Abstract
Gastric cancer (GC) remains a major cause of cancer-related mortality, with limited options for early detection and precision therapy. Collagen family members are increasingly recognized as key structural and regulatory components of the tumor microenvironment.Collagen type X alpha 1 chain (COL10A1) appears among the top overexpressed genes in GC and has been linked with tumorigenesis, but its functional role in GC has not been completely elucidated. The oncogenic potential of COL10A1 was assessed in vitro in GC cell lines using adenoviral-mediated overexpression. Functional assays were further performed to evaluate proliferation, apoptosis, migration, invasion, and epithelial-mesenchymal transition (EMT) markers. Intracellular signaling alterations were analyzed by phosphokinase protein profiling and protein-protein interaction network analysis. COL10A1 overexpression significantly increased proliferation and migration, while reducing GC cell apoptosis. It promotes EMT by up-regulating mesenchymal markers (N-cadherin, Vimentin, Snail/Slug) and suppressing epithelial markers such as E-cadherin and β-catenin. Additionally, COL10A1 overexpression activated oncogenic signaling pathways, including the JNK and MAPK cascades, increasing proliferation and tumorigenic potential. Our results showed that COL10A1 functions as a driver for tumor progression by promoting proliferation, migration, and invasion along with EMT through activation of important oncogenic pathways. These findings highlight its biological role in tumor progression and contribute to a better understanding of GC pathogenesis.
## 1. Introduction
Despite advances in targeted cancer therapy and immunotherapy, GC still represents a major global health concern, being the fifth leading cause of cancer deaths, and the fifth most frequently diagnosed cancer [1]. The poor overall prognosis of GC is primarily attributed to diagnosis in advanced stages, resulting from the late onset of symptoms and the absence of reliable biomarkers with high specificity and sensitivity. The standard curative approach includes gastrectomy and lymphadenectomy, often enhanced by perioperative chemotherapy and radiotherapy [2].
The GC tumor microenvironment is a heterogeneous network of cancer cells, cancerassociated fibroblasts, collagen fibers, and other extracellular matrix components, such as immune cells, endothelial cells, and a complex network of cytokines, chemokines, and growth factors that interact to collectively promote tumor growth, invasion, metastasis, and resistance to therapy [3].
The collagen family represents the most abundant proteins in mammals, containing 28 members that are associated with the occurrence, progression, and prognosis of several cancer types, such as breast, colorectal, gastric, lung, and cervical carcinomas [4]. The involvement of collagens in carcinogenesis appears to occur within the tumor microenvironment, where these proteins form a structural framework that modulates immune responses, angiogenesis, and tumor progression [5].
Our previous studies showed that the collagen family and other proteins associated with the assembly mechanism of collagen fibers and with their degradation play an important role during the carcinogenesis process, and could be important biomarkers that predict poor prognosis in GC [6]. COL10A1 from tissue and its soluble form in plasma were previously identified by us in a series of studies as being associated with early GC and tumor progression [7].
This study aims to elucidate the role of COL10A1 in GC pathogenesis, focusing on the effects associated with its overexpression. Our study provides a detailed characterization of its functional consequences, as this alteration is predominant in the analyzed cohorts and in publicly available datasets (TCGA, GEO, etc.), and is significantly correlated with an aggressive tumor phenotype.
## 2. Results
## 2.1. COL10A1 Is Overexpressed in GC
Bioinformatic analysis of existing data showed that COL10A1 is markedly overexpressed in multiple cancers, with particularly strong upregulation in GC, also known as stomach adenocarcinoma (STAD), compared to normal tissues (Figure 1A,B). Elevated expression is observed across all clinical stages of GC, suggesting its involvement throughout tumor progression (Figure 1C). Survival analysis further demonstrates that patients with high COL10A1 expression have significantly poorer overall survival (Figure 1D). These results suggest that COL10A1 may serve as a prognostic biomarker and therapeutic target in GC.
## 2.2. Upregulation of COL10A1 Expression by Cell Transduction
To evaluate the effect of COL10A1 overexpression on tumor processes, we first analyzed COL10A1 expression in GC cell lines Hs746T, AGS, and NCI-N87. We selected these three gastric cell lines based on our experience with in vitro GC studies and because they represent distinct molecular subtypes of GC. Hs746T cells exhibit a mesenchymal phenotype with high invasive potential [8]; NCI-N87 cells display epithelial features and HER2 overexpression [9], while AGS cells represent a moderately differentiated, intestinal-type model [10]. Our investigation showed that COL10A1 mRNA expression was significantly lower in the Hs746T cell line compared with NCI-N87 (p = 0.0004) and AGS (p = 0.0004) (Figure 2A), therefore we selected this line for subsequent transduction experiments. In the next step, we tested different concentrations of COL10A1 ADV to optimize the cell transduction protocol. The results showed that the use of the COL10A1 ADV system induced a 26-fold increase in COL10A1 gene expression in the Hs746T cell line compared to the control cells treated with NULL ADV (p < 0.0001), at 48 h post-transduction, using COL10A1 ADV stock concentration (Figure 2B). This transduction protocol was selected for the following experiments that analyzed the effects induced by COL10A1 overexpression on the functionality of gastric tumor cells: evaluation of proliferative capacity, apoptosis, invasiveness capacity, EMT markers, and analysis of the main signaling pathways involved in gastric carcinogenesis.
## 2.3. COL10A1 Overexpression Promotes Gastric Cancer Cell Proliferation and Decreases Apoptosis
The upregulation of COL10A1 gene expression increased the proliferation capacity of the Hs746T GC cell line when compared with control cells. Using an MTS-based assay, a significant increase in cell proliferation capacity (48%) was noted mainly at 72 h posttransduction (p = 0.004) (Figure 3A). Consistent with the increased proliferation, a decrease in apoptosis in Hs746T cells treated with COL10A1 ADV compared with untreated cells at 72 h post-transduction was observed using the Incucyte ® Live-Cell Analysis device (Sartorius, Göttingen, Germany) (Figure 3B).
## 2.4. COL10A1 Overexpression Increases Gastric Cancer Cells Migration and Invasion via EMT
The impact of COL10A1 overexpression on cell migration and invasion was evaluated using the QCM ECMatrix Cell Invasion kit (Thermo Fisher Scientific, Waltham, MA, USA), and the results showed an increase in invasiveness by 8% in the case of Hs746T GC cells treated with COL10A1 ADV at 72 h after transduction compared with control cells (Figure 4A). To further assess the impact of COL10A1 overexpression on cell migration and invasion, we also analyzed the expression of the main genes involved in the EMT process, E-cadherin, N-cadherin, Snail + Slug, and Vimentin as important factors involved in promoting metastasis, using the RT-PCR technique. The results obtained showed that in Hs746T cells treated with COL10A1 ADV, at 48 h post-transduction, a significant increase was observed in the expression of the SNAIL, CDH2 (N-cadherin), TIMP, and Vimentin genes, genes that are markers for mesenchymal cells. A concomitant decrease in the expression of CDH1 (E-cadherin) and CTNNB1 (β-catenin) genes, markers for epithelial cells, was also observed (Figure 4B). These results indicate that COL10A1 promotes EMT and enhances the migratory and invasive potential of GC cells.
## 2.5. Analysis of the Effects of COL10A1 Overexpression on Intracellular Signaling Pathways
To investigate the intracellular mechanisms underlying COL10A1-mediated tumor progression, phosphokinase profiling was performed on Hs746T GC cells infected with COL10A1 ADV and compared with control cells. The results revealed significant alterations in the phosphorylation of several proteins (Figure 5A,B). COL10A1 overexpression led to increased phosphorylation and activation of Akt1/2/3, c-Jun, JNK1/2/3, and Lck, while a reduction was observed for eNOS and HSP27. The PPI enrichment analysis showed a high interaction between the activated kinases (STRING p = 0.000427) (Figure 5C,D). The reports showed functional enrichments and positive regulation of the JNK and MAPK cascade, pathways known to drive cell proliferation, survival, and tumorigenic potential.
## 3. Discussion
GC is a heterogeneous disease resulting from the accumulation of numerous genetic and epigenetic alterations, leading to the dysregulation of oncogenic and/or tumor suppressor signaling pathways. Diagnosis in the early stages of the tumor is hindered by the lack of circulating biomarkers with high sensitivity and specificity, and the fact that standard GC diagnosis relies mainly on invasive procedures, such as upper digestive endoscopy. In this context, the development of minimally or non-invasive diagnostic tests, as well as the identification of new biomarkers for GC with high specificity and sensitivity, remains essential. Also, the identification and validation of therapeutic targets could eliminate a large part of the side effects of currently used drugs. In previous studies, our team identified significant increases in the expression of several genes, including COL10A1, in gastric tumors compared to adjacent normal tissues. COL10A1 has also been identified as a potential biomarker for early diagnosis of GC, as its increased expression occurs in early tumor stages and remains elevated during cancer progression [11]. Moreover, in another study, we showed that high plasma levels of COL10A1 are associated with advanced tumor stage in GC patients, and the elevated expression occurs from the beginning of carcinogenesis, in the early stages, and its increased level remains elevated during cancer progression [12]. These results are consistent with other studies that identified an overexpression of COL10A1 in urothelial bladder, breast, colorectal, and pancreatic cancers, associated with tumor progression and poor prognosis [13][14][15][16]. Using high-throughput mRNA sequencing (RNA-seq), Tingting Li et al. identified COL10A1 as the gene with the second highest expression level in both stage I and stage IV of GC [17]. Therefore, in our study, we used the TCGA database to evaluate the expression of COL10A1 in gastric tumor tissue compared with normal adjacent tissue, and we also analyzed the COL10A1 prognosis value in GC patients. The analysis showed that COL10A1 is up-regulated in gastric tumor tissue, and an increased expression is associated with poor prognosis.
Recent studies, using single-cell RNA sequencing (scRNA-seq) analyses, have revealed that COL10A1 is predominantly overexpressed in the tumor stroma of several solid cancers, including breast, pancreatic, and gastrointestinal tumors. Its expression is restricted to matrix-producing cancer-associated fibroblasts (CAFs), rather than cancer cells, and these COL10A1-positive CAFs display immunosuppressive and pro-metastatic properties. In basal cell carcinoma (BCC), scRNA-seq data from two independent datasets identified COL10A1-expressing stromal cells adjacent to infiltrative tumor regions, characterized by extracellular matrix remodeling features [18]. Similarly, in colorectal cancer, a COL10A1-positive fibroblast subpopulation (COL10A1 + Fib) has been associated with tumor progression and poor prognosis in patients [19]. Moreover, in epithelial cancer cells, COL10A1 secreted by CAFs promotes EMT, thereby enhancing migration and invasion, and also induces M2 macrophage polarization, contributing to an immunosuppressive microenvironment [19]. Therefore, we chose to test in vitro the behavior of GC cells in the presence of an external source of COL10A1.
Accordingly, we performed in vitro experiments demonstrating that an increase in COL10A1 expression decreases apoptosis and promotes proliferation, migration, and invasion by activating the expression of several genes involved in EMT, such as SNAIL, CDH2 (N-cadherin), TIMP, and Vimentin. Although significant changes in EMT markers were observed, functional assays of migration and invasion did not reach statistical significance. This may reflect partial EMT states, temporal differences between molecular changes and phenotypic manifestation, or assay sensitivity. Therefore, molecular EMT changes may precede or occur independently of measurable changes in cell motility and invasiveness.
Data available from the literature indicate that COL10A1 overexpression can affect the balance between tumor cell proliferation and apoptosis, thus sustaining the carcinogenesis process in different types of cancer. Thus, COL10A1 upregulation promotes cell proliferation and metastasis and inhibits apoptosis and autophagy in lung cancer cells [4]. In cervical cancer cells, COL10A1 upregulation promotes proliferation, migration, and EMT processes through activation of N-cadherin and Vimentin via TGF-β/Smad signaling [20]. COL10A1 overexpression can also influence immunotherapy response and resistance to radiotherapy and chemotherapy in prostate cancer patients through mechanisms involving endoplasmic reticulum stress [21].
Our findings show that COL10A1 overexpression also induced a significant increase in phosphorylation/activation levels of c-Jun, JNK, and 1/2/3Akt, proteins actively involved in cell proliferation, and a decrease in the level of phosphorylation/activation in the case of eNOS and HSP27 proteins. c-Jun phosphorylation, together with significant enrichment of the MAPK cascade, indicates activation of a pro-mitogenic signaling module that drives cell cycle gene expression via MAPK→AP-1. In parallel, reduced phosphorylation of HSP27-normally a regulator of actin dynamics, stress tolerance, and chaperone functionsuggests cytoskeletal remodeling that favors proliferation and weakens stress defenses. This pattern is consistent with a signaling shift toward dominant ERK activity (with relative suppression of p38/MK2) or increased phosphatase activity, explaining the concurrent increase in p-c-Jun and decrease in p-HSP27. In recent years, JNK has been increasingly recognized as an attractive molecular target for cancer treatment due to its involvement in the regulation of cellular processes associated with carcinogenesis, including proliferation, differentiation and survival of tumor cells. JNK seems to sustain EMT and enhance GC cells invasion and migration [22]. c-Jun is a transcription factor with oncogenic function activated by the Jun N-terminal kinase (JNK), with a central role in cellular signal transduction, positively regulating cell proliferation by inhibiting the expression and function of tumor suppressor genes. c-Jun can also sustain the transcription of genes associated with ECM components and facilitate cell growth and invasion in cancer [23]. The phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) signaling pathway is involved in multiple cellular processes, including cell survival, proliferation, differentiation, metabolism, and cytoskeletal reorganization. Frequent activation of AKT has been reported in approximately 78% of GC cases and is associated with a poor prognosis in patients diagnosed with this type of cancer [24]. Thus, activation of these signaling pathways in Hs746T cells treated with COL10A1 ADV is at the basis of the increased proliferative capacity of cells in which the COL10A1 gene was overexpressed. Although our results provide novel insights into the role of COL10A1 in GC, further validation in additional GC cell lines, including AGS and NCI-N87, as well as GC-CAF co-culture systems, is necessary to strengthen the conclusions and confirm their broader relevance. Future studies should also include pathway inhibition experiments to determine whether the effects of COL10A1 are mediated directly through these signaling pathways.
## 4. Materials and Methods
## 4.1. Analysis of Databases
The expression levels of COL10A1 gene in different types of solid tumors and pathological stages of GC were assessed with UALCAN web tool (http://ualcan.path.uab.edu/, accessed on 26 September 2025), based on TCGA online available RNAseq data [25,26]. Differential mRNA expression analysis includes normal tissues and all TNM stages of GC. The GEPIA web tool (http://gepia.cancer-pku.cn/, accessed on 26 September 2025) was used to show differences in gene expression between tumor and normal tissues using RNA sequencing data from the TCGA and GTEx databases [27]. The prognostic significance of COL10A1 in GC was analyzed using the Kaplan-Meier Plotter web tool (https://kmplot.com/analysis/, accessed on 26 September 2025). The median level of COL10A1 was utilized to categorize 875 patients with GC into high and low COL10A1 groups. The two patient cohorts are compared by a Kaplan-Meier survival plot, and the hazard ratio with 95% confidence intervals and log-rank p value are calculated [28]. The p-value for Student's t-test was set as follows: * p < 0.05, ** p < 0.01,*** p < 0.001.
## 4.2. Cell Culture
Human gastric adenocarcinoma cell lines AGS (catalog no. CRL-1739), NCI-N87 (catalog no. CRL-5822), and Hs746T (catalog no. HTB-135) were purchased from the American Type Culture Collection (Manassas, VA, USA). AGS cells were cultured in Ham's F12 medium (Sigma Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (Biochrom, Cambridge, UK) at 37 • C in 5% CO 2 . The NCI-N87 cells were cultured in RPMI 1640 medium (Sigma Aldrich), and the Hs746T cells were cultured in Dulbecco's Modified Eagle's Medium (Sigma Aldrich), under the same culture conditions.
## 4.3. Cell Transduction
For the experiments of COL10A1 gain of function in GC cell lines, a viral vector-based system was used (adenovirus) containing the COL10A1 gene sequence (COL10A1 ADV) (Applied Biological Materials (abm) Inc., Richmond, BC, Canada), which has the ability to transiently induce the overexpression of the COL10A1 gene. The recombinant adenoviral vector used is based on a replication-incompetent human adenovirus serotype 5 (Ad5) system lacking the E1 and E3 regions but retaining essential packaging signals and inverted terminal repeats. The absence of E1 prevents viral replication and cell cycle progression, while deletion of E3 reduces host immune suppression. Virus production and amplification were carried out in cells that express E1 gene products (such as HEK293 cells). Thus, COL10A1 ADV was amplified in HEK 293 cells, plated at 60-70% confluency. When more than 95% of HEK 293 cells were detached from the dishes, the cells and medium were collected, freeze-thawed three times, and centrifuged at 3000 rpm at room temperature for 10 min. The supernatant was collected and used for cell transduction as COL10A1 ADV stock. Cell transduction was performed in Hs746T cell line and COL10A1 ADV stock, and serial decimal dilutions from 10 -1 to 10 -5 were tested and evaluated at 24, 48, and 72 h intervals to identify the optimal concentration and the time required for transduction. Gene transduction was performed according to the producer protocol, and the transduction efficiency was evaluated by qPCR.
## 4.4. Quantitative Real-Time PCR Analysis
Evaluation of COL10A1 gene expression was performed by the qRT-PCR technique. Cells were harvested 48 h after transduction, and total RNA was extracted with Tri Reagent (Sigma Aldrich). High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Waltham, MA, USA) was used for cDNA synthesis starting from 2 µg RNA while the quantitative RT-PCR reactions were performed on ABI 7300 Real-Time PCR System (Applied Biosystems) using pre-validated Taqman Gene Expression Assays kits (Applied Biosystems) and Maxima SYBR Green/ROX qPCR Master Mix for gene-specific primers (Thermo Fisher Scientific, Waltham, MA, USA). The primers used are obtained from OriGene Technologies Inc. (Rockville, MD, USA), and the primer sequences are described in Table 1. For COL10A1, we used the primer sequences described by Tingting Li et al. [17]. The PCR conditions were set as follows: 95 • C for 10 min; (95 • C for 15 s; 60 • C for 1 min), with 40 cycles and the results were analyzed with RQ study software (7300 System SDS v1.4, Applied Biosystems). The data were normalized to the housekeeping gene GAPDH transcripts and the comparative CT calculation (2 -∆∆Ct method) was employed to obtain relative expression of transcripts, with each reaction performed in triplicate.
## 4.5. Cell Proliferation
The cells were seeded in 96-well culture plates at a density of 6 × 10 3 cells/well, divided into 2 groups: COL10A1 ADV and NULL ADV. CellTiter 96 Aqueous One Solution (Promega, Madison, WI, USA) assay was used to evaluate the gene overexpression effect on cell proliferation at 48, 72, and 96 h after transduction. For each well, 20 µL MTS was added, and the plates were maintained for 4 h at 37 • C. The absorbance was measured at 550 nm using the TECAN GENios reader (Tecan Trading AG, Männedorf, Switzerland). All experiments were conducted in triplicate. Values were calculated as the percentage change induced by COL10A1 ADV as compared to control (mock) cells.
## Gene
Primer Sequence
$$COL10A1 F: 5 ′ -AAGAATGGCACCCCTGTAATGT-3 ′ R: 5 ′ -ACTCCCTGAAGCCTGATCCA-3 ′ TIE F: 5 ′ -GGTCAAGCAACCCAGCCTTTTC-3 ′ R: 5 ′ -CAGGTCATTCCAGCAGAGCCAA-3 ′ TIMP F: 5 ′ -GGAGAGTGTCTGCGGATACTTC-3 ′ R: 5 ′ -GCAGGTAGTGATGTGCAAGAGTC-3 ′ CTNNB1 F: 5 ′ -CACAAGCAGAGTGCTGAAGGTG-3 ′ R: 5 ′ -GATTCCTGAGAGTCCAAAGACAG-3 ′ SNAIL F: 5 ′ -TGCCCTCAAGATGCACATCCGA-3 ′ R: 5 ′ -GGGACAGGAGAAGGGCTTCTC-3 ′ CDH1 F: 5 ′ -GCCTCCTGAAAAGAGAGTGGAAG-3 ′ R: 5 ′ -TGGCAGTGTCTCTCCAAATCCG-3 ′ LCN2 F: 5 ′ -GTGAGCACCAACTACAACCAGC-3 ′ R: 5 ′ -GTTCCGAAGTCAGCTCCTTGGT-3 ′ CDH2 F:5 ′ -CCTCCAGAGTTTACTGCCATGAC-3 R: 5 ′ -GTAGGATCTCCGCCACTGATTC-3 ′ VIM F: 5 ′ -AGGCAAAGCAGGAGTCCACTGA-3 ′ R: 5 ′ -ATCTGGCGTTCCAGGGACTCAT-3 ′$$
## 4.6. Apoptosis
Apoptosis analysis in genetically modified cells was performed by analyzing the expression level of Caspase 3/7, using the Incucyte ® Live-Cell Analysis device (Sartorius, Göttingen, Germany). For this purpose, we used CellEvent Caspase-3/7 Green ReadyProbes Reagent (Thermo Fisher Scientific) for real-time discrimination in culture between live and dead cells. For this analysis, 3500 cells/well were seeded in a 96-well plate, the experiment being performed in triplicate. 24 h after seeding, the treatment with COL10A1 ADV was performed on Hs746T cells, and the results were monitored in real time, comparing the behavior of infected cells with control (mock) cells.
## 4.7. Cell Migration and Invasion
In vitro testing of invasiveness based on the ability of tumor cells to modify extracellular matrix components was performed using the QCM ECMatrix Cell Invasion kit (Thermo Fisher Scientific). For this analysis, 3500 Hs746T cells/well were seeded in a 96-well plate. 48 h after the treatment with COL10A1 ADV cells were starved for 6 h, trypsinized, and resuspended at a density of 3 × 10 5 cells/well in serum-free medium in the 8.0 micron pore dishes inserted in a 24-well plate. In the lower part of the well, a medium containing 10% fetal bovine serum was added, and the system was maintained for 24 h at 37 • C (5% CO 2 ). The cells that had migrated to the lower part of the insert were enzymatically detached, lysed, and stained for 15 min with DNA-binding dye, CyQuant GR (Thermo Fisher Scientific), and the results were obtained by reading the fluorescence at 480/520 nm (excitation/emission) using a Wallac Victor 2 spectrophotometer (Perkin Elmer, Waltham, MA, USA).
## 4.8. Protein Profiler Analysis
Analysis of the main signaling pathways involved in gastric carcinogenesis was performed using the dot-blot technique Human Phospho-Kinase Array Kit (Proteome ProfilerTM Array, R&D Systems, Minneapolis, MN, USA), which allows the simultaneous analysis of the phosphorylation profiles of the main phosphokinases and their protein substrates, proteins that are involved in signaling pathways that are usually altered in tumor cells. The protocol was performed according to the manufacturer's protocol. For immunoblot analysis, 1 × 10 6 cells grown in 6-well plates were transduced with COL10A1 ADV or NULL ADV, and harvested in a lysis protein buffer after 48 h. 400 µg of total proteins diluted in the array buffer were incubated on the nitrocellulose membranes at 4 • C overnight. In the next step, the membranes were washed to remove unbound proteins and were incubated with a cocktail of biotinylated antibodies for 4 h. The final step has involved the repeated washing of the arrays and incubation with a streptavidin-HRP solution. The signals were detected using the chemiluminescent reaction, captured using MicroChemi 4.2 (Bio Imaging Systems, Jackson, MI, USA), and the images were analyzed using ImageJ 1.42 software (National Institute of Health, Bethesda, MD, USA) after subtraction of background levels (negative control) from sample signal levels and normalization to positive control signal to allow comparison between samples. Experiments were performed twice. Protein-Protein Interaction (PPI) Network Analysis was performed using STRING (https://string-db.org/, accessed on 1 October 2025) tool [29].
## 4.9. Statistical Analysis
All data were analyzed using GraphPad Prism 5.0. Results are presented as mean ± SD. Statistical significance was determined using Student's unpaired t-test or one-/two-way ANOVA followed by the appropriate post hoc tests: Tukey for multiple comparisons between GC cell lines, Dunnett for comparing COL10A1 ADV treatments with the NULL ADV control, and Bonferroni for proliferative assays. Differences in cell invasiveness were evaluated using an unpaired t-test. A p-value < 0.05 was considered statistically significant; * (p < 0.05), ** (p < 0.01), *** (p < 0.001), **** (p < 0.0001).
## 5. Conclusions
COL10A1 is markedly overexpressed in GC, with elevated levels that are associated with poor overall survival. Our findings reveal that COL10A1 acts as a functional driver of GC progression by enhancing key oncogenic processes, including proliferation, migration, invasion, and EMT. Moreover, the activation of major oncogenic pathways, such as PI3K/Akt/mTOR and JNK/MAPK signaling cascades, further supports its role in promoting tumor growth and survival. Together, these results highlight COL10A1 as a key element of the tumor microenvironment that contributes to aggressive GC phenotypes through remodeling of cellular signaling and EMT regulation. Overall, our results highlight COL10A1 as a promising molecular marker and potential therapeutic target, warranting additional studies to elucidate and confirm its role in promoting cell migration and invasion.
## References
1. Bray, Laversanne, Sung et al. (2024) "Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries" *CA Cancer J. Clin*
2. Sundar, Nakayama, Markar et al. (2025) "Gastric cancer" *Lancet*
3. Mou, Ge, Sheng et al. (2023) "Research progress on the immune microenvironment and immunotherapy in gastric cancer" *Front. Immunol*
4. Yi, Zhu, Zhu et al. (2024) "Oncogenic mechanisms of COL10A1 in cancer and clinical challenges (Review)" *Oncol. Rep*
5. Wang, Zhang, Qian et al. "Targeting collagen to optimize cancer immunotherapy" *Exp. Hematol. Oncol. 2025*
6. Chivu-Economescu, Necula, Matei et al. (2022) "Collagen Family and Other Matrix Remodeling Proteins Identified by Bioinformatics Analysis as Hub Genes Involved in Gastric Cancer Progression and Prognosis" *Int. J. Mol. Sci*
7. Necula, Matei, Dragu et al. (2022) "Collagen Family as Promising Biomarkers and Therapeutic Targets in Cancer" *Int. J. Mol. Sci*
8. Lee, Kim, Lee et al. (2018) "Selective Cytotoxicity of the NAMPT Inhibitor FK866 Toward Gastric Cancer Cells with Markers of the Epithelial-Mesenchymal Transition, Due to Loss of NAPRT" *Gastroenterology*
9. Sedighian, Abdi, Goleij et al. (2025) "In vitro evidence of HER2-selective cytotoxicity by scFv(Herceptin)-PE-STXA immunotoxin in gastric cancer cells" *Biochem. Biophys. Rep*
10. Guo, Wan, Li et al. "Natural products for gastric carcinoma prevention and treatment: Focus on their antioxidant stress actions in the Correa's cascade" *Phytomedicine*
11. Chivu Economescu, Necula, Dragu et al. (2010) "Identification of potential biomarkers for early and advanced gastric adenocarcinoma detection" *Hepato-Gastroenterol*
12. Necula, Matei, Dragu et al. (2020) "High plasma levels of COL10A1 are associated with advanced tumor stage in gastric cancer patients" *World J. Gastroenterol*
13. Wang, Bai, Zhang et al. (2023) "Prognostic value of COL10A1 and its correlation with tumor-infiltrating immune cells in urothelial bladder cancer: A comprehensive study based on bioinformatics and clinical analysis validation" *Front. Immunol*
14. Zhou, Li, Gu et al. "High expression COL10A1 promotes breast cancer progression and predicts poor prognosis"
15. Kahlert, Shi, Strecker et al. "COL10A1 allows stratification of invasiveness of colon cancer and associates to extracellular matrix and immune cell enrichment in the tumor parenchyma"
16. Peng, Liu, Mao et al. (2024) "Upregulation of collagen type X alpha 1 promotes the progress of triple-negative breast cancer via Wnt/beta-catenin signaling" *Mol. Carcinog*
17. Li, Huang, Shi et al. (2018) "TGF-beta1-SOX9 axis-inducible COL10A1 promotes invasion and metastasis in gastric cancer via epithelial-to-mesenchymal transition" *Cell Death Dis*
18. Esposito, Yerly, Shukla et al. (2024) "COL10A1 expression distinguishes a subset of cancer-associated fibroblasts present in the stroma of high-risk basal cell carcinoma" *Br. J. Dermatol*
19. Hu, Ding, Lou et al. (2025) "A1( + ) fibroblasts promote colorectal cancer metastasis and M2 macrophage polarization with pan-cancer relevance" *J. Exp. Clin. Cancer Res*
20. Sun, Ling, Liu (2022) "Collagen type X alpha 1 promotes proliferation, invasion and epithelial-mesenchymal transition of cervical cancer through activation of TGF-beta/Smad signaling" *Physiol. Int*
21. Cen, Jiang, Lv et al. (2023) "Comprehensive analysis of the biological functions of endoplasmic reticulum stress in prostate cancer" *Front. Endocrinol*
22. Wang, Zhao, Chen et al. (2021) "NKCC1 promotes proliferation, invasion and migration in human gastric cancer cells via activation of the MAPK-JNK/EMT signaling pathway" *J. Cancer*
23. Lee, Kim, Kang et al. (2021) "Upregulation of LAMB1 via ERK/c-Jun Axis Promotes Gastric Cancer Growth and Motility" *Int. J. Mol. Sci*
24. Kang, Chau "Molecular target: Pan-AKT in gastric cancer"
25. Chandrashekar, Karthikeyan, Korla et al. (2022) "UALCAN: An update to the integrated cancer data analysis platform" *Neoplasia*
26. Chandrashekar, Bashel, Balasubramanya et al. (2017) "UALCAN: A Portal for Facilitating Tumor Subgroup Gene Expression and Survival Analyses" *Neoplasia*
27. Tang, Li, Kang et al. (2017) "GEPIA: A web server for cancer and normal gene expression profiling and interactive analyses" *Nucleic Acids Res*
28. Gyorffy "Integrated analysis of public datasets for the discovery and validation of survival-associated genes in solid tumors"
29. Szklarczyk, Gable, Lyon et al. (2019) "STRING v11: Protein-protein association networks with increased coverage, supporting functional discovery in genome-wide experimental datasets" *Nucleic Acids Res*
30. "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=PMC12362977&blobtype=pdf | # Characterization of a MERS-related betacoronavirus in Danish brown long-eared bats (Plecotus auritus)
Camille Johnston, Vithiagaran Gunalan, Hans Baagøe, Anna Fomsgaard, Charlotta Polacek, Morten Rasmussen, Louise Lohse, Thomas Rasmussen
## Abstract
Background Bats are recognized as natural reservoir hosts for numerous viruses and are believed to be the evolutionary origin of alpha-and beta-coronaviruses (CoVs), such as SARS-CoV, SARS-CoV-2, and possibly MERS-CoV. MERSrelated beta-CoVs have been identified in bat species from Africa, America, Asia, and Europe. In this study, we describe the first detection and characterization of a MERS-related beta-CoV in Danish brown long-eared bats (Plecotus auritus).Methods Fecal samples collected through a national surveillance program were screened using pan-CoV RT-qPCRs. Positive samples underwent ORF1b sequencing, microarray analysis and Illumina MiSeq sequencing, followed by metagenomic assembly of full-length genomes. A global phylogenetic tree was used to determine placement within the Coronaviridae family and local maximum likelihood phylogenetic analysis clarified subgroup placement. The receptor-binding potential of the spike protein to human DPP4, ACE2, and bat ACE2 orthologs was assessed through phylogenetic analysis of the receptor-binding domain (RBD), alongside homology modeling and structural analysis.
ResultsThree samples tested positive for CoVs. One sample from a Soprano pipistrelle (Pipistrellus pygmaeus) was identified as alpha-CoV by ORF1b sequencing. The remaining two samples, obtained from a colony of Plecotus auritus, were identified as beta-CoVs, and separate microarray results indicated the presence of a MERS-related CoV. Full genomes were successfully assembled using a metagenomic approach. Phylogenetic analysis placed them within the merbecoviruses, forming a distinct clade with viruses detected in Vespertilionidae bats from Western Europe and East Asia. Analysis of the RBD placed them within the HKU25 clade. Structural modeling suggested hydrogen bonding patterns between the RBD and human/bat ACE2 orthologs or human DPP4, similar to known in vitro complexes, indicating potential receptor binding. Conclusion This is the first report of MERS-related beta-CoVs in bats from Denmark. Phylogenetic analyses reveal that these novel viruses belong to the HKU25 clade, a clade with known ACE2 receptor preference. Experimental validation is needed to confirm the receptor-binding potential, as additional interactions at the RBD-receptor interface may differ from previously described bat-merbecoviruses. Continued surveillance is crucial to identify potential intermediate hosts and assess interspecies transmission risk, with focus on the spike protein, receptor specificity, and binding affinity.
## Background
Bats are recognized as natural reservoir hosts for numerous viruses, with spillover events serving as the foundation for emerging infectious diseases affecting both animals and humans [1]. Bat species (order Chiroptera) represent over 20% of the global mammalian species, with rodent species (order Rodentia) being the only ones that exceed their diversity [2]. Bats are likely the major reservoir of mammalian alpha-and beta-coronaviruses (CoVs), as the genetic diversity of CoVs in bats far exceeds that known for other hosts, and they are believed to be the evolutionary origin of severe acute respiratory syndrome (SARS) CoV and SARS-CoV-2 [3,4], and possibly Middle East respiratory syndrome (MERS) CoV [5,6], which all belong to the genus Betacoronavirus within the Coronaviridae family of the Nidovirales order [7]. Betacoronavirus includes the subgenera Sarbecovirus (SARS-CoV and SARS-CoV-2), Merbecovirus (MERS-CoV), Embecovirus, Hibecovirus and Nobecovirus [7]. Hundreds of merbecoviruses have been found to be circulating among diverse wildlife species across multiple continents [8][9][10][11][12][13][14][15][16][17], including insectivorous bat species from the globally widespread Vespertilionidae and more regionally confined Nycteridae families [6,10,[18][19][20][21][22].
Coronaviruses are enveloped and the viral particles contain a linear, positive-sense RNA of approximately 27-31 kb [23]. Around two-thirds of the genome encodes the replicase genes, which contains two overlapping open reading frames (ORFs), ORF1a and ORF1b. Translation of these genes produce two different length polyproteins via a ribosomal -1 frameshifting [24]. The polyproteins are co-and post-translationally processed into 16 nonstructural proteins. The last third of the genome encodes the structural and accessory proteins, which for MERS-CoV are the spike (S), ORF3, ORF4a, ORF4b, ORF5, envelope (E), membrane (M), nucleocapsid (N) and ORF8b [25].
Several studies have shown that some bat merbecoviruses use the angiotensin converting enzyme 2 (ACE2) receptor for cell entry, which is mediated by the viral spike protein, instead of the dipeptidyl peptidase IV (DPP4) receptor used by MERS-CoV [26][27][28]. Phylogenetic classification, based on the amino acid sequences of the Receptor-Binding Domains (RBDs), delineates at least eight merbecovirus clades: EriCoV, NeoCoV, MOW, HKU5-1, HKU5-2, HKU25, HKU4 and MERS [29][30][31].
Members of the MERS clade utilize DPP4 as their host receptor [32], and several members of the HKU4 clade have also been shown to use DPP4 [15-17, 31, 33, 34]. Members of the NeoCoV, MOW, HKU5-1, HKU5-2 and HKU25 clades have been shown to use ACE2, although with different host-ranges [27-31, 35, 36]. The host receptor for the EriCoV clade is currently unknown [27,31].
In a previous study we presented full-genome sequences within the genus Alphacoronavirus from different Danish bat species obtained from metagenomic next-generation sequencing (mNGS) [37]. In this study we describe the first detection and full-genome characterization of a MERS-related beta-CoV in brown longeared bats (Plecotus auritus) from Denmark.
## Methods
## Collection of samples
Fecal samples from six different bat species were collected as part of a national surveillance program of bats in Denmark (Table S1).
## Screening of samples with pan-CoV RT-PCR assays
The fecal samples were screened using two pan-CoV RT-PCR assays, as previously described [38], using a modified extraction protocol. Briefly, the fecal samples were homogenized in Eagle's medium using a TissueLyser II (QIAGEN, Hilden, Germany) and centrifuged. The RNA was extracted from the supernatant using a MagNA Pure 96 system (Roche, Basel, Switzerland) with the DNA and Viral NA Small Volume Kit and the Viral NA Plasma external lysis S.V. 3.1. protocol. Extracted RNA samples were screened for the presence of CoV RNA using two independent RT-qPCRs (panCoV B and panCoV C,). The products were amplified by RT-qPCR using the Qiagen OneStep RT-PCR kit (QIAGEN, Hilden, Germany) with LightCycler 480 ResoLight Dye (Roche, Basel, Switzerland) with either panCov B or panCov C primers (Table 1) in a final volume of 12.5 µl. For panCoV B the PCR program was as follows: 50 °C for 30 min followed by 95 °C for 15 min, then 5 cycles of 94 °C for 1 min, 40 °C for 1 min and 72 °C for 1 min, followed by 45 cycles of 95 °C for 1 min, 50 °C for 1 min, and 72 °C. The panCoV C was as follows: 50 °C for 30 min followed by 95 °C for 15 °C, then 50 cycles of 94 °C for 30 s, 48 °C for 30 s and 72 °C for 1 min.
## Panvirus microarray
To support the qPCR findings and further characterize detected viruses, selected pan-CoV positive samples were also analyzed with an in-house panvirus microarray assay, as previously described [41,42]. Briefly, the extracted nucleic acids were enriched using unbiased nucleic acid target amplifications and labelled with a fluorophore. The labelled samples were then hybridized to probes covering > 3000 viral genomes from vertebrates and invertebrate viruses, printed on a microarray slide (Agilent Technologies, Santa Clara, CA, USA). Positive hits were calculated using fluorescence intensities from multiple hybridizations within delimited regions.
## Sequencing and assembly
The samples positive in the pan-CoV assays were sequenced using Sanger sequencing and mNGS. The amplicons from the pan-CoV B assay, targeting the ORF1b, were Sanger sequenced as previously described [38]. For NGS, double-stranded cDNA was generated as previously described [37]. Briefly, the samples were treated with DNAse I (Invitrogen, Carlsbad, CA, USA), followed by Superscript III first-strand synthesis kit (Invitrogen) and NEBNext mRNA second-strand synthesis kit (New England Biolabs, Ipswich, MA, USA). The cDNAs were sequenced using MiSeq (Illumina, San Diego, CA, USA) with a modified Nextera XT DNA library protocol with the MiSeq reagent kit v2 (300 cycles), resulting in 2 × 150 bp paired-end reads.
Reads were trimmed using TrimGalore [43], a wrapper for CutAdapt [44], for quality (q26). Reads were then classified using both Kraken2 [45] with the plusPFP database and DIAMOND [46] using the complete Ref-Seq Protein database. Reads that were classified as viral or unknown were used for de novo assembly using both MEGAHIT [47] and metaSPAdes [48]. The resultant contigs were filtered for length ≥ 1000 bp and then mapped using minimap2 [49] to a local viral reference database, which was derived from the Kraken2 plusPFP database (accessed on December 15, 2023). Initially, references were classified using taxoniq [50] and any references categorized under Viruses were subsequently extracted. Any contigs mapping to references belonging to Coronaviridae were extracted and queried using BLASTn. For each contig, the best hit was selected based on the hit with the lowest E-value, highest bitscore and highest percent identity. The most frequent best hit of all the contigs was then selected as the reference sequence. This reference was used as a scaffold, by mapping all contigs against it with minimap2, and then the consensus sequence was extracted using Samtools [51]. As gaps were present, the non-gapped consensus sequences were excised and used as trusted contigs with SPAdes [52] together with the original raw reads to build new contigs. These contigs were then again mapped to the reference sequence, a consensus was extracted, and polished with the trimmed reads using BWA-MEM [53] before a new consensus was extracted. A reference-based consensus sequence was also obtained by mapping the trimmed reads directly to the reference sequence. The consensus sequences were aligned using MAFFT v7.508 with the auto command and manually inspected in Geneious Prime v2024.0.7 (Biomatters INC., Boston, MA, USA) to obtain the final draft genome.
## Gap-closing by Sanger sequencing
A PCR spanning a major gap was produced using primer pairs which were designed based on the draft genome generated above (Table 1). The products were amplified using AccuPrime high-fidelity DNA polymerase (Thermo Scientific, Thermo Fisher Scientific, Waltham, MA, USA) in a final volume of 50 μl using 94 °C for 30 s followed by 35 cycles of 94 °C for 15 s, 55 °C for 30 s and 68 °C for 2 min, with a final extension of 68 °C for 2 min. The initial PCR product was generated using primers 16227-9B_6263F and 16227-9B_7451R, and then used as template for a nested PCR with primers 16227-9B_6354F and 16227-9B_7351R. The PCR product was purified using the GeneJET PCR Purification Kit (Thermo Scientific) according to the manufacturer's instructions. The purified product was sequenced using the Sanger system with a combination of BigDye Terminator v. 3.1. Cycle Sequencing Kit (Applied BioSystems, Waltham, MA, USA) with 10 µM primers, purification using SigmaSpin Post-Reaction columns (Sigma-Aldrich, St. Louis, MO, USA) and an ABI3500 Genetic Analyzer (Applied BioSystems). Sequences were analyzed using Geneious Prime.
## Genome annotation
The assembled genomes were annotated using a modified gb2seq python library [54] with an annotated reference (MG596802.1). The ORF1ab, ORF1a, and ORF1b were annotated manually, by identifying the slippery sequence "TTT AAA C" just prior to the stop-codon in ORF1a.
To produce final genomes, the annotated genes were inspected to identify and correct erroneous frameshifts that were present in the assemblies.
## Phylogenetic analyses
Complete genome sequences belonging to Coronaviridae were downloaded from NCBI and GISAID based on the sequences used by Tan et al. [55]. These were combined with new merbecovirus sequences downloaded from NCBI (accessed February 6, 2024), but excluding those derived from camelids (Table S2). Phylogenies of full-length genomes were constructed using a modified approach previously described [55]. Briefly, initial phylogenies were constructed using an alignment-free phylogenetic reconstruction approach, combining Mash pairwise distances (Mash v2.3 [56]) with Neighbor-Joining tree reconstruction (ape v5.7.1 in R). The genomes in the merbecovirus clade were extracted and then aligned using Augur v14.0.0 [57] with MAFFT v7.508 [58] with a nobecovirus (NC_009021) as outgroup. Genome positions where gaps were assigned to more than 20% of the sequences, were removed from the alignment. The maximum-likelihood tree was performed using IQ-TREE v2.2.2.7 with a GTR + G model, ultrafast boot-strapping (UFBoot) [59] and approximate likelihood-ratio tests (SH-aLRT) [60] with 1000 replicates. All trees were visualized in ggtree v3.6.2 [61].
## Spike receptor-binding domain analysis
Spike sequences belonging to the merbecovirus subgroup and three embecoviruses were downloaded from Genbank and Genbase [62], based on previously published sequences (Table S3) [29][30][31]35]. The spike sequences were translated to amino acids using Geneious Prime v2025.0.3, and were aligned using MAFFT v7.508, using the embecoviruses as outgroup. The RBDs, as defined by the two glycine residues at aa position 372 and 610 in the human MERS reference sequence (JX869059), were then extracted from the alignment. A maximum-likelihood tree was performed using IQ-TREE with WAG + G4 (determined by ModelFinder [63]), 1000 UFBoot replicates and SH-aLRT. The tree was visualized in ggtree v3.6.2.
## Homology modelling and structural analysis
The spike amino acid sequence was extracted from the final draft genome for further analysis. ACE2 orthologs were downloaded from NCBI Orthologs (2024-12-03) for Homo sapiens, Pipistrellus kuhlii and Eptesicus fuscus (NP_068576, XP_008153150, XP_036295422). The mRNA sequence for E. fuscus (NC_072473) was extracted and then subsequently mapped to the genome of P. auritus (GCA_963455305.1) with minimap2 to identify the ACE2 region within P. auritus. The mRNA sequence for this ACE2 region was then extracted, concatenated and translated into amino acids using Geneious Prime. Homology models were generated using appropriate templates obtained from the Protein Data Bank (pdb. org) for MERS-CoV spike (PDB ID: 7M5E), HKU5 RBD, (PDB ID: 9D32), H. sapiens DPP4 (PDB ID: 4L72), Pipistrellus abramus ACE2 (PDB ID: 9D32) either individually or as RBD-receptor complexes using SWISS-MODEL [64]. Models were refined and hydrogen bonding analysis and visualization was performed in USCF ChimeraX 1.5 with the hbonds command [65], allowing for maximum hydrogen-bonding distance (3.3Å) to identify and visualize hydrogen bonds at the RBD-receptor interface. As a reference point, hydrogen bonds in the template structure were used as a point of comparison as these were based on actual interacting RBD-receptor pairs.
## Results
## Samples positive for CoV RNA
Two fecal samples were obtained during the national surveillance program in 2020 from P. auritus bats roosting in bat boxes in the forest of Almindingen, Bornholm, Denmark (Fig. 1). These fecal samples both tested positive in the two pan-CoV RT-PCR assays, panCoV B and panCoV C. One sample had Ct-values of 23.62 and 29.1, respectively, whereas the other had Ct-values of 39.25 and 36.5, respectively. The first sample resulted in a partial ORF1b sequence through Sanger sequencing, which showed 95% identity to two MERS-related beta-CoVs detected in P. kuhlii and Hypsugo savii from Italy [11]. The latter sample was screened against > 3000 vertebrate and invertebrate viral genomes with an in-house panvirus microarray and was found positive for MERS-CoV, hybridizing to the central parts of the genome (Figure S1, Table S4). A third fecal sample from a soprano pipistrelle (Pipistrellus pygmaeus) bat sampled at Svenstrup Gods on Zealand was also found to be positive for CoV by pan-CoV RT-PCR with high Ct-values. The panvirus microarray produced signal for various adenoviruses, but no CoVs. However, the partial ORF1b sequence obtained from the Sanger sequencing indicated the presence of an alpha-CoV with 97% identity to BtCoV/7542-55/P.pyg/DK/2014, as previously described for this bat species [38]. Attempts at NGS only yielded few CoV-specific reads (< 10) and further follow-up on this sample was not pursued.
## Genome assembly
The CoV positive fecal samples from P. auritus were both sequenced by Illumina MiSeq using a metagenomic approach. A total of 3,943,025 and 3,466,578 paired-end reads were classified as viral or unknown, and these were assembled into contigs and filtered for length ≥ 1000 bp, which resulted in 48 and 85 contigs, respectively. Among these contigs, 11 and 16 mapped to viral genomes belonging to the Coronaviridae family, respectively. These contigs were queried using BLASTn and the most frequent best hit was determined to be MAG: Merbecovirus PaGB01 (OQ401251.1), a bat MERS-related -beta-CoV obtained from a P. auritus bat in the UK. This reference was used as a scaffold for the contigs to build consensus sequences. The initial consensus sequences contained gaps, which were closed using raw reads. Referencebased consensus sequences were also obtained using the reference sequence above, and the consensus sequences were aligned and inspected manually to construct the final draft genomes. This resulted in full-length genomes of 29,802 bp (Fig. 2) and 29,626 bp in length, which were annotated, and named BtCoV/16227-9/P.aur/DK/2020 and BtCoV/16227-10/P.aur/DK/2020, respectively. To validate the gap-closing strategy, a nested PCR was designed across a major gap (ca. 680 bp) in the initial consensus sequence of BtCoV/16227-9/P.aur/DK/2020, and sequenced by Sanger. This product was identical to that of the final draft genome. A total of 22,615 and 10,499 reads mapped to these draft genomes, respectively. The BtCoV/16227-9/P.aur/DK/2020 genome contained only 6 nt which were ambiguous, however the BtCoV/16227-10/P.aur/DK/2020 genome contained in total 230 (0.8%) ambiguous nucleotides, where 145 (0.5%) of these were Ns. There were a total of 27 nucleotide differences between the two sequences, discounting ambiguous nucleotides (Table S5). The sequences were deposited in NCBI under accessions no. PP883547 and PP883548, respectively.
## Phylogenetic analyses
A global phylogenetic tree based on alignment-free genetic distances revealed that the two CoVs belonged to the merbecovirus subgroup within the beta-CoVs and subsequent local maximum-likelihood phylogenetic analysis (rooted) determined their placement within the subgroup (Fig. 3). They form a distinct clade together with MERS-CoV-like merbecoviruses isolated from Eptesicus, Hypsugo, Ia, Pipistrellus, Plecotus and Vespertilio spp., all belonging to the Vespertilionidae family of microbats, from Western Europe and East Asia.
The spike amino acid sequences for BtCoV/16227-9/P. aur/DK/2020 and BtCoV/16227-10/P.aur/DK/2020 are identical, except for residue S528 (Figure S2, Table S6), respectively, due to an ambiguity at the nucleotide level for BtCoV/16227-10/P.aur/DK/2020, and therefore only BtCoV/16227-9/P.aur/DK/2020 was used for further analysis. The rooted maximum likelihood phylogenetic analysis of spike RBDs revealed that the BtCoV/16227-9/P.aur/DK/2020 grouped closely together with PaGB01, within the HKU25 merbecovirus RBDs, which contains other bat-merbecoviruses belonging to Hypsugo, Pipistrellus and Vespertilio spp. from Western Europe and East Asia (Fig. 4, Figure S3). Direct comparison of the RBD of BtCoV/16227-9/P.aur/DK/2020 and PaGB01 shows that they are largely identical in this region, except for residues S391T, Q530N, T541I and E549D (Table S6).
## Receptor-binding potential
In order to determine the potential for the RBDs of these two bat-beta-CoVs to bind to different CoV receptors, homology models representing the RBD of BtCoV/16227-9/P.aur/DK/2020 complexed with H. sapiens DPP4, H. sapiens ACE2 or ACE-2 bat orthologs from P. auritus, P. kuhlii, and E. fuscus were built using SWISS-MODEL using appropriate templates obtained from the Protein Data Bank. The resulting models were refined in USCF ChimeraX 1.5 to identify and visualize hydrogen bonds at the RBD-receptor interface (Fig. 5). Based on the results, it was clear that a few residues in the BtCoV/16227-9/P.aur/DK/2020 RBD (S509 and E512) were consistently involved in hydrogen bonding with both bat and human ACE-2 orthologs. A threonine.
Residue at position 502 in the RBD (T502) was also observed to potentially form a hydrogen bond with the E. fuscus ACE-2 model, similar to what was observed in the template structure used for modelling, based on HKU5 RBD and ACE2 from P. abramus [29]. Hydrogen bonding between human DPP4 and the BtCoV/16227-9/P. aur/DK/2020 RBD also showed the same hydrogen bonding as the template structure based on MERS-CoV and human DPP4 [66].
## Discussion
In this study, bat fecal samples from a national surveillance program were screened for alpha-and beta-CoVs. In three samples, CoVs were identified, one fecal sample from a P. pygmaeus bat and two fecal samples from P. auritus. The fecal samples from the bats were initially identified as positive for CoV by RT-qPCR assays and partial ORF1b sequencing. The ORF1b sequence from the P. pygmaeus fecal sample indicated presence of an alpha-CoV as previously described in Danish bats [38], whereas the ORF1b sequence from one of the P. auritus bats indicated presence of a beta-CoV. The use of mNGS enabled us to obtain two full-length genomes of novel bat MERS-related CoVs from the two P. auritus bats. This is the first case description of beta-CoVs (and merbecovirus) in bats in Denmark. Plecotus auritus bats are common and are found in most parts of Denmark, especially on the island of Bornholm [67,68] and it is distributed widely across the Western Paleartic region [69][70][71].
Plecotus auritus are typically forest bats adapted to flying and foraging on insects in cluttered surroundings close to objects, but can also use more open habitats. It is considered a very sedentary species, which in summer normally forages within only a few kilometers from the roosts in hollow trees or buildings. Seasonal movements > 30 km are only rarely observed, and the longest movement recorded was 90 km covered by a female individual in Germany [71]. Plecotus auritus is certainly not among the bat species performing regular migrations over the Baltic Sea, i.e. Pipistrellus or more [70]. Individuals often hang in close clusters in attics, tree holes or bat boxes. Close contact with other bat species is rare, but more spacious roosting sites, such as attics, are sometimes shared with other species [71]. The route by which these P. auritus bats, despite their sedentary behavior and relative isolation on the island of Bornholm, became exposed to merbecoviruses remains unclear. As there is currently no systematic surveillance of these bats, it remains unknown how long merbecoviruses have been circulating within the population. Detection of merbecoviruses is further complicated by the need for sampling at the right time and place, as viral shedding may be intermittent or low. Improved detection tools, such as a PCR assay specifically targeting this merbecovirus could enhance future surveillance efforts and help clarify the prevalence and transmission dynamics of this virus in this relatively isolated bat population. Merbecoviruses circulate among a wide range of bats belonging to the family Vespertilionidae. In Europe, merbecoviruses have been obtained from P. auritus bats in the UK (full-genome) [55] and in Italy (partial genomes) [73]. Both full-length and partial sequences have been obtained from P. kuhlii and H. savi in Italy [11,22,73], and partial genomes sequences in other Pipistrellus species have been obtained in the Netherlands (Pipistrellus pipistrellus), Romania (P. nathusii, P. pipistrellus, and P. pygmaues), and Ukraine (P. nathusii) [20,74]. Partial merbecovirus genome sequences have also been found in Eptesicus serotinus and N. noctula bats in Italy [11,75].
Precise genome reconstruction is imperative, as the consistency and base accuracy of the assembly will influence the outcomes of all downstream analyses [76]. In de Fig. 5 Hydrogen bonding analysis between ACE2 orthologs or DPP4 and BtCoV/16227-9/P.aur/DK/2020 spike receptor-binding domain. Homology models were produced with SWISS-MODEL using appropriate templates and hydrogen bonds (bright green) at the RBD-ACE2 interface were visualized in UCSF ChimeraX and hydrogen-bond-forming residues were shown. The BtCoV/16227-9/P.aur/DK/2020 spike (orange) and P. abramus ACE2 (green) is depicted in (A) and modelling and visualizations were done for the HKU5 RBD-P. abramus ACE2 template (PDB ID: 9D32) (B) and homology models depicting the binding interface between MERS-related RBD and ACE2 orthologs of P. auritus (C), P. kuhlii (D), E. fuscus (E) and H. sapiens (F). Homology models of BtCoV/16227-9/P.aur/DK/2020 spike (orange) and DPP4 (blue) are depicted in (G) and the binding interface between RBD and DPP4 is depicted for H. sapiens DPP4 and MERS-CoV (H) and BtCoV/16227-9/P.aur/DK/2020 (I) novo assembly, sequence reads are assembled into contigs, which can then be gathered into scaffolds or supercontigs, which contain gaps [77]. Some of the challenges in de novo assembly include sequencing error resulting in artifacts being included in the final results and sequencing bias [78], resulting in unequal depth of sequencing across the regions in the genome. These challenges can cause complex de Bruijn graphs, which in general produce results which are not satisfactory [77]. In referencebased approaches, reads are mapped to similar regions of a supplied reference sequence. However, it is important to be aware of reference bias, where reads harboring nonreference genomic variations, can be inaccurately aligned or overlooked by the aligner [79,80]. In this study, we use a two-pronged approach to generate the draft genomes by utilizing both de novo assembly and reference-based mapping, in order to reduce bias by only using one approach.
MERS-CoV infection in humans is primarily zoonotic, originating from camelids [81]. However, increasing evidence suggests that the virus likely originated from bats, as phylogenetic analyses reveal its close relationship to a diverse group of bat-borne CoVs [6,19]. However, the evolutionary origin and the bat species involved are unknown. Several merbecoviruses have been proved able to use the human DPP4 or ACE2 receptors [16,27]. Recent studies have revealed that merbecoviruses from several clades have independently evolved the ability to utilize ACE2 as a receptor, using distinct binding modes [27][28][29][30][31]36]. HKU25 viruses seem to engage ACE2 with a binding pose reminiscent of that observed for HKU5-1, which was also observed for HKU5-2, suggesting a common evolutionary origin of ACE2 utilization for these clades of viruses [29,30,35]. In the preprint by Liu et al. [35], AlphaFold3-predicted structures of HKU25related RBDs, highlighted their similarity to members of the HKU5-1 clade, in terms of overall receptor-binding motifs (RBM) architecture, however PaGB01 harbors a short indel, which is shared by BtCoV/16227-9/P.aur/ DK/2020. They also screened members of the HKU25 clade using an ACE2 ortholog library, which revealed that most HKU25 clade CoVs can engage ACE2 from several bat species and a subset of non-bat mammals. However, no meaningful binding or pseudovirus entry was detected for any ACE2 ortholog tested with PaGB01, including E. fuscus, P. kuhlii, and P. auritus ACE2, the latter which is the host of PaGB01. The ACE2 dependency of PaGB01 cannot be excluded, as polymorphisms in the P. auritus ACE2 allele may have influenced receptor functionality. The study by Tan et al. indicated that PaGB01 could not use human ACE2, human DPP4 or human aminopeptidase N [55].
Examination of hydrogen binding potential between the BtCoV/16227-9/P.aur/DK/2020 spike RBD and bat/ human ACE2 orthologs or human DPP4 suggest similar hydrogen bonding to in vitro complexes and provide a reasonable proxy for binding potential, but such structural approaches do not account for the dynamic nature of receptor-ligand interactions, rely on simplified or arbitrary chemical milieu and overlook post-translational modifications which may play a role in ligand-binding. Given the experimental results from others described above, the binding of BtCoV/16227-9/P.aur/DK/2020 to either ACE2 or DPP4 will have to be independently examined in an experimental setting in order to validate these observations; other interactions at the RBD-receptor interface might also have a significant role to play here and may be different in BtCoV/16227-9/P.aur/DK/2020 compared to previously described bat merbecoviruses. To address this, future studies should include pseudovirus entry assays using the spike proteins in combination with bat and human receptor orthologs. These functional assays will be essential to confirm receptor usage, assess entry efficiency, and better understand the host range and zoonotic potential of this novel virus.
A recent study by Tolentino et al. provides evidence of multiple recombination events in the evolution of merbecoviruses, including one event where the MERS-CoV progenitor acquired a DPP4-binding spike protein, replacing its original ACE2-binding spike region [82]. Recombination in RNA viruses, particularly in positivesense single-stranded RNA viruses like CoVs, is welldocumented [83]. This process frequently occurs among closely related, co-circulating CoVs, facilitating the emergence of novel viruses with altered receptor usage and expanded host ranges, potentially enhancing their fitness and zoonotic potential [27,28,84,85]. While our study did not include a recombination analysis, the identification of novel merbecoviruses in P. auritus bats highlights the relevance of such investigations, as it is plausible that recombination has played a role in their evolutionary history. Future studies on recombination will be important to better understand the genetic diversity, evolutionary dynamics, and potential for cross-species transmission within this group. Reporting MERS-related viruses through the national surveillance program enhances our understanding of pathogen presence at both national and regional levels. Since merbecoviruses pose a potential pandemic threat, continued surveillance of reservoir and intermediate species, such as bats and camelids, is of great importance.
## Conclusion
This study presents the first report of MERS-related beta-CoVs in bats from Denmark. The majority of alpha-and beta-CoVs currently circulating in mammals are thought to have evolutionary origins in ancestral bat CoVs [3]. However, how P. auritus bats, with their sedentary behavior and relative isolation on the island of Bornholm, came in contact with merbecoviruses remains unknown.
Determining whether P. auritus expresses ACE2 or DPP4 receptor orthologs compatible with these viruses -or identifying alternative receptors -will be essential for understanding receptor usage and species tropism in merbecoviruses. The evolutionary history of receptor utilization within the merbecovirus subgroup is still poorly understood. Continued surveillance is vital to identify potential intermediate hosts and to assess the risk of interspecies transmission, with particular attention to the spike protein, receptor specificity, and host cell binding affinity.
## Abbreviations
## References
1. (2015) "Middle East respiratory syndrome-related coronavirus strain Hu/Aseer-KSA" *Rs*
2. (2014) "MK039552 | Middle East respiratory syndrome-related coronavirus isolate Hu"
3. "Middle East respiratory syndrome coronavirus isolate Jeddah_1_2013 KJ614529 | Human betacoronavirus 2c Jordan-N3/2012 isolate MG167 MT576585 | Middle East respiratory syndrome-related coronavirus isolate MA_30" *KJ*
4. "OK573213 | UNVERIFIED: Middle East respiratory syndrome-related coronavirus FDA"
5. "MH734115 | Middle East respiratory syndrome-related coronavirus isolate MERS-CoV"
6. (2016) "MG923477 | Middle East respiratory syndrome-related coronavirus isolate MERS-CoV camel/Nigeria/NV2040"
7. "ON325306 | Middle East respiratory syndrome-related coronavirus isolate Bat"
8. "LC706863 | Merbecovirus sp. VsCoV-kj15 RNA"
9. "VsCoV-a7 RNA LC469308 | Bat coronavirus Vs-CoV-1 KX442564 | Hypsugo bat coronavirus HKU25 isolate YD131305 KX442565 | Hypsugo bat coronavirus"
10. "MG987420 | Middle East respiratory syndrome-related coronavirus isolate NL13892 MG987421 | Middle East respiratory syndrome-related coronavirus isolate NL140455"
11. (2022) "Middle East respiratory syndrome-related coronavirus isolate NL13845 MG021452 | Middle East respiratory syndrome-related coronavirus isolate NL140422 OQ297692 | MAG: Middle East respiratory syndrome-related coronavirus isolate GZ/L165"
12. "Ej-CoV-3 RNA LC706865 | Merbecovirus sp. EjCoV-3 RNA"
13. "ON325307 | Middle East respiratory syndrome-related coronavirus isolate BatCoV/V"
14. "Middle East respiratory syndrome-related coronavirus isolate Bat-CoV/H"
15. "MG596803 | Middle East respiratory syndrome-related coronavirus isolate Bat-CoV/P"
16. "MK907286 | Erinaceus hedgehog coronavirus HKU31 strain"
17. "MK907287 | Erinaceus hedgehog coronavirus HKU31 strain Rs13 MN879331 | UNVERIFIED: Erinaceus amurensis coronavirus isolate cw_6 MZ293757 | Hipposideros bat coronavirus isolate PREDICT"
18. Btvs-Betacov, Sc (2013)
19. "VsCoV-kj15 RNA LC706864 | Merbecovirus sp. VsCoV-a7 RNA LC469308 | Bat coronavirus Vs-CoV-1 KX442564 | Hypsugo bat coronavirus HKU25 isolate YD131305 KX442565 | Hypsugo bat coronavirus"
20. "MG987420 | Middle East respiratory syndrome-related coronavirus isolate NL13892 MG987421 | Middle East respiratory syndrome-related coronavirus isolate NL140455"
21. (2022) "Middle East respiratory syndrome-related coronavirus isolate NL13845 MG021452 | Middle East respiratory syndrome-related coronavirus isolate NL140422 OQ297692 | MAG: Middle East respiratory syndrome-related coronavirus isolate GZ/L165"
22. "Ej-CoV-3 RNA LC706865 | Merbecovirus sp. EjCoV-3 RNA"
23. "ON325307 | Middle East respiratory syndrome-related coronavirus isolate BatCoV/V"
24. "Middle East respiratory syndrome-related coronavirus isolate Bat-CoV/H"
25. "MG596803 | Middle East respiratory syndrome-related coronavirus isolate Bat-CoV/P"
26. (2013) "| ErinCoV-15-19 MK907286 | HKU31-F6 OM451213 | HKU31-HeB-MO1 KC869678 | NeoCoV KX574227 | PDF2180 JX899383 | NsGHA2010 ON325306 | MOW15-22 OQ405399 | PnNL2018B LC469308 | VsCoV-1 ON325308 | VmSL2021 ON325307 | VmSL2020 MG596802 | Hsavi-2011 MG596803 | Pkhuli-2011 LC706864 | VsCoV-a7 KJ473821 | Bt-CoV-SC"
27. Letko, Seifert, Olival et al. (2020) "Bat-borne virus diversity, spillover and emergence" *Nat Rev Microbiol*
28. Burgin, Colella, Kahn et al. (2018) "How many species of mammals are there?" *J Mammal*
29. Hu, Ge, Wang et al. (2015) "Bat origin of human coronaviruses" *Virol J*
30. Temmam, Vongphayloth, Baquero et al. (2022) "Bat coronaviruses related to SARS-CoV-2 and infectious for human cells" *Nature*
31. Memish, Mishra, Olival et al. (2013) "Middle East respiratory syndrome coronavirus in bats, Saudi Arabia" *Emerg Infect Dis*
32. Anthony, Gilardi, Menachery et al. (2017) "Further evidence for bats as the evolutionary source of middle east respiratory syndrome coronavirus. mBio"
33. Woo, De Groot, Haagmans et al. (2023) "ICTV virus taxonomy profile: coronaviridae 2023" *J Gen Virol*
34. Yang, Du, Liu et al. (2014) "Receptor usage and cell entry of bat coronavirus HKU4 provide insight into bat-to-human transmission of MERS coronavirus" *Proc Natl Acad Sci*
35. Lau, Li, Tsang et al. (2013) "Genetic characterization of betacoronavirus lineage C viruses in bats reveals marked sequence divergence in the spike protein of pipistrellus bat coronavirus HKU5 in Japanese Pipistrelle: implications for the origin of the novel Middle East respiratory sy" *J Virol*
36. Ithete, Stoffberg, Corman et al. (1910) "Close relative of human Middle East respiratory syndrome coronavirus in bat" *South Africa. Emerg Infect Dis*
37. Moreno, Lelli, De Sabato et al. (2017) "Detection and full genome characterization of two beta CoV viruses related to Middle East respiratory syndrome from bats in Italy" *Virol J*
38. Haagmans, Dhahiry, Reusken et al. (2014) "Middle east respiratory syndrome coronavirus in dromedary camels: an outbreak investigation" *Lancet Infect Dis*
39. Sabato, Bartolo, Marco et al. (2020) "Can coronaviruses steal genes from the host as evidenced in Western European hedgehogs by EriCoV genetic characterization?" *Viruses*
40. He, Hou, Zhao et al. (2022) "Virome characterization of game animals in China reveals a spectrum of emerging pathogens" *Cell*
41. Wang, Qi, Yuan et al. (2014) "Bat origins of MERS-CoV supported by bat coronavirus HKU4 usage of human receptor CD26" *Cell Host Microbe*
42. Chen, Yang, Si et al. (2023) "A bat MERS-like coronavirus circulates in pangolins and utilizes human DPP4 and host proteases for cell entry" *Cell*
43. Luo, Wang, Yang et al. (2018) "Discovery of novel bat coronaviruses in South China that use the same receptor as Middle East respiratory syndrome coronavirus" *J Virol*
44. Yang, Wu, Ren et al. (2014) "MERS-related betacoronavirus in Vespertilio superans bats" *Emerg Infect Dis*
45. Corman, Ithete, Richards et al. (2014) "Rooting the phylogenetic tree of Middle East respiratory syndrome coronavirus by characterization of a conspecific virus from an African bat" *J Virol*
46. Annan, Baldwin, Corman et al. (2013) "Human betacoronavirus 2c EMC/2012-related viruses in bats, Ghana and Europe" *Emerg Infect Dis*
47. Anthony, Ojeda-Flores, Rico-Chávez et al. (2013) "Coronaviruses in bats from Mexico" *J Gen Virol*
48. Lelli, Papetti, Sabelli et al. (2013) "Detection of coronaviruses in bats of various species in Italy" *Viruses*
49. De Groot, Baker, Baric et al. (2012) "Family: coronaviridae"
50. Sola, Almazán, Zúñiga et al. (2015) "Continuous and discontinuous RNA synthesis in coronaviruses" *Annu Rev Virol*
51. Van Boheemen, De Graaf, Lauber et al. (2012) "Genomic characterization of a newly discovered coronavirus associated with acute respiratory distress syndrome in humans" *mBio*
52. Ma, Liu, Xiong et al. (2023) "Identification of ACE2 as the entry receptor for two novel European bat merbecoviruses" *bioRxiv*
53. Xiong, Cao, Ma et al. (2022) "Close relatives of MERS-CoV in bats use ACE2 as their functional receptors" *Nature*
54. Ma, Liu, Xiong et al. (2023) "Broad host tropism of ACE2-using MERS-related coronaviruses and determinants restricting viral recognition" *Cell Discov*
55. Park, Liu, Lee et al. (2025) "Molecular basis of convergent evolution of ACE2 receptor utilization among HKU5 coronaviruses" *Cell*
56. Chen, Zhang, Li et al. (2025) "Bat-infecting merbecovirus HKU5-CoV lineage 2 can use human ACE2 as a cell entry receptor" *Cell*
57. Catanzaro, Wu, Fan et al. (2024) "ACE2 from Pipistrellus Abramus bats is a receptor for HKU5 coronaviruses" *bioRxiv*
58. Raj, Mou, Smits et al. (2013) "Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC" *Nature*
59. Xia, Wang, Cui et al. (2024) "Isolation and characterization of a pangolinborne HKU4-related coronavirus that potentially infects human-DPP4transgenic mice" *Nat Commun*
60. Tse, Hou, Mcfadden et al. (2023) "A MERS-CoV antibody neutralizes a pre-emerging group 2c bat coronavirus" *Sci Transl Med*
61. Liu, Park, Ma et al. (2025) "ACE2 Utilization of HKU25 Clade MERS-related coronaviruses with broad geographic distribution" *bioRxiv*
62. Ma, Liu, Park et al. (2025) "Multiple independent acquisitions of ACE2 usage in MERS-related coronaviruses" *Cell*
63. Lazov, Belsham, Bøtner et al. (2021) "Full-genome sequences of alphacoronaviruses and astroviruses from myotis and pipistrelle bats in Denmark" *Viruses*
64. Lazov, Chriél, Baagøe et al. (2018) "Detection and characterization of distinct Alphacoronaviruses in five different bat species in Denmark" *Viruses*
65. Escutenaire, Mohamed, Isaksson et al. (2007) "SYBR green real-time reverse transcription-polymerase chain reaction assay for the generic detection of coronaviruses" *Arch Virol*
66. Vijgen, Moës, Keyaerts et al. (2008) "A pancoronavirus RT-PCR assay for detection of all known coronaviruses"
68. Erlandsson, Rosenstierne, Mcloughlin et al. (2011) "The microbial detection array combined with random Phi29-amplification used as a diagnostic tool for virus detection in clinical samples" *PLoS One*
69. Rosenstierne, Mcloughlin, Olesen et al. (2014) "The microbial detection array for detection of emerging viruses in clinical samples -a useful panmicrobial diagnostic tool" *PLoS One*
70. Krueger, James, Ewels et al. (2023) "sclamons FelixKrueger/TrimGalore: V0.6.10 -add default decompression path"
71. Martin (2011) "Cutadapt removes adapter sequences from high-throughput sequencing reads" *EMBnet.journal*
72. Wood, Lu, Langmead (2019) "Improved metagenomic analysis with Kraken 2" *Genome Biol*
73. Buchfink, Reuter, Drost (2021) "Sensitive protein alignments at tree-oflife scale using diamond" *Nat Methods*
74. Li, Liu, Luo et al. (2015) "MEGAHIT: an ultra-fast singlenode solution for large and complex metagenomics assembly via succinct de bruijn graph" *Bioinformatics*
75. Nurk, Meleshko, Korobeynikov et al. (2017) "Metaspades: a new versatile metagenomic assembler" *Genome Res*
76. Li (2018) "Minimap2: pairwise alignment for nucleotide sequences" *Bioinformatics*
77. "Taxoniq: taxon information query -fast, offline querying of NCBI taxonomy and related data Available online"
78. Danecek, Bonfield, Liddle et al. (2021) "Twelve years of SAMtools and BCFtools" *Gigascience*
79. Prjibelski, Antipov, Meleshko et al. (2020) "Using SPAdes De Novo Assembler" *Curr Protoc Bioinforma*
80. Li (2013) "Aligning sequence reads, clone sequences and assembly contigs with BWA-MEM"
81. Jones, Muehlemann, Gieraths "Gb2seq: A Python Library and Command-Line Scripts for Working with GenBank Genome Sequences and Annotations Available online"
82. Tan, Trew, Peacock et al. (2023) "Genomic screening of 16 UK native bat species through conservationist networks uncovers coronaviruses with zoonotic potential" *Nat Commun*
83. Ondov, Treangen, Melsted et al. (2016) "Mash: fast genome and metagenome distance estimation using MinHash" *Genome Biol*
84. Huddleston, Hadfield, Sibley et al. (2021) "Augur: a bioinformatics toolkit for phylogenetic analyses of human pathogens" *J Open Source Softw*
85. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
86. Hoang, Chernomor, Haeseler et al. (2018) "UFboot2: improving the ultrafast bootstrap approximation" *Mol Biol Evol*
87. Guindon, Dufayard, Lefort et al. (2010) "New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0" *Syst Biol*
88. Yu, Smith, Zhu et al. (2017) "GGTREE : an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data" *Methods Ecol Evol*
89. Bu, Zheng, Zhao et al. (2024) "GenBase: a nucleotide sequence database" *Genomics Proteomics Bioinf*
90. Kalyaanamoorthy, Minh, Wong et al. (2017) "Modelfinder: fast model selection for accurate phylogenetic estimates" *Nat Methods*
91. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-model: homology modelling of protein structures and complexes" *Nucleic Acids Res*
92. Meng, Goddard, Pettersen et al. (2023) "UCSF chimerax: tools for structure building and analysis" *Protein Sci*
93. Wang, Shi, Jiang et al. (2013) "Structure of MERS-CoV spike receptor-binding domain complexed with human receptor DPP4" *Cell Res*
94. Baagøe, Flagermus (2007) "Dansk Pattedyratlas"
95. Elmeros, Fjederholt, Møller et al. (2024) "Opdatering Af: Håndbog Om Dyrearter På Habitatdirektivets Bilag IV. Del 2 -Odder Og Flagermus"
96. Bats (2019) "Handbook of the Mammals of the World"
97. Ancillotto, Russo (2023) "Brown Long-Eared Bat Plecotus Auritus (Linnaeus, 1758)"
98. Dietz, Helversen, Nill et al. (2009)
99. Ahlén, Baagøe, Bach (2009) "Behavior of Scandinavian bats during migration and foraging at sea" *J Mammal*
100. Leopardi, Desiato, Mazzucato et al. (2023) "One health surveillance strategy for coronaviruses in Italian wildlife" *Epidemiol Infect*
101. Reusken, Lina, Pielaat et al. (2010) "Circulation of group 2 coronaviruses in a bat species common to urban areas in western Europe" *Vector-Borne and Zoonotic Diseases*
102. Benedictis, Marciano, Scaravelli et al. (2014) "Alpha and lineage C betacov infections in Italian bats" *Virus Genes*
103. Denton, Lugo-Martinez, Tucker et al. (2014) "Extensive error in the number of genes inferred from draft genome assemblies" *PLoS Comput Biol*
104. Liao, Li, Zou et al. (2019) "Current challenges and solutions of de Novo assembly" *Quant Biol*
105. Rodrigue, Malmstrom, Berlin et al. (2009) "Whole genome amplification and de novo assembly of single bacterial cells" *PLoS One*
106. Lunter, Goodson (2011) "Stampy: a statistical algorithm for sensitive and fast mapping of illumina sequence reads" *Genome Res*
107. Brandt, Aguiar, Bitarello et al. (2015) "Mapping bias overestimates reference allele frequencies at the HLA genes in the 1000 genomes project phase I data" *G3 Genes|Genomes|Genetics*
108. Memish, Cotten, Meyer et al. (2013) "Human infection with MERS coronavirus after exposure to infected camels, Saudi Arabia" *Emerg Infect Dis*
109. Tolentino, Lytras, Ito et al. (2024) "Recombination analysis on the receptor switching event of MERS-CoV and its close relatives: implications for the emergence of MERS-CoV" *Virol J*
110. Patiño-Galindo, Rabadan (2021) "Global patterns of recombination across human viruses" *Mol Biol Evol*
111. Forni, Cagliani, Sironi (2020) "Recombination and positive selection differentially shaped the diversity of betacoronavirus subgenera" *Viruses*
112. Wells, Bonavita, Navarrete-Macias et al. (2023) "The coronavirus recombination pathway" *Cell Host Microbe* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12869191&blobtype=pdf | # Frontiers Editorial O ce, Frontiers Media SA, Switzerland *CORRESPONDENCE
Yanping Zhang, Yulong Gao, Changjun Liu, Lan Wang, Wang Zhang, Gao Liu, © Li, Yu Lan, Wang, Gao, Liu Wang, This, Kai Li, Zhenghao Yu, Xingge Lan, Yanan Wang, Xiaole Qi, Hongyu Cui, Li Gao, Xiaomei Wang
## Abstract
Complete genome analysis reveals evolutionary history and temporal dynamics of Marek's disease virus. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12691671&blobtype=pdf | # PEDV NSP8 inhibits IFN-III production induced by MAVS through downregulation of PEX13
Jinxiu Lou, Huixin Zhu, Zhen Yang, Ping Jiang, Gongguan Liu, Xianwei Wang
## Abstract
Porcine epidemic diarrhea virus (PEDV), an enteropathogenic coronavirus that causes severe intestinal disease in piglets, employs sophisticated strategies to subvert host antiviral immunity. While type III interferons (IFN-III) play a pivotal role in mucosal defense at intestinal epithelial barriers, the mechanisms underlying PEDV evasion of IFN-III signaling remain poorly understood. Given that peroxisomes serve as critical platforms for IFN-III signaling and that their abundance influences immune activation, we investigated the role of pexophagy in PEDV-mediated immune evasion. We demonstrated that the PEDV nonstructural protein NSP8 functions as a potent inhibitor of mitochondrial antiviral-signaling protein (MAVS)-dependent IFN-III produc tion. Functional analyses revealed that NSP8 significantly reduces peroxisomal protein levels and promotes pexophagy. Mechanistically, mass spectrometry identified a direct interaction between NSP8 and PEX13. NSP8 also induces dose-dependent degradation of PEX13 via the autophagy-lysosomal pathway. This downregulation of PEX13 triggers ubiquitination of the peroxisomal import receptor PEX5, facilitating its recognition by the autophagy receptor NBR1 and the ubiquitin ligase PEX2, thereby promoting autophagic clearance of peroxisomes. Collectively, our findings reveal a novel immune evasion strategy in which PEDV exploits NSP8 to disrupt peroxisome homeostasis by targeting PEX13, thereby dismantling MAVS-dependent IFN-III antiviral signaling through pexophagy.IMPORTANCE Porcine epidemic diarrhea virus (PEDV) NSP8 is a highly conserved protein that plays a crucial role in viral replication. Investigating the functional mech anisms of NSP8 contributes to a deeper understanding of PEDV pathogenesis and supports the development of antiviral strategies against coronaviruses. In this study, we elucidate how NSP8 suppresses type Ⅲ interferon (IFN-Ⅲ) production by promoting pexophagy through the downregulation of PEX13. We demonstrate that NSP8 directly interacts with PEX13 and enhances the ubiquitination of PEX5, leading to reduced peroxisome abundance and impaired mitochondrial antiviral-signaling protein (MAVS)mediated IFN-Ⅲ signaling. These findings suggest that NSP8 hijacks the PEX13-depend ent pexophagy pathway as a means of evading host antiviral defenses. This work provides critical insights into the interplay between viral proteins and host cellular machinery and highlights the NSP8-PEX13 axis as a promising target for therapeu tic interventions aimed at enhancing antiviral immunity against PEDV and related coronaviruses.
accessory protein (open reading frame 3, ORF3), and four structural proteins: spike (S), envelope (E), membrane (M), and nucleocapsid (N) (2,3). The innate immune response serves as the host's first line of defense against viral infection. Consequently, many viruses, including PEDV, have evolved sophisticated strategies to modulate or evade host innate immunity during infection (4,5). Previous studies have reported that the NSP8 of severe acute respiratory syndrome coronavirus 2 and porcine deltacoronavi rus mediate evasion of the innate immune response during infection (6)(7)(8). However, the role of PEDV NSP8 in regulating host immunity remains unexplored.
Innate immune recognition of pathogen-associated molecular patterns triggers the production of interferons (IFNs), key antiviral cytokines that mediate host defense against viral infections (9,10). Among them, type III interferons (IFN-λ/IFN-III) play a critical role in mucosal immunity (11), particularly at epithelial barriers, which serve as primary replication sites of PEDV. The mitochondrial antiviral-signaling protein (MAVS) acts as a central adaptor in the RIG-I-like receptor (RLR) pathway, transmitting signals from viral RNA sensors to promote IFN-III production (9,12). Notably, MAVS is local ized not only to mitochondria but also to peroxisomes, membrane-bound organelles traditionally known for lipid metabolism and reactive oxygen species (ROS) detoxifica tion (13,14). Emerging evidence suggests that peroxisome-associated MAVS contributes to the rapid induction of IFN-III and enhances antiviral defenses (9,12). However, viruses have evolved strategies to counteract these responses, and PEDV encodes multiple nonstructural proteins (NSPs) that antagonize IFN production (15,16). While NSP8 is implicated in PEDV replication and immune evasion, its precise role and mechanism remain poorly defined. In particular, it is unclear how PEDV manipulates peroxisomemediated innate immune signaling to facilitate immune escape.
Autophagy, a conserved cellular degradation pathway that plays dual roles during viral infection: it can eliminate viral components (virophagy) or be exploited by viruses to degrade host immune regulators (17,18). The selective autophagic degradation of peroxisomes, termed pexophagy, is tightly regulated by proteins, such as PEX13, a peroxisomal membrane protein essential for peroxisome biogenesis and maintenance (19)(20)(21). Although autophagy is known to intersect with innate antiviral signaling, the specific interplay between viral proteins, pexophagy, and IFN-III suppression remains unexplored.
In this study, we uncover a novel immune evasion mechanism in which PEDV NSP8 disrupts MAVS-dependent IFN-III signaling pathway by promoting peroxisome degra dation. We demonstrated that NSP8 directly interacts with PEX13, downregulates its expression, and induces pexophagy, thereby impairing peroxisomal MAVS-mediated IFN-III production. These findings not only identify a unique immune evasion strategy employed by PEDV but also underscore the critical role of peroxisome homeostasis in antiviral innate immunity. Here, we aimed to determine whether, and how, PEDV NSP8 modulates peroxisome dynamics to inhibit the production of IFN-III.
## RESULTS
## PEDV inhibits the antiviral innate immune response
PEDV employs several strategies to inhibit the production of IFN to circumvent host antiviral defenses, facilitating rapid viral replication (5,22,23). The IFN-III family includes IFN-λ1, λ2, λ3, and λ4; however, Zhang et al. reported a potential absence of IFN-λ2 expression in swine. Therefore, this study specifically quantified the mRNA levels of IFN-λ1, λ3, and λ4. To investigate the impact of PEDV on the antiviral innate immune response, IPEC-J2 cells were infected with PEDV, and whole-cell lysates were collected at 24 hours post-infection (hpi) for analysis by quantitative real-time PCR (RT-qPCR). The results showed that infection with PEDV at different multiplicities of infection (MOI) significantly suppressed the production of IFN-λ1, IFN-λ3, and IFN-λ4 in IPEC-J2 cells (Fig. 1A through C). To further explore the temporal effects of PEDV infection on IFN-III expression, IPEC-J2 cells were infected with PEDV (MOI = 10) and harvested at designa ted time points for RT-qPCR analysis. The results demonstrated a transient upregulation of IFN-III mRNA expression at 3 hpi, followed by a time-dependent suppression as the infection progressed (Fig. 1D through F). These findings indicate that PEDV infection ultimately inhibits IFN-III expression in intestinal epithelial cells.
## PEDV NSP8 inhibits MAVS-mediated production of IFN-III
To determine which PEDV proteins are involved in the regulation of IFN-III signaling, cells were co-transfected with an IFN-λ1 luciferase reporter construct and expression plasmids encoding individual viral proteins known to contribute to immunosuppres sion. A dual-luciferase reporter assay was then performed to evaluate changes in IFN-λ1 promoter activity. The results demonstrated that nearly all tested PEDV proteins significantly inhibited the production of IFN-III, with the exception of NSP14. Notably, NSP8 exhibited the most potent inhibitory effect (Fig. 2A). Therefore, the role of NSP8 was selected for further investigation in this study.
To further determine whether NSP8 is involved in regulating the antiviral innate immune response, varying amounts of Flag-NSP8 were transfected into IPEC-J2 cells for 24 h and then harvested for RT-qPCR analysis. The results showed that NSP8 suppressed the mRNA expression levels of IFN-λ1, IFN-λ3, and IFN-λ4 (Fig. 2B through D). To explore how NSP8 attenuates IFN-III-mediated antiviral responses, we examined its effect on key components of the RLRs, specifically RIG-I, MDA5, MAVS, TBK1, IRF3, and IKKε by evaluating their ability to activate IFN-λ1 promoter activity in the presence or absence of NSP8. While these signaling molecules individually activated the IFN-λ1 promoter, co-expression with NSP8 markedly inhibited MAVS-mediated activation, exhibited only weak suppression of RIG-I and MDA5-mediated activation, and had no inhibitory effect on TBK1-, IRF3-, or IKKε-induced IFN-λ1-luciferase activity (Fig. 2E through J). To further confirm the regulatory effect of NSP8 on MAVS expression, 293T cells were co-trans fected with Flag-NSP8 and HA-MAVS for 24 h, followed by a western blotting assay. Immunoblot analysis demonstrated that NSP8 significantly reduced MAVS protein levels (Fig. 2K). Additionally, co-transfection of 293T and IPEC-J2 cells with a fixed amount of HA-MAVS and increasing doses of Flag-NSP8 plasmid revealed that NSP8 inhibited MAVS expression in a dose-dependent manner (Fig. 2L andM). Collectively, these findings suggest that PEDV NSP8 suppresses type III IFN production by downregulating MAVSmediated signaling.
## NSP8 reduces peroxisome abundance
A previous study reported that IFN-III-mediated antiviral signaling in intestinal epithelial cells primarily depends on peroxisomes (12). To assess the influence of NSP8 on the expression of peroxisome-related proteins, peroxisomal membrane protein 70 (PMP70) and catalase were used as peroxisomal markers (24). IPEC-J2 cells and LLC-PK1 cells were transfected with varying amounts of Flag-NSP8. The results showed that NSP8 decreased the expression of endogenous PMP70, catalase, and MAVS in a dose-dependent manner in IPEC-J2 cells and LLC-PK1 (Fig. 3A andB). Therefore, we speculated that NSP8 induces the degradation of peroxisomes, leading to a decrease in MAVS localized in peroxisomes. To test this hypothesis, the peroxisomes were extracted from IPEC-J2 cells transfected with pXJ41 or Flag-NSP8. The results confirmed that NSP8 reduces the concentration of MAVS and related proteins on peroxisomes (Fig. 3C). To further investigate the effect of NSP8 on peroxisome abundance, IPEC-J2 cells were transfected with Flag-NSP8, and the numbers and morphology of peroxisomes were examined. In cells transfected with the empty vector pXJ41, peroxisomes were abundantly distributed throughout the cytoplasm. In contrast, the number of peroxisomes and MAVS located on peroxisomes was significantly reduced in NSP8-transfected cells (Fig. 3D). These results indicate that NSP8 significantly reduces peroxisome abundance in cells.
To further explore whether PEDV similarly modulates peroxisome abundance, IPEC-J2 cells and LLC-PK1 cells were infected with PEDV at different MOIs, and cell lysates were harvested at 24 hpi. The results demonstrated that the expression levels of endoge nous PMP70, catalase, and MAVS were significantly decreased in PEDV-infected cells in a dose-dependent manner (Fig. 3E andF). To further investigate whether PEDV-medi ated suppression of peroxisome-associated proteins and MAVS occurs specifically on peroxisomes, the peroxisomes were isolated from PEDV-infected IPEC-J2 cells. The results demonstrated that PEDV reduces the protein levels of MAVS and related proteins on peroxisomes (Fig. 3G). Finally, the abundance of peroxisomes in PEDV-infected IPEC-J2 cells was assessed. Confocal microscopy revealed a significant reduction in PMP70 expression in PEDV-infected cells compared with uninfected cells (Fig. 3H). Moreover, another peroxisome marker, GFP-Ser-Lys-Leu (SKL), was used to evaluate peroxisome number. The SKL motif, a peroxisomal matrix-targeting signal peptide, was fused to the C-terminus of GFP to direct it to peroxisomes. Consistent with PMP70 staining, GFP fluorescence was markedly reduced in PEDV-infected cells compared with uninfected controls (Fig. 3I). These results demonstrate that NSP8 suppresses MAVS by reducing the abundance of peroxisomes. dose-dependent manner. 293T cells (L) or IPEC-J2 cells (M) seeded in 6-well plates were transfected with HA-MAVS (3 µg) together with empty vector or increasing doses of Flag-NSP8 (3, 4, or 5 µg), respectively. At 24 hpt, the cells were harvested for western blotting analysis with antibodies against HA-tag, Flag-tag, and actin. At 24 hpt, the cells were harvested for western blotting analysis with antibodies against Flag-tag, PMP70, catalase, MAVS, and actin. (C) Whole cell lysate or peroxisomes were isolated from pxj41-transfected or Flag-NSP8-transfected IPEC-J2 cells and analyzed by western blotting using the indicated antibodies.
(D) Reduction in the abundance of peroxisomes in NSP8-transfected cells. IPEC-J2 cells grown on coverslips were transfected with Flag-NSP8 or the pXJ41 empty vector for 24 h and then incubated with anti-PMP70 (peroxisome marker) antibody, anti-MAVS antibody, and anti-Flag antibody for immunostaining. (E-F) PEDV degrades peroxisome-associated proteins PMP70 and catalase in a dose-dependent manner. IPEC-J2 cells and LLC-PK1 cells seeded in 6-well plates were infected with PEDV. At 24 hpi, the cells were harvested for western blotting analysis with antibodies against PEDV-N, PMP70, catalase, MAVS, and actin. (G) Whole cell lysate or peroxisomes were isolated from mock-infected or PEDV-infected IPEC-J2 cells and analyzed by western blotting using the indicated antibodies.
(H) Reduction in the abundance of peroxisomes in PEDV-infected cells. IPEC-J2 cells grown on coverslips were infected with PEDV for 24 h and then incubated with anti-PMP70 (peroxisome marker) antibody and anti-PEDV N antibody for immunostaining. (I) IPEC-J2 cells were transfected with GFP-SKL for 24 h, and then samples were collected at 24 hpi with PEDV, followed by incubation with anti-PEDV N antibody for immunostaining.
## PEDV NSP8 downregulates the production of IFN-III via the pexophagy pathway
A previous study reported that the reduction in peroxisome abundance is primarily regulated through selective autophagy, specifically pexophagy (25). To investigate whether NSP8-induced peroxisome reduction is associated with pexophagy, IPEC-J2 cells and LLC-PK1 cells were transfected with increasing amounts of Flag-NSP8 and analyzed by western blotting to assess the expression of autophagy markers LC3 and p62. The results demonstrated that a dose-dependent increase in the LC3-II/LC3-I ratio and a cells and LLC-PK1 cells seeded in 12-well plates were transfected with Flag-NSP8 (2, 2.5, 3 µg). At 24 hpt, the cells were harvested for western blotting analysis with antibodies against LC3, P62, Flag-tag, and actin. (C) IPEC-J2 cells were co-transfected with GFP-RFP-SKL and Flag-NSP8 or pXJ41 for 24 h, followed by incubation with anti-Flag antibody for immunostaining. (D-F) Baf-A1 can reverse the suppression of IFN-III production by NSP8. IPEC-J2 cells were transfected with pXJ41 (2 µg) or Flag-NSP8 (2 µg). After 8 h, cells were DMSO-treated or treated with Baf-A1 (25 nM). At 24 hpt, the cells were subsequently stimulated with 0.5 µg/mL poly(I:C) for 12 h before harvesting for RT-qPCR analysis. Date are means ± SDs from three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns: not significant (P > 0.05).(G) Baf-A1 reverses NSP8-mediated suppression of MAVS on peroxisomes in IPEC-J2 cells. IPEC-J2 cells were transfected with pXJ41 or Flag-NSP8. After 8 h, cells were DMSO-treated or treated with Baf-A1. After 24 h, the cells were harvested, and peroxisomes were extracted using a kit. Subsequently, western blotting was performed with the corresponding antibodies for analysis. (H) IPEC-J2 cells were transfected with GFP-RFP-SKL for 24 h, and then samples were collected at 24 hpi with PEDV, followed by incubation with anti-PEDV N antibody for immunostaining. (I) IPEC-J2 cells were transfected with GFP-RFP-SKL for 24 h, then mock-infected or infected with PEDV. Cells were treated with DMSO or Baf-A1 (25 nM) at 2 hpi with PEDV. After another 22 hpi, cells were incubated with anti-PEDV N antibody for immunostaining. concurrent decrease in p62 levels in both cell types (Fig. 4A andB) suggested that NSP8 activates the autophagy pathway. To further determine whether NSP8 specifically induces pexophagy, cells were co-transfected with either the empty vector (pXJ41) or Flag-NSP8 and the RFP-GFP-SKL reporter construct, a dualfluorescence probe target ing peroxisomes via the PTS1 signal. Confocal microscopy was employed to monitor changes in GFP fluorescence, which is quenched in acidic lysosomes, while RFP remains stable. In cells transfected with Flag-NSP8, the GFP signal was significantly reduced, resulting in fewer overlapping GFP/RFP signals (Fig. 4C). This result indicated that NSP8 promotes peroxisome degradation via pexophagy. To determine whether NSP8-induced suppression of IFN-III is mediated via pexophagy, the autophagy inhibitor Bafilomycin A1 (Baf-A1) was applied to pXJ41 or Flag-NSP8-transfected cells. RT-qPCR results demonstra ted that Baf-A1 reversed the NSP8-induced suppression of IFN-III (IFN-λ1, IFN-λ3, and IFN-λ4) expression (Fig. 4D through F). Finally, to determine whether NSP8-mediated inhibition of MAVS occurs through pexophagy, IPEC-J2 cells transfected with either pXJ41 or Flag-NSP8 were treated with Baf-A1, and then the peroxisomes were isolated. The results demonstrated that Baf-A1 alleviated NSP8-induced suppression of MAVS in peroxisomes (Fig. 4G).
Autophagy and the ubiquitin-proteasome system are the two major intracellu lar degradation pathways. While individual proteins are typically degraded via the proteasome, organelles, such as peroxisomes, are degraded via the lysosomal pathway. To assess which pathway mediates peroxisome degradation during PEDV infection, IPEC-J2 cells were treated with the proteasome inhibitor MG132 or the autophagy inhibitors Baf-A1 and 3-MA. In both mock-and PEDV-infected cells, degradation of PMP70 and catalase was unaffected by MG132 but was blocked by Baf-A1 and 3-MA (Fig. S1), confirming that PEDV-induced peroxisome degradation is mediated by autophagy. To further validate whether PEDV infection induces pexophagy, the RFP-GFP-SKL reporter was transfected into PEDV-infected cells. Confocal microscopy revealed that the GFP fluorescence was quenched in acidic lysosomes, resulting in fewer overlapping GFP/RFP signals. Many PEDV-infected cells displayed RFP signals that did not co-localize with GFP, indicative of active pexophagy (Fig. 4H). Importantly, treatment with Baf-A1 rescued the suppression of the RFP/GFP ratio in PEDV-infected IPEC-J2 cells (Fig. 4I). Taken together, these findings suggest that PEDV NSP8 suppresses the expression of MAVS on peroxisomes and the production of IFN-III by promoting peroxisome degradation via pexophagy.
## PEDV NSP8 interacts with PEX13
To identify host proteins involved in NSP8-triggered pexophagy, a mass spectrometry (MS)-based proteomic screening was conducted. Notably, PEX13, a key component of the peroxisomal matrix protein import machinery (26,27), was the only peroxi some-associated protein significantly enriched in NSP8 immunoprecipitates. To validate the results of MS, a co-immunoprecipitation (Co-IP) assay was performed. 293T cells were co-transfected with Flag-NSP8 and pCAGGS-HA-PEX13 plasmids, followed by immunoprecipitation using anti-Flag antibodies. The Co-IP results confirmed that NSP8 interacts with PEX13 (Fig. 5A). A reverse Co-IP experiment was subsequently performed using anti-HA antibodies, which further demonstrated that NSP8 could be efficiently co-immunoprecipitated with PEX13 (Fig. 5B). Together, the forward and reverse Co-IP results support a specific interaction between PEDV NSP8 and PEX13 under overexpres sion conditions. To assess whether this interaction occurs endogenously during PEDV infection, IPEC-J2 cells infected with PEDV (MOI = 10) for 12 hpi were subjected to Co-IP using anti-NSP8 antibodies. As shown in Fig. 5C, endogenous PEX13 was suc cessfully co-immunoprecipitated with NSP8 in PEDV-infected cells, confirming that the interaction occurs in a physiological context. Confocal microscopy further supported this finding by demonstrating strong colocalization between NSP8 and PEX13 (Fig. 5D andE). As a coronavirus, PEDV utilizes NSP8 as a key component of its viral replica tion complex (NSP7/NSP8/NSP12/dsRNA), and NSP8 can bind RNA templates to initiate the synthesis of complementary oligonucleotides and exhibits secondary RNA-depend ent RNA polymerase activity (28,29). Therefore, the interaction between NSP8 and PEX13 may occur either within the context of the replication complex on endoplasmic reticulum (ER). To confirm this, the confocal microscopy was used to show whether HA-PEX13 colocalized with NSP8 in the ER. As shown in Fig. 5F, the PEX13 protein with NSP8 was located in the ER. As a negative control, in the mock-infected group, there was no co-localization between PEX13 and the ER. Therefore, these results hypothesize that NSP8 in the PEDV replication complex recruits PEX13 to form the ER-peroxisome membrane contact sites.
To map the domain of NSP8 responsible for the interaction, a series of NSP8 truncation mutants wase generated based on structural predictions using SWISS-MODEL. The tertiary structure of NSP8 was predicted to contain a long α-helix (N2), a random coil region (N3), and a short α-helix (C), with the remaining region designated as N1. Co-IP analysis showed that the N3 domain of NSP8 was responsible for the interaction with PEX13. Deletion of the N3 domain (GFP-NSP8-ΔN3) abolished this interaction (Fig. 5G andH), confirming the importance of this region in mediating binding.
## PEDV NSP8 induces pexophagy and inhibits IFN-III production by degrading PEX13
To further investigate the relationship between NSP8 and PEX13, IPEC-J2 cells were transfected with a fixed amount of pCAGGS-HA-PEX13 and increasing amounts of Flag-NSP8. Western blotting revealed that NSP8 reduced the expression of overex pressed PEX13 in a dose-dependent manner (Fig. 6A). In a separate experiment, IPEC-J2 cells were transfected with increasing doses of Flag-NSP8, and endogenous PEX13 levels were analyzed by western blotting. As shown in Fig. 6B, NSP8 suppressed endogenous PEX13 expression in a dose-dependent manner. Similarly, PEDV infection also signifi cantly decreased PEX13 expression in a dose-dependent manner (Fig. 6C). These results collectively demonstrated that NSP8 mediates the degradation of PEX13. Next, we explored the pathway responsible for NSP8-induced PEX13 degradation. As shown in Fig. 6D, treatment with the autophagy inhibitors 3-MA and Baf-A1 reversed the degradation of PEX13, whereas the proteasome inhibitor MG132 had no effect. This indicates that NSP8 promotes the degradation of PEX13 via the autophagy pathway. (4 µg) for 6 h, after which the medium was refreshed again. At 36 hpt, cells were processed for IP using either anti-Flag magnetic beads (A) or anti-HA magnetic beads (B). Both whole-cell lysates (input) and immunoprecipitated proteins were immunoblotted with antibodies targeting Flag-tag, HA-tag, and actin. (C) IPEC-J2 cells seeded in 60 mm dishes were infected with PEDV or mock. At 24 hpi, cells were processed for IP using anti-NSP8. Both input and immunoprecipitated proteins were immunoblotted with antibodies targeting NSP8, PEX13, and actin. (D) PEDV NSP8 colocalizes with PEX13 in the cytoplasm.
IPEC-J2 cells were sequentially transfected with Flag-NSP8 (4 µg) along with either the empty vector or HA-PEX13 (2 µg). At 36 hpt, cells were stained with anti-HA (green) and anti-Flag (red) antibodies and subjected to confocal microscopy. Nuclei Previous studies have demonstrated that depletion of PEX13 can trigger pexophagy (19,20). To further investigate whether NSP8-induced pexophagy is dependent on the degradation of PEX13, we generated IPEC-J2 wild type (WT), PEX13 knockdown (KD), and PEX13 overexpression (OE) cell lines (Fig. 6E andF). These cells were transfected with either the empty vector pXJ41 or Flag-NSP8, along with the RFP-GFP-SKL reporter, and fixed at 24 h post-transfection. Confocal microscopy revealed that in WT cells, NSP8 overexpression significantly reduced the RFP/GFP signal ratio compared with vectortransfected controls. In PEX13-KD cells, the RFP/GFP ratio was further decreased, while in PEX13-OE cells, GFP signal quenching was absent (Fig. 6G). These results demonstrate that NSP8 induces pexophagy by downregulating PEX13 expression.
To evaluate the effect of the NSP8-PEX13 interaction on IFN-III signaling, increas ing amounts of NSP8 were transfected into IPEC-J2-WT, IPEC J2-PEX13 KD, and IPEC-J2-PEX13 OE cells, and MAVS expression was assessed. NSP8 significantly suppressed MAVS expression in both WT and KD cells in a dose-dependent manner. However, this inhibitory effect was abolished in PEX13-OE cells (Fig. S2A). In addition, poly(I:C)-induced expression of IFN-λ1 was significantly higher in PEX13-OE cells compared to WT cells, while it was markedly reduced in PEX13-KD cells (Fig. S2B). Overall, these findings indicate that PEDV NSP8 suppresses MAVS expression and inhibits IFN-III production by promoting pexophagy through PEX13 degradation.
## NSP8 triggers pexophagy by downregulating PEX13 and inducing PEX5 ubiquitination
Recent studies have demonstrated that depletion of PEX13 leads to the accumulation of ubiquitinated PEX5 on peroxisomal membranes, which in turn triggers ubiquitindependent pexophagy (30,31). To explore whether NSP8 regulates PEX5 ubiquitina tion through its effects on PEX13, 293T cells were co-transfected with GFP-PEX13 or the empty vector, along with pXJ41-MYC-NSP8 or its vector control, HA-tagged ubiquitin (HA-UB), and Flag-tagged PEX5. As shown in Fig. 7A, overexpression of PEX13 significantly reduced the polyubiquitination of PEX5 protein. However, co-expression of NSP8 and PEX13 reversed this effect, enhancing PEX5 ubiquitination compared to PEX13 overexpression alone. Interestingly, while PEX5 was found to interact with PEX13, no direct interaction was observed between PEX5 and NSP8. These findings suggest that NSP8 relieves PEX13-mediated suppression of PEX5 ubiquitination. To further elucidate the mechanism underlying NSP8-induced pexophagy via ubiquitina ted PEX5 accumulation, we performed a measure of three critical pexophagy-related proteins in PEX13-KD cells: NBR1, an autophagy receptor responsible for recognizing ubiquitinated peroxisomes and recruiting them to nascent autophagosomes; PEX2, a peroxisomal E3 ubiquitin ligase mediating peroxisomal protein ubiquitination during pexophagy induction; and PEX5, a well-characterized ubiquitination target in pexoph agy. As demonstrated in Fig. 7B, knockdown of PEX13 significantly increased PEX5 ubiquitination, and this effect was further enhanced by NSP8 overexpression. Co-immu noprecipitation assays revealed that PEX5 interacted with both NBR1 and PEX2, and Flag-NSP8 (2, 2.5, or 3 µg). At 24 hpt, cells were collected for western blotting using antibodies against Flag-tag, PEX13, and actin. (C) PEDV degrades PEX13 in a dose-dependent manner. IPEC-J2 cells seeded in 12-well plates were transfected with HA-PEX13 (1 µg). At 24 hpt, the cells were infected with mock or increasing MOIs of PEDV (MOI = 1, 5, 10). At 24 hpi, the cells were harvested for western blotting analysis using antibodies against HA-tag, PEDV-N, and actin. (D) Effects of MG132, Baf-A1, and 3-MA on PEX13 degradation induced by NSP8. IPEC-J2 cells seeded in 12-well plates were staggered-transfected with HA-PEX13 (1 µg) and Flag-NSP8 (3 µg). At 12 hpt, cells were treated with DMSO, MG132 (10 nM), 3-MA (2 mM), or Baf-A1 (25 nM). After another 24 h, cells were harvested for western blotting analysis using antibodies against HA-tag, Flag-tag, and actin. (E) Western blotting analysis of lysates from IPEC-J2 cells infected with sgPEX13 or pLentiCRISPR-V2. β-actin was used as the loading control. (F) Western blotting analysis of lysates from IPEC-J2 cells infected with pCDH-HA-PEX13 or pCDH-HA. these interactions were strengthened in PEX13-KD cells compared with WT cells (Fig. 7B). Interference efficiency of siRNAs targeting PEX5, PEX2, NBR1 (siNBR1-2), and ATG7 (siATG7-2) was validated, and the most efficient siRNAs were selected for further experiments (Fig. 7C through F). ATG7, essential for autophagosome formation, served as a positive control for autophagy inhibition. To investigate whether these pexophagy-rela ted proteins are involved in NSP8-mediated peroxisome loss due to PEX13 degradation, both WT and PEX13-KD IPEC-J2 cells were transfected with siRNAs targeting PEX2, PEX5, NBR1, and ATG7, followed by transfection with either Flag-NSP8 or the empty vector. Cells were fixed and analyzed by confocal microscopy to assess peroxisome abundance via PMP70 immunostaining. As expected, NSP8 reduced peroxisome abundance in both WT and PEX13-KD cells. However, silencing PEX2, PEX5, NBR1, or ATG7 significantly reversed the NSP8-induced loss of peroxisomes in PEX13-KD cells (Fig. 7G). Taken together, these data suggest that NSP8 triggers pexophagy by downregulating PEX13 expression, which in turn induces PEX5 ubiquitination and leads to the recruitment of pexophagy machinery, culminating in peroxisome degradation.
## DISCUSSION
PEDV is one of the primary pathogens responsible for diarrhea in piglets, character ized by high morbidity and mortality. The virus primarily invades pigs through the digestive tract, targeting small intestinal epithelial cells for replication. The intestinal mucosa mounts a type III interferon response as the first line of antiviral defense at mucosal surfaces (32,33). However, PEDV evades host immunity, notably by suppressing IFN-III production. This study aimed to elucidate the molecular mechanisms underlying PEDV-mediated evasion of IFN-III responses, thereby identifying potential therapeutic targets for combating PEDV infection. Given that PEDV causes acute intestinal disease, we hypothesized that it disrupts the innate immune defenses of intestinal epithelial cells, particularly IFN-III signaling. Consistent with this hypothesis, our data revealed that PEDV infection inhibits the production of IFN-III in IPEC-J2 cells. Mechanistically, PEDV NSP8 acts as a potent immune suppressor, inhibiting the promoter activity of IFN-III by downregulating the expression of MAVS. Both type I and type III interferons are activated through the RLR signaling pathway (34). While mitochondrial MAVS induces IFN-I production, peroxisomal MAVS is primarily responsible for IFN-III induction (35). Peroxisomes are now recognized as key subcellular platforms for antiviral signaling due to MAVS localization (12,36). Similar to our findings, viruses, such as hepatitis C virus and human cytomegalovirus, exploit this system by targeting peroxisomal MAVS to block IFN-III responses (36)(37)(38)(39)(40), consistent with our findings.
Beyond MAVS localization, peroxisome abundance and plasticity are also critical for regulating immune signaling (41). For example, flaviviruses like dengue and West Nile virus reduce peroxisome numbers by targeting the peroxisomal biogenesis factor PEX19, leading to diminished IFN-III production (42). Odendall et al. also found that increasing peroxisome abundance enhances IFN-III expression (9). A prior study revealed that PEDV NSP1 suppresses IFN-III production by reducing peroxisome abundance (16). were probed using antibodies against Flag-tag, MYC-tag, HA-tag, GFP-tag, and actin. (B) NSP8 promotes the ubiquitination of PEX5 and enhances the interaction of PEX5 with NBR1 and PEX2 in PEX13 knockdown cells. WT or KD PEX13 cells were transfected with MYC-NSP8 (12 µg), Flag-PEX5 (6 µg), and HA-Ub (6 µg). At 36 hpt, cells were processed for IP with anti-Flag magnetic beads. Input and precipitated proteins were probed using antibodies against Flag-tag, MYC-tag, HA-tag, PEX13, NBR1, PEX2, and actin. (C-F) IPEC-J2 cells were transfected with siCtrl, siPEX5, siPEX2, siNBR1, or siATG7. At 36 hpt, cells were harvested for western blotting analysis using antibodies against actin, PEX5, PEX2, NBR1, or ATG7. (G) The NSP8-mediated reduction in peroxisome abundance was completely abolished in cells with knockdown (KD) of PEX5, PEX2, NBR1, or ATG7.
Representative images of KD PEX13 or WT cells treated with the indicated siRNA and transfected with pXJ41 or Flag-NSP8 were immunostained for the peroxisomal marker PMP70 at 36 hpt.
Consistent with this, our study revealed that both PEDV infection and NSP8 overexpres sion decreased the expression of peroxisome-associated proteins and overall peroxisome abundance. This loss is mediated via pexophagy, the selective autophagic degradation of peroxisomes. Supporting this, we demonstrated that PEDV infection and NSP8 overex pression both activate autophagy and pexophagy, which contribute to the suppression of MAVS-mediated IFN-III production.
To investigate the molecular mechanism by which NSP8 induces pexophagy, we performed mass spectrometry-based screening and identified PEX13 as a direct binding partner. This interaction was further confirmed through Co-IP and confocal microscopy assays. PEX13 is a key component of the peroxisomal matrix import machinery and plays a protective role against peroxisomal degradation. PEX13 can recognize proteins with the PTS1 signal and mediate their entry into peroxisomes (43). The absence of PEX13 leads to ER stress (44). Research has found that PEX13 serves as a novel factor regulating pexophagy, and reducing the protein level of PEX13 is an effective and intrinsic way to induce pexophagy (20). Loss of PEX13 leads to the accumulation of ubiquitinated PEX5 on peroxisomes, a hallmark trigger of pexophagy. Our findings revealed that both PEDV infection and NSP8 overexpression suppress PEX13 expression, and this down regulation occurs via the autophagy pathway. Other studies have found that during Porcine Deltacoronavirus (PDCOV) infection, the SIRT5 protein regulates the desucciny lation of the PDCOV M protein, which in turn leads to an increase in the ubiquitina tion of PEX5 and initiates pexophagy (45). Newcastle disease virus infection triggers excessive ROS production, activating the phosphorylation and peroxisomal localiza tion of ataxia-telangiectasia mutated (ATM). Activated ATM promotes the interaction between the peroxisomal receptor PEX5, driving pexophagy (46). Consistent with this, further mechanistic studies revealed that NSP8 reduces the pool of PEX13-bound PEX5, facilitating its ubiquitination. Ubiquitinated PEX5 subsequently recruits the autophagy receptor NBR1, promoting autophagosome formation (20,47). Notably, PEX2, an E3 ubiquitin ligase, is also involved in mediating PEX5 ubiquitination during pexophagy (30). We found that NSP8 enhances PEX5 ubiquitination and strengthens its interactions with both PEX2 and NBR1. Importantly, the NSP8-induced reduction in peroxisome abundance was abolished when PEX5, PEX2, NBR1, or ATG7 were silenced, highlight ing their essential roles in this pathway. These findings establish that NSP8 induces pexophagy by downregulating PEX13, a novel regulator of peroxisomal autophagy. This disruption impairs MAVS-mediated signaling. Supporting this, NSP8-mediated suppres sion of MAVS was exacerbated in PEX13 knockdown cells but alleviated in PEX13-overex pressing cells. Furthermore, IFN-III expression was significantly reduced in PEX13-KD cells and elevated in PEX13-OE cells compared to wild-type controls.
In summary, our study reveals a previously unrecognized mechanism by which PEDV NSP8 hijacks the host autophagy machinery to degrade peroxisomes through a PEX13-dependent pexophagy pathway (Fig. 8). This process undermines peroxisomal MAVS signaling and suppresses IFN-III production, allowing PEDV to evade antiviral responses. These insights provide a potential basis for developing antiviral strategies targeting the NSP8-PEX13-pexophagy axis.
## MATERIALS AND METHODS
## Cell lines, viruses, and plasmids
Vero cells (ATCC CCL-81) and 293T cells (ATCC CRL-3216) were cultured in Dulbecco's modified Eagle's medium (DMEM) (Corning, USA) supplemented with 10% fetal calf serum (Gibco, USA) and 50 IU/mL penicillin-streptomycin (30-002-CI, Corning, USA). IPEC-J2, a continuous line of epithelial cells derived from the jejunum of a 12-h-old, colostrum-deprived, mixed breed piglet (48), was also maintained in DMEM (Corning, USA). LLC-PK1 cells (ATCC CL-101) were cultured in MEM (Corning, USA). All cells were incubated at 37°C in a humidified incubator with 5% CO 2 .
The PEDV variant strain MSCH (GenBank accession no. MT683617) was isolated and maintained in our laboratory. The virus was propagated in Vero cells and IPEC-J2 cells with 6 µg/mL trypsin (Sigma-Aldrich, USA) as previously described (49).
Plasmids used in this study included pIFN-λ1(-225 to -36), HA-PEX13, GFP-PEX13, MYC-NSP8, GFP-SKL, RFP-GFP-SKL, Flag-PEX5, and various NSP8 and PEX13 trunca tion constructs (all constructed in-house using standard molecular cloning proce dures), except for GFP-NSP8-N3, which was synthesized by Tsingke Biotechnology (Beijing, China). Additional expression plasmids, including HA-RIG-I, HA-MDA5, HA-MAVS, HA-IRF3, Flag-TBK1, Flag-IKKε, HA-UB, and pRL-TK, were kindly provided by Dr. Xing Liu. Plasmids encoding PEDV proteins were generated as previously described (50).
## Antibodies and chemicals
Rabbit antibodies against LC3, P62, MAVS, catalase, PEX2, PMP70, and PEX13 were purchased from Proteintech (USA). Mouse antibodies against GFP and Flag were obtained from Abmart (China), while anti-NBR1 and anti-PEX5 antibodies were acquired from Santa Cruz Biotechnology (USA). Rabbit anti-HA and anti-ATG7 antibodies were sourced from Cell Signaling Technology (CST) (USA). Monoclonal antibodies against the PEDV N protein and polyclonal antibodies against NSP8 were preserved in our labo ratory. Horseradish peroxidase (HRP)-conjugated secondary antibodies against rabbit or mouse IgG were obtained from Bioss (China). For chemical treatments, cells were incubated with DMSO, proteasome inhibitors, or autolysosome inhibitors, ensuring that the final DMSO concentration did not exceed 2%. The proteasome inhibitor MG132 and the autophagy inhibitors 3-methyladenine (3-MA) and bafilomycin A1 (Baf-A1) were purchased from Selleck (USA).
## Plasmids and molecular clones
Plasmid transfections in 293T cells were carried out using Lipofectamine 3000 (Thermo Fisher Scientific, L3000015), while transfections in Vero and IPEC-J2 cells were performed using Lipo8000 (Bryotime, C0533). For siRNA transfection, Lipofectamine 3000 was used to transfect siRNAs into IPEC-J2 cells following the manufacturer's instructions. All transfection procedures were conducted in accordance with the recommended protocols provided by the respective manufacturers. The sequences of the sgRNAs and siRNAs used in this study are listed in Table 1.
## Peroxisome isolation
Peroxisome isolation was performed from IPEC-J2 cells using the Peroxisome Isolation Kit (Sigma, PEROX1). Briefly, 2 × 10⁸ cells were harvested, washed in phosphatebuffered saline, and centrifuged at 250 × g for 5 min. The cell pellet was resuspended in perox isome extraction buffer [5 mM 3-(N-morpholino) propanesulfonic acid, pH 7.65, 0.25 M sucrose, 1 mM EDTA, 0.1% ethanol, and protease inhibitor cocktail], vortexed, and homogenized with a 7 mL Dounce homogenizer. The homogenate was centrifuged sequentially at 1,000 × g and 2,000 × g for 10 min each. The supernatant was then centrifuged at 25,000 × g for 20 min to isolate the crude peroxisomal fraction, which was further purified by density gradient centrifugation using Optiprep. After centrifugation at 100,000 × g for 1.5 h, the purified peroxisomes were collected for analysis.
## Western blotting analysis
Cells were lysed using radioimmunoprecipitation assay lysis buffer (Beyotime, China) on ice for 30 min. The lysates were separated by SDS-PAGE and subsequently transferred onto nitrocellulose membranes. Membranes were then blocked with 10% nonfat milk in PBST for 2 h at room temperature (RT), followed by washing with PBST and incuba tion with the appropriate primary antibodies overnight at 4°C. Afterward, membranes were incubated with HRP-conjugated secondary antibodies for 1 h at RT. Protein bands were visualized using an enhanced chemiluminescence kit (Tanon, China), and signal intensities were captured and analyzed using a Tanon 5200 chemiluminescence imaging system (Tanon, China).
## Quantitative real-time PCR
Cells were washed once with PBS, and total RNA was extracted using the Total RNA Kit I (Omega Bio-Tek, USA) according to the manufacturer's instructions. For reverse transcription, 1 µg of RNA was reverse transcribed using HiScript qRT SuperMix (Vazyme, Relative gene expression was calculated using the comparative cycle threshold (CT) method, with GAPDH used as the internal control for normalization of both host and viral gene expression. The primers used in this study are listed in Table 2.
## Lentivirus production and transduction
Lentiviral particles were produced by transient co-transfection of the packaging plasmids pMD2.G (Addgene #12259), psPAX2 (Addgene #12260), and plentiCRISPRv2-sgRNA into HEK293T cells using polyethylenimine transfection reagent (Yeasen, China). Medium was 12 hours post-transfection (hpt), and viral supernatants were collected 48 and 60 hpt and stored at -80°C. Target cells were transduced with lentivirus in the presence of 10 µg/mL polybrene.
## Co-immunoprecipitation assay
For Co-IP, cells were transfected with plasmids for 24 h, lysed in NP40 cell buffer containing PMSF (protease inhibitor, Beyotime, China), and centrifuged. Supernatants were incubated with anti-Flag (Sigma, USA), anti-HA (CST, USA), or anti-GFP (Proteintech, USA) magnetic beads. Beads were washed with PBST and eluted in 50 mM glycine buffer (pH 2.8). The eluted proteins were analyzed by immunoblotting using appropriate antibodies.
## Confocal microscopy
For immunofluorescence, cells were cultured on 15 mm glass-bottom dishes (Nest Biotechnology, China), fixed with 4% paraformaldehyde for 10 min at RT, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 2% BSA in PBS for 1 h. Primary antibodies were applied overnight at 4°C, followed by incubation with Alexa Fluor 488-conjugated secondary antibodies (Proteintech, China) for 1 h at RT. Nuclei were then stained with DAPI (Biosharp, China) for 10 min. Fluorescence images were captured using a Nikon A1 confocal microscope (Japan), and two or three channels were recorded either sequentially or simultaneously while avoiding signal overlap.
## Dual luciferase assay
For the dual luciferase assay, 293T cells grown in 24-well plates were co-transfected with the IFN-λ1-Luc reporter plasmid, pRL-TK internal control plasmid, and various expression plasmids or control vectors using Lipofectamine 3000, according to the manufacturer's instructions. At 36 hpt, cells were harvested and luciferase activity was measured using a dual-luciferase reporter assay kit according to standard procedures.
## Statistical analysis
Data from at least three independent experiments were analyzed using one-way or two-way analysis of variance followed by Tukey's post hoc test for multiple comparisons (GraphPad Prism Software Inc., San Diego, CA, USA). Results are expressed as the mean
## References
1. Jung, Saif (2015) "Porcine epidemic diarrhea virus infection: etiology, epidemiology, pathogenesis and immunoprophylaxis" *Vet J*
2. Kocherhans, Bridgen, Ackermann et al. (2001) "Completion of the porcine epidemic diarrhoea coronavirus (PEDV) genome sequence" *Virus Genes*
3. Lee (2015) "Porcine epidemic diarrhea virus: an emerging and reemerging epizootic swine virus" *Virol J*
4. Wu, Li, Tian et al. (2023) "Broad antagonism of coronaviruses nsp5 to evade the host antiviral responses by cleaving POLDIP3" *PLoS Pathog*
5. Xu, Gao, Zhang et al. (2023) "Porcine epidemic diarrhea virus antagonizes host IFN-λ-mediated responses by tilting transcription factor STAT1 toward acetylation over phosphorylation to block its activation" *mBio*
6. Li, Lai, Qiu et al. (2024) "Hyperacety lated microtubules assist porcine deltacoronavirus nsp8 to degrade MDA5 via SQSTM1/p62-dependent selective autophagy" *J Virol*
7. Zhang, Yang, Pan et al. (2023) "SARS-CoV-2 Nsp8 suppresses MDA5 antiviral immune responses by impairing TRIM4-mediated K63-linked polyubiquitination" *PLoS Pathog*
8. Deng, Zheng, Nan et al. (2023) "SARS-CoV-2 NSP8 suppresses type I and III IFN responses by modulating the RIG-I/MDA5, TRIF, and STING signaling pathways" *J Med Virol*
9. Odendall, Dixit, Stavru et al. (2014) "Diverse intracellular pathogens activate type III interferon expression from peroxisomes" *Nat Immunol*
10. Mesev, Ledesma, Ploss (2019) "Decoding type I and III interferon signalling during viral infection" *Nat Microbiol*
11. Donnelly, Kotenko (2010) "Interferon-lambda: a new addition to an old family" *J Interferon Cytokine Res*
12. Dixit, Boulant, Zhang et al. (2010) "Peroxisomes are signaling platforms for antiviral innate immunity" *Cell*
13. Lodhi, Semenkovich (2014) "Peroxisomes: a nexus for lipid metabolism and cellular signaling" *Cell Metab*
14. He, Dean, Lodhi (2021) "Peroxisomes as cellular adaptors to metabolic and environmental stress" *Trends Cell Biol*
15. Zhang, Shi, Yoo (2016) "Suppression of type I interferon production by porcine epidemic diarrhea virus and degradation of CREB-binding protein by nsp1" *Virology (Auckland)*
16. Zhang, Ke, Blikslager et al. (2018) "Type III interferon restriction by porcine epidemic diarrhea virus and the role of viral protein nsp1 in IRF1 signaling" *J Virol*
17. Choi, Bowman, Jung (2018) "Autophagy during viral infection -a double-edged sword" *Nat Rev Microbiol*
18. Viret, Duclaux-Loras, Rozières et al. (2021) "Selective autophagy receptors in antiviral defense" *Trends Microbiol*
19. Lee, Sumpter, Zou et al. (2017) "Peroxisomal protein PEX13 functions in selective autophagy" *EMBO Rep*
20. Demers, Riccio, Jo et al. (2023) "PEX13 prevents pexophagy by regulating ubiquitinated PEX5 and peroxisomal ROS" *Autophagy*
21. Meinecke, Cizmowski, Schliebs et al. (2010) "The peroxisomal importomer constitutes a large and highly dynamic pore" *Nat Cell Biol*
22. Li, Yang, Zhu et al. (2020) "Porcine epidemic diarrhea virus and the host innate immune response" *Pathogens*
23. Ding, Fang, Jing et al. (2014) "Porcine epidemic diarrhea virus nucleocapsid protein antagonizes beta interferon production by sequestering the interaction between IRF3 and TBK1" *J Virol*
24. Zheng, Chen, Liu et al. (2022) "Ubiquitin ligase MARCH5 localizes to peroxisomes to regulate pexophagy" *J Cell Biol*
25. Oku, Sakai (2010) "Peroxisomes as dynamic organelles: autophagic degradation" *FEBS J*
26. Wang, Li, Chai et al. (2019) "Pex13 and Pex14, the key components of the peroxisomal docking complex, are required for peroxisome formation, host infection and pathogenicity-related morphogenesis in Magnaporthe oryzae" *Virulence*
27. Gao, Skowyra, Feng et al. (2022) "Protein import into peroxisomes occurs through a nuclear pore-like phase" *Science*
28. Gao, Yan, Huang et al. (2020) "Structure of the RNA-dependent RNA polymerase from COVID-19 virus" *Science*
29. Biswal, Diggs, Xu et al. (2021) "Two conserved oligomer interfaces of NSP7 and NSP8 underpin the dynamic assembly of SARS-CoV-2 RdRP" *Nucleic Acids Res*
30. Sargent, Van Zutphen, Shatseva et al. (2016) "PEX2 is the E3 ubiquitin ligase required for pexophagy during starvation" *J Cell Biol*
31. Law, Bronte-Tinkew, Pietro et al. (2017) "The peroxisomal AAA ATPase complex prevents pexophagy and development of peroxisome biogenesis disorders" *Autophagy*
32. Mordstein, Neugebauer, Ditt et al. (2010) "Lambda interferon renders epithelial cells of the respiratory and gastrointestinal tracts resistant to viral infections" *J Virol*
33. Minkoff (2023) "Innate immune evasion strategies of SARS-CoV-2" *Nat Rev Microbiol*
34. Onoguchi, Yoneyama, Takemura et al. (2007) "Viral infections activate types I and III interferon genes through a common mechanism" *J Biol Chem*
35. He, Huang, Nie et al. (2023) "MAVS integrates glucose metabolism and RIG-I-like receptor signaling" *Nat Commun*
36. Magalhães, Ferreira, Gomes et al. (2016) "Peroxisomes are platforms for cytomegalovirus' evasion from the cellular immune response" *Sci Rep*
37. Bender, Reuter, Eberle et al. (2015) "Activation of type I and III interferon response by mitochondrial and peroxisomal MAVS and inhibition by hepatitis C virus" *PLoS Pathog*
38. Ferreira Ana, Magalhães, Camões et al. (2016) "Hepatitis C virus NS3-4A inhibits the peroxisomal MAVSdependent antiviral signalling response" *J Cell Mol Med*
39. Ferreira, Rita, Gouveia et al. (2022) "Human cytomegalovirus vMIA inhibits MAVS oligomerization at peroxisomes in an MFF-dependent manner" *Front Cell Dev Biol*
40. Ashley, Abendroth, Mcsharry et al. (2019) "Interferonindependent innate responses to cytomegalovirus" *Front Immunol*
41. Cook, Moreno, Beltran et al. (2019) "Peroxisome plasticity at the virus-host interface" *Trends Microbiol*
42. You, Hou, Malik-Soni et al. (2015) "Flavivirus infection impairs peroxisome biogenesis and early antiviral signaling" *J Virol*
43. Gaussmann, Peschel, Ott et al. (2024) "Modulation of peroxisomal import by the PEX13 SH3 domain and a proximal FxxxF binding motif" *Nat Commun*
44. Rishi, Bhatia, Secondes et al. (2020) "Hepatocytespecific deletion of peroxisomal protein PEX13 results in disrupted iron homeostasis" *Biochim Biophys Acta Mol Basis Dis*
45. Li, Tang, Lai et al. (2025) "SIRT5mediated desuccinylation of the porcine deltacoronavirus M protein drives pexophagy to enhance viral proliferation" *PLoS Pathog*
46. Jiang, Qu, Kan et al. (2025) "Pexophagy-driven redox imbalance promotes virusinduced ferroptosis" *Cell Rep*
47. Wang, Wang, Zhang et al. (2020) "The autophagic degradation of cytosolic pools of peroxisomal proteins by a new selective pathway" *Autophagy*
48. Schierack, Nordhoff, Pollmann et al. (2006) "Characterization of a porcine intestinal epithelial cell line for in vitro studies of microbial pathogenesis in swine" *Histochem Cell Biol*
49. Wang, Bai, Liu et al. (2020) "Tomatidine inhibits porcine epidemic diarrhea virus replication by targeting 3CL protease" *Vet Res*
50. Zhu, Li, Bai et al. (2022) "A systemic study of subcellular localization of porcine epidemic diarrhea virus proteins" *Pathogens* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12697183&blobtype=pdf | # Phages infecting Clostridium sporogenes
Joanna Steczynska, Kaleb Jackson, Kelly Williams
## Abstract
Clostridium sporogenes is a spore-forming anaerobe found ubiquitously in the environment associated with food spoilage. It is also a gut commensal. We report isolation of phages able to infect C. sporogenes, with 30 unique genomes falling into three congeneric species.
KEYWORDS phage, C. sporogenes, ClostridiumC lostridium sporogenes is a common human gut commensal and is associated with food spoilage. It is also a non-toxigenic model organism for Clostridium botulinum (1). Here, we report the isolation of 30 unique phages for this organism, whose genomes have been sequenced and annotated. They serve as a useful resource for the develop ment of tools to control C. sporogenes and related species.Multiple sewage samples were harvested from the Livermore Water Reclamation Plant (Livermore, CA, USA) wastewater facility on September 20, 2024. The supernatants were filtered (0.45 mm pore size) and enriched immediately with C. sporogenes ATCC 15579. Growth was carried out under anaerobic conditions at 37°C (Don Whitley, A35 chamber, 5% H 2 ) in Reinforced Clostridial Medium (Sigma-Aldrich). Resulting plaques were medium to large with clear centers and hazy, irregular edges, except for CS19, which exhibited large, very hazy plaques. Forty-three plaques were purified three times, and genomic DNA was extracted from high titer stocks (>10 8 pfu/mL; Phage DNA Isolation kit, Norgen Biotek). Libraries were prepared using the Illumina DNA Prep kit and sequenced using the MiSeq V3 150-cycle kit in paired-end mode (Illumina).Reads were obtained for 42 phages and assembled using SPAdes v3.15.5 (2). Each assembly yielded a single large (~38 kbp) high-coverage contig. Circularization of these contigs was performed by manually trimming one of the ~55 bp terminal repeats typically left for DNA circles by SPAdes, and circularity of the product was verified using ReadStepper with the raw reads (3). Some of the 42 resulting genome sequences were identical, comprising 30 unique sequences (Table 1; Fig. 1A). The genomes all had close (with hits ≥8,000 bits, percent identity ranging from 91.6 to 100, and length from 4,433 to 48,556) BLASTN relationships with each other and with a GenBank entry (NC_019924.1) for another C. sporogenes phage, phi8074-B1, but showed no hits to any other viral GenBank entries nor any of the 8630 reference prokaryotic virus genomes listed in the International Committee on Taxonomy of Viruses (ICTV) document MSL40.v1 (4). Phylogenetic analysis is shown in Fig. 1A. Pharokka (5) v1.7.5 (flags: -g prodigal --dnapler) was used for genome annotation, yielding no tRNA genes. Gene annotations (numerous tail genes, lack of integrase or repressor genes) indicated a virulent tailed phage (Fig. 1B), that is, the taxonomic class Caudoviricetes. No lower taxonomic ranks could be assigned due to the failure to match any ICTV reference genomes. However, the sequences for the 30 new genomes and for phi8074-B1 were submitted to the VIRIDIC website (6), which placed them into a single genus cluster and four species clusters (one containing only phi8074-B1) (Table 1; Fig. 1A). Default parameters were used for all software unless otherwise specified.
## References
1. Koukou, Stergioti, La Cour et al. (2022) "Clostridium sporogenes as surrogate for proteolytic C. botulinum -Development and validation of extensive growth and growth-boundary model" *Food Microbiol*
2. Bankevich, Nurk, Antipov et al. (2012) "SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing" *J Comput Biol*
3. Hudson, Bent, Meagher et al. (2014) "Resistance determinants and mobile genetic elements of an NDM-1-encoding Klebsiella pneumoniae strain" *PLoS One*
4. Lefkowitz, Dempsey, Hendrickson et al. (2018) "Virus taxonomy: the database of the International Committee on Taxonomy of Viruses (ICTV)" *Nucleic Acids Res*
5. Bouras, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics*
6. Moraru, Varsani, Kropinski (2020) "VIRIDIC-A novel tool to calculate the intergenomic similarities of prokaryote-infecting viruses" *Viruses*
7. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
8. Kozlov, Darriba, Flouri et al. (2019) "RAxML-NG: a fast, scalable and user-friendly tool for maximum likelihood phylogenetic inference" *Bioinformatics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12773709&blobtype=pdf | # Ebola virus matrix protein VP40 triggers inflammatory responses linked to the ebolavirus virulence
Satoko Yamaoka, Zein M'hamdi, L Wang, Vail Swenson, Kri Mcnally, Shao-Chia Lu, Reema Singh, Stephanie Saundh, Brady Zell, Sonja Best, Michael Barry, Angela Rasmussen, Hideki Ebihara, Adolfo Garcia-Sastre
## Abstract
SignificanceThis study investigates how the Ebola virus (EBOV) triggers excessive inflammation, which is a critical factor in the fatal Ebola virus disease (EVD). We found that the EBOV matrix protein, VP40, leads to strong and prolonged activation of NF-κB in nonimmune cells. This suggests a mechanism for the widespread inflammation observed in EVD patients. Notably, VP40 from highly virulent EBOV induced a stronger inflammatory response compared to those from less virulent ebolaviruses. Therefore, this study proposes that VP40 functions as a virulence determinant among ebolaviruses by mediating inflammatory activation, addressing the intriguing and unresolved question of pathogenicity differences among various ebolaviruses.Uncontrolled systemic inflammatory responses are a critical pathological feature of fatal Ebola virus (EBOV) infection. While some inflammatory responses may originate from mononuclear phagocytes (MNPs), nonimmune cells vastly outnumber MNPs and may be an important source of inflammation. Here, we demonstrated that highly virulent EBOV induced a high and sustained pro-inflammatory response compared to less virulent ebolaviruses in non-MNPs through TLR4-independent NF-κB activation. We identified the EBOV matrix protein VP40 as a potent activator of NF-κB in non-MNPs, whose intrinsic inflammatory activation ability is higher than VP40 proteins from less virulent ebolaviruses. This suggests that VP40 is a virulence determinant inducing distinct degrees of pro-inflammatory responses among ebolaviruses. Mechanistically, VP40 activated the NF-κB signaling pathway, primarily via TNFR1 using a ligand-independent mechanism. These findings reveal mechanisms that may drive systemic inflammation and promote EBOV pathogenesis, suggesting potential therapeutic strategies to mitigate immune dysregulation in severe EBOV infections.
Ebola virus disease (EVD), caused by Ebola virus (EBOV), is one of the most severe viral diseases, with case fatality rates reaching as high as 90% (1). The largest EVD outbreak in West Africa from 2013 to 2016 was devastating, with more than 28,600 people infected and more than 11,300 deaths (1 , 2). Clinical studies indicate a strong correlation exists between uncontrolled pro-inflammatory activation and EVD fatality (3 -6). Uncontrolled pro-inflammatory responses associated with excessive production of various pro-inflammatory cytokines and chemokines, known as the cytokine storm, is a critical pathological feature of severe EVD (3 , 5 -9). This leads to systemic endothelial dysfunction, vascular leakage, and coagulation abnormalities, impairing the immune response against the viral infection (10 , 11 ).
Several previous studies have demonstrated that immune cells, such as macrophages, dendritic cells (DCs), and T cells, play important roles in producing pro-inflammatory mediators during EBOV infection (8 , 12 -14). In mononuclear phagocytes (MNPs), inflammatory activation is initiated by EBOV surface glycoprotein (GP 1,2 ) or by shed GP binding to toll-like receptor 4 (TLR4) (14 -20), which activates the nuclear factor kappa B (NF-κB) pathway to drive pro-inflammatory responses (21 , 22). Several studies have shown that EBOV GP binding to TLR4 triggers a temporary pro-inflammatory response in MNPs (8 , 16) which may involve negative feedback loops that regulate an appropriate inflammatory response.
TLR4 is only expressed on a small subset of host cells including MNPs (23 ). Given this, inflammatory activation by GP through TLR4 may not explain the entire breadth and strength of uncontrolled inflammation during EBOV infection. This concept is supported by the observation that TLR4 inhibitors only partially reduce serum levels of some cytokines and chemokines in EBOV-infected mice (20) and that TLR4 knockout mice are still fully susceptible to lethal EBOV infection (20). Thus, EBOV-mediated induction of an uncontrolled pro-inflammatory response must involve as yet unidentified mechanisms in addition to the TLR4 engagement by GP.
In this study, we demonstrate that the EBOV matrix protein VP40 triggers sustained activation of NF-κB signaling via a tumor necrosis factor receptor (TNFR)-dependent mechanism in nonimmune target cells lacking TLR4 expression. This finding represents a TLR4-independent mechanism for amplifying and sustaining pro-inflammatory responses in the host by EBOV. Intriguingly, NF-κB activation induced by VP40 from EBOV is significantly higher than that elicited by VP40 from less virulent ebolaviruses such as Bundibugyo virus and Reston virus. This suggests that VP40 may be a determinant of ebolavirus species-specific virulence. Our findings provide critical insights into the molecular mechanism leading to the induction of uncontrolled pro-inflammatory responses underlying EVD pathogenesis. Understanding these mechanisms is crucial for developing effective therapeutic strategies against severe EVD. Furthermore, since VP40 plays a central role in virion formation and is a prominent vaccine antigen candidate next to GP (24), identifying the regions of VP40 related to pathogenicity could help in the design of safer and effective vaccine antigens.
## Results
## EBOV Infection Induces Significantly Stronger Pro-Inflammatory
Responses Compared to Other Ebolaviruses Through a TLR4-Independent Mechanism. The genus Orthoebolavirus consists of six virus species, each represented by a single type virus: EBOV, Sudan virus (SUDV), Bundibugyo virus (BDBV), Taï Forest virus (TAFV), Reston virus (RESTV), and Bombali virus (BOMV) (2,25). While EBOV and SUDV are highly virulent ebolaviruses, BDBV and TAFV are often referred to as moderately or less virulent ebolaviruses. RESTV is a unique ebolavirus thought not to cause clinically apparent disease in humans, although it can cause fatal disease in cynomolgus macaques (26). No human cases of BOMV infection have been reported thus far, and experimental infection in type I interferon receptor knockout (IFNAR -/-) mice and HLA-A2-transgenic NOD-scid-IL-2γ receptor-knockout (NSG-A2) mice reconstituted with human hematopoiesis suggest a low pathogenic potential of BOMV in humans (27,28) (Fig. 1A). We first compared the ability of EBOV, BDBV, and RESTV to induce a pro-inflammatory response in 293 cells which do not express detectable levels of TLR4 (18, 29) (Fig. 1B) and have been suggested to originate from the adrenal glands (30), one of the main target organs for EBOV. EBOV infection induced a time-dependent increase in the release of pro-inflammatory chemokines IL-8 and MIP-1β, as well as pro-inflammatory cytokine TNF-α (Fig. 1C). Notably, the levels of IL-8, MIP-1β, and TNF-α were consistently higher in cells infected with EBOV when compared to those induced by BDBV and RESTV (Fig. 1C). No statistically significant difference in virus replication was observed among the three ebolaviruses, while RESTV replication may be somewhat delayed at early time point (Fig. 1D). These results indicate that highly virulent EBOV triggers the pro-inflammatory activation to a greater extent than the less virulent BDBV or nonpathogenic RESTV through a mechanism independent of TLR4.
## NF-κB Is Critically Involved in the Varying Degrees of Pro-
Inflammatory Activation Observed Across Ebolaviruses. We next performed RNA sequencing (RNA-seq) to investigate global transcriptomic changes in 293 cells infected with EBOV and RESTV, ebolaviruses with distinct pathogenicity in humans. A multidimensional scaling (MDS) plot revealed wide separation in gene expression profiles between EBOV and RESTV infections at 48 h postinfection (hpi), indicating substantially different global expression profiles in terms of directionality and magnitude of expression (Fig. 2A). Indeed, a smaller number of both upregulated and downregulated differentially expressed genes (DEGs) were observed in RESTV infection compared to EBOV infection at 48 hpi when both were compared to uninfected cells (Fig. 2B and Dataset S1). Gene set enrichment analysis (GSEA) utilizing the Hallmark gene sets demonstrated pathways related to inflammation, such as TNFα Signaling via NFκB, were enriched in both EBOV and RESTV infections, however a greater number of genes were enriched in response to EBOV infection (Fig. 2C and SI Appendix, Fig. S1A). Functional analysis with Ingenuity Pathway Analysis (IPA) predicted numerous inflammatory signaling pathways, such as IL-6, IL-1, TNFR1, TNFR2 Signaling, and NFκB Activation by Viruses, to be more strongly activated in EBOV than in RESTV (Fig. 2D and Dataset S2). In addition, the IPA Upstream Analysis, supported by subsequent DE analyses, indicated that NF-κB-driven inflammatory gene expression was more enriched upon EBOV infection than RESTV (Fig. 2E and SI Appendix, Fig. S1 B and C and S2 A and B).
Given the established crucial role of NF-κB in driving pro-inflammatory activation (21 , 22), we proceeded to experimentally verify NF-κB activation induced by ebolavirus infection. A NF-κB-responsive luciferase reporter assay demonstrated a time-dependent increase in NF-κB activity induced by EBOV, significantly higher than that induced by BDBV and RESTV at 72 hpi (Fig. 2F). While BDBV infection induced a modest increase in NF-κB activity over time, RESTV infection did not elicit any discernible increase in NF-κB activity. The central role of NF-κB in pro-inflammatory responses induced by EBOV was further visually demonstrated using a network assembled from transcriptome IPA results, focusing on the "NFκB Activation by Viruses" pathway (Fig. 2G). TLR4-related pathways did not meet enrichment or activation criteria (Dataset S2), verifying a lack of expression in the 293 cells used in this analysis. Together, these results suggest that NF-κB is a key regulator for driving different degrees of pro-inflammatory responses between EBOV and other ebolaviruses.
## EBOV Matrix Protein VP40 Activates NF-κB and Induces the Production of Pro-Inflammatory Mediators in Non-MNP Target
Cells. To examine which EBOV protein(s) might mediate NF-κB activation, we individually expressed seven structural EBOV proteins (NP, VP35, VP40, GP, VP30, VP24, and L) in 293 cells and assessed NF-κB activation ability of each viral protein. Expression of each EBOV protein was confirmed by Western blotting with its specific antibody (SI Appendix, Fig. S3). Among the seven viral proteins, the matrix protein VP40 was the only EBOV protein that activated NF-κB (Fig. 3A and SI Appendix, Fig. S4). VP40-mediated NF-κB activation was also verified in two other human cell lines, Huh7 and HepG2 hepatocytes, which also represent main target cell type for EBOV infection and lack detectable TLR4 expression (31,32) (Fig. 3B and SI Appendix, Fig. S5). Notably, while the magnitude of NF-κB activation was similar following expression of VP40 and TRAF6 used as a key signal transducer in the NF-κB pathway (33), the NF-κB activation induced by VP40 remained elevated for at least 72 h posttransfection (hpt), whereas TRAF6 induced transient NF-κB activation that peaked by 24 hpt and subsequently declined (Fig. 3C).
Importantly, VP40-expressing 293 cells demonstrated the release of pro-inflammatory mediators in a pattern generally resembled that observed in EBOV-infected cells (Fig. 3D ); the expression levels of IL-8, MIP-1β, and TNF-α were all significantly higher in cells expressing VP40 compared to the control group transfected with an empty vector, while PDGF-BB, MCP-1, and IP-10 were also increased. We quantified the mRNA levels of key inflammatory mediators, such as IL-8, MIP-1β, and TNF-α, and found significant upregulation in cells expressing VP40 compared to the control group (Fig. 3E). These findings highlight the significant role of VP40 in pro-inflammatory activation through NF-κB in non-MNP target cells.
## The Distinct Inflammatory Phenotypes of Ebolaviruses Correlate
with Their Respective Abilities of VP40 for NF-κB Activation. We next examined whether VP40 proteins from different ebolaviruses activate NF-κB to different degrees. Notably, NF-κB reporter activity induced by EBOV VP40 was significantly higher than that elicited by VP40 from other ebolaviruses (Fig. 4A). While EBOV VP40 induced dose-dependent NF-κB activation, such dose dependency was not observed for VP40 from other ebolaviruses, particularly for RESTV VP40. Even at the maximum tested dose (2 μg VP40-expression plasmid), NF-κB reporter activity induced by RESTV VP40 did not reach the level achieved by EBOV VP40 (Fig. 4B). The robust ability of VP40 to activate NF-κB was conserved across several EBOV variants, including Mayinga, Kikwit, and Makona-C07 (SI Appendix, Fig. S6). These results indicate that there is an association between virulence/ inflammatory phenotypes of the ebolaviruses and their respective VP40 for NF-κB activation.
## The N-Terminal Highly Variable Region of VP40 Determines
Its Ability to Activate NF-κB. Alignment of the amino acid sequences of VP40 from all six ebolaviruses revealed a high sequence similarity across the protein (SI Appendix, Fig. S7), with the exception of a highly variable region (HVR) located at the N terminus, amino acid (aa) positions 21 to 44 (Fig. 4C). Intriguingly, EBOV VP40 chimeras possessing the HVR of BDBV or RESTV VP40 (EVP40 HVR-BDBV and EVP40 HVR-RESTV ) showed a reduced ability to activate NF-κB, to a degree similar to wild-type BDBV and RESTV VP40 (Fig. 4D). Likewise, BDBV and RESTV VP40 chimeras possessing the EBOV HVR (BVP40 HVR-EBOV and RVP40 HVR-EBOV ) activated NF-κB to a degree comparable to wild-type EBOV VP40 (Fig. 4D). The mutant EBOV VP40, VP40 ∆1-20aa , which lacks the first 20 aa residues at the N terminus, including the two classical late domain motifs, 7-PTAP-10 and 10-PPEY-13, retained its ability to activate NF-κB (SI Appendix, Fig. S8 A andB). These findings indicate the critical role of the 24 aa HVR within EBOV VP40 in activating NF-κB, while the two classical late domains are dispensable. There are two primary signaling pathways, both canonical and noncanonical, which are known to activate NF-κB (21, 22) (Fig. 5A). We initially investigated which NF-κB subunit(s) translocate to the nucleus upon EBOV VP40 expression (Fig. 5B). Overexpression of positive control TRAF6 led to the nuclear localization of all tested NF-κB subunits, including p65, RelB, cRel, p52, and p50, verifying the involvement of both canonical and noncanonical pathways for NF-κB activation (33). In contrast, in cells expressing VP40, p65, cRel, and p50 were observed to translocate to the nucleus, while RelB and p52 did not show such localization. This suggests that VP40-mediated NF-κB activation is driven by the canonical pathway, but not the noncanonical pathway. Furthermore, a significant increase in p52 expression levels was observed with TRAF6 or NIK overexpression, but not with VP40 (SI Appendix, Fig. S9A), thereby ruling out the possibility of VP40 activating the noncanonical NF-κB pathway via the NIK-IKKα axis. The p65:p50 heterodimer, with p65 acting as a key transcription factor, is recognized as the primary form of NF-κB that activates pro-inflammatory genes via the canonical pathway (34 ). We examined whether the p65 NF-κB subunit is necessary for NF-κB activation induced by VP40. Depleting p65 through siRNA transfection in 293 cells markedly reduced NF-κB activity in cells expressing VP40 (Fig. 5C). Taken together, these findings indicate that VP40 activates the p65-dependent canonical NF-κB pathway.
## IKK Activation Is Essential for NF-κB Signaling Triggered by EBOV
VP40. We next investigated the upstream signaling events in the canonical NF-κB pathway, including IKK activation followed by Iκβα degradation (Fig. 5A). The expression level of Iκβα was significantly reduced in the presence of VP40 at 24 hpt compared to the negative controls, such as empty vector or EBOV GP expression (Fig. 5D). The time course analysis further revealed that the Iκβα reduction became detectable starting from 18 hpt, becoming more pronounced by 24 hpt (SI Appendix, Fig. S9B), despite an increase in Iκβα mRNA levels over time (SI Appendix, Fig. S9C). In contrast, the reduction of Iκβα level was not notable upon expression of RESTV VP40 (SI Appendix, Fig. S9D), nor in the expression of EBOV VP40 chimeras, such as EVP40 BDBV-HVR and EVP40 RESTV-HVR (SI Appendix, Fig. S9E). The expression of EVP40 ∆1-20 reduced Iκβα levels to a similar level as wild-type EBOV VP40 (SI Appendix, Fig. S9E). Although western blotting did not clearly detect phosphorylation of Iκβα and IKKα/β in the cells expressing EBOV VP40, two supplementary experiments revealed the critical roles of Iκβα phosphorylation and IKKα/β kinase activity in EBOV VP40-mediated NF-κB activation. First, the expression of a kinase-inactive, dominant negative IKKβ mutant (K44M) significantly impaired the NF-κB activity induced by VP40 as well as TRAF6 (Fig. 5E). Second, replacing the endogenous, wild-type Iκβα by overexpression of a canonical phosphorylation site-deficient Iκβα mutant (SS32/36AA) in Iκβα CRISPR-Cas9 knockout cells resulted in a complete inhibition of VP40-mediated NF-κB activation (Fig. 5F). These findings strongly suggest that VP40-induced NF-κB activation is mediated by IKKα/β kinase and Iκβα phosphorylation in the canonical NF-κB pathway. The absence of phosphorylation on Iκβα and IKKα/β on our western blotting results may be attributed to their expression levels, which could result from their rapid degradation, or some technical factors, such as limitations in detection sensitivity. The involvement of Iκβα reduction in NF-κB activation was further validated by an observed negative correlation between Iκβα levels and NF-κB reporter activity in the VP40-expressing cells (Fig. 5G). Moreover, cytoplasmic and nuclear fractionation assay demonstrated that coexpression of Iκβα with VP40 resulted in a reduction of p65 nuclear accumulation compared to VP40 expression alone (Fig. 5H). Together, these results suggest that VP40 triggers IKK activation followed by Iκβα phosphorylation and degradation, leading to the pro-inflammatory gene expression regulated by NF-κB.
TNFR1 Contributes to the NF-κB Activation Induced by EBOV VP40. Multiple cellular receptors, such as TNFR superfamily members (TNFRSF) (35), interleukin receptors (ILRs) (36), and NOD-like receptors (NLRs) (37) initiate signal transduction cascades after ligand binding to activate the IKK-p65 axis in the canonical NF-κB signaling pathway. IPA Upstream Analysis filtered by cytokines indicated the involvement of TNF in inflammatory cascade activation during EBOV infection (Fig. 6A). In addition, TNF was predicted to be a much stronger upstream regulator in EBOV infection when compared to RESTV (Fig. 6A). The IPA analyses also indicated a pronounced role of TNFR1 in NF-κB activation after EBOV infection (SI Appendix, Fig. S10). Together, these findings indicate that the TNF signaling pathway is activated during EBOV infection.
To examine whether TNFR1 is involved in the NF-κB activation induced by EBOV VP40, we generated TNFR1 knockout cells using CRISPR-Cas9 gene editing and performed NF-κB reporter assays. Notably, VP40-induced NF-κB activity was significantly reduced in TNFR1 knockout cells compared to a negative control cell line that was edited with CRISPR-Cas9 targeting the Adeno-Associated Virus Integration Site 1 (AAVS1) (Fig. 6 B , Right and SI Appendix, Fig. S11 A andC). Despite a significant reduction, TNFR1 knockout did not completely abolish VP40-mediated NF-κB activation, as seen in TNFα-mediated NF-κB activation (Fig. 6 B , Left). Based on this result, we also knocked out another TNFRSF member, lymphotoxin β receptor (LTβR), which was reported to be expressed in 293 cells (35). This additional knockout of LTβR further reduced the NF-κB activity induced by VP40 (SI Appendix, Fig. S11 B andC). An IPA custom network with genes associated with VP40-mediated NF-κB activation reaffirmed that TNFR1, LTβR, NF-κB and their interacting proteins, including inflammatory cytokines, such as TNF, are upregulated after EBOV infection compared to RESTV (SI Appendix, Fig. S11D). These findings suggest that several TNFRSF members may additively contribute to the VP40-mediated NF-κB activation in EBOV-infected cells.
VP40-Mediated NF-κB Activation by TNFR1 Is Independent of Its Ligand TNFα. Given that TNFR1 activation is typically triggered by binding of its ligand TNFα (38), we subsequently tested the contribution of TNFα in the VP40-mediated NF-κB activation. Four clones of TNFα CRISPR-Cas9 knockout cell lines were established and the knockout effect was confirmed by flow cytometry staining and genomic DNA sequencing (SI Appendix, Figs. S12 andS13D). There was no statistically significant reduction in VP40-mediated NF-κB activation in these TNFα knockout cell lines (Fig. 6C). We next examined whether TNFα secreted from VP40-expressing cells contributes additively to the activation of NF-κB. To test this, cells were treated with TNFα neutralizing antibody (NAb) (passage 0; p0 cells), and then their cell supernatants were used to stimulate TNFR1 on the newly seeded cells (passage 1; p1 cells). Positive control TNFα induced NF-κB activation, which was markedly reduced in both p0 and p1 cells when TNFα was combined with TNFα NAb (Fig. 6D). In contrast, TNFα NAb had no negative effect on VP40-mediated NF-κB activation in p0 cells (Fig. 6E). Additionally, NF-κB activity in p1 cells, which received supernatants from VP40expressing p0 cells, was significantly lower compared to that in the VP40-expressing p0 cells (Fig. 6E). In conclusion, these findings demonstrate that, despite the significant contribution of TNFR1, VP40-mediated NF-κB activation is not triggered by autocrine and paracrine TNFα signaling through TNFR1. This strongly suggests that EBOV VP40 activates the NF-κB pathway primarily through TNFR1 but with a ligand-independent mechanism.
## Discussion
Here, we report a mechanism for the activation of a sustained pro-inflammatory response by EBOV. Systemic, sustained, and dysregulated inflammatory responses induced by EBOV infection are critical drivers of fatal disease progression in EVD (3 , 5 -9 ). Importantly, our study demonstrates that the EBOV matrix protein VP40 induces sustained pro-inflammatory responses by activating the canonical NF-κB signaling pathway, primarily via TNFR1 with a TNFα ligand-independent mechanism. Of note, the ability of EBOV VP40 to activate pro-inflammatory responses was found in several non-MNP target cells, including cells derived from human adrenal glands (30) and hepatocytes. While monocytes and DCs are critical as the initial target cells for EBOV infection (10 , 31 , 39 , 40), subsequent migration of virus to target organs (e.g. liver, lymph nodes, spleen, adrenal gland) results in robust viral replication and systemic immunopathology (31 , 32 ). Organ-associated inflammatory dysregulation has been observed in various lethal EVD models (32 , 41 , 42), suggesting the significance of non-MNP target cells in amplifying systemic inflammatory responses in the late phase of the disease. Moreover, TNFR1, a receptor involved in VP40-mediated NF-κB activation, is a ubiquitous membrane receptor, and its expression can be found in various cell types in vivo (38). This contrasts with TLR4-a receptor involved in GP-mediated NF-κB activationwhich is mainly expressed in MNPs (23). Our study suggests that the coordinated mechanism of EBOV GP-mediated activation of inflammatory responses in MNPs and VP40-mediated amplification of inflammatory responses in non-MNP cells could be a driving factor in the development of a cytokine storm during EBOV infection in vivo.
The present study also proposes that VP40 is a potential virulence factor in determining distinct degrees of pro-inflammatory responses among ebolaviruses. Several clinical and experimental studies have suggested distinct patterns of inflammatory activation induced by each ebolavirus with different pathogenic potentials in humans. For instance, the expression levels of pro-inflammatory mediators in patients infected with SUDV or BDBV are shown to be lower than those in fatalities infected by EBOV (43 , 44 ). Moreover, a significantly weaker inflammatory response was observed in BDBV-or RESTV-infected human peripheral blood mononuclear cells, compared to EBOV infection (45 , 46 ). A potential role for GP in virus-specific induction of inflammation was suggested by Olejnik et al., demonstrating that, unlike EBOV GP, RESTV GP does not trigger TLR4 signaling in primary human monocyte-derived macrophages (47). This finding suggests that GP-mediated TLR4 activation might be a critical factor in determining the different inflammatory activations induced by EBOV or RESTV in macrophages. However, the finding described herein that VP40 derived from the most virulent ebolavirus, EBOV, exhibits a greater ability to activate pro-inflammatory responses via NF-κB signaling than VP40 derived from other less virulent ebolaviruses, such as RESTV and BDBV, in nonimmune cells. This may provide a mechanism for systemic inflammation and strongly suggests that VP40 serves as a key determinant for species-specific differences in inflammatory activation and virulence among ebolaviruses in humans. Nevertheless, we believe that VP40 alone does not determine pathogenicity; rather, it likely acts together with other viral proteins such as GP, VP35, and VP24 to shape the overall virulence phenotype.
To date, only limited studies have been reported regarding ligand-independent TNFR1 activation (35 , 48 -54). Depletion of ESCRT (endosomal sorting complex required for transport) proteins results in a clustering of internalized TNFR1 and LTβR at endosomes, initiating the NF-κB signaling cascade in a ligandindependent manner (35). Importantly, EBOV VP40 interacts with multiple ESCRT proteins via its late domains or an uncharacterized mechanism and recruits ESCRT proteins from endosomes to the plasma membrane for efficient budding (55 -57 ). Thus, it is possible that the interaction between the host ESCRT machinery and VP40 affects the dynamics of TNFRSF at endosomes, thereby triggering the ligand-independent NF-κB signaling activation. The mapping analyses in our study strongly suggest that the domain spanning 21 to 44 aa within EBOV VP40 is a key candidate involved in this mechanism. Investigating this hypothesis by assessing the cellular distribution of TNFRSF in the presence of VP40, as well as conducting a comprehensive interactome analysis targeting the VP40 HVR, will be a focus of future research. Elucidating the molecular mechanism by which the HVR contributes to NF-κB activation will be key to understanding why, as observed with VP40 from EBOV and SUDV, NF-κB activity does not correlate directly with overall sequence homology.
We acknowledge the inherent limitations of using immortalized cell lines to model a complex human disease. The initial transcriptomic analyses in this study were performed in HEK293 cells which, as a transformed embryonic cell line, do not fully recapitulate the physiology of primary target cells in vivo. However, this cell line was chosen strategically for its established lack of TLR4 expression, which provided a controlled background to investigate the TLR4-independent inflammatory mechanisms that were the central focus of our study. Crucially, our key findings regarding VP40-driven, NF-κB-mediated inflammation were subsequently validated in other biologically relevant cell types, including human hepatocyte-derived Huh7 and HepG2 cells (Fig. 3B ), confirming that this mechanism is not a cell line-specific artifact. Nevertheless, future studies utilizing primary cells or organoids, or in vivo pathogenesis studies, will be essential for exploring the impact of VP40-mediated inflammation within the context of a multicellular tissue environment. In summary, we propose a molecular mechanism of sustained pro-inflammatory activation mediated by EBOV VP40 and a previously unrecognized function for VP40 as a virulence determinant among ebolaviruses. Although previous reports have suggested that EBOV VP40 may regulate host cellular responses, including inflammatory and antiviral responses (58 -60), our study may offer a potential molecular mechanism for the induction of uncontrolled inflammatory responses linked to ebolavirus pathogenesis. Further investigation into the molecular basis of the host-EBOV interaction involved in EBOV-induced inflammatory activation may advance the development of therapeutic approaches targeting the interaction interface between host and viral proteins that trigger pathogenic, sustained inflammatory responses in severe EVD.
## Materials and Methods
Cells and Transfections. 293 (CRL-1573), 293-TLR4/MD2 (BEI Resources), Huh7 (a kind gift from Dr. Yoshiharu Matsuura, Osaka University), HepG2 (ATCC HB-8065), and Vero E6 cells (ATCC CRL-1586) were maintained in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin (PS) (growth medium). Transient transfection was performed with Transit-LT1 (Mirus) unless mentioned otherwise. Details of the plasmids are provided in SI Appendix.
## pnas.org
Viruses and Biosafety Statement. EBOV (variant Mayinga), BDBV, and RESTV (variant Pennsylvania) were propagated in Vero E6 cells. Virus infectivity titers (FFUs) were determined by counting the number of infected cell foci using an indirect immunofluorescent antibody assay using a rabbit polyclonal anti-VP40 antibody as a primary antibody (61), as previously described (62). All work with infectious ebolaviruses was performed under biosafety level 4 conditions at the Rocky Mountain Laboratories Integrated Research Facility (Hamilton, MT) in accordance with standard operating protocols approved by the Rocky Mountain Laboratories Institutional Biosafety Committee (IBC).This study was approved by the Mayo Clinic IBC and the NIH RML IBC. It also underwent Dual Use Research of Concern (DURC) screening by both committees and NIH grant oversight, and no DURC-related concerns were identified.
Luciferase Reporter Assays. For NF-κB-responsive luciferase reporter assays with individual protein expression, 293 cells (6 × 10 4 cells), Huh7 (5 × 10 4 cells), and HepG2 (4 × 10 5 cells) were seeded in 24-well plate 1 d before transfection. Cells were transfected with a pNFκB-luc (0.25 μg) and a pRL-TK (0.04 μg) together with 0.5 μg of expression plasmids unless mentioned otherwise. Cells were lysed using Passive Lysis Buffer (PLB) (Promega) at the indicated time points, and luciferase activities were measured using DLR System (Promega).The results are presented as relative light units (RLU), calculated as the ratio of firefly luciferase activity to Renilla luciferase activity.
For NF-κB luciferase reporter assays with virus infection, 293 cells (1 × 10 5 cells) were seeded in 24-well plate 1 d before transfection. Cells were transfected with a pNFκB-luc (0.25 μg) and a pRL-TK (0.04 μg). Next day, cells were infected with ebolaviruses at a multiplicity of infection (MOI) of 1. After 1 h adsorption with tilting every 15 min, cells were washed once and 1 mL of DMEM supplemented with 3% FBS was added to the cells. Cells were lysed using PLB at 24, 48, and 72 hpi, and luciferase activities were measured using DLR System. Cell supernatants were harvested at 0, 24, 48, and 72 hpi, and used for virus titration.
Western Blotting. Cell lysates were prepared in either RIPA lysis buffer containing 1% NP-40, NE-PER Nuclear and Cytoplasmic Extraction reagents (Thermo Scientific), or PLB. An equivalent amount of proteins was subjected to SDS-PAGE and proteins were transferred onto PVDF membranes.The primary and secondary antibodies are shown in the SI Appendix, Table S2.
Nuclear and Cytoplasmic Fractionation Assay. 293 cells (3.5 × 10 5 cells) were seeded in 6-well plate 1 d before transfection. Cells were transfected with the indicated expression plasmids (2 μg) and were harvested with NE-PER Nuclear and Cytoplasmic Extraction reagents at 24 hpt. The extracted fractions were utilized for protein detection by western blotting, with LaminA/C and β-tubulin serving as controls for nuclear and cytoplasmic protein, respectively.
Gene Silencing by siRNAs. siRNA-mediated gene silencing was performed using Dharmacon ON-TARGETplus SMARTpool siRNA. 293 cells (1 × 10 5 cells) were seeded in 12-well plate. Immediately after cell seeding, cells were transfected using DharmaFECT 1 (Dharmacon) with a 25 nM siRNA targeting NF-κB p65 (RelA) subunit, GAPDH, or nontargeting negative control. Next day, the cells were transfected with a pNFκB-luc (0.5 μg) and a pRL-TK (0.08 μg) together with a pCAGGs-EBOV VP40 (1 μg). Forty-eight hours later, cells were harvested in PLB and cell lysates were used for DLR assay and western blotting.
Gene Knockout by CRISPR-Cas9. 293 cells (3 × 10 5 cells) were seeded in 6-well plate 1 d before transfection. Cells were transfected with a lentiCRISPR v2 encoding gRNA targeting Iκβα, TNFR1, or TNFα (0.3 μg). Cells transfected with lentiCRISPR v2-Iκβα or -TNFR1 were passaged several times under 1 μg/mL of puromycin selection. Monoclonal cell lines were established by limiting dilution. The methods used for gene knockout confirmation are described in SI Appendix.
TNFα Neutralization Assay. 293 cells (6 × 10 4 cells) were seeded in 24-well plate 1 d before transfection. For TNFα treatment as a control, cells were transfected with a pNFκB-luc (0.25 μg) and a pRL-TK (0.04 μg). Next day, recombinant human TNFα (R&D Systems) (0.05 ng) and purified anti-human TNFα antibody (BioLegend) (12.5 μg) were mixed into 0.5 mL of growth medium and the complex was incubated for 1 h at room temperature. The supernatant from plasmidtransfected cells was then removed and replaced with medium containing the TNFα-TNFα NAb complex. Cell supernatant and cell lysates were harvested at 6 h posttreatment. For VP40 transfection, cells were transfected with a pNFκB-luc (0.25 μg), a pRL-TK (0.04 μg), and pCAGGs-EBOV VP40 (0.5 μg). Immediately after transfection, TNFα NAb (0.05 ng) was added into the medium of transfected cells. Cell supernatant and cell lysates were harvested at 48 h posttreatment. Cell supernatant was stored in -80 °C until use.
For transferring supernatants, 293 cells (6 × 10 4 cells) were seeded in 24-well plate 1 d before transfection and were transfected with a pNFκB-luc (0.25 μg) and a pRL-TK (0.04 μg). Next day, the supernatant was removed and replaced with previously harvested cell supernatant. Cell lysates were harvested at 8 h posttreatment and were used for DLR assays.
qRT-PCR. 293 cells (4 × 10 4 cells) were seeded in 6-well plate 1 d before transfection. Cells were transfected with a pCAGGs-EBOV VP40 (2 μg) and were lysed using Trizol at 24 hpt for RNA extraction. RNA was purified using the Direct-zol RNA MiniPrep Kit (Zymo Research) with in-column DNAseI treatment according to the manufacturer's directions. The RNA of IL-8 (CXCL8), TNF-α, and MIP-1β (CCL4) was quantified by qRT-PCR using the iTaq Universal Probes One-Step Kit (BioRad). Details of the PCR conditions are available in SI Appendix.
RNA-Seq and Transcriptomic Analysis. 293 cells (2 × 10 5 cells) were seeded in 12-well plate 1 d before infection. Cells were infected with EBOV (variant Mayinga), RESTV (variant Pennsylvania) at MOI of 1, or mock-infected with culture medium.The infection experiment was independently repeated three times (n = 3). Cells were inactivated using Trizol at 24 and 48 hpi, and samples were sent to the University of Saskatchewan where the RNA extraction was performed with the Zymo Direct-Zol kit per the manufacturer's protocol. Detailed descriptions of the processing and analytical methods are available in SI Appendix.
Cytokine Quantification. For protein expression, 293 cells (2 × 10 5 cells) were seeded in 12-well plate 1 d before transfection. Cells were transfected with a pCAGGs-EBOV VP40 or -empty vector (1 μg) and supernatants were harvested at 24, 48, and 72 hpt.
For virus infection, 293 cells (2 × 10 5 cells) were seeded in 12-well plate 1 d before infection. Cells were infected with EBOV, BDBV, or RESTV at MOI of 1 and supernatants were harvested at 24, 48, and 72 hpi. Cytokine levels were quantified using Bioplex human cytokine/chemokine 12-plex assay (BioRad) with Luminex Magpix instrument.Analyzed cytokines were as follows: TNF-α, IL-6, IL-1RA, IL-1β, IL-17 (IL-17A), IFN-γ, IL-8 (CXCL8), MIP-1β (CCL4), PDGF-BB, MCP-1 (CCL2), IP-10 (CXCL10), RANTES (CCL5).
Statistical Analysis. All experiments were performed as at least three independent experiments. Statistical analyses were performed using either the t test or one-way ANOVA with GraphPad Prism 10 version 10.0.2.
Data, Materials, and Software Availability. Raw FASTQ files for RNA-seq have been deposited to the NCBI Sequence Read Archive and can be accessed via BioProject PRJNA1040271 (63). The step-by-step analysis, codes and parameters used for the transcriptomics data analysis are present at https://github.com/ rasmussen-lab/Ebola-VP40 (64). All other data are included in the manuscript and/or supporting information.
ACKNOWLEDGMENTS. We thank Dr. Logan Banadyga from the University of Manitoba, and Dr. Andrea Marzi and Dr. Atsushi Okumura from the Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases (NIAID), NIH for their insightful discussions and for providing the plasmids used for NF-κB-responsive luciferase reporter assays. We are also grateful to Friederike Feldmann from the Rocky Mountain Laboratories, NIAID, NIH for her invaluable assistance with BSL-4 work. We also thank Carla M. Weisend from Mayo Clinic for her exceptional support in managing the project budget and overseeing laboratory operations throughout this research. We acknowledge Dr. Jie Sun and Dr. Chaofan Li from the University of Virginia for their crucial support in analyzing the preliminary transcriptomic data of VP40-expressing cells. This work was supported in part by the National Institute of Allergy and Infectious Diseases (R01 AI134937) and the Division of Intramural Research, NIAID, NIH. This work was also partly supported by grant T32 AI132165 from the NIH awarded to V.A.S., and Grant JP22fk0108624 from the Japan Agency for Medical Research and Development (AMED) awarded to H.E. The opinions, interpretations, conclusions, and recommendations are those of the authors and are not necessarily endorsed by the NIH and National Institute of Infectious Disease, Japan Institute for Health Security.
## References
1. Clinic "59840; f Department of Immunology"
2. Yamaoka, Ebihara (2021) "Pathogenicity and virulence of ebolaviruses with species-and variantspecificity" *Virulence*
3. Jacob (2020) "Ebola virus disease" *Nat. Rev. Dis. Primers*
4. Reynard (2019) "Immune parameters and outcomes during Ebola virus disease" *JCI Insight*
5. Kerber (2018) "Kinetics of soluble mediators of the host response in Ebola virus disease" *J. Infect. Dis*
6. Baize (2002) "Inflammatory responses in Ebola virus-infected patients" *Clin. Exp. Immunol*
7. Wauquier, Becquart, Padilla et al. (2010) "Human fatal zaire ebola virus infection is associated with an aberrant innate immunity and with massive lymphocyte apoptosis" *PLoS Negl. Trop. Dis*
8. Kerber (2014) "Analysis of diagnostic findings from the European mobile laboratory in Guéckédou" *J. Infect. Dis*
9. Gupta, Mahanty, Ahmed et al. (2001) "Monocyte-derived human macrophages and peripheral blood mononuclear cells infected with Ebola virus secrete MIP-1alpha and TNF-alpha and inhibit poly-IC-induced IFN-alpha in vitro" *Virology*
10. Bixler, Goff (2015) "The role of cytokines and chemokines in filovirus infection" *Viruses*
11. Bray, Geisbert (2005) "Ebola virus: The role of macrophages and dendritic cells in the pathogenesis of Ebola hemorrhagic fever" *Int. J. Biochem. Cell Biol*
12. Feldmann (1996) "Filovirus-induced endothelial leakage triggered by infected monocytes/ macrophages" *J. Virol*
13. Ströher (2001) "Infection and activation of monocytes by Marburg and Ebola viruses" *J. Virol*
14. Gibb, Norwood, Woollen et al. (2002) "Viral replication and host gene expression in alveolar macrophages infected with Ebola virus (Zaire strain)" *Clin. Diagn. Lab. Immunol*
15. Younan (2017) "Ebola virus binding to Tim-1 on T lymphocytes induces a cytokine storm" *mBio*
16. Escudero-Perez, Volchkova, Dolnik et al. (2014) "Shed GP of Ebola virus triggers immune activation and increased vascular permeability" *PLoS Pathog*
17. Wahl-Jensen (2005) "Role of Ebola virus secreted glycoproteins and virus-like particles in activation of human macrophages" *J. Virol*
18. Martinez, Valmas, Basler (2007) "Ebola virus-like particle-induced activation of NF-kappaB and Erk signaling in human dendritic cells requires the glycoprotein mucin domain" *Virology*
19. 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*
20. Lai, Strange, Wong et al. (2017) "Ebola virus glycoprotein induces an innate immune response in vivo via TLR4" *Front. Microbiol*
21. Younan, Ramanathan, Graber et al. (2017) "The toll-like receptor 4 antagonist Eritoran protects mice from lethal filovirus challenge" *mBio*
22. Ahmed, Williams, Hannigan (2015) "Transcriptional activation of inflammatory genes: Mechanistic insight into selectivity and diversity" *Biomolecules*
23. Liu, Zhang, Joo et al. (2017) "NF-κB signaling in inflammation" *Signal Transduct. Target. Ther*
24. Vaure, Liu (2014) "A comparative review of toll-like receptor 4 expression and functionality in different animal species" *Front. Immunol*
25. Marzi (2011) "Vesicular stomatitis virus-based Ebola vaccines with improved cross-protective efficacy" *J. Infect. Dis*
26. Kuhn (2019) "ICTV virus taxonomy profile: Filoviridae" *J. Gen. Virol*
27. Geisbert, Strong, Feldmann (2015) "Considerations in the use of Nonhuman primate models of Ebola virus and Marburg virus infection" *J. Infect. Dis*
28. Bodmer (2023) "In vivo characterization of the novel ebolavirus Bombali virus suggests a low pathogenic potential for humans" *Emerg. Microbes Infect*
29. Fletcher (2023) "Pathogenicity of lloviu and bombali viruses in type I interferon receptor knockout mice" *J. Infect. Dis*
30. Medvedev, Vogel (2003) "Overexpression of CD14, TLR4, and MD-2 in HEK 293T cells does not prevent induction of in vitro endotoxin tolerance" *J. Endotoxin Res*
31. Stepanenko, Dmitrenko (2015) "HEK293 in cell biology and cancer research: Phenotype, karyotype, tumorigenicity, and stress-induced genome-phenotype evolution" *Gene*
32. Geisbert (2003) "Pathogenesis of Ebola hemorrhagic fever in cynomolgus macaques: Evidence that dendritic cells are early and sustained targets of infection" *Am. J. Pathol*
33. Jankeel (2020) "Early transcriptional changes within liver, adrenal gland, and lymphoid tissues significantly contribute to Ebola virus pathogenesis in Cynomolgus macaques" *J. Virol*
34. Walsh, Lee, Choi (2015) "Tumor necrosis factor receptor-associated factor 6 (TRAF6) regulation of development, function, and homeostasis of the immune system" *Immunol. Rev*
35. Giridharan, Srinivasan (2018) "Mechanisms of NF-κB p65 and strategies for therapeutic manipulation" *J. Inflamm. Res*
36. Mamińska (2016) "Escrt proteins restrict constitutive NF-κB signaling by trafficking cytokine receptors" *Sci. Signal*
37. Laue (2000) "Stimulation of endogenous GH and interleukin-6 receptors selectively activates different Jaks and Stats, with a Stat5 specific synergistic effect of dexamethasone" *J. Endocrinol*
38. Jiang (2020) "Activation of the pattern recognition receptor NOD1 augments colon cancer metastasis" *Protein Cell*
39. Dostert, Grusdat, Letellier et al. (2019) "The TNF family of ligands and receptors: Communication modules in the immune system and beyond" *Physiol. Rev*
40. Connolly (1999) "Pathogenesis of experimental Ebola virus infection in guinea pigs" *J. Infect. Dis*
41. Bosio (2003) "Ebola and Marburg viruses replicate in monocyte-derived dendritic cells without inducing the production of cytokines and full maturation" *J. Infect. Dis*
42. Ebihara (2011) "Host response dynamics following lethal infection of rhesus macaques with Zaire ebolavirus" *J. Infect. Dis*
43. Cilloniz (2011) "Functional genomics reveals the induction of inflammatory response and metalloproteinase gene expression during lethal Ebola virus infection" *J. Virol*
44. Hutchinson, Rollin (2007) "Cytokine and chemokine expression in humans infected with Sudan Ebola virus" *J. Infect. Dis*
45. Gupta, Macneil, Reed et al. (2012) "Serology and cytokine profiles in patients infected with the newly discovered Bundibugyo ebolavirus" *Virology*
46. Gupta, Goldsmith, Metcalfe et al. (2010) "Reduced virus replication, proinflammatory cytokine production, and delayed macrophage cell death in human PBMCs infected with the newly discovered Bundibugyo ebolavirus relative to Zaire ebolavirus" *Virology*
47. Xiang (1999) "Comparison of cellular gene expression in Ebola-Zaire and Ebola-Reston virusinfected primary human monocytes" *Nat. Genet*
48. Olejnik (2017) "Ebolaviruses associated with differential pathogenicity induce distinct host responses in human macrophages" *J. Virol*
49. Aguadé-Gorgorió (2020) "TNFR2 is required for RIP1-dependent cell death in human leukemia" *Blood Adv*
50. Karathanasis (2020) "Single-molecule imaging reveals the oligomeric state of functional TNFαinduced plasma membrane TNFR1 clusters in cells" *Sci. Signal*
51. Saveljeva, Mc Laughlin, Vandenabeele et al. (2015) "Endoplasmic reticulum stress induces ligand-independent TNFR1-mediated necroptosis in L929 cells" *Cell Death Dis*
52. Lewis, Valley, Sachs (2012) "TNFR1 signaling is associated with backbone conformational changes of receptor dimers consistent with overactivation in the R92Q TRAPS mutant" *Biochemistry*
53. Ozsoy, Sivasubramanian, Wieder et al. (2008) "Oxidative stress promotes ligand-independent and enhanced ligand-dependent tumor necrosis factor receptor signaling" *J. Biol. Chem*
54. Sheikh, Antinore, Huang et al. (1998) "Ultraviolet-irradiation-induced apoptosis is mediated via ligand independent activation of tumor necrosis factor receptor 1" *Oncogene*
55. Colombani, Andersen (2023) "Drosophila TNF/TNFRs: At the crossroad between metabolism, immunity, and tissue homeostasis" *FEBS Lett*
56. (2003) "In vivo oligomerization and raft localization of Ebola virus protein VP40 during vesicular budding" *Proc. Natl. Acad. Sci. U.S.A*
57. Martin-Serrano, Zang, Bieniasz (2001) "HIV-1 and Ebola virus encode small peptide motifs that recruit Tsg101 to sites of particle assembly to facilitate egress" *Nat. Med*
58. Silvestri (2007) "Involvement of vacuolar protein sorting pathway in Ebola virus release independent of TSG101 interaction" *J. Infect. Dis*
59. Rizk, Basler, Guatelli (2017) "Cooperation of the Ebola virus proteins VP40 and GP(1, 2) with BST2 to activate NF-κB independently of virus-like particle trapping" *J. Virol*
60. Pleet (2018) "Ebola virus VP40 modulates cell cycle and biogenesis of extracellular vesicles" *J. Infect. Dis*
61. Le, Spearman, Waggoner et al. (2022) "Ebola virus protein VP40 stimulates IL-12-and IL-18dependent activation of human natural killer cells" *JCI Insight*
62. Tsuda (2015) "An improved reverse genetics system to overcome cell-type-dependent Ebola virus genome plasticity" *J. Infect. Dis*
63. Ebihara (2000) "Pathogenicity of Hantaan virus in newborn mice: Genetic reassortant study demonstrating that a single amino acid change in glycoprotein G1 is related to virulence" *J. Virol*
64. Rasmussen (2023) "Impact of VP40 on inflammatory gene expression" *NCBI BioProject*
65. Rasmussen (2024) "Ebola-VP40" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737603&blobtype=pdf | # High-Fidelity and Cost-Effective Engineering of SARS-CoV-2
Marco Olguin-Nava, Thomas Hennig, Charlene Börtlein, Patrick Bohn, Uddhav Ambi, Alexander Gabel, Lina Günter, Anne-Sophie Gribling-Burrer, Nora Schmidt, Neva Caliskan, Lars Dölken, Mathias Munschauer, Redmond Smyth
## Abstract
Efficient reverse genetics systems are essential for understanding SARS-CoV-2 pathogenesis, host-virus interactions, and potential therapeutic interventions. Here, we developed a cost-effective PCR-based reverse genetics platform that splits the SARS-CoV-2 genome into only six bacterial plasmids, enabling cloning, manipulation, and the rescue of recombinant SARS-CoV-2 (rSARS-CoV-2) with high fidelity and high viral titers after a single passage. Using this system, we generated and characterized spike protein mutants Y453F and N501Y, as well as a U76G mutation in the 5 ′ -UTR. Y453F showed reduced replication kinetics, lower cell binding, and diminished fitness, while N501Y exhibited comparable replication and fitness, highlighting the distinct effects of these spike protein mutations. The U76G mutation is located within a novel NSP9 binding site in the 5 ′ -UTR and leads to impaired RNA synthesis and reduced viral replication efficiency, suggesting an important role in transcription and replication. Our findings highlight the robustness and adaptability of this reverse genetics system, providing a versatile, cost-effective tool for studying SARS-CoV-2 mutations and their effects on replication and fitness, with potential applications in vaccine and therapeutic development.
## 1. Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of the respiratory illness COVID-19, which emerged in December 2019 and has since unleashed a pandemic that continues to impact human health [1]. SARS-CoV-2 possess a large (30 kb), capped and polyadenylated positive sense genome comprised of numerous open reading frames (ORFs) and flanked by two terminal untranslated regions (UTRs). The 5 ′ and 3 ′ UTRs are around 250 to 360 bp and highly structured [2,3]. The 5 ′ -terminal region contains two overlapping ORFs, ORF1a and ORF1b, which encode for two polyproteins, pp1a and pp1ab, that are cleaved into 16 non-structural proteins (NSPs) with multiple activities required for viral infection [2,4]. Pp1ab is translated using a minus 1 programmed ribosomal frameshift event to produce the RNA-dependent RNA polymerase (RdRp). Meanwhile, the 3 ′ -terminal region (one-third of the genome) encodes the structural proteins which are expressed from subgenomic mRNAs (sgmRNAs) [5]. Like many RNA viruses, SARS-CoV-2 exhibits rapid rates of evolution [6][7][8][9]. Mutations arising in the SARS-CoV-2 genome, especially in "variants of concern", have been shown to impact infectivity, transmissibility or evasion of the immune system compared to the original strain [7,10,11]. Consequently, despite the availability of highly effective vaccines [12][13][14], high infection rates persist worldwide.
A reverse genetics system (RGS) is a key tool required for molecular virology as it enables the generation of recombinant viruses to study various aspects of viral biology, including interactions with host cells [15][16][17]. RGS can be employed to investigate mutant virus behavior, screen antiviral drugs, develop therapeutic strategies, facilitate diagnostics [18][19][20][21][22][23][24], and generate attenuated viruses for vaccine development [15,[25][26][27]. Despite their utility for virus research, RGS have been challenging to establish for viruses with large genomes, such as SARS-CoV-2. Cloning long viral sequences into bacterial plasmids is hindered by issues of instability and toxicity [16]. Furthermore, in vitro transcription of long RNA transcripts and their efficient transfection into mammalian cells pose a significant technical challenge [16].
Bacterial Artificial Chromosome (BAC) [23,28] and Yeast Artificial Chromosome (YAC) [29,30] are high-capacity vectors capable of propagating very large viral sequences in bacteria and yeast, respectively. Transfection or electroporation of BAC and YAC has been used for the rescue of SARS-CoV-2 infectious molecular clones [18,19,23,31]. However, BAC and YAC are challenging to construct, and their subsequent genetic manipulation is more complex compared to that of high-or low-copy small bacterial plasmids. Alternatively, the SARS-CoV-2 genome can be partitioned into fragments, cloned into bacterial plasmids, and reassembled either in vitro or in cells [22,[32][33][34][35][36][37]. While this approach facilitates genetic manipulation using widely established techniques, it requires precise reassembly of the DNA fragments to form a contiguous genome. High-fidelity reassembly of SARS-CoV-2 genome has been achieved using Type IIS restriction enzymes, which cleave outside their recognition sequences for scarless and directional assembly, in a strategy known as Golden Gate-like assembly [22,[32][33][34]. Alternatively, circular polymerase extension reaction (CPER) generates overlapping SARS-CoV-2 genome fragments, which are subsequently assembled into the complete genome in a PCR-like reaction [35,36,38].
Here, we established a SARS-CoV-2 reverse genetics system based on the splitting and cloning of the viral genome into six individual fragments, which were subsequently reassembled by a PCR-like reaction (Figure 1). We divided and cloned the full SARS-CoV-2 genome into six plasmids, one of the smallest fragment numbers reported to date, enabling stable propagation and genetic manipulation using well-established cloning methodologies [39]. The PCR assembled SARS-CoV-2 genome was then transfected into mammalian cells using polyethylenimine (PEI), a low-cost and easily accessible reagent. We applied this strategy to generate two rSARS-CoV-2 spike mutant viruses, Y453F and N501Y, which were first identified in variants of concern [7]. In addition, we generated an rSARS-CoV-2 containing the non-coding U76G mutation in the 5 ′ UTR of the SARS-CoV-2 genome. The recombinant viruses were characterized, revealing distinct phenotypic effects when tested for replication, cell binding, competition assays and subgenomic mRNA synthesis. The SARS-CoV-2 genome is divided into six fragments (F1 1-5682, F2 5682-8869, F3 8864-14,487, F4 14,483-17,992, F5 17,948-24,119, F6 24,089-29,891) and cloned into bacterial plasmids. Each fragment is amplified with 20 bp overlapping ends to facilitate the assembly of the complete viral genome by PCR. Following assembly, the SARS-CoV-2 DNA is transfected into mammalian cells to rescue the recombinant virus for molecular characterization.
## 2. Materials and Methods
## 2.1. Cells
All cell lines were cultured at 37 • C in a humidified atmosphere containing 5% CO 2 in Dulbecco ′ s Modified Eagle Medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 100 U/mL penicillin-streptomycin (Gibco, Thermo Fisher Scientific, Waltham, MA, USA). Vero E6 TMPRSS2 cells (African green monkey kidney; a generous gift from S. Pöhlmann), A549 ACE2 cells (human lung carcinoma epithelial cells; a generous gift from A. Pichlmair), and HEK293 cells (human embryonic kidney; a generous gift from U. Fischer) were used in this study. HEK293 ACE2 cells were generated by retroviral transduction based on the plasmid pWPI-ACE2-puro (a generous gift from A. Pichlmair), followed by puromycin selection. All cell lines were routinely tested and confirmed to be free of mycoplasma contamination.
## 2.2. Plasmids
The full genome of SARS-CoV-2 WT was split into 6 fragments (from 6.2 to 3.2 kb) which were amplified using as a template the YAC pCC1BAC-HIS3-SARS-CoV-2 (GenBank MN996528.1) and for SARS-CoV-2 GFP an extra fragment using as a template the YAC pCC1BAC-HIS3-SARS-CoV-2-GFP (BioProject: PRJNA615319; BioSample: SAMN14450690; Sample name: GFP-2_rSARS-CoV-2), both clones were kindly provided by Prof. Dr. Volker Thiel. The amplification was performed using specific primers (Supplementary Table S1) and high-fidelity PrimeSTAR GXL DNA polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan). After the generation of 6 fragments, fragment 4 (3.5 kb) was cloned into the high copy number plasmid (PUC19) using restriction enzymes BamHI and KpnI (New England Biolabs, Ipswich, MA, USA). Meanwhile the remaining six fragments (6.2 to 3.2 kb) were cloned into pUA66 to improve stability. The following restriction enzymes were used during the cloning: BamHI and KpnI (New England Biolabs, Ipswich, MA, USA) for fragment 1 and fragment 2, BamHI HF and SacI (NEB) for fragment 3, XbaI and BamHI (New England Biolabs, Ipswich, MA, USA) for fragment 5 and XmaI (New England Biolabs, Ipswich, MA, USA) for fragment 6.
The following plasmids were then modified: for plasmid F1 the initial promoter T7 was removed and replaced by a CMV promoter. In the case of fragment F6 WT and F6 GFP a linker sequence was added to the 3 ′ UTR using restriction enzymes KpnI and Esp3I (New England Biolabs, Ipswich, MA, USA). The linker was de novo synthesized (IDT) and contains the next elements: (1) Hepatitis delta virus ribozyme (HDVr), (2) SV40 poly (A) signal and (3) a spacer sequence of 364 bp.
Mutations Y453F and N501Y were inserted in plasmid F5 by Directed Site Mutagenesis using specific primers (Supplementary Table S1) and high-fidelity PrimeSTAR GXL DNA polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan). Similar strategy was followed for the mutation T76G in plasmid F1.
An individual plasmid containing the nucleoprotein (N) of SARS-CoV-2 was generated using the clone pCC1BAC-HIS3-SARS-CoV-2 as a template. The N sequence of 1260 bp was amplified using specific primers (Supplementary Table S1) and cloned downstream a CMV promoter into the vector PUC19.
Plasmids were propagated in NEB stable competent E. coli (New England Biolabs, Ipswich, MA, USA) grown overnight at 30 • C. All plasmids were subjected to Sanger sequencing using a set of primers (Supplementary Table S1), only a single silent point mutation C28103A located in plasmid 6 WT was detected compared to the sequence of SARS-CoV-2 isolate Wuhan-Hu-1 (GenBank MN996528.1).
## 2.3. PCR Amplification of SARS-CoV-2 Individual Fragments
Each individual SARS-CoV-2 fragment was amplified from their respective plasmids using an exclusive pair of primers (Supplementary Table S1) and high-fidelity PrimeSTAR GXL DNA polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan), followed by gel isolation with NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Düren, North Rhine-Westphalia, Germany) following the manufacturer ′ s recommendations. PCR conditions consisted in 0.05 U of PrimeSTAR GXL polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan), 250 nM of each primer, 200 µM of each dNTP and 1 × PrimerSTAR GXL buffer in a total volume of 50 µL. Cycling conditions were initial denaturation for 2 min at 98 • C, followed by 35 cycles for 10 sec at 98 • C, 15 sec at 55 • C, and 10 min at 68 • C, followed by a final extension for 15 min at 68 • C. Amplicon quality was checked on 1% agarose gel post-stained in EtBr.
## 2.4. SARS-CoV-2 Genome Assembly by Polymerase Extension
SARS-CoV-2 genome was assembled using the following PCR-based protocol (Takara Bio Inc. Kusatsu, Shiga, Japan). Each fragment possesses complementary ends of 20 nucleotides overlap used for the assembly. The six fragments were mixed in equimolar amount of 0.1 pM, 2 µL of PrimeSTAR GXL DNA polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan), 200 µM of each dNTP, 1x GXL buffer into a final volume of 50 µL.
Two initial cycling conditions for SARS-CoV-2 WT and SARS-CoV-2 GFP were tested: We performed an optimization of the number of cycles using 6, 9 or 12 cycles. The condition using 12 cycles performed the best results. The same conditions were used to generate all rSARS-CoV-2 viruses, exchanging the corresponding fragment with the fragment containing the desired mutation. Assembly reactions were then used for electroporation, nucleofection or transfection without any purification step.
## 2.5. DNA Electroporation
Unpurified SARS-CoV-2 GFP DNA assembly and N plasmid were electroporated into HEK293T and HEK293T ACE2 cells. Briefly, 1 × 10 6 HEK293T or HEK293T ACE2 cells were resuspended in cold Opti-MEM (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) completed with ATP (Thermo Fisher Scientific, Waltham, MA, USA) and glutathione (Sigma-Aldrich, St. Louis, MO, USA). Then, the cell suspension was mixed with 50 µL of SARS-CoV-2 GFP DNA assembly and 200 ng of N plasmid. The mixture was transferred into a 4 mm electroporation cuvette (Bio-Rad Laboratories, Hercules, CA, USA) and placed into the Gene Pulser Xcell electroporation system (Bio-Rad Laboratories, Hercules, CA, USA). A single electrical pulse with settings of 270 V at 950 µF was performed. Next, the mixture was incubated for 5 min at RT and later transferred into a Falcon tube containing 1.5 mL Dulbecco ′ s Modified Eagle Medium DMEM (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 100 U/mL of penicillin-streptomycin (Gibco, Thermo Fisher Scientific, Waltham, MA, USA). Cells were incubated under 5% CO 2 and 37 • C conditions. After 24 h, cells were trypsinized and seeded over a monolayer of 5 × 10 5 Vero E6 TMPRSS2 cells. Cells were incubated under 5% CO 2 and 37 • C conditions and observation for GFP expression was performed every 24 h for 10 days.
## 2.6. DNA Nucleofection
Unpurified SARS-CoV-2 GFP DNA assembly and N plasmid were nucleofected into HEK293T or HEK293T ACE2 cells using program CM-130 in the 4D-Nucleofector System (Lonza, Basel, Switzerland). Then, 2 × 10 5 HEK293T or HEK293T ACE2 cells were resuspended in 20 µL of 4D-Nucleofector Solution (Lonza, Basel, Switzerland), mixed with 50 µL of SARS-CoV-2 GFP DNA assembly and 200 ng of N plasmid. The mixture was placed in Nucleocuvette Vessel (Lonza, Basel, Switzerland) and processed using program CM-130. After the end of program, cells were resuspended in 100 µL warm Dulbecco ′ s Modified Eagle Medium DMEM (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) with 10% fetal bovine serum (Sigma-Aldrich, St. Louis, MO, USA) and 100 U/mL of Penicillin-Streptomycin (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) and then were incubated under 5% CO 2 and 37 • C conditions. After 24 h, cells were trypsinized and seeded over a monolayer of 5 × 10 5 Vero E6 TMPRSS2 cells. Cells were incubated under 5% CO 2 and 37 • C conditions and observation for GFP expression was performed every 24 h for 10 days.
## 2.7. DNA Transfection Using Commercial Transfection Reagents
Reverse transfection of SARS-CoV-2 DNA assembly and N plasmid into HEK293T and HEK293T ACE2 cells was performed using different commercial reagents. Briefly, 50 µL of SARS-CoV-2 GFP DNA assembly and 200 ng of N plasmid were mixed with Lipofectamine 2000 (Thermo Fisher Scientific), Lipofectamine 3000 (Thermo Fisher Scientific), TransIT 2020 (Mirus Bio, Madison, WI, USA) and TransIT LT1 (Mirus Bio, Madison, WI, USA following the manufacturer ′ s recommendations. The mixture was added in a 6 well plate and 1 × 10 6 HEK293T or HEK293T ACE2 cells were dropped on the top of the DNA: transfection reagent mixture. Cells were incubated under 5% CO 2 and 37 • C conditions. After 24 h, cells were trypsinized and seeded over a monolayer of 5 × 10 5 Vero E6 TMPRSS2 cells. Cells were incubated under 5% CO 2 and 37 • C conditions and observation for GFP expression was performed every 24 h for 10 days.
## 2.8. Virus Rescue by DNA Transfection Using PEI
Virus rescue was accomplished by reverse transfection of SARS-CoV-2 DNA assembly and N plasmid into HEK293T ACE2 cells. As an initial test different conditions were tested. Briefly, 25 µL of SARS-CoV-2 GFP DNA assembly without purification and 200 ng of N plasmid were mixed with 7.2 or 14.4 µL of polyethylenimine (PEI) (Polysciences, Warrington, PA, USA) or 50 µL of SARS-CoV-2 GFP DNA without purification and 200 ng of N plasmid were mixed with 14.4 or 28.8 µL of PEI. Then, the mixture was placed in a 6 well plate and after 10 min of incubation at RT, 1 × 10 6 HEK293T or HEK293T ACE2 cells were dropped on the top of the DNA: PEI mixture. Cells were incubated under 5% CO 2 and 37 • C conditions. After 24 h, cells were trypsinized and seeded over a monolayer of 5 × 10 5 Vero E6 TMPRSS2 cells. Cells were incubated under 5% CO 2 and 37 • C conditions and observation for GFP expression was performed every 24 h for 10 days.
For subsequent rescues of rSARS-CoV-2 WT and mutant viruses, the only condition used was the reverse transfection of 50 µL of SARS-CoV-2 DNA assembly mixed with 200 ng of N plasmid and 28.8 µL of PEI into HEK293T ACE2 cells. Recombinant viruses were amplified once on Vero E6 TMPRSS2 to generate viral stocks.
## 2.9. Production of Stocks for rSARS-CoV-2
First, 8 × 10 6 Vero E6 TMPRSS2 cells were infected using rSARS-CoV-2 at a MOI 0.1 for 1 h at 37 • C. After, the inoculum was removed and 15 mL of DMEM supplemented with 5% FCS, 100 U/mL of penicillin-streptomycin was added. Then, 48 hpi supernatant was removed and centrifugated at 3000 × g for 10 min. Without disturbing the cell pellet, supernatant was taken and 500 µL aliquots were generated. Each aliquot was immediately storage at -70 • C.
## 2.10. Viral Infections
In general, rSARS-CoV-2 inoculums were prepared in DMEM supplemented with 1% FCS. Before the infection, Vero E6 TMPRSS2 or A549 ACE2 cells were washed once with PBS and incubated with the respective inoculum for 1 h at 37 • C with gentle shaking every 10 min. The inoculum was removed and fresh DMEM supplemented with 5% FCS, 100 U/mL of penicillin-streptomycin was added to the cells.
## 2.11. Replication Kinetics for rSARS-CoV-2
Replication kinetics for the different rSARS-CoV-2 were performed using Vero E6 TMPRSS2 or A549 ACE2 cells with a MOI 0.01 PFU/cell. Samples collected at 8, 24, 28 and 72 hpi were titrated in duplicate by plaque assay. Statistical analysis comparison was performed using two-sided unpaired Student ′ s t-test with Prism 7 (GraphPad Software Inc, San Diego, CA, USA).
## 2.12. Plaque Assay
Cell supernatant containing virus was 10-fold serial diluted in DMEM 1% FCS, inoculated onto TMPRSS2-Vero E6 cell monolayer in duplicate, incubated at 37 • C for 1 h. After the incubation, the inoculum was removed and the cell monolayer was overlayed with 0.6% (w/v) methylcellulose (Carl Roth GmbH, Karlsruhe, Germany) in MEM (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 25 mM of HEPES, 0.44% NaHCO 3 , 2 mM of GlutaMAX (Gibco, Thermo Fisher Scientific, Waltham, MA, USA), 100 U/mL of penicillin-streptomycin and 5% FCS and incubated at 37 • C. After 3 days, cells were fixed and stained with 2x staining solution (0.23% crystal violet, 8% formaldehyde, 10% ethanol) directly to the medium for 24 h. Cells were washed twice with H 2 O and plaques enumerated to determine viral titers.
## 2.13. RNA Extraction, cDNA Synthesis and qPCR
Total RNA was extracted with Trizol (Invitrogen, Carlsbad, CA, USA) using the manufacturer ′ s recommendations and the totality of the viral RNA was treated with Turbo DNase (Thermo Fisher Scientific, Waltham, MA, USA) for 30 min at 37 • C. Following DNase treatment, RNA was column purified using NTC buffer and the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel, Düren, North Rhine-Westphalia, Germany), according to the manufacturer ′ s instructions.
The RNA was then reverse transcribed using SSIV (Thermo Fisher Scientific, Waltham, MA, USA) with a set of random hexamers (Integrated DNA Technologies, Coralville, IA, USA). Quantitative real-time PCR (qRT-PCR) was performed using PowerUP SYBR green (Thermo Fisher Scientific, Waltham, MA, USA) according to manufacturer ′ s instructions. A standard curve for N and RdRp was made using serial dilutions of plasmids F6 and F4, respectively. The oligonucleotides used to amplify N, RdRp, TMPRSS2, ACE2 and 18S rRNA are described in Supplementary Table S1. The level of each RNA was determined by CFX96 Touch Real-Time PCR Detection System (Bio-Rad) with the cycling condition: 50 • C for 2 min, 95 • C for 2 min, followed by 40 cycles of 95 • C for 15 s and 60 • C for 30 s, finishing with melt profile analysis. The software used for data statistical analysis is Prism 7 (GraphPad Software Inc, San Diego, CA, USA).
## 2.14. cDNA Synthesis and Sanger Sequencing for Spike Mutants and 5 ′ UTR Mutant
After RNA extraction and DNase treatment, the RNA was reverse transcribed for 4 h using SSIV (Thermo Fisher Scientific, Waltham, MA, USA) with a set of reverse SARS-CoV-2 primers (Supplementary Table S1). An amplicon of 2.4 kb was produced using the previous cDNA and a specific pair of primers (Supplementary Table S1) for the T76G, Y453F and N501Y using the following conditions: 1 µL of diluted 1/10 RT reaction with 0.05 µL of PrimeSTAR GXL polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan), 250 nM of each primer, 200 µM of each dNTP and 1 x PrimerSTAR GXL buffer in a total volume of 25 µL. Cycling conditions were initial denaturation for 1.5 min at 98 • C, followed by 35 cycles for 10 s at 98 • C, 15 s at 55 • C, and 3 min at 68 • C, with a final extension for 5 min at 68 • C. Amplicon quality was checked on 1% agarose gel post-stained in EtBr.
PCR products were purified using NucleoSpin Gel and PCR Clean-up (Macherey-Nagel, Düren, North Rhine-Westphalia, Germany) according to manufacturer ′ s recommendations. Spike mutant PCR products for Y453F and N501Y were analyzed by Sanger sequencing using the specific primer (TTCAGCCCCTATTAAACAGCCTGCACGTGT), meanwhile PCR product for T76G was analyzed by Sanger sequencing with the specific primer (GGCAAAACGCCTTTTTCAACTTC).
## 2.15. Nanopore Sequencing and Bioinformatics Analysis
Supernatant of infected Vero E6 TMPRRS2 cells were treated for RNA extraction using Trizol (Invitrogen, Carlsbad, CA, USA) based on the manufacturer ′ s recommendations, followed by Ethanol/Sodium Acetate precipitation and resuspension in RNase-free H 2 O. Then, 10 µg of RNA were treated with Turbo DNase (Thermo Fisher Scientific, Waltham, MA, USA) for 30 min at 37 • C. Following DNase treatment, RNA was column purified using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel, Düren, North Rhine-Westphalia, Germany) with NTC buffer according to the manufacturer ′ s instructions.
Reverse transcription was performed using in-house purified MarathonRT. pET-6xHis-SUMO-MarathonRT encoding MarathonRT was a gift from Anna Pyle (Addgene plasmid # 109029; http://n2t.net/addgene:109029 (accessed on 1 May 2025); RRID: Addgene_109029). Viral RNA was reverse transcribed using a mix of reverse primers listed in Table S1. Specifically, RNA was mixed with 0.5 mM dNTPs, 5 µM of each primer in 9 µL total volume and denatured for 5 min at 65 • C. Samples were placed on ice for 2 min and reverse transcription was initiated by adding 40 U of MarathonRT in 50 mM Tris-HCl pH 8.3, 200 mM KCl, 20% glycerol (v/v), 1 mM MnCl 2 , 4 U of RNasin in a 20 µL total volume. Samples were incubated for 4 h at 42 • C. Controls lacking reverse transcriptase were carried out as above, with the omission of the MarathonRT enzyme.
The RT product was subsequently used in a set of 14 separate PCR reactions to produce 2.4 kb amplicons that covered the full genome of SARS-CoV-2. PCR amplification conditions were 1 µL of diluted 1/10 RT reaction with 0.05 U of PrimeSTAR GXL polymerase (Takara Bio Inc. Kusatsu, Shiga, Japan), 250 nM of each primer, 200 µM of each dNTP and x1 PrimerSTAR GXL buffer in a total volume of 25 µL, using the cycling conditions in Section 2.14. Amplicon quality was checked on 1% agarose gel post-stained in EtBr.
PCR products were pooled and purified using Mag-Bind Totalpure NGS beads (Omega Bio-tek, Norcross, GA, USA) by addition of 0.6 × volumes of beads followed by light agitation for 5 min at room temperature. Beads were pelleted on a magnetic rack (Invitrogen, Carlsbad, CA, USA), followed by removal of supernatant and 2 washes with 100 µL freshly prepared 70% ethanol. Finally, beads were air dried for 3-5 min (until appearance changed from glossy to rough) and DNA was eluted by addition of 30 µL H 2 O, followed by 5 min incubation at room temperature.
DNA concentration was quantified via Nanodrop, and 300 ng of pooled product was taken into the Nanopore native ligation barcoding library preparation. Specifically, the DNA in 11.5 µL in a PCR tube was end-repaired by addition of 1.75 µL NEB Ultra II End Repair Buffer (New England Biolabs, Ipswich, MA, USA) and 0.75 µL Enzyme Mix followed by thorough mixing via pipetting, incubated for 5 min at room temperature and 5 min at 65 • C. Next, 1 µL of end-repaired DNA was transferred into a new PCR tube, followed by addition of 0.75 µL H 2 O, 1.25 µL native ligation barcode (ONT SQK-NBD114-96) and 3 µL Blunt/TA Ligase Master Mix (New England Biolabs, Ipswich, MA, USA). The reaction was mixed by pipetting, incubated for 20 min at room temperature, and terminated by addition of 1 µL EDTA (SQK-NBD114-96). The barcoded DNA samples were then pooled and purified by addition of 0.4 volumes of SPRI beads (SQK-NBD114-96). For binding, they were incubated for 5 min with mixing, then pelleted on a magnetic rack. Next, beads were washed twice by resuspension and re-pelleting in 200 µL Short Fragment Buffer (SFB, ONT SQK-NBD114-96) before a final wash step with 100 µL 80% ethanol. After a brief drying on air, the DNA was then eluted by addition of 35 µL and incubation for 10 min at 37 • C with mixing. Finally, 30 µL of the barcoded DNA was ligated onto 2.5 µL Native Ligation Adapter (NA, ONT SQK-NBD114-96) by addition of 12.5 µL NEBNext Quick Ligation Reaction Buffer (New England Biolabs, Ipswich, MA, USA) and 5 µL high conc. T4 DNA Ligase (New England Biolabs, Ipswich, MA, USA). After incubation for 20 min at room temperature, 20 µL of Ampure XP Beads (SQK-NBD114-96) were added, followed by incubation for 10 min at room temperature with mixing, pelleting of beads in a magnetic rack, and two washes with 125 µL SFB (SQK-NBD114-96). After removal of wash buffer, the beads were resuspended in 7 µL Elution Buffer (SQK-NBD114-96), incubated for 10 min at 37 • C, followed by pelleting and transfer of the supernatant into a 1.5 mL DNA LoBind tube. The prepared library was quantified with AccuClear Ultra High Sensitivity dsDNA kit (Biotium, Fremont, CA, USA) and 14 ng (approximately 10 fmol) were loaded onto a Kit 14 Flongle flow cell (FLO-FLG114). Sequencing data were acquired with MinKNOW version 22.12.7 (4 kHz sampling) for the rescue verification and 23.07.8 (5 kHz sampling) for the competition assay, respectively.
All data were subsequently basecalled and demultiplexed with dorado v0.5.0 (minqscore 11) and aligned to the respective SARS-CoV-2 wildtype (NC045512.2) or SARS-CoV-2 GFP (BioProject: PRJNA615319; BioSample: SAMN14450690) reference sequence with LAST v1450. The generated maf files were then converted into sorted and indexed bam files using samtools v1.16.1, and mutation frequencies for each position with a coverage of at least 100 were quantified with perbase v0.8.5. The generated tables were then parsed into pandas (v1.4.3) dataframes. Figures were generated with the python package plotly v5.9.0.
## 2.16. Western Blot
For Western blot analysis, 5 × 10 5 cells were rinsed with PBS and lysed in 100 µL Laemmli buffer (2% SDS, 10% Glycerol, 60 mM Tris, 0.01% (w/v) bromphenol blue, 50 mM DTT) and lysates were sheared by passing through a syringe. Proteins were separated by SDS PAGE in NuPAGE 4-12% Bis-Tris Protein Gels (Thermo Fisher Scientific, Waltham, MA, USA) and transferred to a nitrocellulose membrane using the iBlot dry blotting system (Thermo Fisher Scientific, Waltham, MA, USA). Membranes were incubated with anti-N (cat. GTX135357, Cell Signaling Technology, Danvers, MA, USA), anti-ACE2 (cat. 66699, Proteintech Group, Rosemont, IL, USA), anti-TMPRSS2 (cat. PA5-14264, Thermo Fisher Scientific, Waltham, MA, USA), or anti-Actin (cat. sc-47778, Santa Cruz Biotechnology, Dallas, TX, USA) antibodies. We used the following secondary antibodies: IRDye 800CW goat anti-rabbit IgG (LI-COR Biosciences, Lincoln, NE, USA), IRDye 680RD goat antirabbit IgG (LI-COR), IRDye 800CW donkey anti-goat IgG (LI-COR Biosciences) and IRDye 800CW goat anti-mouse IgG (LI-COR Biosciences) using the iBind Automated Western System (Thermo Fisher Scientific, Waltham, MA, USA). The membranes were imaged in the Odyssey Clx Infrared Imager System (LI-COR Biosciences). Statistical analysis comparison between the blots were performed using one-way ANOVA with Prism 7 (GraphPad Software Inc, San Diego, CA, USA).
## 2.17. Binding Assay
First, 5 × 10 5 Vero E6 TMPRSS2 and A549 ACE2 cells were washed twice with cold PBS and then incubated with rSARS-CoV-2 WT, rSARS-CoV-2 Y453F or rSARS-CoV-2 N501Y using at MOI 0.1 or 0.01 PFU/cell for1 h at 4 • C with gentle shaking every 10 min. Then, the inoculum was removed, and cells were washed twice with cold PBS and cells were lysed with TRIZOL (Invitrogen, Carlsbad, CA, USA). RNA extraction, RT and qPCR for 18S rRNA and N were performed as described before. To calculate differences in RNA expression we used the DDCT method versus 18S rRNA. Statistical analysis comparison between the bindings of each RDP spike mutant against rSARS-CoV-2 WT was performed using one-way ANOVA with Prism 7 (GraphPad Software Inc, San Diego, CA, USA).
## 2.18. Competition Assay
Competition assay was performed infecting a monolayer of 5 × 10 5 Vero E6 TMPRSS2 or A549 ACE2 cells with 5 different virus combinations at a MOI 0.1 PFU/cell: (1) rSARS-CoV-2 WT, (2) rSARS-CoV-2 Y453F, (3) rSARS-CoV-2 N501Y, (4) rSARS-CoV-2 WT plus rSARS-CoV-2 Y453F or 5) rSARS-CoV-2 WT plus rSARS-CoV-2 N501Y for 1 h at 4 • C with gently shake every 10 min. After, the inoculum was removed, fresh DMEM supplemented with 5% FCS, 100 U/mL of penicillin-streptomycin was added to the cells. Cells were incubated under 5% CO 2 and 37 • C conditions.
After 2 days, cell monolayers were lysed using TRIZOL. RNA extraction and RT using Marathon RT were performed as described before. Next, each sample was PCR amplified to produce a 2.4 kb amplicon using primers (Fw primer: ACAAATCCAATTCAGTTGTCTTC-CTATTC and Rv primer: TGTGTACAAAAACTGCCATATTGCA). Library preparation, nanopore sequencing and data analysis were carried out as described in the section for nanopore sequencing and bioinformatics analysis. For both cell lines Vero E6 TMPRSS2 and A549 ACE2, two independent experiments with technical replicates were performed.
## 2.19. gRNA and sgRNA Quantification
First, 5 × 10 5 Vero E6 TMPRSS2 cells were infected with rSARS-CoV-2 WT or rSARS-CoV-2 at a MOI 0.01 PFU/cell collecting samples at 8, 24, 48 and 72 hpi. After RNA extraction and DNase treatment described as before, RNA was reverse transcribed using SSIV (Thermo Fisher Scientific, Waltham, MA, USA) with a set of random hexamers (Integrated DNA Technologies). To specifically analyze viral sgRNAs and gRNA, we used qRT-PCR with previously designed primers [40]: a forward primer that binds within the SARS-CoV-2 leader sequence and a specific reverse primer for the ORF1a RNA, the M mRNA or the N mRNA (Supplementary Table S1). qRT-PCR was performed using PowerUP SYBR green (Thermo Fisher Scientific, Waltham, MA, USA) according to manufacturer ′ s instructions. To calculate differences in RNA expression we used the DDCT method versus 18S rRNA. Statistical analysis comparison between rSARS-CoV-2 U76G against rSARS-CoV-2 WT 8 h as a control was performed using two-sided unpaired Student ′ s t-test with Prism 7 (GraphPad Software Inc, San Diego, CA, USA).
## 2.20. Covalent RNA Immunoprecipitation Sequencing (cRIP-Seq)
Covalent RNA immunoprecipitation sequencing (cRIP-seq) was performed as previously reported [40]. Briefly, 2.4 × 10 6 A549 ACE2 cells were infected with rSARS-CoV-2 WT or rSARS-CoV-2 U76G at MOI 0.1 PFU/cell. At 48 hpi, culture media was removed, cells were rinsed and scraped in cold PBS. Following centrifugation (200× g, 8 min, 4 • C), the supernatant was completely removed and the cell pellet lysed in 2x Co-IP buffer (100 mM Tris-HCl pH 7.4, 300 mM NaCl, 2% (v/v) IGEPAL CA-630, 1% sodium deoxycholate, 0.5 mM TCEP, EDTA-free Protease Inhibitor Cocktail (Sigma-Aldrich, St. Louis, MO, USA). After incubation for 30 min at room temperature, an equal volume of water was added and lysis was completed by sonication (2 kJ with 10% amplitude, 0.7 s on/2.3 s off). Fresh lysates were immediately used for immunoprecipitation (IP) and sequencing library preparation as described in cRIP-seq method [40]. Briefly, limited RNase digestion with RNase I was used to trim unprotected RNA followed by IP with Anti-NSP9 antibody (cat. GTX135732-100, GeneTex, Irvine, CA, USA). Afterwards, SDS-PAGE was applied to separate IP and size-matched input (SMI) samples followed by transfer to a nitrocellulose membrane. The expected size range was excised and Proteinase K (New England Biolabs, Ipswich, MA, USA) was employed to release protein-bound RNA. After conversion of IP and SMI RNA into cDNA libraries and amplification with 15 PCR cycles, and sequenced with a Illumina NextSeq instrument(Illumina, San Diego, CA, USA) with a read length of 2 × 40 nucleotides.
## 2.21. cRIP-Seq Analysis
cRIP paired-end reads were adapter-and quality trimmed using cutadapt (v1.18) [41] Reads with a total length less than 18 nt were discarded. A custom java program was applied that simultaneously identified and clipped the remaining unique molecular identifier (UMI) associated with each read pair. The trimmed reads were aligned to the genomes of Chlorocebus sabaeus (Ensembl release 111) genome and SARS-CoV-2 (NC_045512.2, GenBank: MN908947.3) using STAR (v2.7.10a) [42] with the parameters-outFilterScoreMinOverLread 0-outFilterMatchNminOverLread 0-outFilterMatchNmin 0-alignSoftClipAtReferenceEnds No-alignSJoverhangMin 8-alignSJDBoverhangMin 1-outFilterMismatchNoverLmax 0.04-scoreDelOpen-1-alignIntronMin 20-alignIntronMax 3000-alignMatesGapMax 3000-alignEndsType EndToEnd. PCR duplicates were removed with the UMI-aware deduplication functionality of Picard ′ s MarkDuplicates. The resulting aligned reads were separated based on their strand orientation. Regions with enriched protein binding were identified for each strand with MACS2 [43] using the parameters -g 29903 -s 31 --keep-dup all --nomodel --dmin 25 --summits --scale-to small --shift 25 --nolambda --extsize 5 --max-gap 20 --min-length 5. The identified MACS2 peaks were additionally filtered by calculating the enrichment of strand specific reads within each peak over all remaining strand specific mapped reads between IP and SMI. A statistically significant enrichment relative to SMI control was calculated by a one-sided Fisher ′ s exact test. The resulting p values were corrected with the Benjamini-Yekutieli [44] procedure and only peaks with an adjusted p value < 0.05 were considered for further downstream analysis.
Crosslinking sites were defined as the first nucleotide in 5 ′ -direction at the 5 ′ -end of a R2 read that overlaps a significantly enriched peak. Coverage of each crosslinking site in IP and SMI was calculated as the number of R2 reads sharing the same 5 ′ -end. Only crosslinking sites in wildtype samples with at least 20 reads of coverage in IP were included in further analyses. To calculate the enrichment of IP over SMI or wildtype over mutant, Fisher ′ s exact test was applied in the same manner as described for identifying enriched peaks in IP over SMI. The resulting p values were corrected using the Benjamini-Yekutieli procedure. Adjusted p values below a threshold of 0.05 were considered statistically significant.
## 2.22. Microscopy
Because SARS-CoV-2 is considered as a BioSafety Level 3 (BSL3) pathogen, all the images of live cells provided in this study were performed inside the BSL3 laboratory using a Revolve R4 microscope (ECHO, San Diego, CA, USA).
## 2.23. Figure Design and Generation
All figures presented in this work were designed and assembled using web-based platform Biorender (BioRender, Toronto, ON, Canada).
## 3. Results
## 3.1. Dividing and Rebuilding a Long Viral Genome
The 30 kb genome of SARS-CoV-2 is among the largest of any known RNA viruses, presenting significant technical challenges for its handling and genetic manipulation [16,[45][46][47]. Therefore, as a first step, we decided to divide the genome into smaller regions that could be handled and propagated in plasmids. Bacterial plasmids ensure the safe storage of the SARS-CoV-2 genome E. coli, without the need for specialized protocols, and can be manipulated using well-established mutagenesis methodologies [39]. This approach also facilitates the assembly of the full genome SARS-CoV-2 and shuffling of mutations within different fragments, potentially speeding the generation of mutant recombinant viruses.
Six fragments ranging from 6.2 to 3.2 kb were amplified using as a template the SARS-CoV-2 isolate Wuhan-Hu-1 (GenBank MN996528.1 from pCC1BAC-HIS3-SARS-CoV-2). An extra fragment containing a Turbo GFP reporter sequence instead of the ORF7, was amplified using the YAC pCC1BAC-HIS3-SARS-CoV-2 GFP (BioProject: PRJNA615319; BioSample: SAMN14450690; Sample name: GFP-2_rSARS-CoV-2). One fragment (3.5 kb) was inserted into a high copy number plasmid (pUC19) meanwhile the reminding six (6.2 to 3.2 kb) were cloned into a low copy number plasmid (pUA66). The rationale for employing a low-copy-number vector arose from the observed instability of viral sequences in the high-copy-number vector, characterized by recurrent recombination events, bacterial sequence insertions, and loss of the viral sequence from the plasmid. A cytomegalovirus (CMV) promoter (598 bp) was added to the upstream region of the 5 ′ UTR in fragment F1, replacing the previous T7 promoter. This new promoter enables transcription of the viral genome by cellular RNA polymerase II in the nucleus. A linker sequence was also incorporated after the 3 ′ UTR in plasmids F6 WT and F6 GFP. The linker comprises three distinct elements: (1) a hepatitis delta virus ribozyme (HDVr) to produce a specific 3 ′ end in the viral RNA, (2) a simian virus 40 PolyA signal (SV40 polyA) to enhance transcription termination and promote RNA stability, and (3) a 364 bp spacer sequence to create an intermediate area between the extremes of the viral genome during the circularization assembly [38] (Supplementary Figure S1A). After successfully cloning all fragments, we sought to verify the authenticity of the SARS-CoV-2 sequence. Apart from a silent point mutation at position C28103A located in Fragment 6 WT, our sequence was identical to the original genome of SARS-CoV-2 isolate Wuhan-Hu-1.
We chose a PCR-based strategy for the reassembly of the full SARS-CoV-2 genome, which has been previously described for different RNA viruses [24,35,36,38,48]. This method relies on generating SARS-CoV-2 genome fragments with a 20 bp overlap between adjacent fragments, enabling their assembly during PCR. After amplifying and purifying the individual fragments (Figure 2A), each of them was mixed in an equimolar amount of 0.1 pMol. During temperature cycling, in the presence of a high-fidelity DNA polymerase and dNTPs, complementary fragments anneal based on the 20 bp overlap, facilitating the formation of a complete circular genome. To optimize the assembly, we tested two cycling conditions: (1) 2B). Considering that the primary discrepancy between the two assembly protocols lay in the number of cycles employed (12 and 20 cycles), and recognizing that the optimal condition was achieved with fewer cycles (12 cycles), we proceeded to refine the assembly reaction by evaluating 6 and 9 cycles. However, these conditions yielded incomplete assemblies, evidenced by the presence of multiple bands indicating intermediate products and unused initial fragments (Supplementary Figure S1B,C).
Taking into account all the aforementioned factors, we opted for a PCR-based assembly protocol with an initial denaturation at 98 • C for 30 s, 12 cycles of 10 s at 98 • C, 20 s at 55 • C and 10 min at 68 • C, and a final elongation for 12 min at 68 • C, as the primary strategy for assembling complete SARS-CoV-2 genomes.
## 3.2. Rescue of Recombinant SARS-CoV-2
Unpurified SARS-CoV-2 GFP DNA assembly alongside an expression plasmid coding for N protein was transfected into HEK293T cells overexpressing angiotensin-converting enzyme 2 (HEK293T ACE2) [18,32,34]. HEK293T cells are well-suited for the uptake of external genetic material, and ACE2 expression enhances their susceptibility to SARS-CoV-2 infection [49,50]. To rescue rSARS-CoV-2 GFP, we tested a range of transfection methods, including electroporation, nucleofection, and lipid-based reagents such as Lipofectamine 2000, Lipofectamine 3000, TransIT 2020, TransIT LT1, and polyethylenimine (PEI). Then, 24 h after transfection, HEK293 ACE2 cells were detached with trypsin and seeded onto a monolayer of Vero E6 cells overexpressing transmembrane serine protease 2 (Vero E6 TMPRSS2), which are highly susceptible and permissive to infection, serving as a strategy to enhance the isolation of rSARS-CoV-2 [50][51][52]. We examined GFP expression, as a marker of rescue, every 24 h for 10 days in a BSL3 facility. Remarkably the condition that yielded infectious rSARS-CoV-2 GFP most consistently and cost-effectively was when HEK293T ACE2 cells were transfected using a low-cost reagent, PEI (Supplementary Figure S2). Further optimizations were conducted using varying amounts of DNA combined with different DNA: PEI ratios. Ultimately, the optimal rescue condition was found to be the transfection of 50 µL of SARS-CoV-2 GFP DNA assembly mixed with 200 ng of N plasmid and 28.8 µL of PEI (Supplementary Figure S2).
Using this approach, we successfully obtained infectious rSARS-CoV-2 WT and rSARS-CoV-2 GFP, which induced cytopathic effects (CPE) or displayed GFP expression, respectively (Figure 2C). Titers of 2.9 × 10 5 PFU/mL for rSARS-CoV-2 WT and 2.5 × 10 5 PFU/mL for rSARS-CoV-2 GFP were achieved after a single passage, and further analysis revealed that plaque morphology and size were similar for both viruses (Supplementary Figure S3). Next, Vero E6 TMPRSS2 cells were infected separately with each virus at a MOI of 0.01 PFU/cell, and supernatants were collected at 8, 24, 48 and 72 h post infection (hpi). Growth kinetics of both recombinant viruses were not statistically different demonstrating that the rescue of SARS-CoV-2 was successful and that the substitution of ORF7 for reporter GFP did not affect the replication of SARS-CoV-2 in Vero E6 TMPRSS2 cells during the period evaluated (Figure 2D).
In conclusion, we rescued rSARS-CoV-2 by transfecting SARS-CoV-2 PCR-based DNA assembly along with an N expression plasmid into a co-culture of HEK293T ACE2 and VERO E6 TMPRSS2 cells using PEI, a cost-effective transfection reagent. The successful rescue using PEI with unpurified PCR product highlights the high yield and robustness of the assembly reaction, as it does not require specialized or high-cost transfection reagents to achieve infectious titers.
## 3.3. Analysis of the Full Genome of Recombinant SARS-CoV-2 Using Nanopore Sequencing
To assess the accuracy of the RGS and the stability of the genome after passage, we sequenced recombinant viruses after two passages using Oxford Nanopore Sequencing (Figure 3A). RNA from the supernatants of passage 1 and passage 2 of rSARS-CoV-2 WT and rSARS-CoV-2 GFP grown in Vero E6 TMPRSS2 cells were isolated, reverse transcribed, and amplified into overlapping DNA amplicons. We generated fourteen 2.4 kb amplicons using two panels of primers based on the Artic sequencing protocol for SARS-CoV-2 (Figure 3B). Each of these fourteen amplicons was purified, pooled, multiplexed, barcoded, and sequenced using an Oxford Nanopore MinION device.
The sequences of all recombinant viruses were compared to the reference sequences. Mutations with a frequency exceeding 50% were examined, revealing that the initial passage of the four independently rescued rSARS-CoV-2 WT had a nucleotide accuracy of 99.9933%. The second passage of rSARS-CoV-2 WT maintained a nucleotide accuracy of 99.9866%. Similarly, both the first and second passages of rSARS-CoV-2 GFP showed high-fidelity rescue, with a nucleotide accuracy of 99.9967% (Table 1). At the amino acid level, we identified four non-synonymous mutations (P23L and A1527V in NSP3, I382V in NSP4 and T71M in N) in one of the four independently rescued rSARS-CoV-2 WT during the first passage, resulting in an amino acid fidelity of 99.959%. The remaining three rescued rSARS-CoV-2 WT did not exhibit any non-synonymous mutations in the first passage. In the second passage, three additional mutations arose (T1180S and A1305G in NSP3 and A701T in S), although the amino acid accuracy remained high at 99.969%. On the other hand, only synonymous mutations were detected in rSARS-CoV-2 GFP in both passages (Figure 3C and Supplementary Table S2). Collectively, our RGS for SARS-CoV-2 demonstrated a rescue accuracy exceeding 99.9% for both nucleotide and amino acid sequences. Moreover, the system enables the rescue of stable rSARS-CoV-2, as evidenced by the minimal mutations observed between passages 1 and 2.
## 3.4. Generation of SARS-CoV-2 Spike Mutants
The transmembrane spike (S) protein is responsible for SARS-CoV-2 entry into the host cell by binding to the ACE2 receptor, initiating the infection process. The S protein region exhibits the highest number of non-synonymous mutations in the entire SARS-CoV-2 genome [7,53,54]. Spike mutations are highly relevant since they are responsible for influencing the virus host range, tissue tropism, antibody escape and pathogenesis. Although multiple studies have evaluated the impact of some of these mutations, most experiments have been conducted using only purified proteins or pseudoviruses expressing the S protein. As a result, these studies may not fully account for the presence of other factors or cell-virus interactions that could influence the natural course of infection [55][56][57][58][59][60][61][62].
Considering all the above, we generated two recombinant viruses containing either the N501Y spike mutation (present in the Alpha, Beta, Gamma and Omicron variants) and Y453F spike mutation (found in the Cluster 5 variant). These mutations, located directly at the interface between the receptor-binding domain (RBD) of the S protein and ACE2, have been linked to increased ACE2-binding affinity [7,58]. First, plasmid F5 was modified by site-directed mutagenesis [39] to introduce the A22920T and A23053T nucleotide exchanges, corresponding to the Y453F and N501Y mutations (Figure 4A). The genome DNA assembly for both mutants was performed under the conditions previously described (Figure 4B,C). We obtained preparations for each of the recombinant viruses with titers of 3.4 x10 5 PFU/mL for rSARS-CoV-2 Y453F and 5.0 × 10 5 PFU/mL for rSARS-CoV-2 N501Y, and confirmed the successful introduction of each mutation in the four recombinant viruses by Sanger sequencing (Figure 4D). Similarly, to the previously rescued rSARS-CoV-2 WT, CPE was observed (Figure 4E) and plaque morphology and size were comparable across all viruses (Supplementary Figure S3).
## 3.5. Biological Impact of rSARS-CoV-2 Spike Mutants
ACE2 has been identified as the main entry point for several coronaviruses, including SARS-CoV, MERS-CoV, and SARS-CoV-2 [1,52,63,64]. Given the association between the RBD spike mutations Y453F and N501Y with enhanced ACE2 binding affinity, improved viral fitness, higher transmission rates, and increased viral load [7,9,57,58,65], we investigated their impact on ACE2 binding affinity, replication kinetics, and competition assays. For this analysis, two different cell lines were used: Vero E6 TMPRSS2 and A549 ACE2. Vero E6 cells are epithelial kidney cells that are susceptible and permissive for SARS-CoV-2 infection due to the presence of ACE2 receptor on their membrane [52,66]. In contrast, A549 cells are epithelial lung cells that are susceptible to SARS-CoV-2 infection but poorly permissive due to the absence of ACE2 receptor [50,52]. However, overexpression of ACE2 in A549 cells renders them both susceptible and permissive to SARS-CoV-2 infection [67,68]. Since we did not have precise information on the levels of ACE2 receptor in those cell lines, we first examined expression levels by RT-qPCR and Western blot. In A549 ACE2 cells, we observed ACE2 mRNA expression six orders of magnitude higher than that observed in Vero E6 TMPRSS2 cells (Supplementary Figure S4A). Despite the variation in ACE2 mRNA expression, no significant difference was found in ACE2 protein levels between Vero E6 TM-PRSS2 and A549 ACE2 cells (Supplementary Figure S4B). Notably, Vero E6 TMPRSS2 cells express a monkey ACE2 (mACE2) whereas A549 cells express a human ACE2 (hACE2). We therefore performed a protein alignment to compare these proteins and assess their identity. Bioinformatics tools revealed a 94.66% identity between the two proteins. Importantly, all amino acids involved in interactions with spike residues 501 and 453 were completely identical between the two proteins (Supplementary Figure S4C,D) [50,55,[69][70][71].
After confirming similar ACE2 protein levels and a high degree of protein identity between the two cell lines, we evaluated the binding affinities of rSARS-CoV-2 spike mutants. Binding assays were carried out by incubating a monolayer of cells with virus at 4 • C, during which the biochemical interaction between S protein and ACE2 receptor can occur, but viral entry is inhibited [72,73]. Vero E6 TMPRSS2 and A549 ACE2 cell lines were separately incubated with recombinant SARS-CoV-2 viruses: rSARS-CoV-2 WT, rSARS-CoV-2 Y453F or rSARS-CoV-2 N501Y using a MOI of 0.1 PFU/cell at 4 • C for 1 h. After incubation, unbound virus particles were washed away with ice-cold PBS, while the viruses that had interacted with ACE2 remained attached. The amount of virus attached to the cell monolayer was quantified via RT-qPCR targeting RdRp (Supplementary Figure S5A). Binding efficiency of each spike mutant was compared to rSARS-CoV-2 WT. Interestingly, rSARS-CoV-2 Y453F exhibited reduced binding efficiency relative to rSARS-CoV-2 WT in both cell lines (Figure 5A,B). Meanwhile, rSARS-CoV-2 N501Y showed no significant difference in binding efficiency compared to rSARS-CoV-2 WT in either cell line (Figure 5A,B). Additionally, no significant differences in binding were observed between the two cell lines for any of the recombinant viruses tested (Supplementary Figure S5B-D).
Next, we evaluated the replication kinetics of spike mutant viruses in Vero E6 TM-PRSS2 and A549 ACE2 cells. Both cell lines were separately infected with rSARS-CoV-2 WT, rSARS-CoV-2 Y453F or rSARS-CoV-2 N501Y at a MOI of 0.01 PFU/mL and samples were collected at 8, 24, 48 and 72 hpi for titration by plaque assay (Figure 5C,D). We consistently observed lower viral titers at each time point in A549 ACE2 compared to Vero E6 TMPRSS2 cells, indicating that A549 ACE2 cells have lower permissibility to SARS-CoV-2 compared to Vero E6 TMPRSS2 cells, as previously reported [50,52,74]. Additionally, rSARS-CoV-2 Y453F showed slower replication, with lower titers at 24 hpi in Vero E6 TMPRSS2 cells and at both 24 and 48 hpi in A549 ACE2 cells. However, by 72 hpi the titers of the Y453F mutant reached levels comparable to those of rSARS-CoV-2 WT. In contrast, rSARS-CoV-2 N501Y displayed similar replication kinetics to rSARS-CoV-2 WT in both cell lines (Figure 5C,D).
To further assess whether the introduced RBD spike mutants conferred a growth advantage or disadvantage, we conducted competition assays in which Vero E6 TMPRSS2 or A549 ACE2 cells were co-infected at a MOI 0.1 PFU/cell with rSARS-CoV-2 WT plus rSARS-CoV-2 Y453F or rSARS-CoV-2 N501Y. As controls, we also infected both cell lines with rSARS-CoV-2 WT, rSARS-CoV-2 Y453F and rSARS-CoV-2 N501Y separately. At 48 hpi, RNA was extracted from supernatants and subsequently prepared for Oxford Nanopore sequencing, as previously described (Supplementary Figure S6A). In cells infected with only one virus, we observed a correlation of over 90% between each recombinant virus and the corresponding nucleotide position analyzed (Supplementary Figure S6B,C). Result from the competition assays showed that rSARS-CoV-2 WT virus outperformed the rSARS-CoV-2 Y453F mutant in both cell lines. In Vero E6 TMPRSS2 cells, WT accounted for 86.65% of the distribution compared to 13.35% for Y453F (Figure 5E). Similarly, in A549 ACE2 cells, rSARS-CoV-2 WT had a presence of 67.20% versus 32.80% for rSARS-CoV-2 Y453F (Figure 5F). Meanwhile, rSARS-CoV-2 N501Y and rSARS-CoV-2 WT exhibited similar distributions during competition assays in Vero E6 TMPRSS2 cells, with rSARS-CoV-2 N501Y accounting for 51.16% and rSARS-CoV-2 WT for 48.84%, indicating no significant advantage for the mutant (Figure 5G). However, in A549 ACE2 cells, the competition assay showed that rSARS-CoV-2 WT virus replicated slightly better than rSARS-CoV-2 N501Y, with 58.82% for WT and 41.17% for N501Y, although this difference was not statistically significant (Figure 5H).
Taken together, our results revealed distinct phenotypes for the N501Y and Y453F mutants regarding ACE2 binding and viral infectivity. The rSARS-CoV-2 Y453F mutant exhibited decreased binding compared to the rSARS-CoV-2 WT virus, likely contributing to its slower replication kinetics and lower fitness in competition assays. In contrast, the rSARS-CoV-2 N501Y mutant showed comparable binding efficiency, replication kinetics and similar fitness in competition assays to the rSARS-CoV-2 WT in both cell lines.
## 3.6. A Putative Role for Anti-Leader Sequence in SARS-CoV-2 Replication
During the replication cycle of SARS-CoV-2, a set of subgenomic RNAs (sgRNAs) are generated to produce viral structural and non-structural proteins. sgRNAs synthesis starts at the 3 ′ end of the genome and employs a discontinuous transcription mechanism to generate negative-sense sgRNAs (-sgRNAs), which are later transcribed into positivesense subgenomic RNAs (+sgRNAs) [3,5]. Template-switching is proposed as the main mechanism underlying discontinuous transcription. This process requires a 50 bp leader transcription regulatory sequence (TRS-L) in the 5 ′ UTR and eight body transcription regulatory sequences (TRS-B) distributed across the genome upstream of each viral subgenomic mRNA (sgmRNA). Although the precise molecular mechanism remains unclear, base-pairing interactions between the TRS-L and the TRS-B upstream of each sgmRNA likely facilitate the fusion of the leader and body sequences. This fusion generates at least eight sgRNAs, which vary in length depending on the TRS-B location but share identical 5 ′ and 3 ′ ends [2, 3,5,68,75].
The viral protein NSP9 is an RNA-binding protein that is suggested to play a role in RNA priming and capping during transcription, although its primary functions remain unclear [40,68,[76][77][78][79][80][81][82]. A recent study using covalent RNA immunoprecipitation (cRIP) demonstrated that NSP9 is covalently attached to the 5 ′ ends of SARS-CoV-2 RNAs. In addition to the annotated genome ends, cRIP identified an unusual covalent linkage at position 76 in the negative-sense viral RNA [40]. Interestingly, this site is located adjacent to the TRS-L, suggesting that NSP9 may influence discontinuous transcription or template switching. To test this hypothesis, we modified plasmid F1 to induce a T76G point mutation into the plasmid containing the SARS-CoV-2 5 ′ UTR by site-directed mutagenesis [39]. rSARS-CoV-2 U76G was successfully rescued following the same conditions previously described, producing CPE after 48 hpi in Vero E6 TMPRSS2 cells (Figure 6A-C). Sanger sequencing analysis of the rSARS-CoV-2 T76G virus confirmed the presence of the intended mutation (Figure 6D). Titration of the viral stock via plaque assay indicated rescue efficiencies of 7.0 x10 5 PFU/mL for rSARS-CoV-2 U76G. Notably, rSARS-CoV-2 U76G displayed smaller plaque size compared to rSARS-CoV-2 WT (Supplementary Figure S3). We then evaluated the replication kinetics of rSARS-CoV-2 U76G virus alongside rSARS-CoV-2 WT. Using a MOI of 0.01 PFU/cell, we infected Vero E6 TMPRSS2 cells and collected samples at 8, 24, 48, and 72 hpi (Figure 6E). rSARS-CoV-2 U76G showed a striking growth deficiency at each time point evaluated compared to rSARS-CoV-2 WT, with the highest significant difference at 24 hpi. Subsequently, we asked whether the mutation U76G next to the TRS-L could impair the synthesis of gRNA or sgRNAs, potentially explaining the reduced titers observed in the replication kinetic assays. To test this, Vero E6 TMPRSS2 cells were infected at a MOI 0.1 PFU/cell with either rSARS-CoV-2 WT or rSARS-CoV-2 U76G, and samples were collected at 8, 24, 48 and 72 hpi. RNA levels of gRNA and two positive sense sgRNAs were quantified using RT-qPCR with primers targeting the leaderbody junction for ORF1a gRNA and the M and N sgRNA junctions [40]. Compared to rSARS-CoV-2 WT, rSARS-CoV-2 U76G exhibited significantly reduced levels of ORF1ab gRNA and M sgRNA at 24 and 48 hpi. Notably, N sgRNA levels were also decreased during rSARS-CoV-2 U76G infection at 24 hpi and at 72 hpi compared to rSARS-CoV-2 WT (Figure 6F).
Given the observed reductions in gRNA and sgRNA levels, along with the proposed role of NSP9 in transcription and its previously identified binding at U76, we next evaluated NSP9 binding at position 76 in the negative-sense viral RNA and other key sites previously reported [40]. Vero E6 TMPRSS2 cells were infected with either rSARS-CoV-2 WT or rSARS-CoV-2 U76G at a MOI of 0.1 PFU/cell, and samples were collected at 24 hpi for covalent RNA immunoprecipitation sequencing (cRIP-seq). Sequence reads covalently linked to NSP9 were analyzed, separating them based on their origin from the positive or negative viral RNA strand, and peaks significantly enriched relative to a size-matched input (SMI) control were identified (Figure 7A). By analyzing NSP9 binding at the 5 ′ ends of both positive sense and negative sense viral RNA, we identified a significant reduction in NSP9 binding at position 76 in the negative-sense viral RNA during rSARS-CoV-2 U76G infection compared to rSARS-CoV-2 WT. Additionally, NSP9 binding was observed at positions 74 to 79 in the negative-sense viral RNA in Vero E6 TMPRSS2 cells, sites not previously detected in A549 cells [40]. Notably, NSP9 binding at these additional sites was also reduced during rSARS-CoV-2 U76G infection relative to rSARS-CoV-2 WT (Figure 7B and Supplementary Figure S7).
Overall, our findings show that the U76G mutation in the 5 ′ UTR of SARS-CoV-2 significantly impairs NSP9 binding to position 76 in the negative-sense viral RNA, as well as at additional sites. Notably, the reduced linkage to positive stranded RNA at nucleotide 1 is consistent with reduced positive stranded RNA priming and synthesis. These reductions correlate with observed deficiencies in gRNA and sgRNA synthesis, diminished growth kinetics and smaller plaque sizes compared to the wild-type virus, suggesting that the binding of NSP9 to these regions impacts viral transcriptional and replication.
## 4. Discussion
RGSs allow the rescue and manipulation of recombinant DNA and RNA viruses, facilitating the study of their biological properties and improving our understanding of emerging viral pathogens. Here, we established a versatile and cost-effective PCR-based reverse genetic system for engineering SARS-CoV-2. We achieved this through three key steps: dividing and cloning the SARS-CoV-2 genome into six stable bacterial plasmids, employing a PCR-based assembly strategy, and conducting low-cost transfection of the assembled genome into mammalian cells. This approach enabled accurate and efficient rescue of rSARS-CoV-2.
Previously, the cloning and storage of long SARS-CoV-2 sequences in bacterial plasmids proved to be unstable [16]. Although yeast artificial chromosomes (YACs) and bacterial artificial chromosomes (BACs) allow for the cloning and manipulation of the entire SARS-CoV-2 genome and mitigate issues of plasmid instability and toxicity, these systems require significant technical expertise and do not support the simultaneous introduction of multiple mutations [17,83]. To address these challenges, we divided the SARS-CoV-2 genome into six fragments, one of the lowest fragment number reported to date, and cloned them into low-copy vectors for propagation in E. coli and [17,35,36]. This strategy not only improved storage stability but also facilitated the use of well-established cloning methodologies for precise and simultaneous genetic engineering [39].
Several strategies for reconstituting SARS-CoV-2 from fragments have been described, including in-yeast genome assembly [18], Golden Gate-like assembly [22,32,34] and polymerase extension reaction [35,36] (Supplementary Figure S8). In our hands, Golden Gatelike assembly failed to produce the desired results, primarily yielding incomplete viral genome assemblies and limiting the availability of a complete genome for mammalian cell transfection. Instead, we employed a PCR-based assembly strategy, which aligns genome fragments with 20 bp overlaps to ensure efficient and accurate reconstruction of the SARS-CoV-2 genome. This approach not only resulted in high-yield genome assembly but was also shorter and more straightforward.
In previous PCR-based assembly systems for SARS-CoV-2, introducing a desired mutation involves the generation of an extra fragment [35,36], which could lead to subsequent issues during the assembly process. In our system, mutations are introduced directly into the plasmids without increasing the number of fragments needed for the final assembly. Furthermore, viruses with multiple mutations can be easily generated by combining fragments containing the desired mutations. An alternative PCR-based assembly method has been reported, utilizing DNA fragments generated directly from reverse-transcribed RNA without cloning [36]. While this strategy circumvents plasmid toxicity or instability in bacteria, generating fragments directly via reverse transcription carries the risk of introducing unintended mutations due to the error rate of reverse transcriptase, typically around 10 -4 . In contrast, generating fragments using plasmids as templates and employing a high-fidelity DNA polymerase, with a substantially lower error rate of 10 -6 , minimizes the likelihood of unintended mutations.
Our genome assembly design included key regulatory elements, reported by others, to enable efficient viral RNA transcription directly within mammalian cells. These features include a CMV promoter to drive RNA transcription by RNA polymerase II in the nucleus, an HDV ribozyme to generate the authentic viral 3 ′ end with a polyA tail, and an SV40 polyA signal to ensure proper transcription termination [24,35,36,38,48]. This streamlined design allowed the assembled genome to be directly transfected into mammalian cells. Although this approach precludes the use of T7-mediated in vitro transcription-thereby preventing the generation of infectious RNA in cell-free reactions-it purposefully bypasses the expensive and complex in vitro RNA production and transfection steps associated with handling large, unstable RNA transcripts [18,22,32,34]. Instead, we demonstrated that transfection with PEI, a low-cost transfection reagent, enabled the rescue of high titers of rSARS-CoV-2 even after just one passage. Recently, a new RGS strategy for SARS-CoV-2, known as Infectious Subgenomic Amplicon (ISA) allowed the rescue of rSARS-CoV-2 by directly transfecting overlapping fragments, thus bypassing the need for assembly [37,84]. However, despite its apparent simplicity, this strategy requires the use of expensive transfection agents and involves a longer rescue time post-transfection, including additional viral passages (Supplementary Figure S9). Nanopore sequencing revealed high-fidelity rescue after 2 passages (~99.99% nucleotide accuracy and 99.97% amino acid accuracy), with only 1 to 4 nucleotide mutations detected across the full genome (Table 1). This fidelity is comparable to or exceeds that of established CPER-based systems, which have been reported to exhibit 3 to 5 nucleotide substitutions in similar rescue experiments. Most observed mutations were transitions (Table S2), consistent with the error profile of polymerases, and their presence at high frequencies suggest their selection during passaging in cell culture, as reported previously in other systems [35,36,85].
We used our RGS to explore the impact of different mutations on SARS-CoV-2 infection. We first focused on the RBD of the S protein due to its ongoing and rapid evolution [7,53,54] and generated two different RBD S protein mutant viruses: rSARS-CoV-2 Y453F and rSARS-CoV-2 N501Y. The mutations, Y453F and N501Y, emerged early in the pandemic and have been reported to enhance the binding of S protein to the ACE2 receptor [7,57,58]. ACE2 has been reported as the principal receptor for some coronaviruses like SARS-CoV and SARS-CoV-2 [1,52,63,64]. The Y453F mutation was first identified in humans with contact to infected minks [7] and is located in the RBD, interacting directly with ACE2 [55,56]. This mutation has been associated with increased binding of the isolated RBD to hACE2 in biophysical [55] and deep mutational screening assays [58]. However, our results showed a significant decrease in the binding of rSARS-CoV-2 Y453F to Vero E6 TMPRSS2 and A549 ACE2 cells compared to rSARS-CoV-2 WT. Moreover, it exhibited slightly slower replication kinetics and reduced fitness during competition assays against rSARS-CoV-2 WT in both cell lines. Consistent with our findings, the Y453F mutation has been reported to enhance binding of S protein primarily to the mink ACE2 receptor [86] and attenuate SARS-CoV-2 replication in human cells [56,87]. Together with the fact that Y453F mutation has disappeared from circulating human SARS-CoV-2 strains, our data support the hypothesis that this mutation represents a mink-specific adaptation that ′ spilled back ′ to humans.
The second mutation, N501Y, has been widely investigated due to its early emergence and its presence in several variants of concern, including Alpha, Beta, Gamma, and Omi-cron [7,53,54]. The N501Y mutation has been associated with enhanced fitness, higher transmission, and increased viral load [7,9,57,65]. Although the N501Y mutation reduces neutralization by certain RBD-specific antibodies [7,88,89], most studies agree that its enhanced fitness is due to an increased affinity of the RBD of the S protein for the hACE2 receptor [57][58][59][60][61][62]. In our study, we observed similar cell binding, replication kinetics, and fitness during competition assays of rSARS-CoV-2 N501Y compared to rSARS-CoV-2 WT in both cell lines evaluated. Although our results may appear contradictory to previous reports, most studies showing increased binding of N501Y S protein to hACE2 were conducted through computational modeling or in vitro experiments focusing solely on the RBD domain, rather than in the context of full S protein [59][60][61][62]. To our knowledge, only two studies employing pseudotyped virus particles have characterized the N501Y mutation [56,90]. The first study found no difference in infectivity compared to rSARS-CoV-2 WT during infection of Vero cells [90], while the second reported only a small increase in infectivity of HIV-1 N501Y in HEK ACE2 cells [56]. Additionally, an rSARS-CoV-2 N501Y virus was shown to replicate better than rSARS-CoV-2 WT in vivo (in hamsters) and in human airway epithelial (HAE) cells; however, the fitness difference in Vero E6 cells was not as considerable [57]. This suggests that in cell lines with high ACE2 surface expression, receptor binding may not be the rate-limiting step for viral entry, thereby masking the affinity advantage of N501Y that is observed. Furthermore, this study evaluated the N501Y mutation in the context of the D614G mutant, while our study assessed the N501Y mutant in the context of the ancestral Wuhan strain, which does not contain the D614G mutation that rapidly became dominant in circulating human strains [22,91,92]. Previous studies have highlighted the importance of epistasis in SARS-CoV-2 evolution [57,92]. Therefore, we speculate that the fitness advantage of N501Y may be dependent on the presence of the D614G mutation, and its introduction into the ancestral background (as performed here) yields a neutral phenotype due to the absence of these synergistic effects.
Taken together, our data highlight the complexity of exploring the relative fitness of S protein mutants, and it becomes evident that various approaches and techniques need to be combined to enable meaningful evaluation of viral fitness. Biochemical, biophysical and cell culture-based assays using only the RBD domain of S protein are of fundamental importance but may fail to capture the effects of spike mutations on ACE2 affinity in the context of the full S protein, the interaction with ACE2 receptor, or the presence of co-receptors like NRP1 [91]. Even when mutants are assessed in the context of the authentic virus, characterization studies should also consider the effects of epistasis, a powerful force affecting the evolution of protein sequences [92][93][94], as seen with D614G mutation, which is known to shift the epistatic landscape of the S protein of SARS-CoV-2 [57,70,71].
Last, we explored a protein-interaction site adjacent to the TRS-L in the 5 ′ -UTR of SARS-CoV-2. The viral protein NSP9 has been reported to be NMPylated by NSP12 and this modification has been hypothesized as a first step for priming and capping functions during RNA transcription [40,68,76,79,80,82,95]. Supporting these roles, a recent study reported the covalent binding of NSP9 to the 5 ′ ends of both positive sense RNA and negative sense RNA of SARS-CoV-2 [40]. Additionally, an interesting third binding site of NSP9 was reported: a single adenine residue located at position 76 of negative sense viral RNA [40]. Due to this nucleotide being located adjacent to the TRS-L, we wondered whether this unexpected position might have relevance in the process of discontinuous transcription or template switching. Upon rescuing the mutant rSARS-CoV-2 U76G, we observed a marked reduction in NSP9 binding to position 76 on the negative-sense RNA through cRIP-seq. This reduced binding was associated with a replication defect compared to rSARS-CoV-2 WT when assessing virus growth kinetics. We also detected lower levels of ORF1a gRNA, as well as M sgRNA and N sgRNA, indicating that the U76G mutation disrupts both gRNA and sgRNA synthesis. Although we did not observe a differential effect on the ratio of gRNA to sgRNA levels that would specifically suggest a role in discontinuous transcription, the mutation U76G and NSP9 binding at this position proved important for efficient transcriptional processes and viral replication.
Furthermore, cRIP-seq analysis identified two additional NSP9 binding sites at positions 77 and 78 on the negative-sense RNA, observed during infection in Vero E6 TMPRSS2 cells but not previously reported in A549 cells [40]. The presence of uracil at all positions (76, 77, and 78) suggests a potential "flexibility" in NSP9 ′ s priming position, which may vary depending on the cell line and warrants further study. Importantly, NSP9 binding to positions 77 and 78 was also diminished in the rSARS-CoV-2 U76G mutant compared to rSARS-CoV-2 WT. These reductions across adjacent positions suggests that the U76G mutation may alter NSP9 ′ s binding dynamics in regions potentially involved in priming and hypothesized to be critical for non-canonical discontinuous transcription [5] or anti-leader synthesis [40]. However, the existence of a short "anti-leader" containing RNA remains speculative and requires further exploration. To our knowledge, no prior studies have explored the effects of a U76G mutation, including its interaction with NSP9 or other viral or cellular proteins. While our cRIP-seq data demonstrate a clear change in covalent binding upon mutation of this region, further research is essential to define the precise structural determinants of this interaction and fully elucidate the mechanisms and implications of this mutation for SARS-CoV-2 replication, as well as NSP9 ′ s broader regulatory role in viral RNA synthesis.
In conclusion, we have developed a low-cost and efficient SARS-CoV-2 reversegenetics strategy that streamlines the generation of viral mutants and enables rapid studies on vaccine efficacy, immune escape, and complex virus-host interactions.
## Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/v17121604/s1, Supplementary Figure S1. General overview and design of PCR-based assembly for SARS-CoV-2 DNA. (A) Schematic representation of SARS-CoV-2 WT and rSARS-CoV-2 GFP. Some modifications were introduced into the SARS-CoV-2 genome. First, a CMV promoter was added to the upstream region of the 5' UTR in fragment F1. Next, a linker sequence was introduced after the 3' UTR in fragments F6 WT and F6 GFP consisting of: (1) Hepatitis delta virus ribozyme: cleavage of the phosphodiester bound for a precise 3' end of the viral RNA;
(2) SV40 Poly a terminator: ending of transcription to avoid the generation of concatemeric RNA and (3) a Spacer sequence: 350 bp to create an intermediate area between SV40 and the CMV transcription promoter during the circularization assembly. (B) DNA fragments of SARS-CoV-2. PCR products of 7 fragments of SARS-CoV-2, including the fragment for the GFP reporter, using bacterial plasmids as template. (C) Optimization of assembly for SARS-CoV-2 DNA. Three different assembly conditions (12, 9 and 6 polymerase extension cycles) were tested using an equimolar mixture of 6 DNA fragments to assemble the 30 kb SARS-CoV-2 genome.; Supplementary Figure S2. Gene delivery for rescue of rSARS-CoV-2. SARS-CoV-2 DNA assembly and N plasmid were nucleofected, electroporated or lipofected using different transfection reagents into HEK293T ACE2 cells. After 24 h, cells were trypsinized and seeded over a monolayer Vero E6 TMPRSS2 cells. After 10 days, expression of GFP was evaluated as a sign of rescue of rSARS-CoV-2 GFP.; Supplementary Figure S3. S1. List of oligonucleotides used in the present study.; Supplementary Table S2. List of mutations during rSARS-CoV-2 sequencing with more 50% presence. at accession number PRJEB105095 at the European Nucleotide Archive. Additional information and materials are available from the corresponding author upon reasonable request.
## References
1. Hu, Guo, Zhou et al. (2021) "Characteristics of SARS-CoV-2 and COVID-19" *Nat. Rev. Microbiol*
2. Lan, Allan, Malsick et al. (1128) "Secondary structural ensembles of the SARS-CoV-2 RNA genome in infected cells" *Nat. Commun*
3. Brant, Tian, Majerciak et al. (2021) "SARS-CoV-2: From its discovery to genome structure, transcription, and replication" *Cell Biosci*
4. Wang, Horby, Hayden et al. (2020) "A novel coronavirus outbreak of global health concern" *Lancet*
5. Wang, Jiang, Feng et al. (2021) "The SARS-CoV-2 subgenome landscape and its novel regulatory features" *Mol. Cell*
6. Duarte, Ketcheson, Eguíluz et al. (2022) "Alam, I. Rapid evolution of SARS-CoV-2 challenges human defenses" *Sci. Rep*
7. Harvey, Carabelli, Jackson et al. (2021) "SARS-CoV-2 variants, spike mutations and immune escape" *Nat. Rev. Microbiol*
8. Singh, Pandit, Mcarthur et al. (2021) "Evolutionary trajectory of SARS-CoV-2 and emerging variants" *Virol. J*
9. Thakur, Sasi, Pillai et al. "SARS-CoV-2 Mutations and Their Impact on Diagnostics, Therapeutics and Vaccines" *Front. Med*
10. Sonnleitner, Prelog, Sonnleitner et al. (2022) "Cumulative SARS-CoV-2 mutations and corresponding changes in immunity in an immunocompromised patient indicate viral evolution within the host" *Nat. Commun*
11. Shannon, Selisko, Le et al. (2020) "Rapid incorporation of Favipiravir by the fast and permissive viral RNA polymerase complex results in SARS-CoV-2 lethal mutagenesis" *Nat. Commun*
12. Polack, Thomas, Kitchin et al. (2020) "Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine" *N. Engl. J. Med*
13. Baden, El Sahly, Essink et al. (2021) "Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine" *N. Engl. J. Med*
14. Voysey, Costa Clemens, Madhi et al. (2021) "Single-dose administration and the influence of the timing of the booster dose on immunogenicity and efficacy of ChAdOx1 nCoV-19 (AZD1222) vaccine: A pooled analysis of four randomised trials" *Lancet*
15. Stobart, Moore (2014) "RNA virus reverse genetics and vaccine design" *Viruses*
16. Almazán, Sola, Zu et al. (2014) "Coronavirus reverse genetic systems: Infectious clones and replicons" *Virus Res*
17. Wang, Peng, Jin et al. (2022) "Reverse genetics systems for SARS-CoV-2" *J. Med. Virol*
18. Thao, Labroussaa, Ebert et al. (2020) "Rapid reconstruction of SARS-CoV-2 using a synthetic genomics platform" *Nature*
19. Luna, Thao, Le Pen et al. (2021) "Replication and single-cycle delivery of SARS-CoV-2 replicons" *Science*
20. Fahnøe, Pham, Fernandez-Antunez et al. "Versatile SARS-CoV-2 Reverse-Genetics Systems for the Study of Antiviral Resistance and Replication" *Viruses*
21. Beitzel, Hulseberg, Palacios (2019) "Reverse genetics systems as tools to overcome the genetic diversity of Lassa virus" *Curr. Opin. Virol*
22. Hou, Okuda, Edwards et al. (2020) "SARS-CoV-2 Reverse Genetics Reveals a Variable Infection Gradient in the Respiratory Tract" *Cell*
23. Ye, Chiem, Park et al. (2020) "Rescue of SARS-CoV-2 from a single bacterial artificial chromosome" *MBio*
24. Edmonds, Van Grinsven, Prow et al. (2013) "A Novel Bacterium-Free Method for Generation of Flavivirus Infectious DNA by Circular Polymerase Extension Reaction Allows Accurate Recapitulation of Viral Heterogeneity" *J. Virol*
25. Neumann, Fujii, Kino et al. (2005) "An improved reverse genetics system for influenza A virus generation and its implications for vaccine production" *Proc. Natl. Acad. Sci*
26. Nogales, Martínez-Sobrido (2017) "Reverse genetics approaches for the development of influenza vaccines" *Int. J. Mol. Sci*
27. Jung, Lee, Seong (2010) "Reverse genetic platform for inactivated and live-attenuated influenza vaccine" *Exp. Mol. Med*
28. Tischer, Kaufer (2012) "Viral bacterial artificial chromosomes: Generation, mutagenesis, and removal of mini-F sequences" *J. Biomed. Biotechnol*
29. Kouprina, Leem, Solomon et al. (2003) "Segments missing from the draft human genome sequence can be isolated by transformation-associated recombination cloning in yeast" *EMBO Rep*
30. Marschall, Malik, Larin (1999) "Transfer of YACs up to 2.3 Mb intact into human cells with polyethylenimine" *Gene Ther*
31. Rihn, Merits, Bakshi et al. (2021) "A plasmid DNA-launched SARS-CoV-2 reverse genetics system and coronavirus toolkit for COVID-19 research" *PLoS Biol*
32. Xie, Lokugamage, Zhang et al. (2021) "Engineering SARS-CoV-2 using a reverse genetic system" *Nat. Protoc*
33. Ju, Zhu, Wang et al. (2021) "A novel cell culture system modeling the SARS-CoV-2 life cycle" *PLoS Pathog*
34. Shi, Menachery, Leduc et al. (2020) "An Infectious cDNA Clone of SARS-CoV-2" *Cell Host Microbe*
35. Torii, Ono, Suzuki et al. (2021) "Establishment of a reverse genetics system for SARS-CoV-2 using circular polymerase extension reaction" *Cell Rep*
36. Amarilla, Sng, Parry et al. "A versatile reverse genetics platform for SARS-CoV-2 and other positive-strand RNA viruses" *Nat. Commun. 2021*
37. Mélade, Piorkowski, Touret et al. (2022) "A simple reverse genetics method to generate recombinant coronaviruses" *EMBO Rep*
38. Setoh, Prow, Peng et al. (2017) "De Novo Generation and Characterization of New Zika Virus Isolate Using Sequence Data from a Microcephaly Case" *mSphere*
39. Sambrook, Russell (2012) "Molecular Cloning: A Laboratory Manual"
40. Schmidt, Ganskih, Wei et al. (2023) "SND1 binds SARS-CoV-2 negative-sense RNA and promotes viral RNA synthesis through NSP9" *Cell*
41. Martin (2011) "Cutadapt removes adapter sequences from high-throughput sequencing reads" *EMBnet.J*
42. Dobin, Davis, Schlesinger et al. (2013) "STAR: Ultrafast universal RNA-seq aligner" *Bioinformatics*
43. Zhang, Liu, Meyer et al. (2008) "Model-based analysis of ChIP-Seq (MACS)" *Genome Biol*
44. February, Academy (2001) "The Control of the False Discovery Rate in Multiple Testing Under Dependency" *Ann. Stat*
45. Lindenbach (2022) "Reinventing positive-strand RNA virus reverse genetics" *Adv. Virus Res*
46. Thiel, Herold, Schelle et al. (2001) "Infectious RNA transcribed in vitro from a cDNA copy of the human coronavirus genome cloned in vaccinia virus" *J. Gen. Virol*
47. Nouda, Minami, Kanai et al. "Development of an entirely plasmid-based reverse genetics system for 12-segmented double-stranded RNA viruses"
48. Setoh, Prow, Rawle et al. (2015) "Systematic analysis of viral genes responsible for differential virulence between American and Australian West Nile virus strains" *J. Gen. Virol*
49. Bayati, Kumar, Francis et al. (2021) "SARS-CoV-2 infects cells after viral entry via clathrin-mediated endocytosis" *J. Biol. Chem*
50. De Souza, Le Bideau, Boschi et al. (2022) "Choosing a cellular model to study SARS-CoV-2" *Front. Cell. Infect. Microbiol*
51. Matsuyama, Nao, Shirato et al. (2020) "Enhanced isolation of SARS-CoV-2 by TMPRSS2-expressing cells" *Proc. Natl. Acad. Sci*
52. 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*
53. Carabelli, Peacock, Thorne et al. (2023) "SARS-CoV-2 variant biology: Immune escape, transmission and fitness" *Nat. Rev. Microbiol*
54. Markov, Ghafari, Beer et al. (2023) "The evolution of SARS-CoV-2" *Nat. Rev. Microbiol*
55. Bayarri-Olmos, Rosbjerg, Johnsen et al. (2021) "The SARS-CoV-2 Y453F mink variant displays a pronounced increase in ACE-2 affinity but does not challenge antibody neutralization" *J. Biol. Chem*
56. Motozono, Toyoda, Zahradnik et al. (2021) "SARS-CoV-2 spike L452R variant evades cellular immunity and increases infectivity" *Cell Host Microbe*
57. Liu, Liu, Plante et al. (2022) "The N501Y spike substitution enhances SARS-CoV-2 infection and transmission" *Nature*
58. Starr, Greaney, Hilton et al. (2020) "Deep Mutational Scanning of SARS-CoV-2 Receptor Binding Domain Reveals Constraints on Folding and ACE2 Binding" *Cell*
59. Zahradník, Marciano, Shemesh et al. "SARS-CoV-2 variant prediction and antiviral drug design are enabled by RBD in vitro evolution" *Nat. Microbiol. 2021*
60. Tian, Tong, Sun et al. "N501y mutation of spike protein in SARS-CoV-2 strengthens its binding to receptor ACE2. Elife 2021"
61. Han, Su, Zhang et al. "Molecular insights into receptor binding of recent emerging SARS-CoV-2 variants" *Nat. Commun. 2021*
62. Barton, Macgowan, Kutuzov et al. "Effects of common mutations in the SARS-CoV-2 Spike RBD and its ligand the human ACE2 receptor on binding affinity and kinetics"
63. Cui, Li, Shi (2019) "Origin and evolution of pathogenic coronaviruses" *Nat. Rev. Microbiol*
64. Lamers, Haagmans (2022) "SARS-CoV-2 pathogenesis" *Nat. Rev. Microbiol*
65. Chakraborty (2022) "E484K and N501Y SARS-CoV 2 spike mutants Increase ACE2 recognition but reduce affinity for neutralizing antibody" *Int. Immunopharmacol*
66. Plante, Liu, Liu et al. (2021) "Spike mutation D614G alters SARS-CoV-2 fitness" *Nature*
67. Zhou, Thao, Hoffmann et al. "SARS-CoV-2 spike D614G change enhances replication and transmission" *Nature*
68. Yan, Huang, Ge et al. (2022) "A mechanism for SARS-CoV-2 RNA capping and its inhibition by nucleotide analog inhibitors" *Cell*
69. Volz, Hill, Mccrone et al. (2021) "Evaluating the Effects of SARS-CoV-2 Spike Mutation D614G on Transmissibility and Pathogenicity"
70. Hou, Chiba, Halfmann et al. (2020) "SARS-CoV-2 D614G variant exhibits efficient replication ex vivo and transmission in vivo" *Science*
71. Zhang, Cai, Xiao et al. (2021) "Structural impact on SARS-CoV-2 spike protein by D614G substitution" *Science*
72. Kim, Jang, Lee et al. (2022) "Trypsin enhances SARS-CoV-2 infection by facilitating viral entry" *Arch. Virol*
73. Haywood (1994) "Virus receptors: Binding, adhesion strengthening, and changes in viral structure" *J. Virol*
74. Shang, Wan, Luo et al. "Cell entry mechanisms of SARS-CoV-2"
75. Kim, Lee, Yang et al. (2020) "The Architecture of SARS-CoV-2 Transcriptome" *Cell*
76. Park, Osinski, Hernandez et al. (2022) "The mechanism of RNA capping by SARS-CoV-2" *Nature*
77. Slanina, Madhugiri, Bylapudi et al. "Coronavirus replication-transcription complex: Vital and selective NMPylation of a conserved site in nsp9 by the NiRAN-RdRp subunit"
78. Slanina, Madhugiri, Wenk et al. (2023) "Conserved Characteristics of NMPylation Activities of Alpha-and Betacoronavirus NiRAN Domains" *J. Virol*
79. Wang, Svetlov (2021) "Artsimovitch, I. NMPylation and de-NMPylation of SARS-CoV-2 nsp9 by the NiRAN domain" *Nucleic Acids Res*
80. Longhi, Rancurel, Snijder et al. (2004) "The severe acute respiratory syndromecoronavirus replicative protein nsp9 is a single-stranded RNA-binding subunit unique in the RNA virus world" *Proc. Natl. Acad. Sci*
81. El-Kamand, Du Plessis, Breen et al. (2022) "A distinct ssDNA/RNA binding interface in the Nsp9 protein from SARS-CoV-2" *Proteins*
82. Small, Fedorova, Olinares et al. (2023) "Structural and functional insights into the enzymatic plasticity of the SARS-CoV-2 NiRAN domain" *Mol. Cell*
83. Kurhade, Xie, Shi (2023) "Reverse genetic systems of SARS-CoV-2 for antiviral research" *Antivir. Res*
84. Kipfer, Hauser, Lett et al. (2023) "Rapid cloning-free mutagenesis of new SARS-CoV-2 variants using a novel reverse genetics platform"
85. Furusawa, Yamayoshi, Kawaoka (2023) "The accuracy of reverse genetics systems for SARS-CoV-2: Circular polymerase extension reaction versus bacterial artificial chromosome" *Influenza Other Respi. Viruses*
86. Ren, Lan, Ju et al. (2021) "Mutation Y453F in the spike protein of SARSCoV-2 enhances interaction with the mink ACE2 receptor for host adaption" *PLoS Pathog*
87. Zhou, Peacock, Brown et al. (2022) "Mutations that adapt SARS-CoV-2 to mink or ferret do not increase fitness in the human airway" *Cell Rep*
88. Lu, Chu, Zhang et al. (2021) "The impact of spike N501Y mutation on neutralizing activity and RBD binding of SARS-CoV-2 convalescent serum"
89. Collier, De Marco, Ferreira et al. (2021) "Sensitivity of SARS-CoV-2 B.1.1.7 to mRNA vaccine-elicited antibodies" *Nature*
90. Tang, Gao, Wu et al. (2021) "Characterization of SARS-CoV-2 Variants N501" *Viral Infectivity. Front. Cell. Infect. Microbiol*
91. Kong, Montano, Corley et al. "Neuropilin-1 Mediates SARS-CoV-2 Infection of Astrocytes in Brain Organoids, Inducing Inflammation Leading to Dysfunction and Death of Neurons"
92. Bloom, Neher (2023) "Fitness effects of mutations to SARS-CoV-2 proteins" *Virus Evol*
93. Starr, Greaney, Hannon et al. (2022) "Shifting mutational constraints in the SARS-CoV-2 receptor-binding domain during viral evolution" *Science*
94. Moulana, Dupic, Phillips et al. (2022) "Compensatory epistasis maintains ACE2 affinity in SARS-CoV-2 Omicron BA.1" *Nat. Commun*
95. Yan, Ge, Zheng et al. (2021) "Cryo-EM Structure of an Extended SARS-CoV-2 Replication and Transcription Complex Reveals an Intermediate State in Cap Synthesis" *Cell*
96. "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=PMC12724194&blobtype=pdf | # ZIKV infection causes placental inflammation through activating PANoptosis
Shuqing Zhang, Delin Chen, Kexin Zhang, Minjie Liu, Hao Liang, Linyue Liang, Jingyao Liang, Minqi Liang, Shu An, Mingcui Lyu, Junying Zhu, Shangwei Li, Dingwen Hu, Xun Zhu, Jueheng Wu, Zhenjian He, Mengfeng Li
## Abstract
Infection by the Zika virus (ZIKV) during pregnancy can cause congenital Zika syndrome (CZS) or other central nervous system conditions in infants, underscoring the importance of understanding the role of ZIKV-induced placental damage in the development of CZS. Here, we established a ZIKV-infected pregnant mouse model and examined the pathological changes in the placenta following ZIKV infection. We found that ZIKV infection induces severe placental inflammation associated with trophoblast PANoptosis. Specifically, ZIKV activates RIG-I and recruits ASC, caspase-1, NLRP3, caspase-8, and RIPK1 to form a PANoptosome complex, leading to the activa tion of PANoptosis. Moreover, the combined use of Z-VAD-FMK with GSK872 or with the RIG-I inhibitor RIG012 robustly suppressed ZIKV-induced cell death, attenuated the inflammatory response in trophoblast cells and the placenta induced by ZIKV infection. Collectively, we elucidate a previously unrecognized mechanism by which ZIKV infection causes severe placental inflammation by activating PANoptosis and provide a foundation for the potential application of anti-PANoptotic therapy against ZIKV-associated diseases. IMPORTANCE Zika virus (ZIKV), a mosquito-borne virus, has caused significant disease in humans during outbreaks over the last decade. Currently, there is no approved preventive vaccine or specific therapeutic drug against ZIKV. The World Health Organi zation declared a Public Health Emergency of International Concern regarding micro cephaly and other neurological disorders caused by ZIKV during pregnancy in 2016, highlighting the importance of understanding the role of the maternal-fetal barrier in this viral disease. The mechanism by which ZIKV causes placental pathogenesis, however, remains unclear. In this study, our data elucidate a previously unrecognized mecha nism underlying ZIKV infection that causes severe placental inflammation by activating PANoptosis. Furthermore, we propose a treatment that effectively inhibits ZIKV-induced PANoptosis and attenuates the inflammatory response in trophoblast cells in vitro and in vivo.
While ZIKV infection generally causes mild symptoms, such as fever, rashes, arthralgia, myalgia, and conjunctivitis (8), maternal infection during pregnancy can lead to severe complications such as an increased risk of premature birth, miscarriage, stillbirth, and CZS. Infants with CZS exhibit microcephaly or other congenital malformations, including brain calcifications, limb contractures, high muscle tone, eye abnormalities, and hearing loss (5,(9)(10)(11). Transmission of ZIKV from mother to fetus occurs in all trimesters of pregnancy, with the highest risk in the first trimester (12,13). Clinical evidence has demonstrated the presence of ZIKV in the placenta, amniotic fluid, infant's blood, cord blood, and cerebrospinal fluid (14). Despite the extensive research on ZIKV in the past decades, the pathogenesis of ZIKV, particularly the mechanisms for its vertical transmis sion and placental pathological alterations, remains largely unclear.
PANoptosis is a unique modality of programmed cell death (PCD) that can occur during infection and inflammation, which is usually mediated by a complex known as PANoptosome. PANoptosis shares cellular features and molecular mechanisms with apoptosis, pyroptosis, and necroptosis but cannot be accounted for by any of them alone (15)(16)(17). Of note, apoptosis proceeds through either an extrinsic or intrinsic pathway and is a caspase-8-or caspase-9-dependent form of cell death, which is activated by executioner caspases such as caspase-3 and caspase-7 (18,19), whereas pyropto sis is a lytic form of pro-inflammatory cell death that can be initiated by inflammasome activation-mediated caspase-1 cleavage of gasdermin D (GSDMD) (20). Other executioners, such as gasdermin E (GSDME), cleaved by caspase-3, can also activate pyroptosis (21). Another form of programmed cell death, necroptosis, is activated by proteins containing a receptor-interacting protein homotypic interaction motif, including receptor-interacting serine/threonine-protein kinase 1 (RIPK1) and receptor-interacting serine/threonine-protein kinase 3 (RIPK3). Activation of RIPK3 further phosphorylates mixed lineage kinase domain-like (MLKL), the executioner of necroptosis (22). Previous studies have reported that innate immune sensors or regulators can drive the assembly of the PANoptosome to induce PANoptosis, such as Z-conformation nucleic acid binding protein 1 (ZBP1), absent in melanoma 2 (AIM2), transforming growth factor-β-activated kinase 1 (TAK1), and RIPK1 (23). PANoptosis has been demonstrated to be involved in various infectious and inflammatory diseases and can be induced by diverse triggering factors ranging from viruses to fungi, acting as an important host immune defense against pathogen infection (17,24,25). Notably, its excessive activation causes severe inflammatory pathological damage to the host. For example, influenza A virus (IAV) infection activates ZBP1-mediated PANoptosis and induces inflammatory damage in the lungs while limiting IAV replication (24,(26)(27)(28). Similarly, SARS-CoV-2 infection induces the expression of pro-inflammatory cytokines tumor necrosis factor alpha (TNF-α) and gamma interferon (IFN-γ), which in turn activate PANoptosis, leading to a prolonged and sustained cytokine storm, resulting in severe lung inflammation and pathological damage (29). It is noteworthy that placentas infected by ZIKV develop severe inflammatory response and cell death (9,30), prompting us to further ask whether PANoptosis plays a role in the process and consequently leads to ZIKV-related placental diseases.
In this study, our data show that ZIKV infection induces PANoptosis and inflammation in trophoblast cells and placentas. The combined use of a pan-caspase inhibitor Z-VAD-FMK with the RIPK3 inhibitor GSK872 or with the RIG-I inhibitor RIG012 can effectively inhibit ZIKV-induced PANoptosis and attenuate the inflammatory response in trophoblast cells in vitro as well as in vivo.
## RESULTS
## ZIKV replicates in the placenta and causes severe pathological changes
Type I interferon receptor-deficient (Ifnar1 -/-) mice, a commonly used animal model for pathogenic viruses, including ZIKV, were employed in this study. To investigate the pathogenesis of ZIKV in the placenta and fetal demise during pregnancy, Ifnar1 -/- dams with a C57BL/6 background were inoculated intraperitoneally with ZIKV (1 × 10 5 plaque-forming units [PFUs]) at embryonic day 7.5 (E7.5) and euthanized at E15.5 (Fig. 1A), as previously described (31). The experiment showed that ZIKV infection resulted in lower body weight in pregnant dams compared with the mock controls from 4 to 7 days post-infection (dpi) (Fig. 1B), and viral loads in maternal organs and fetal heads were quantified using a quantitative real-time RT-PCR (qRT-PCR) assay. The results showed that ZIKV RNA was detected in all examined tissues, with the highest copy numbers of viral RNA in placentas and fetal heads (Fig. 1C andD). The pregnant dams were sacrificed, and embryo resorption in uterine horns after ZIKV infection was observed in contrast to mock-infected controls (Fig. 1E). Further separating the fetuses and placentas from the uterine horns, we found that 76% of the fetuses obtained from ZIKV-infected dams exhibited abnormalities, indicating a significant uterine growth restriction (Fig. 1F andG). Placentas from ZIKV-infected dams also showed abnormal morphology and tissue necrosis when compared with the mock-infected controls (Fig. 1G), suggesting that the observed abnormalities of the placenta and fetus were closely associated with ZIKV infection.
We next sought to determine whether inflammation was present in the placentas upon ZIKV infection. Using a qRT-PCR assay, high expression levels of pro-inflammatory cytokine genes, including IL6, IL15, IFNγ, TNFα, CXCL15, IL18, and IL1β, were confirmed in ZIKV-infected placentas, suggesting that ZIKV infection may trigger placental inflammation (Fig. 1H). Further histopathological analysis showed that ZIKV-infected placen tas were smaller than mock-infected placentas (Fig. 1I), consistent with previously reported cases of congenital ZIKV infection (32). The results also showed that the area of the labyrinthine layer in mouse placentas, which is believed to be responsi ble for mother-fetus nutrient and gas exchanges, was particularly reduced after ZIKV infection (Fig. 1I). In addition, ZIKV infection also led to uneven distribution of blood vessels, smaller vessel lumen, and tissue hemorrhage, partially in the labyrinthine layer. Furthermore, large numbers of monocytes infiltrated the labyrinthine layer (Fig. 1I). To assess whether cell death was caused by ZIKV infection in placental tissue, we further examined the activation of molecules key to various forms of programmed cell death, such as caspase-1, GSDMD, GSDME, and NLRP3 for pyroptosis, caspase-3 and caspase-8 for apoptosis, and MLKL phosphorylation and RIPK3 phosphorylation for necroptosis activation (33). Our results revealed that all these molecular markers, i.e., cleaved caspase-1, cleaved-GSDMD, cleaved-GSDME, cleaved-caspase-3, cleaved-cas pase-8, NLRP3, phosphorylation of MLKL, phosphorylation of RIPK3, and RIG-I, were significantly increased following ZIKV infection, suggesting that ZIKV might have caused PANoptosis activation in the placentas (Fig. 1J).
## ZIKV infection induces PANoptosis in trophoblast cells
The above in vivo data prompted us to further examine and characterize ZIKV-induced PANoptosis in cultured human placental trophoblast cells in vitro. Human trophoblast cell lines JEG-3 and HTR-8 were infected with ZIKV at a multiplicity of infection (MOI) of 5, respectively. Propidium iodide (PI) staining assay showed that the proportions of PI-positive cells significantly elevated after ZIKV infection in both JEG-3 and HTR-8 lines (Fig. 2A andB), and lactate dehydrogenase (LDH) in cellular supernatant also increased after ZIKV infection, together suggesting that the cells underwent cytolytic cell death (Fig. 2C). To further demonstrate the PANoptotic features of ZIKV-induced cytological changes, we assessed the activation of proteins known to be key to pyroptosis, apoptosis, and necroptosis, respectively. As shown in Fig. 2D, elicited the activation of pyroptosis mediators (GSDMD, GSDME, and caspase-1), apoptosis mediators (caspase-8, caspase-9, caspase-3, and caspase-7), and necroptosis mediators (MLKL and RIPK3), robustly suggesting that ZIKV infection can activate almost all key molecules involved in pyroptotic, apoptotic, and necroptotic pathways in trophoblast cells, which are essential for PANoptosis to occur. These results were consistently observed in ZIKV-infected primary human placental trophoblast cells (Fig. 2E through H). Collectively, these findings suggest that ZIKV infection induces PANoptosis in trophoblast cells.
## ZIKV infection promotes interaction among RIG-I, ASC, and caspase-8 to form PANoptosis-mediating PANoptosome
Previous studies have shown that PANoptosis can be initiated by nucleic acid receptors, such as ZBP1 and AIM2, when recognizing exogenous nucleic acid stimuli (34,35).
To identify the initiating receptor for the activation of PANoptosis induced by ZIKV in trophoblast cells, we first employed small-interfering RNA (siRNA) silencing strategy to screen a pool of nucleic acid receptors, including ZBP1, AIM2, TLR3, TLR7, TLR8, RIG-I, MDA5, and LGP2 by evaluating changes in their effects on ZIKV-induced trophoblast death, respectively (Fig. S1). Our results showed that knocking down RIG-I expression inhibited the cell death most potently compared with the siRNA negative control and other nucleic acid receptor groups (Fig. 3A andB) and reduced PI-positive cells and LDH release triggered by ZIKV infection (Fig. 3C through E). Moreover, the activation of PANoptosis mediator proteins, such as caspase-1, caspase-3, and MLKL, could all be suppressed when RIG-I was knocked down, indicating that RIG-I may be a sensing receptor for ZIKV-induced PANoptosis in trophoblasts (Fig. 3F).
To further identify the specific pathway via which RIG-I activates PANoptosis, we performed co-immunoprecipitation assays to test whether RIG-I interacts with key proteins functioning in pyroptotic, apoptotic, and necroptotic pathways. Our results revealed that caspase-8 was coprecipitated with RIG-I and RIPK1 upon ZIKV infection (Fig. 3G), and that ASC physically binds with RIG-I, caspase-1, and NLRP3 (Fig. 3G), suggesting that ZIKV infection could promote interaction among caspase-8, RIPK1, RIG-I, NLRP3, caspase-1, and ASC to form the PANotosome, which is essential for the activation of PANoptosis.
## Z-VAD-FMK and GSK872 treatment inhibits PANoptosis and inflammation in trophoblast cells
Next, we investigated whether PANoptosis can be a potential therapeutic target for placental pathogenic inflammation. Several inhibitors known to suppress key proteins of the PANoptosis pathway were tested for their therapeutic effects in trophoblast cells, among which were included selective caspase 1 inhibitor VX765 (10 µM), caspase 3 inhibitor Z-DEVD-FMK (30 µM), pan caspase inhibitor Z-VAD-FMK (40 µM), and specific inhibitor of RIPK3 phosphorylation GSK872 (3 µM). The above inhibitors were applied to treat ZIKV-infected JEG-3 cells, and the results showed that co-treatment with Z-VAD-FMK (40 µM) and GSK872 (3 µM) could significantly inhibit cell death triggered by ZIKV infection. The inhibitory effect was far more potent than those of separate treatments
## 48 hpi). (G) Comparative analysis of LDH release between mock-and ZIKV-infected primary human trophoblast (MOI = 5, hpi). (H) The expression levels of
PANoptosis-associated proteins in primary human trophoblast were examined by immunoblotting analysis (MOI = 5, 48 hpi). All data are presented as mean ± SD, Student's t test, *P < 0.05, **P < 0.01, and ***P < 0.001. with VX765, Z-DEVD-FMK, Z-VAD-FMK, or GSK872 alone (Fig. 4A andB). In addition, Z-VAD-FMK and GSK872 co-treatment remarkably reduced the number of PI-positive cells and LDH release induced by ZIKV infection (Fig. 4C through E). Furthermore, activation of molecules key to the induction of pyroptosis, apoptosis, or necroptosis was robustly inhibited after the Z-VAD-FMK and GSK872 co-treatment, indicating that PANoptosis is repressed by Z-VAD-FMK and GSK872 (Fig. 4F). Moreover, the mRNA expression levels of pro-inflammatory cytokines, such as IL8, IL1β, TNFα, IFNα, IFNγ, IL15, IL13, and CXCL5, were decreased by the combined treatment of Z-VAD-FMK and GSK872 (Fig. 4G). Together, our data suggest that ZIKV-induced PANoptosis and inflammation in trophoblast cells could be antagonized by a combination of Z-VAD-FMK and GSK872 inhibitors.
## Z-VAD-FMK and GSK872 treatment attenuates PANoptosis and inflammation in the placentas of pregnant dams
To further look into the in vivo therapeutic effects of the Z-VAD-FMK and GSK872 combination in our animal model, pregnant dams intraperitoneally infected with 1 × 10 5 PFUs of ZIKV or mock-infected were co-treated with Z-VAD-FMK (10 mg/kg of body weight, i.p.) and GSK872 (10 mg/kg of body weight, i.p.). As shown in Fig. 5A andB, when all pregnant dams were sacrificed at E15.5 of pregnancy to obtain placentas and fetuses, ZIKV infection caused significant weight loss in pregnant dams. However, the illness could be attenuated by the combined Z-VAD-FMK/GSK872 treatment. Further analysis of the morphology of the fetuses revealed that while the incidence of abnormality in ZIKV-infected fetuses was 38.46%, pregnant dams treated with the Z-VAD-FMK and GSK872 combination displayed a significantly lower rate (17.64%) of fetal abnormalities (Fig. 5C andD). In addition, histopathological examination of the placentas showed that the pathological changes caused by ZIKV infection, such as uneven distribution of blood vessels, narrowing of vascular space, tissue hemorrhage, and infiltration of monocytes, were alleviated after combined treatment with Z-VAD-FMK and GSK872 (Fig. 5E). Notably, no adverse effects on maternal survival, tissue histology, fetal development, or placental morphology were observed in pregnant dams treated with control vehicle or Z-VAD-FMK and GSK872 without ZIKV infection (Fig. S2A through D; Fig. 5E). Moreover, after the Z-VAD-FMK/GSK872 co-treatment, proteolytic cleavages of both caspase-1 and caspase-3 and the phosphorylation level of MLKL were found to be repressed in the placentas of ZIKV-infected pregnant dams compared to that in the control vehicle group, indicating that the combined Z-VAD-FMK and GSK872 treatment can indeed suppress PANoptosis induced by ZIKV infection in vivo (Fig. 5F). Furthermore, we also found that Z-VAD-FMK plus GSK872 treatment abrogated the increase in mRNA levels of IL6, IL15, IFNγ, TNFα, CXCL15, IL18, and IL1β in ZIKV-infected placentas (Fig. 5G). Together, the above data suggest that pan-caspase inhibitor Z-VAD-FMK, when used in combination with RIPK3 phosphorylation inhibitor GSK872, suppresses PANoptosis and attenuates inflammatory placental damages in pregnant dams.
## RIG-I inhibitor RIG012 alleviates the placental inflammation caused by ZIKV infection
As our above experiments identified RIG-I as a key sensor driving ZIKV-induced PANoptosis in trophoblast cells, we next further evaluated the therapeutic potential of RIG-I inhibition using the inhibitor RIG012 in ZIKV-infected Ifnar1 -/-dams. As shown in Fig. 6A through E, when Ifnar1 -/-dams were inoculated intraperitoneally with 1 × 10 5 PFUs of ZIKV at E7.5 and treated with either RIG012 (10 mg/kg, i.p.) or carrier solvent (PBS/DMSO). The RIG012 treatment alleviated the morphological defects in both fetuses and placentas after ZIKV infection (Fig. 6A) and reduced the fetal abnormality rate from 60% to 26.67% (Fig. 6B). Placental weight was also partially restored in RIG012-treated pregnant dams following ZIKV infection (Fig. 6C). In addition, the mRNA levels of IL6, IL15, IFNγ, TNFα, CXCL15, IL18, and IL1β in ZIKV-infected placentas were significantly decreased by RIG012 treatment (Fig. 6D). Furthermore, this experimental therapy attenuated the pathological changes caused by ZIKV infection in placentas (Fig. 6E). Of note, no apparent adverse effects on maternal survival, tissue histology, fetal development, or placental morphol ogy were detected in pregnant dams treated with control vehicle or RIG012 without ZIKV infection (Fig. S2A through D; Fig. 6E). Collectively, these data demonstrate that the RIG-I inhibitor RIG012 may ameliorate the pathological and inflammatory effects of the placenta caused by ZIKV infection.
## DISCUSSION
Our present study demonstrated that ZIKV infection induces PANoptosis in placental trophoblast cells, leading to severe placental inflammation. We further elucidated that ZIKV infection drives PANoptosome formation, with RIG-I nucleating the complex by recruiting ASC, caspase-1, NLRP3, caspase-8, and RIPK1 to activate PANoptosis. We also found that inhibition of PANoptosis by treatment with pan-caspase inhibitor Z-VAD-FMK and RIPK3 inhibitor GSK872 together or RIG-I inhibitor RIG012 can attenuate placental inflammatory damages caused by ZIKV infection. These findings reveal a previously unidentified mechanism by which ZIKV infection triggers PANoptosis in trophoblast cells and provide potential therapeutic targets and new treatment options for ZIKV-associated diseases.
During pregnancy, ZIKV can be transmitted across the placenta to the developing fetus, leading to potential teratogenic consequences in the fetus, which may include a wide spectrum of structural anomalies, functional impairments, and clinical sequelae that usually manifest at birth or in early life (36). The fatality rate of infants born with CZS was 10% (37). Evidence of ZIKV infection in placental tissue from one miscar riage case showed heterogeneous chorionic villi with calcification, fibrosis, perivillous fibrin deposition, patchy intervillositis, and focal villitis (9). Chronic villous inflammation, edema, and trophoblastic epithelium lesions were observed in the placental tissue of another miscarriage caused by ZIKV infection (38). These previous findings suggest that ZIKV infection during pregnancy might be able to induce severe pathogenic a hope of using a pyroptosis inhibitor to suppress mother-fetus transmission of ZIKV. Nevertheless, in our preliminary assessment for the therapeutic effect of caspase-1 inhibitor VX765, its inhibition of cell death and inflammation caused by ZIKV infection was suboptimal, suggesting that ZIKV may also induce other forms of inflammation-associated cell death. On such a backdrop, our current study identifies that the placenta infected with ZIKV not only undergoes pyroptosis but also is subject to apoptosis and necroptosis, which meets the characteristics of a highly interconnected inflammatory cell death process known as PANoptosis. We have isolated and cultured human primary trophoblasts and confirmed this phenomenon in vitro. Direct validation may be performed in the future using ZIKV-infected placental samples from pregnancies, or further validation could be conducted using human placental tissues or trophoblastderived organoids. Moreover, only one Asian lineage ZIKV strain was used in the present study; whether other strains, including prototypical African lineages, can similarly induce placental PANoptosis remains to be investigated. Pathologically, PANoptotic cell death and its associated inflammatory damage in the placenta following ZIKV infection can contribute to intrauterine growth restriction and adverse fetal outcomes, such as reduced crown-rump length, lower fetal weight, and widespread developmental impairments. In our experimental model, pregnant individuals suffering from ZIKV infection frequently exhibited fetal abnormalities, some of which were severe and fatal. Mechanistically, placental inflammation can compromise nutrient transport and oxygen supply and consequently contribute to developmental defects. Notably, neurological abnormalities in the fetus likely arise from both direct viral infection of neural cells and indirect inflammatory effects. In rare cases where fetuses are carried to term, they may exhibit congenital CZS, characterized by microcephaly, intracranial calcifications, and other neurological disorders. In such a context, while our study provides novel insights into the role of PANoptosis in ZIKV-induced placental damage, further study is needed to fully elucidate the long-term neurological conse quences of this process. Moreover, inflammation driven by PANoptosis may be intricately linked to immune dysfunctions, including cytokine storms, a breakdown in immune tolerance, and autoimmune disorders in various other viral infections such as those by SARS-CoV-2 (24,27,29). Similarly, the success of ZIKV infection is usually associated with a whole variety of dysregulated immune responses for the virus to evade immune suppression, and the development of such immune dysregulation usually contributes to the pathogenesis of viral diseases. Therefore, this form of inflammatory cell death might represent a critical driver of immune dysregulation, which exacerbates disease patho genesis, and dysregulation of PANoptosis and PANoptosis-mediated inflammation can be causative of other immunopathic conditions. Hence, understanding the mediating mechanisms for ZIKV-associated PANoptosis might facilitate the development of new targeted therapeutic strategies.
Interestingly, our data demonstrate that combining pan-caspase inhibitor Z-VAD-FMK and RIPK3 inhibitor GSK872 effectively suppressed ZIKV-induced PANoptosis and attenuated the inflammatory conditions in vitro, with superior efficacy to either agent alone. Consistently, in vivo studies confirmed that the dual inhibitor combination alleviates ZIKV-induced placental inflammatory damage. These findings identify PANoptosis as a key driver of ZIKV-induced placental damage and reveal that its pharmacological inhibition-concurrently targeting pyroptosis, apoptosis, and necroptosis-can potentially be a promising therapeutic strategy. This approach provides translatable benefits by mitigating placental inflammation and improving fetal outcomes in affected pregnancies. However, it remains to be determined whether the therapeutic effects of these inhibitors are solely due to PANoptosis inhibition or also involve immune system modulation and off-target effects.
Of particular note, PANoptosis can be achieved by activation of a unique inflammatory PCD pathway during viral infection, which could expose the virus to the immune system and limit virus replication. However, hyperactivation of PANoptosis can also lead to systemic inflammation and pathological conditions (24). PANoptosis activation is mediated by the formation of a protein complex termed PANoptosome, which renders a molecular scaffold that allows for coordinated interactions and activation of three modes of programmed cell death, i.e., pyroptosis, apoptosis, and necroptosis. Thus far, three PANoptosome complexes with different sensors and regulators have been identified, namely, the ZBP1-, AIM2-, and RIPK1-PANoptosome, respectively (15). The ZBP1 PANoptosome has been found to sense the viral ribonucleoprotein complexes of IAV that trigger PANoptosis (27), whereas the AIM2 PANoptosome has been identified as a double-stranded DNA virus sensor to induce PANoptosis (34). Of note, the RIPK1-PANoptosome assembles in the absence or under the inhibition of TAK1 and during Yersinia infection (40). In PANoptosome, caspase-8 and ASC are two key scaffold proteins, and caspase-8 has been previously suggested to promote apoptosis and also found to participate in activating pyroptosis by cleaving the GSDM family proteins (19). In addition, ASC is a key scaffold protein involved in the formation of the inflammasome and mediates pyroptosis and has also been reported to be an essential adaptor for the formation of PANoptosome complexes (27,41). In this study, we uncover that RIG-I, a cytoplasmic dsRNA sensor, can recruit caspase-8, RIPK1, ASC, caspase-1, and NLRP3, forming a more complex PANoptosome to activate PANoptosis following ZIKV infection. Knockdown or inhibition of RIG-I leads to suppression of PANoptosis and inflammatory cytokine release, suggesting that RIG-I may be crucial for ZIKV-induced inflammation. Nevertheless, it remains unclear how RIG-I functions to mediate the ZIKV activation of PANoptosome, and future studies are warranted to determine whether other RNA viruses can also activate RIG-I and form PANoptosome to induce PANoptosis. Furthermore, while our study demostrates activation of all three programmed cell death pathwayspyroptosis, apoptosis, and necroptosis-during ZIKV infection, the precise roles and relative contributions of each to PANoptosis remain to be fully delineated.
Given that RIG-I is a cytosolic RNA sensor, it is conceivable that the ZIKV genomic RNA can be recognized by RIG-I within the cell. A key objective is therefore to deter mine whether ZIKV triggers PANoptosis through this RIG-I-dependent recognition. Alternatively, the potential role of ZIKV-encoded proteins in modulating RIG-I to induce PANoptosis also merits exploration. On the other hand, RIG-I is important in innate (and adaptive) immunity; it recognizes viral RNA within the host cells and activates down stream proteins such as MAVS and TBK1, leading to the production of type I IFN and eliciting the innate immune response. Our in vivo data show that ZIKV infection actually induces PANoptosis in the placenta of Ifnar1 -/-C57BL/6 dams, which lack a functional type I IFN signaling pathway. Although further validation using genome editing or pharmacological intervention is warranted, it indirectly indicates that RIG-I promotes ZIKV-induced PANoptosis not through the canonical type I IFN signaling pathway. Besides, our present data show that RIG-I inhibition effectively suppresses PANoptosis in the placenta caused by ZIKV infection in vivo. Although these preclinical findings are encouraging, further studies are required to evaluate the clinical translational potential of RIG012 or its analogues.
## MATERIALS AND METHODS
## Cell culture and virus
JEG-3 cells (TCHu195, the Cell Bank of the Chinese Academy of Sciences, Shanghai, China), HEK293T cells, and Vero cells were maintained at 37°C with 5% CO 2 in DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA), 10 mM L-glutamine (HyClone, Logan, UT, USA), 100 µg/mL streptomycin, and 100 U/mL penicillin. HTR-8/SVneo cells (CRL-3271, ATCC, Manassas, VA, USA) were maintained at 37°C with 5% CO 2 in RPMI-1640 (Gibco, Grand Island, NY, USA) supplemented with 10% FBS, 100 µg/mL streptomycin, and 100 U/mL penicillin. All cell lines were tested negative for mycoplasma contamination. The Asian lineage ZIKV SZ01 strain (GenBank accession no. KU866423) was propagated in Vero cells. Virus stocks were titrated by plaque assays on Vero cells.
## In vitro ZIKV infection of trophoblast cells
Human primary trophoblast cells were isolated from healthy term placentas according to a previously described method (42). Human primary trophoblast cells or JEG-3 cells were seeded in 6-well plates at 5 × 10 5 cells per well. After 24 h, the cells were washed with PBS and then incubated with ZIKV at the indicated MOI for 2 h at 37°C. Subsequently, the inoculum was removed, and cells were washed with PBS and then replaced with fresh medium. Mock-infected cells were incubated with the culture supernatant from uninfected Vero cells. For inhibitor treatment experiments, cells were incubated with ZIKV in the presence of the indicated inhibitors for 2 h at 37°C, washed once with PBS, and maintained in fresh medium supplemented with the corresponding inhibitors. At designated time points after treatment and ZIKV challenge, cellular proteins or RNA were harvested for subsequent analysis according to experimental requirements.
## ZIKV-induced placental inflammation model
Ifnar1 -/-C57BL/6 female and male mice (8-12 weeks old) were mated overnight at a 2:1 ratio. The following morning, female mice with a viscous vaginal plug were marked as E0.5 of pregnancy. Pregnant dams were randomly divided into two groups and then challenged with 1 × 10 5 PFUs of ZIKV or an equal volume of Vero cell culture supernatant by intraperitoneal injection at E7.5 of pregnancy, respectively. The weight and behavior of pregnant dams were recorded daily until E15.5. To evaluate the therapeutic effect of combined Z-VAD-FMK and GSK872 or RIG012, pregnant dams were intraperitoneally pretreated with combined Z-VAD-FMK (10 mg/kg) and GSK872 (10 mg/kg) or RIG012 (10 mg/kg) diluted in PBS/DMSO at E6.5 before infection. Pregnant dams were then intraperitoneally inoculated with 1 × 10 5 PFUs of ZIKV at E7.5 of pregnancy. Two hours after infection, mice were intraperitoneally injected with Z-VAD-FMK/GSK872 or RIG012, with subsequent maintenance doses administered every 2 days from E7.5 through E13.5. Vehicle-treated mice that received injections of PBS/DMSO served as a control. All pregnant dams were sacrificed at E15.5 of pregnancy, and both maternal and fetal tissues were collected for subsequent molecular analyses and histopathological examination.
## Histopathological evaluation of placenta using hematoxylin and eosin staining
Placenta tissues were fixed in 4% paraformaldehyde for 24 h at room temperature, followed by overnight dehydration. Subsequently, the tissues were embedded in paraffin and sectioned using a microtome. Tissue sections were rehydrated and stained with Mayer's hematoxylin and ethanolic eosin. Following dehydration, the hematoxylin and eosin (H&E)-stained sections were microscopically examined for histologic changes. Images were obtained using an AxioScan.Z1 and analyzed with the ZEN software (Carl Zeiss MicroImaging).
## Assessment of cell death by propidium iodide staining
Cultured cells on coverslip were washed twice using PBS and stained with 5 µg/mL Hoechst 33342 (Beyotime, Shanghai, China) at the ratio of 1:1,000 at 37°C in dark for 10 min and then restained in PI (Beyotime, Shanghai, China) staining solution (1.5 µM in PBS) at 4°C in dark for 30 min. The stained coverslip was rinsed briefly with PBS to remove unbound dye and subsequently dried to mount in an antifade reagent. Images were obtained using a Zeiss Axio Observer Z1 microscope (Carl Zeiss MicroImaging). PIand Hoechst-positive cells were analyzed by ImageJ software.
## Lactate dehydrogenase release assay
LDH released into cell culture supernatants was measured using the CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega) according to the manufacturer's instructions. Briefly, cells were seeded in a 24-well plate overnight, followed by infection with ZIKV or mock control. Supernatants and cell lysates were collected after 48 hpi and were transferred to a 96-well plate. An equal volume of CytoTox 96 Reagent is added to each well and incubated for 30 min. Stop Solution is added, and the absorbance signal is measured at 490 nm in a plate reader. The LDH activity in the culture supernatant was expressed as a percentage of total LDH in the cell lysates.
## siRNA synthesis and transfection
Control siRNA and gene-specific siRNAs were purchased from RiBoBio (Guangzhou, China). The siRNA was delivered into the cells by using Lipofectamine 3000 transfec tion reagent (Invitrogen) according to the manufacturer's instructions. The targeting sequences of siRNA for human indicated genes are as follows: ZBP1-siRNA: GGAACATCATTACAAGACA TLR3-siRNA: GCACGAATTTGACTGAACT TLR7-siRNA: GGGTATCAGCGTCTAATAT TLR8-siRNA: GAACGGAAATCCCGGTATA RIG-I-siRNA: TAGTAATGCTGGTGTAATT MDA5-siRNA: GGACAAGCTTCTAGTTAGA
LGP2-siRNA: GGGATCCTGTGGTCATCAA AIM2-siRNA: GCAACGTGCTGCACCAAAA
## Short hairpin RNA-mediated gene knockdown in trophoblast cells
Two short hairpin RNAs (shRNAs) targeted to RIG-I were cloned into the pLVX-shRNA vector. Approximately 4 µg shRNA plasmid DNA, 2 µg packaging plasmid pMD2.G, and 4 µg pSPAX2 were co-transfected into HEK293T cells using Lipofectamine 3000 transfection reagent (Invitrogen) according to the manufacturer's instructions. Viral supernatants were collected 48 h post-transfection, filtered through a 0.45 µm filter to remove cell debris, and then inoculated into JEG-3 cells for another 48 h. Stable cell lines were selected via DMEM containing 1.5 µg/mL puromycin (Sigma, St. Louis, MO, USA). Western blotting was employed to verify RIG-I knockdown efficiency. shRNA sequences were as follows: RIG-I-shRNA1: 5′-TAGTAATGCTGGTGTAATT-3′ ; RIG-I-shRNA2: 5′-CCGGCA CAGAAGTGTATAT-3′.
## Western blotting analysis
The harvested tissues or cells were lysed with RIPA lysis buffer (Millipore, Bedford, MA, USA) containing a cocktail of protease and phosphatase inhibitors (Sigma, St. Louis, MO, USA). Protein concentrations of cell lysates were determined with a bicinchoninic acid protein assay (Thermo Fisher Scientific, Rochester, NY, USA). Protein samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane (Roche, Indianapolis, IN, USA). Nonfat milk (5%) in Tris-buffered saline (20 mM Tris-HCl [pH 7.6] and 135 mM NaCl) containing 0.1% Tween 20 was used to block nonspecific antibody binding sites for 1 h at room tem perature. Next, the membranes were incubated with the following primary antibodies, respectively, overnight at 4°C: anti-Caspase1/P20/P10 (1:1,000, Proteintech, 22915-1-AP), anti-Caspase-1(p20) (1:1,000, AdipoGen, AG-20B-0042), anti-GSDMD (1:1,000, Abcam, ab210070, ab209845), anti-GSDME (1:1,000, Abcam, ab215191), anti-Caspase3 (1:2,000, Cell Signaling Technology, #9662), anti-Caspase7 (1:2,000, Cell Signaling Technology, #9492), anti-Caspase8 (1:2,000, Cell Signaling Technology, #9746), anti-Caspase8 (1:2,000, ABclonal, A0215), anti-Caspase9 (1:1,000, Cell Signaling Technology, #9502), caspase-9 p10 (H-83) (1:500, Santa Cruz Biotechnology, sc-7885), anti-Phospho-MLKL (1:1,000, Cell Signaling Technology, #37333, Beverly, MA, USA), anti-MLKL (phospho S358) (1:1,000, Abcam, ab187091) , 1:2,000, ab243142RIP (D94C12) XP Rabbit mAb (1:1,000, Cell Signaling Technology, #3493, Cambridge, England), anti-Phospho-RIP3 (1:1,000, Cell Signaling Technology, #93654, #91702), anti-RIP3 (1:1,000, Proteintech, 68786-2-Ig), anti-Phospho-RIPK1 (1:1,000, Proteintech, 66854-1-Ig), anti-ZIKV E (1:2,000, GeneTex Inc., GTX133314), anti-AIM2 (1:2,000, Proteintech, 66902-1-Ig), anti-ZBP1 (1:2,000, Proteintech, 13285-1-AP), anti-RIG-I/DDX58 (1:2,000, Proteintech, 20566-1-AP), anti-LGP2 (1:2,000, Proteintech, 11355-1-AP), anti-TLR8 (1:2,000, Proteintech, 67317-1-AP), anti-TLR3 (1:1,000, ABclonal Technology, A11778), anti-TLR7 (1:1,000, ABclonal Technology, A0991), anti-TLR7 (ABclonal Technology, Wuhan, China), anti-DHX58 (1:1,000, ABclonal Technol ogy, A8257), anti-MAVS (1:1,000, Cell Signaling Technology, #24930), anti-TBK1 (1:1,000, Cell Signaling Technology, #3013), anti-NLRP3 (1:1,000, Cell Signaling Technology, #15101), anti-ASC (1:1,000, Santa Cruz Biotechnology, sc-271054), and anti-GAPDH (1:20,000, Proteintech, 60004-1-Ig). Subsequently, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature, and signals were detected by enhanced chemiluminescence using a commercial kit (Tanon, Shanghai, China) according to the manufacturer's protocols.
## Co-immunoprecipitation assay
JEG-3 cells were infected with ZIKV for 48 h or mock infected, and the cells were lysed with protein lysis buffer (25 mM HEPES, 150 mM NaCl, 1 mM EDTA, 2% glycerol, 1% NP-40, and a cocktail of protease and phosphatase inhibitors). The lysates were incubated with the indicated antibodies or IgG antibody overnight at 4°C. The precipi tates were washed five times with wash buffer (20 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 2% glycerol, and 0.1% NP-40) every 5 min. The immunocomplexes were resuspended in sampling buffer and examined by western blotting analysis.
## Quantitative real-time reverse transcription PCR
Total RNA was extracted from cultured cells or tissues with the TRIzol reagent (Invitro gen, Carlsbad, CA, USA) according to the manufacturer's instructions. First-strand cDNA synthesis was conducted using random hexamer primers, and qPCR was performed by using the ChamQ SYBR qPCR Master Mix (Vazyme, Nanjing, China) or AceQ qPCR probe Master Mix (Vazyme, Nanjing, China). All readings were normalized to the level of glyceraldehyde-3-phosphate dehydrogenase mRNA. The sequences of primer sets used for qRT-PCR are shown in Tables S1 andS2.
## Statistical analysis
Statistical analyses were performed using GraphPad Prism (version 10.1.2). Data are presented as the mean ± standard deviation from at least three independent biological replicates (the specific n values are provided in the figure legends). Chi-square test, Student's t-test, and one-way ANOVA with multiple comparisons were used to deter mine significance level and are indicated in the figure legends. Statistical significance is indicated as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.
## References
1. Dick, Kitchen, Haddow (1952) "Zika virus. I. Isolations and serological specificity" *Trans R Soc Trop Med Hyg*
2. Giraldo, Gonzalez-Orozco, Rajsbaum (2023) "Pathogenesis of Zika virus infection" *Annu Rev Pathol*
3. (2025) *Full-Length Text Journal of Virology*
4. Lin, Yip, Huang et al. (2018) "Zika virus structural biology and progress in vaccine development" *Biotechnol Adv*
5. Musso, Ko, Baud (2019) "Zika virus infection -after the pandemic" *N Engl J Med*
6. Marbán-Castro, Goncé, Fumadó et al. (2021) "Zika virus infection in pregnant women and their children: a review" *Eur J Obstet Gynecol Reprod Biol*
8. Duffy, Chen, Hancock et al. (2009) "Zika virus outbreak on Yap Island, Federated States of Micronesia"
9. (2025) "Countries & territories at risk for Zika"
10. Pierson, Diamond (2018) "The emergence of Zika virus and its new clinical syndromes" *Nature*
11. Martines, Bhatnagar, Keating et al. (2015) "Notes from the field: evidence of Zika virus infection in brain and placental tissues from two congenitally infected newborns and two fetal losses--Brazil" *MMWR Morb Mortal Wkly Rep*
12. Mlakar, Korva, Tul et al. (2016) "Miscarriage associated with Zika virus infection" *N Engl J Med*
13. Hoen, Schaub, Funk et al. (2018) "Pregnancy outcomes after ZIKV infection in French territories in the Americas" *N Engl J Med*
14. Shapiro-Mendoza, Rice, Galang et al. (2016) "Pregnancy outcomes after maternal Zika virus infection during pregnancy -U.S. Territories"
15. Nogueira, Júnior, Estofolete et al. (2018) "Adverse birth outcomes associated with Zika virus exposure during pregnancy in São José do Rio Preto" *Brazil. Clin Microbiol Infect*
16. Chen, Gullett, Tweedell et al. (2023) "Innate immune inflammatory cell death: PANoptosis and PANoptosomes in host defense and disease" *Eur J Immunol*
17. Pandian, Kanneganti (2022) "PANoptosis: a unique innate immune inflammatory cell death modality" *J Immunol*
18. Place, Lee, Kanneganti (2021) "PANoptosis in microbial infection" *Curr Opin Microbiol*
19. Kesavardhana, Malireddi, Kanneganti (2020) "Caspases in cell death, inflammation, and pyroptosis" *Annu Rev Immunol*
20. Jiang, Qi, Li et al. (2021) "Caspase-8: a key protein of cross-talk signal way in "PANoptosis" in cancer" *Int J Cancer*
21. Ding, Wang, Liu et al. (2016) "Pore-forming activity and structural autoinhibition of the gasdermin family" *Nature*
22. Wang, Gao, Shi et al. (2017) "Chemotherapy drugs induce pyroptosis through caspase-3 cleavage of a gasdermin" *Nature*
23. Weinlich, Oberst, Beere et al. (2017) "Necroptosis in development, inflammation and disease" *Nat Rev Mol Cell Biol*
24. Gullett, Tweedell, Kanneganti (2022) "It's all in the PAN: crosstalk, plasticity, redundancies, switches, and interconnectedness encompassed by PANoptosis underlying the totality of cell deathassociated biological effects" *Cells*
25. Nguyen, Kanneganti (2022) "PANoptosis in viral infection: the missing puzzle piece in the cell death field" *J Mol Biol*
26. Malireddi, Karki, Sundaram et al. (2021) "Inflammatory cell death, PANoptosis, mediated by cytokines in diverse cancer lineages inhibits tumor growth" *Immunohor izons*
27. Thapa, Ingram, Ragan et al. (2016) "DAI senses influenza A virus genomic RNA and activates RIPK3-dependent cell death" *Cell Host Microbe*
28. Malireddi, Kesavardhana, Kanneganti (2019) "ZBP1 and TAK1: master regulators of NLRP3 inflammasome/pyroptosis, apoptosis, and necroptosis (PAN-optosis)" *Front Cell Infect Microbiol*
29. Zheng, Karki, Vogel et al. (2020) "Caspase-6 is a key regulator of innate immunity, inflammasome activation, and host defense" *Cell*
30. Karki, Sharma, Tuladhar et al. (2021) "Synergism of TNF-α and IFNγ triggers inflammatory cell death, tissue damage, and mortality in SARS-CoV-2 infection and cytokine shock syndromes" *Cell*
31. Rabelo, Souza, Salomão et al. (2018) "Placental inflammation and fetal injury in a Rare Zika case associated with Guillain-Barre syndrome and abortion" *Front Microbiol*
32. Yockey, Jurado, Arora et al. (2018) "Type I interferons instigate fetal demise after Zika virus infection" *Sci Immunol*
33. Miner, Cao, Govero et al. (2016) "Zika virus infection during pregnancy in mice causes placental damage and fetal demise" *Cell*
34. Bedoui, Herold, Strasser (2020) "Emerging connectivity of programmed cell death pathways and its physiological implications" *Nat Rev Mol Cell Biol*
35. Lee, Karki, Wang et al. (2021) "AIM2 forms a complex with pyrin and ZBP1 to drive PANoptosis and host defence" *Nature*
36. Zheng, Kanneganti (2020) "The regulation of the ZBP1-NLRP3 inflammasome and its implications in pyroptosis, apoptosis, and necroptosis (PANoptosis)" *Immunol Rev*
37. Paixao, Cardim, Costa et al. (2022) "Mortality from congenital Zika syndrome -nationwide cohort study in Brazil" *N Engl J Med*
38. Costa, Cardim, Teixeira et al. (2020) "Case fatality rate related to microcephaly congenital Zika syndrome and associated factors: a nationwide retrospective study in Brazil" *Viruses*
39. (2025) *Full-Length Text Journal of Virology*
40. Noronha L De, Zanluca, Azevedo et al. (2016) "Zika virus damages the human placental barrier and presents marked fetal neurotropism" *Mem Inst Oswaldo Cruz*
41. Zhao, Li, Ashraf et al. (2022) "Zika virus causes placental pyroptosis and associated adverse fetal outcomes by activating GSDME"
42. Malireddi, Kesavardhana, Karki et al. (2020) "RIPK1 distinctly regulates Yersinia-induced inflammatory cell death" *PANoptosis. Immunohorizons*
43. Fernandes-Alnemri, Wu, Yu et al. (2007) "The pyroptosome: a supramo lecular assembly of ASC dimers mediating inflammatory cell death via caspase-1 activation" *Cell Death Differ*
44. Motomura, Morita, Yamamoto et al. (2025) "Isolation of pure primary term human cytotro phoblasts and their differentiation into syncytiotrophoblast-like cells as an ex vivo model of the human placenta" *Placenta* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12152991&blobtype=pdf | # North American-origin influenza A (H10) viruses in Eurasian wild birds (2022-2024): implications for the emergence of human H10N5 virus
Jiaying Wu, Xiaoqing Zhang, Yubo Zhao, Shunyuan Zhang, Yanhai Wang, Wenxue Yang, Haizhou Liu, Jiang Feng, Wenzhuo Tan, Ke Wang, Qianqian Shi, Qichao Wei, Jianqing Sun, Yuan Zhang, Jianjun Chen
## Abstract
During our surveillance of avian influenza viruses (AIVs) in wild birds across China, H10Nx viruses were isolated from diverse migratory flyways between 2022 and 2024. We identified one wild-bird H10N5 strain that shared a common ancestor with the human H10N5 virus in multiple gene segments. Phylogenetic and molecular dating revealed the origin and evolution of H10N5, highlighting the need for continued monitoring.
In January 2024, a case of mixed human infection with H3N2 and H10N5 was reported in Zhejiang Province, China, marking the first recorded instance of human infection with avian influenza A(H10N5) [1]. Phylogenetic analysis revealed that the hemagglutinin (HA) and neuraminidase (NA) genes of the human H10N5 virus originated from wild bird influenza viruses, with the HA gene belonging to the North American lineage and the NA gene to the Eurasian lineage [2]. However, we know very little about the origin of H10N5. To date, human infections with other H10 subtypes, including H10N3, H10N7, and H10N8, have also been reported [3][4][5]. Although no evidence of human-to-human transmission has been observed, ongoing surveillance of H10 viruses remains critical. Additionally, human infections with H7N9, H7N4, and H3N8 viruses, which have genomes partially or entirely derived from wild bird AIVs, have been documented in recent years [6][7][8]. These findings highlight the ongoing potential threat posed by AIVs from wild birds, underscoring the need for continuous monitoring. Here, we report the circulation of North American lineage H10 viruses among wild birds along distinct migratory flyways in China from 2022 to 2024. Notably, one of these viruses appears to share a common ancestor with the human H10N5 strain.
Since 2013, we have conducted surveillance for AIVs among migratory birds in Qinghai Lake (Qinghai Province) and the Chenhu Wetland (Hubei Province), which are situated on different migratory flyways (Appendix Figure 1). Sampling at Qinghai Lake, conducted during the breeding season of migratory birds (May to August), yielded samples primarily in the summer (Appendix Table 1). In contrast, sampling at the Chenhu Wetland occurred during wintering season, from December to March of the following year (Appendix Table 2). Prior to 2022, no H10 subtype strains were detected at either site. However, in 2022 and 2024, H10 subtype strains were isolated from wild birds at both Qinghai Lake and the Chenhu Wetland. At Qinghai Lake, five strains of the H10N7 virus (0.6%, 5/726) were isolated in June 2022, and nine H10 subtype AIV strains (0.79%, 9/1131) were isolated in July 2023. These included H10N1 (n = 2), H10N5 (n = 3), H10N8 (n = 1), and H10N9 (n = 3). Additionally, one H10N5 virus was isolated from the Chenhu Wetland in January 2023 and another in February 2024. We obtained wholegenome sequences of these 16 viral strains using nextgeneration sequencing (Appendix Table 3).
We compared the genomic sequence similarity of the 16 H10 isolates with the human-derived H10N5 virus. The HA gene similarity among these 17 strains ranged from 80 to 100%, indicating that the HA genes belonged to different origins. The internal gene segments, including polymerase basic protein 2 (PB2), polymerase basic protein 1 (PB1), polymerase acidic (PA), matrix protein (MP), and nucleoprotein (NP), exhibited high diversity, suggesting that these genes originated from multiple sources. Notably, the HA, NA, PA, and NP genes of the human H10N5 strain displayed high nucleotide similarity with those of A/ wild bird/Hubei/YYH40/2024 (H10N5) (YYH40), a strain isolated from the Chenhu Wetland in February 2024. This finding suggests that the human H10N5 strain and the wild bird H10N5 strain (YYH40) may share a common ancestral virus.
We conducted a phylogenetic analysis of these strains alongside relevant sequences from public databases to explore the relationships between the H10 viruses identified in this study, the human H10N5 strain, and other related viruses (Appendix Figure 2). The phylogenetic tree of the HA gene divided the viruses into North American and Eurasian lineages. The human H10N5 virus clustered with the YYH40 strain on a distinct branch within the North American lineage. This branch also included five H10N7 strains isolated from Qinghai Lake in June 2022, one H10N5 strain from the Chenhu Wetland isolated in January 2023, and other wild bird H10 viruses isolated in East Asia and Bangladesh from 2019 to 2023. In contrast, the nine H10 strains isolated from Qinghai Lake in July 2023 belonged to the Eurasian lineage (Figure 1(A)). In the NA gene tree, all five H10N5 strains from this study grouped within the Eurasian lineage. Notably, two H10N5 strains from Qinghai Lake and two from the Chenhu Wetland clustered closely with the human H10N5 virus. For the internal protein genes, the 16 H10 isolates were distributed across multiple branches within the Eurasian lineage, predominantly consisting of wild bird strains. In the PA and NP gene trees, the human H10N5 virus clustered closely with the YYH40 strain. These findings suggest that North American lineage H10 subtype viruses are circulating among Eurasian migratory birds and may have contributed to the emergence of the human H10N5 strain.
Molecular dating was employed to estimate the timing of reassortment events leading to the emergence of H10N5 viruses (Appendix Figure 3, Figure 1(B)). H10 viruses of North American lineage appear to have been introduced into East Asian wild birds via the East Asia-Australasia migration flyway between December 2017 and November 2018. Subsequently, these viruses underwent reassortment with Eurasian lineage wild bird AIVs. Novel reassorted viruses circulated among wild birds on this flyway during 2018-2019 and eventually spread to the Central Asian flyway. It can be inferred that the most recent common ancestor (MRCA) of the human H10N5 virus likely emerged in wild birds in November 2023 before spilling over to humans in late 2023. The YYH40 strain, isolated in February 2024, is estimated to have an MRCA that emerged in wild birds in April 2023.
Amino acid analysis revealed that HA proteins of the H10 isolates and the human H10N5 strain possess a single basic amino acid at the cleavage site, consistent with the characteristic of low-pathogenic AIVs (Appendix Table 5). At the 220-loop binding site, there were no mutations indicative of a binding preference for α-2,6-linked human sialic acid receptors [9]. Furthermore, residues Q591, E627, and D701 in the PB2 protein suggest these H10 viruses have not yet adapted for efficient replication in mammalian hosts [10][11][12]. The non-structural 1 (NS1) proteins of these viruses contain a C-terminal ligand sequence Glu-Ser-Glu-Val (ESEV) motif [13]. Overall, these findings suggest that the risk of H10 virus infections in wild birds posing a significant threat to the general public remains low.
## Conclusions
During our active surveillance of AIVs in wild birds, we isolated seven North American lineage H10 strains in China between 2022 and 2024, which reflects intercontinental circulation of North American lineage AIVs within Asia, along with reassortment events with Eurasian lineage AIVs, leading to the emergence of the human H10N5 strain.
The crossing of North American and Eurasian lineage AIVs across continents and establishing stable circulation is rare. However, such events have become more frequent in recent years. In 2022, a novel poultry H3N8 virus was reported to infect humans in China, with the NA gene of the human H3N8 virus originating from North American lineage viruses circulating in wild birds [8]. More recently, clade 2.3.4.4b Eurasian H5N1 viruses crossed the Atlantic flyway and spread to North America by the end of 2021 [14]. This virus then reassorted with North American wild bird strains, generating genotype B3.13, which led to poultry outbreaks and spread to U.S. dairy cows over several months [15]. Given these observations, it is crucial to continue monitoring whether H10N5 or other H10Nx viruses persist in wild birds and whether they might acquire the potential to cause sporadic human infections.
## References
1. He, Gong, Chen (2023) "A retrospective investigation of a case of dual infection by Avian-Ori gin Influenza A (H10N5) and seasonal influenza A (H3N2) viruses -Anhui Province, People's Republic of China" *China CDC weekly*
2. Yuan, Zhang, Jiang (2024) "Epidemiology and evolution of human-origin H10N5 influenza virus" *One Health*
3. Dai, Zhao, Xia (2024) "Phylogenetic and mutational analysis of H10N3 avian influenza A virus in China: potential threats to human health" *Front Cell Infect Microbiol*
4. Arzey, Kirkland, Arzey (2012) "Influenza virus A (H10N7) in chickens and poultry abattoir workers" *Australia. Emerg Infect Dis*
5. To, Tsang, Chan (2014) "Emergence in China of human disease due to avian influenza A(H10N8)-cause for concern" *J Infect*
6. Wang, Chen, Wang (2022) "Serological evidence of human infection with avian influenza A(H7N9) virus: a systematic review and meta-analysis" *J Infect Dis*
7. Huo, Cui, Chen (2018) "Severe human infection with a novel avian-origin influenza A(H7N4) virus" *Science Bulletin*
8. Yang, Sun, Gao (2022) "Human infection of avian influenza A H3N8 virus and the viral origins: a descriptive study" *Lancet Microbe*
9. Vachieri, Xiong, Collins (2014) "Receptor binding by H10 influenza viruses" *Nature*
10. Bai, Lei, Song (2024) "Amino acids in the polymerase complex of shorebird-isolated H1N1 influ enza virus impact replication and host-virus interactions in mammalian models" *Emerg Microbes Infect*
11. Liu, Qi, Bao (2024) "Novel H10N3 avian influenza viruses: a potential threat to public health" *Lancet Microbe*
12. Nieto, Pozo, Vidal-García (2017) "Identification of rare PB2-D701N mutation from a patient with severe I nfluenza: contribution of the PB2-D701N mutation to the pathogenicity of human influenza" *Front Microbiol*
13. Zielecki, Semmler, Kalthoff (2010) "Virulence determinants of avian H5N1 influenza A virus in mammalian an d avian hosts: role of the C-terminal ESEV motif in the viral NS1 protein" *J Virol*
14. Kandeil, Patton, Jones (2023) "Rapid evolution of A(H5N1) influenza viruses after intercontinental sp read to North America" *Nat Commun*
15. Oguzie, Marushchak, Shittu (2024) "Avian influenza A(H5N1) virus among Dairy Cattle" *Emerg Infect Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12680181&blobtype=pdf | # Neisseria meningitidis serogroup B causing invasive disease, Italy, 2010-2021
Paola Vacca, Fenicia Vescio, Fortunato D'ancona, Cecilia Fazio, Arianna Neri, Anna Carannante, Luigina Ambrosio, Florigio Lista, Silvia Fill, Andrea Ciammaruconi, Antonella Fortunato, Paola Stefanelli
## Abstract
Data availability statement: Whole genome sequencing (WGS) data generated and analyzed in this study are publicly available in the PubMLST repository (only accession numbers are listed in S1 Table Supllementary Material). The clinical and epidemiological dataset used in this study cannot be made
## Introduction
Invasive meningococcal disease (IMD) remains a global public health problem, mainly caused by six (A, B, C, W, X and Y) of the twelve recognised Neisseria meningitidis serogroups [1]. Traditionally, N. meningitidis is associated with severe clinical manifestations such as meningitis, sepsis, or both. However, its clinical spectrum can be broader, including atypical or milder presentations ─ such as pneumonia, pericarditis, epiglottitis, urethritis, gastrointestinal symptoms, pharyngitis, and conjunctivitis─ which can be relatively rare and often unrecognized or underestimated in clinical practice [2,3].
IMD can affects individuals of all age groups and health conditions, although children younger than 4 years of age and patients with specific comorbidities and immunocompromising condition are at highest risk [1][2][3][4]. The case fatality rate (CFR) remains quite high worldwide (5-15%), and 20% of survivors may suffer from longterm disability [1][2][3][4]. Vaccination and antibiotic treatment (including chemoprophylaxis of close contacts of cases) remain the best strategies for the prevention and control of meningococcal disease [1][2][3][4].
Since the introduction of meningococcal vaccines against serogroups A, C, W, and Y, the epidemiology of IMD has changed geographically and temporally, resulting in a shift in the serogroups distribution [1][2][3][4][5]. As a consequence, N. meningitidis serogroup B (MenB) has become the main cause of IMD in several European countries, predominantly affecting infants, adolescents and young adults [1][2][3][4][5].
In Italy, meningococcal vaccination policies are defined within the National Immunization Plan (Ministero della Salute, Piano Nazionale Prevenzione Vaccinale. https://www. salute.gov.it/new/it/tema/vaccinazioni/piano-nazionale-prevenzione-vaccinale/ Last visit, 9 July 2025). Possibly, regional differences in vaccine introduction timing, target age groups, and implementation status (routine vs. recommended) need to be considered.
In 2005, MenC vaccination was introduced as recommended at national level, targeting infants and adolescents [6]. Overtime, the MenACYW conjugate vaccine has gradually replaced MenC vaccine and is following the recommendation currently in place for individuals from 12 months of age and for adolescents aged 12-18 years (https://www.salute.gov.it/new/it/tema/vaccinazioni/calendario-vaccinale/).
Regarding to serogroup B, two vaccines ─ the four component meningococcal serogroup B vaccine (4CMenB) and the bivalent factor H-binding protein vaccine (MenB-fHbp) ─ are available in Italy [7]. The 4CMenB vaccine was licensed in 2013 by both the European Medicines Agency (EMA) and the Italian Medicines Agency (AIFA) and it was progressively introduced in the regional immunisation plan [7]. According to official coverage data provided by the Italian Ministry of Health for the period corresponding to this study, vaccination coverage for MenB at 24 months of age was 66.3% in 2020 and 79.7% in 2021, respectively (https://www.salute.gov.it/ new/sites/default/files/imported/C_17_bancheDati_38_1_7_file.pdf).
Currently, the 4CMenB is recommended and actively offered free of charge to all infants under one year of age, following a four dose schedule between the 3 rd and 15 th month of life (https://www.salute.gov.it/new/it/tema/vaccinazioni/
## Competing interests:
The authors have declared that no competing interests exist. calendario-vaccinale/). The bivalent MenB-fHbp vaccine was approved by the AIFA in 2017 and is recommended for individuals aged over 10 years (https://www.aifa.gov.it/sites/default/files/DETERMINA_NP_TRUMENBA_1366.pdf).
The public health measures implemented during COVID-19 pandemic had led to a decline in the circulation of N. meningitidis, resulting in a reduction of MenB IMD incidence across the EU/EAA (from 0.3 cases per 100 000 population in 2017 to 0.08 cases per 100 000 population in 2021) [5]. However, MenB remained the predominant serogroup and accounted for 74% of IMD cases in infants aged <1 year in Europe in 2021 [8].
The study aims to describe the epidemiological trend of MenB IMD in Italy from 2010 to 2021 and to characterize the phenotypic and genotypic profiles of invasive strains.
## Materials and methods
The Italian national surveillance system for invasive meningococcal disease In Italy, IMD surveillance is part of the National Surveillance System (NSS) for Invasive Bacterial Diseases (IBD) and is coordinated by Istituto Superiore di Sanità (ISS) with the support of the Italian Ministry of Health.
The National Reference Laboratory (NRL) of the ISS receives bacterial isolates and/or biological samples (blood and cerebrospinal fluid) from laboratory-confirmed IMD cases by local laboratories, to perform serogroup identification/confirmation, antimicrobial susceptibility testing and molecular investigations.
For each IMD case, epidemiological information, including patient's information (age, clinical picture, outcome of disease, and nationality) are routinely collected. Here, data are reported and analysed in a pseudo anonymous and aggregated form. A national report is published annually and made available at https://www.iss.it/sn-mbi-rapporti-iss.
MenB cases refer to both culture and PCR-confirmed cases of invasive meningococcal diseases, collected by NRL of ISS from 2010 to 2021.
The study comprises exclusively invasive meningococcal diseases of serogroup B
## Statistical analysis
Epidemiological data of IMD cases were extracted from the MaBI platform (https://mabi.iss.it/) dedicated to the National Surveillance System of IBD. Variables included in the analysis were: age groups (<1, 1-4, 5-14, 15-24, 25-49, 50-64, ≥ 65 years), seasonality (winter, summer, spring and autumn), and geographical area of residence (North, Centre and South of Italy). Year was included in the models as a categorical variable. Missing values for serogroup were considered Missing At Random (MAR). As previously described [9], to reclassify missing serogroup values, a multinomial model were used to estimate adjusted rate ratios considering outcome, age group, geographical area, year, and seasonality as predictors.
To assess the predictive power of the multinomial model, a multiclass Area Under the Curve (AUC) was calculated (AUC = 0.79; 95% CI: 0.77, 0.81).
The number of reported cases per 100 000 inhabitants (incidence rate on a yearly basis) will be referred in the text as "incidence". It should be noted, however, that this value represents the rate of reporting to the system itself, being influenced by the proportion of cases of invasive bacterial disease in which laboratory confirmation and characterization of the etiological agent have been performed.
Crude, age-specific, and age-standardized incidences of MenB were calculated in the dataset with imputed missing data using the Eurostat European population 2010-2021 (Last access: April 8, 2024 https://ec.europa.eu/eurostat/ databrowser/product/page/TPS00001).
From 2007 to 2021 it was not mandatory to report the outcome including death of the cases to the surveillance system without obligation to update on the final outcome (i.e., after hospital discharge) as the result of a long-term follow-up and therefore an underestimate of CFR was likely occur Moreover, the serogroup was not defined for all cases.
CFR of IMD by serogroups and binomial exact 95% CI were calculated. CFRs were age-standardized using the Eurostat European population 2010-2021.
IMD cases with missing outcome data (N° = 504) were excluded from this analysis. Statistical analyses were conducted using R version 4.2.33 (Project for Statistical Computing. Last access: April 8, 2024 https://www.r-project.org/).
## Microbiological analysis
Meningococcal isolates were grown on Thayer Martin agar plate with 2% IsoVitalex (Oxoid, Ltd) incubated in 5% CO 2 at 37°C. Serogroup identification/confirmation was obtained by slide agglutination with commercial antisera (Remel Europe, Ltd, UK) or by multiplex PCR [10]. Antimicrobial susceptibility to cefotaxime, ceftriaxone, ciprofloxacin, penicillin G and rifampicin was assessed with the Minimum Inhibitory Concentration (MIC) using Etest (Biomerieux, Sweden) and MIC test strip (Liofilchem, Diagnostici, Italy) on Mueller Hinton agar plates (Oxoid, Ltd) supplemented with 5% sheep blood. Clinical breakpoints were those recommended by the European Committee Antimicrobial Susceptibility Testing (EUCAST v. 14.0) [11].
## Molecular analysis
Meningococcal DNA was extracted using QIAamp mini kit (Qiagen, Hilden, Germany), following the manufacturer's procedure. Species and genogroup identification of biological samples (blood and cerebrospinal fluid) was performed by RT-PCR using MenSerogroup kit (Diagenode, Belgium).
WGS was performed on NextSeq500 or MiSeq sequencers (Illumina, CA, USA). For each isolate 1.5 ng of DNA was used to prepare libraries using the Nextera XT DNA protocol according to the manufacturer's instructions. The High Output Kit v2 (300 cycles) and the v3 Reagent kit (600 cycles) were used for NextSeq 500 and MiSeq, respectively (Illumina, CA, USA). Quality check of the raw sequence data was performed using FastQC software [12]. High-quality bases (Q score >25) were retained and reads were trimmed using the software Sickle [13]. De novo assembly was performed with ABySS software version 1.5.2 (K parameter = 63) [14].
The de novo assembled genomes were uploaded to the BIGSdb platform (https://pubmlst.org/bigsdb), and analysed via a hierarchical gene-by-gene annotation approach. According to the designation tools included in the Neisseria PubMLST website (http://pubmlst.org/neisseria/), isolates were characterized by the finetype of two outer membrane proteins -the Variable Region (VRs) VR1 and VR2 of the porin A (PorA) and the VR of the Ferric enterobactin transport (FetA) -and by the Multilocus Sequence Type (MLST). The combination of capsular group, finetype and MLST defines the genotypic profile, as follows: capsular group: PorA (P1). VR1, VR2: FetA (F)VR: sequence type (ST) (clonal complex). Alleles of the main antimicrobial resistance target genes were also analysed: gyrA, DNA gyrase subunit A, penA, penicillin binding protein 2 (PBP2), and rpoB, RNA polymerase β chain.
Genome comparisons were performed on the available MenB genomes collected between 2012 and 2021 using the Genome Comparator tool in the PubMLST database. The resulting distance matrix was visualized as a Neighbor-Net network in Split Tree4 (version 4.13.1). Incomplete loci were automatically removed from the distance matrix calculation for the neighbour-net graphs.
For the culture-negative meningococci, PCR and Sanger sequencing were performed to determine the genotypes, as previously described [15].
The MenDeVAR index [16], a genomic-based tool (available at pubmlst.org/bigsdb?db = pubmlst_neisseria_isolates), integrates genomic information on vaccine antigens with evidence from published serological studies (MATS, MESURE, hSBA) and was used to analyze a subsample of MenB available genomes. As defined, the "exact match" category includes isolates carrying at least one antigen identical to a vaccine variant; "cross-reactive" refers to isolates containing at least one antigenic variant shown in experimental studies to cross-react with vaccine variants, and the "none" category includes variants lacking reactivity in experimental data, while "insufficient data" indicates variants for which experimental evidence is not available.
This study was conducted using data from the National Surveillance System of invasive meningococcal diseases established by law. The samples were collected for the hospitalization routine and no ethical approval was required 1, the frequency distribution by age group, geographical area, year, and seasonality was similar in the original dataset compared to those with imputed missing.
## Results
## Invasive meningococcal disease in
Fig 1 shows the difference between MenB incidence calculated using data from the original NSS dataset (here indicated as MenB IMD NSS) and the adjusted values based on the multinomial model (here indicated as MenB IMD Adj). In comparison, MenB IMD Adj values were higher than MenB IMD NSS in the first half of the study period. From 2015, the values were more similar.
During the study period, the highest MenB IMD incidence was recorded in 2019 (0.13 in 2019 per 100 000 inhabitants). In 2020, due to COVID-19 pandemic, the incidence of IMD dropped dramatically, bringing the MenB IMD incidence to 0.06 per 100 000 inhabitants ( Meningitis was the main clinical presentation (53%; 373/704). Sepsis and sepsis plus meningitis were diagnosed in 46% (321/704) and in 18% (127/704) of patients, respectively.
Other clinical presentations (e.g., pneumonia, arthritis, septic arthritis, peritonitis) occurred in 11 patients. A total of 504 IMD cases with unknown outcome were excluded from the analysis. Overall, 203 patients affected by IMD and with a confirmed serogroup died during the study period, of which 64 were due to MenB. The observed CFR for all IMD (obs. CFR) was 15.14% (95% CI: 13.55%-16.84) while for MenB IMD was 9.29% (95% CI: 7.53%-11.31) (Table 2). As reported in Table 2, the obs. CFR varied among serogroups; the highest value was reported in IMD due to MenC (26.23%; 95% CI: 7.53%-11.31) followed by MenW (21.67%; 95% CI: 13.32%-32.22), NG (11.11%; 95% CI: 2.01%-31.03), MenY (10.13%; 95% CI: 6.46%-14.97). For comparison, hereby, the CFR was age-standardized (std. CFR) using the European population. The std. CFR for all IMD was 14.95% (95% CI, 12.96%-17.15) (Table 2), slightly lower than the obs. CFR (15.14%, 95% CI: 13.55%-16.84).
For IMD caused by MenB, the std CFR was slightly higher than the obs. CFR (10.77%, 95% CI, 8.3%-13.77) (Table 2) but lower than other serogroups.
Std CFR was not calculated for Men A, MenX and NG due to the low number of cases (Table 2).
## Antimicrobial susceptibility test and characterization of target genes
Antimicrobial susceptibility was assessed on 322 culture-positive MenB isolates. All isolates were susceptible to cefixime (MIC ≤ 0.125 mg/L) and to ceftriaxone (MIC ≤ 0.125 mg/L). They were also susceptible to penicillin G (MIC < 0.25 mg/L), although 172 MenB isolates showed a MIC values close to the resistance breakpoint (MIC values ranging between ≥0.094 and ≤0.25 mg/L). Two isolates were resistant, with MIC values of 0.38 mg/L for both.
Thirty-seven penA alleles were identified, of which penA14 (N = 31), penA1 (N = 23) and penA9 (N = 21), were the most frequent. The 2 penicillin-resistant MenB isolates harboured penA9 and penA12, respectively, and were both characterized by polymorphisms in the C-terminal region of penicillin binding protein 2 (F504L, A510V, I515V, H541N, and I566V).
All isolates were susceptible to ciprofloxacin (MIC ≤ 0.03 mg/L), except for one resistant isolate from 2018 (MIC value of 0.25 mg/L). Molecular analysis of the target gene gyrA identified the following alleles as prevalent: gyrA4 (N = 90), gyrA2 (N = 41) and gyrA3 (N = 17). For the ciprofloxacin resistant isolate, the presence of the T91I amino acid substitution in the protein sequence, encoded by the gyrA gene (allele gyrA212) confirmed the antimicrobial resistant phenotype.
MenB isolates were susceptible to rifampicin (MIC < 0.25 mg/L), with the exception of one isolate from 2012, which was resistant to rifampicin (MIC value of 2 mg/L).
Of 15 rpoB alleles, rpoB28 (N = 50), rpoB18 (N = 30), and rpoB4 (N = 20) were the most frequent. The remaining alleles (rpoB1, rpoB2, rpoB7, rpoB9, rpoB31, rpoB34, rpoB38, rpoB40, rpoB72, rpoB73, rpoB85) were poorly represented.
## Molecular profiles of MenB
A total of 477 MenB were studied at molecular level. Meningococci were grouped into 20 ccs, of which the most common were: cc41/44 (N = 101), cc162 (N = 86), cc32 (N = 49), cc213 (N = 43), cc269 (N = 28), cc461 (N = 23), cc865 (N = 18), cc11 (N = 12), cc1572 (N = 7), (Table 3).
The remaining isolates belonged to 11 different ccs (cc18, cc23, cc35, cc60, cc167, cc174, cc198, cc334, cc1136, cc1157, cc4821) that were represented by one or two isolates.
For 39 MenB, ccs were not yet assigned according to the current MLST typing scheme (unknown, UNK), while for 47 MenB no genetic information could be obtained.
MenB:cc41/44 represented 31 different STs, where ST-414 (N = 12) and ST-1403 (N = 10) were the most common, followed by ST-1194 and ST-41 (N = 5 isolates each). Other STs were represented by one or two isolates. In contrast, in MenB:cc162, 9 STs were identified. ST-162 was prevalent (N = 63), the remaining STs (ST-5573, ST-8087, ST-9293, ST-9465, ST-10175, ST-10812, ST-12193, ST-13870) were represented by few isolates (Table 3).
Meningococci cc32, cc213, cc269, cc461, and cc865, mainly belonged to ST-32, ST-213, ST-269, ST-1946, and ST-3327 respectively.
Among the main hypervirulent ccs identified, cc11 was found in 12 MenB. In particular, 7 were sporadic cases that occurred between 2011 and 2019; 5 MenB:cc11 constituted an outbreak that arose in Sardinia in 2018 [18].
Although cc41/44 was the most frequent cc, its prevalence has decreased over time (Fig 3). In contrast, cc162, which accounted for 9% of MenB cases in 2010, increased to 25% in 2019, becoming the predominant cc ─ except in 2017 and 2020, when cc213 and cc32, respectively, were the most prevalent (Fig 3). A total of 134 finetypes (the combination of PorA VR1, VR2 and FetA VR) were identified, with P1.22,14:F3-6 (N° = 49) in cc162 being the most frequent. Overall, considering the high genetic variability of MenB, the most common genotype was B:P1.22,14:F3-6:ST-162(cc162), representing for 10% (50/477) of the identified genotypes.
A comparative genomic analysis was performed on 201 MenB genomes collected between 2012 and 2021 (Fig 4). The phylogenetic tree obtained by cgMLST analysis was reconstructed on the estimated allelic distances and revealed a star-like topology where strains belonging to the same ccs clustered together (Fig 4). Of note, MenB:cc11 isolates (N° = 5) associated with a putative capsule switched strain, from serogroup C to B, and responsible for an outbreak in 2018 [17], formed a single phylogenetic network of the main lineage (Fig 4).
Figs 5 and6 show the MenDeVAR analyses for 4CMenB and MenB-fHbp, respectively, performed on 201 meningococcal genomes per year and clonal complex from 2013 to 2021.
Among the different clonal complexes, cc41/44 and cc32 showed the highest proportion of genomes with an "exact match" to the vaccine variants in the MenDeVAR analysis for 4CMenB across all years. For most genomes belonging to other clonal complexes, including cc162, the MenDeVAR analysis for 4CMenB indicated "insufficient data". As previously described, cc213 was the clonal complex with no compatibility ("none" in the MenDeVAR analysis for 4CMenB) due to different antigenic variants (Fig 5).
The MenDeVAR analysis for MenB-fHbp (Fig 6) suggested that cc213 was the only clonal complex with an "exact match" to the vaccine variants. Most genomes carried cross-reactive antigenic variants, with the exception of a single fHbp-negative genome. For both analyses, "insufficient data" results were obtained for the majority of the genomes.
## Discussion
Hereby, the MenB IMD in Italy from 2010 to 2021 and the phenotypic and genotypic profiles of invasive strains are described. Italy is considered a country with a low incidence of IMD in the overall population [4,5].
MenB has been the predominant cause of IMD in Italy over the last decade, as in many other countries [5]. Exceptionally, during 2015-2016, the incidence of MenC exceeded that of MenB due to an outbreak occurred in Tuscany [18].
Since 2020, preventive measures implemented to reduce the impact of the SARS-CoV-2 pandemic have significantly contributed to the reduction in the incidence of meningococcal disease and MenB cases (0.14 per 100 000 inhabitants in 2019, 0.06 per 100 000 inhabitants in 2020, 0.02 per 100 000 inhabitants in 2021; data available on https://atlas.ecdc. europa.eu/public/index.aspx).
More recently, in 2022, the notification rate for MenB in Italy was 0.06 per 100 000 inhabitants; similar rates were observed in Finland (0.05), Germany (0.09), Greece (0.04), Hungary (0.08), Norway (0.09), Portugal (0.04), and Sweden (0.05) (https://atlas.ecdc.europa.eu/public/index.aspx).
As also reported by other countries [1][2][3][4], the incidence of MenB IMD in Italy varied across the age groups. However, data from the national surveillance system for IMD (2020-2022) [19] described an increase in the percentage of IMD cases caused by MenB in young adults, from 75% in 2021 to 84.6% in 2022 [19]. In contrast, the age group 1-4 years, which was the second most affected age group until 2020, has shown a progressive decline [19].
Regarding CFR by serogroup, a comparison between the observed and age-standardized CFRs, using the Eurostat European population, revealed that the observed CFR by serogroup was slightly higher than the age-standardized CFRs, except for MenB, which showed a lower std. CFR compared to other serogroups. This result is likely due to the differences Although most IMD cases in Italy were sporadic, several outbreaks, mostly caused by MenC, occurred in our country over the last 20 years [18,[20][21][22]. However, in 2018, an outbreak caused by a capsular switched MenB:cc11 occurred in Sardinia, for which active immunization against MenB was implemented to control the outbreak [17].
In contrast, several MenB outbreaks have been previously reported in Europe and the USA in closed and organization-based settings [23][24][25][26]. In Europe, MenB outbreaks were reported in the Republic of Ireland (2010-2013) [25], in France (2012-2013) [24], and in the United Kingdom (2016-2017) [23]. In the USA, the number of MenB outbreaks increased in the last decade [26]. From 2013 to 2018, ten university-based outbreaks caused by MenB were recorded, mainly involving university student communities and their contacts [26]. In these cases, MenB vaccines were used in response to the outbreaks, including chemoprophylaxis for close contacts of case-patients [26].
Antimicrobial susceptibility testing confirmed that MenB isolates were susceptibility to the antimicrobials used for therapy and chemoprophylaxis.
Half of the MenB isolates tested in this study showed penicillin G MIC values close to the resistance breakpoint (ranging between ≥0.094 and ≤0.25 mg/L) and according to the new EUCAST guideline [11], with the exception of 2 resistant isolates, the remaining MenB isolates are considered susceptible to penicillin G.
Consistent with findings in other countries [4], antimicrobial resistant in meningococci remains low in Italy, and associated with sporadic cases.
Molecular and phylogenetic analyses confirm the high genetic variability that distinguishes MenB from other meningococcal serogroups circulating in Italy. Overall, in this study we observed a high number of STs grouped in 20 ccs, of which 5 hypervirulent ccs (cc41/44, cc162, cc32, cc213, cc269) responsible for 65% of MenB IMD cases between 2010 and 2021. Furthermore, compared to other ccs, cc41/44 showed the highest genetic heterogeneity, with 32 STs and 26 finetypes combinations, as also suggested by the phylogenetic analysis.
Of note, molecular characterization revealed a shift in the distribution of ccs over the study period. Cc41/44 was predominant until 2014, then it decreased in favour to cc162, which has become the prevalent in the country. A slight increase of cc213 and cc865 was also observed in the last four years. This scenario was quite different when compared with other European countries where cc269, cc41/44, cc32, cc213, cc461 the main ccs responsible of endemic MenB IMD cases [27] and cc162 was less represented, accounting for only 2.5% [27]. The exception was the Greece, that similarly to our results, showed the cc162 predominant among MenB causing invasive diseases [28]. This situation appears to be the result of multiple factors that need to be taken into account and also suggest the need for further specific studies focusing on the evaluation of functional immunogenicity for this emerging ST, along with integrated analyses combining spatiotemporal data on strain distribution and vaccine coverage. Moreover, suboptimal or heterogeneous vaccine uptake, transmission dynamics, and population movements, may also have contributed to the spread of cc162.
The MenDeVAR index is used to identify potential matches with available vaccine antigens [16]. For 4CMenB, cc41/44, which was one of the most frequent clonal complexes, and cc32 showed the highest proportion of genomes with an "exact match."
However, it is important to highlight that the MenDeVAR results need to be interpreted in the context of the temporal fluctuations in the prevalence of clonal complexes observed over the years, particularly for cc41/44 and cc162, wich represented the prevalent clonal complexes observed during the study period.
The MenDeVAR analysis for MenB-fHbp indicated that cc213 showed the best correspondence with the vaccine antigens. Both MenDeVAR analyses classified the majority of genomes as having "insufficient data." Overall, these results do not allow a clear definition of the exact match between genomes and vaccine antigens, particularly for the most recently obtained isolates.
Nevertheless, the possibility of evaluating these data in silico may be useful for gathering further scientific evidence, especially for genomes classified as "cross-reactive" or "insufficient data."
From this perspective, strengthening genomic analyses within surveillance systems is essential. Some limits of the study should be mentioned. Firstly, the data here presented are based on reports from Regions and Autonomous Provinces (PA) submitted to the MaBI platform for invasive bacterial diseases voluntary. Information on individual vaccination status in the MaBi platform is not often systematically reported. Secondly, the surveillance system only includes laboratory-confirmed IMD cases, and molecular characterization was performed on most, but not all, meningococcal isolates. As a result, genomic information was available only for a subset of IMD cases. Finally, the outcome of each IMD case is mandatory since 2022 and without obligation to update on the final outcome (i.e., after hospital discharge) as the result of a long-term follow-up. This could lead to an underestimation of CFR by serogroup.
In conclusion, MenB remained the predominant cause of IMD in Italy during the study period. Molecular investigations of genotypic profiles highlight the dynamic nature and genetic diversity of this serogroup, which continues to evolve. Molecular monitoring is essential to identify the emergence of new variants with specific virulence characteristics.
## References
1. Epidemiology, Health, Pescara et al. "Caterina Vocale, Microbiology Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna; Marina Busetti and Roberta Fais, Microbiology Unit"
2. Camporese, Microbiology, Department
3. Vulcano, Fontana, Scognamiglio et al. "Francesca Crea and Erika Coppo"
4. Lombardi, Seremi, Al ; G. D'alessandro "Regional reference Centre for Infectious Disease Surveillance" *Internal Medicine of Excellence*
5. Palermo, Gian, Rossolini et al. ") for the collaboration in the National Surveillance System of Invasive Bacterial Diseases and for the data collection at Istituto Superiore di Sanità"
6. Parikh, Campbell, Bettinger et al. (2020) "The everchanging epidemiology of meningococcal disease worldwide and the potential for prevention through vaccination" *J Infect*
7. Bai, Borrow, Bukovski et al. (2019) "Prevention and control of meningococcal disease: Updates from the Global Meningococcal Initiative in Eastern Europe" *J Infect*
8. Bobde, Sohn, Bekkat-Berkani et al. (2024) "The Diverse Spectrum of Invasive Meningococcal Disease in Pediatric and Adolescent Patients: Narrative Review of Cases and Case Series" *Infect Dis Ther*
9. Acevedo, Bai, Borrow et al. (2019) "The Global Meningococcal Initiative meeting on prevention of meningococcal disease worldwide: Epidemiology, surveillance, hypervirulent strains, antibiotic resistance and high-risk populations" *Expert Rev Vaccines*
10. Pardo De Santayana, Tin Htar, Findlow et al. (2023) "Epidemiology of invasive meningococcal disease worldwide from 2010-2019: a literature review" *Epidemiol Infect*
11. Vacca, Fazio, Neri et al. (2024) "Antimicrobial susceptibility profiles and genotyping of Neisseria meningitidis of serogroup C, Italy, 2000-2020" *Front Microbiol*
12. Palmieri, Moscara, Tafuri et al. (2024) "Policies for the immunization against serogroup B meningococcus for adolescents immunized during the first two years of life: A mini review" *Hum Vaccin Immunother*
13. (2023) "Invasive meningococcal disease: Annual epidemiological report for 2021. Stockholm: ECDC"
14. Pezzotti, Bellino, Riccardo et al. (2007) "Vaccine preventable invasive bacterial diseases in Italy: A comparison between the national surveillance system and recorded hospitalizations" *Vaccine*
15. Zhu, Wang, Wen et al. (2012) "Development of a multiplex PCR assay for detection and genogrouping of Neisseria meningitidis" *J Clin Microbiol*
16. (2024) "European Committee on Antimicrobial Susceptibility Testing. Breakpoint tables for interpretation of MICs and zone diameters"
17. Andrews, Fastqc (2010) "A quality control tool for high throughput sequence data"
18. Joshi, Fass (2011) "Sickle: a sliding-window, adaptive, quality-based trimming tool for FastQ files"
19. Simpson, Wong, Jackman et al. (2009) "ABySS: a parallel assembler for short read sequence data" *Genome Res*
20. Birtles, Hardy, Gray et al. (2005) "Multilocus sequence typing of Neisseria meningitidis directly from clinical samples and application of the method to the investigation of meningococcal disease case clusters" *J Clin Microbiol*
21. Rodrigues, Jolley, Smith et al. (2020) "Meningococcal Deduced Vaccine Antigen Reactivity (MenDe-VAR) Index: a Rapid and Accessible Tool That Exploits Genomic Data in Public Health and Clinical Microbiology Applications" *J Clin Microbiol*
22. Stefanelli, Fazio, Vacca et al. (2019) "An outbreak of severe invasive meningococcal disease due to a capsular switched Neisseria meningitidis hypervirulent strain B:cc11" *Clin Microbiol Infect*
23. Stefanelli, Miglietta, Pezzotti et al. (2015) "Increased incidence of invasive meningococcal disease of serogroup C/clonal complex 11" *Euro Surveill*
24. Superiore, Sanità (2023) "Sorveglianza nazionale delle malattie batteriche invasive"
25. Fazio, Neri, Tonino et al. (2007) "Characterisation of Neisseria meningitidis C strains causing two clusters in the north of Italy in" *Euro Surveill*
26. Stefanelli, Fazio, Neri et al. (2012) "Cluster of invasive Neisseria meningitidis infections on a cruise ship" *Euro Surveill*
27. Fazio, Daprai, Neri et al. (2019) "Reactive vaccination as control strategy for an outbreak of invasive meningococcal disease caused by Neisseria meningitidis C:P1.5-1,10-8:F3-6:ST-11(cc11)" *Euro Surveill*
28. Clark, Lucidarme, Angel et al. (2019) "Outbreak strain characterisation and pharyngeal carriage detection following a protracted group B meningococcal outbreak in adolescents in South-West England" *Sci Rep*
29. Thabuis, Tararbit, Taha et al. (2016) "Community outbreak of serogroup B invasive meningococcal disease" *Euro Surveill*
30. Mulhall, Brehony, Connor et al. (2016) "Resolution of a Protracted Serogroup B Meningococcal Outbreak with Whole-Genome Sequencing Shows Interspecies Genetic Transfer" *J Clin Microbiol*
31. Hao, Holden, Wang et al. (2018) "Distinct evolutionary patterns of Neisseria meningitidis serogroup B disease outbreaks at two universities in the USA" *Microb Genom*
32. Vogel, Taha, Vazquez et al. (2013) "Predicted strain coverage of a meningococcal multicomponent vaccine (4CMenB) in Europe: a qualitative and quantitative assessment" *Lancet Infect Dis*
33. Tzanakaki, Hong, Kesanopoulos et al. (2014) "Diversity of Greek meningococcal serogroup B isolates and estimated coverage of the 4CMenB meningococcal vaccine" *BMC Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC11929253&blobtype=pdf | Ritu Gaur
## Virology Journal
Correction: C-28 linker length modulates the activity of second-generation HIV-1 maturation inhibitors In this article [1], the given and family names of Yuvraj KC were incorrectly structured as K.C. Yuvraj. The name was displayed correctly in all versions at the time of publication.
The original article has been corrected.
$$K. C.$$
## References
1. Yuvraj, Singh, Datta (2025) "C-28 linker length modulates the activity of second-generation HIV-1 maturation inhibitors" *Virol J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548442&blobtype=pdf | # Positive correlation between structural disorder of the HIV-1 Gag N-terminal segment and progeny virus particle formation
Takaaki Koma, Osamu Kotani, Keisuke Kamba, Kei Miyakawa, Taiki Morita, Hiromi Nakamura, Masaru Yokoyama, Naoya Doi, Tomoyuki Kondo, Takashi Nagata, Akihide Ryo, Akio Adachi, Akira Ono, Masato Katahira, Masako Nomaguchi, Hironori Sato
## Abstract
A class of unstructured peptide segments termed disordered regions plays a crucial role in the regulation of protein structure and function. Although the N-terminal region of the matrix (MA) domain of the HIV-1 Gag precursor protein is composed of an unstructured peptide, the mechanisms underlying the regulation of the unstruc tured state relating to control of viral phenotypes remain unclarified. We examined, in association, the structural, evolutionary, and biological roles of the MA N-terminal region via mutagenesis. Molecular dynamics simulation of a full-length Gag dimer model suggested that an amino acid residue at position 9 in the MA N-terminal region (MA-9) participates in the Gag dimerization. Information entropy analysis indicated that the MA-9 residue is variable in nature, but the hydrophobic amino acid substitution is evolutionarily maladaptive. Disordered region prediction study suggested that single hydrophobic amino acid substitutions at MA-9 reduce the disordered state of the MA N-terminal region. Consistently, NMR analysis indicated that such substitution reduces motional dynamics of the MA N-terminal region and alters the conformation of the MA domain. A site-directed mutagenesis study showed that hydrophobic amino acid substitutions at the MA-9 residue impair, to different degrees, the elementary and overall processes of virus particle formation in the cells. Importantly, the level of the virus particle formation was positively correlated with the level of disorder of the MA N-terminal region. These results indicate that the maintenance of structural disorder and dynamics of the Gag N-terminal segment is regulated by the MA-9 residue and critical for maintaining the optimal production of HIV-1 particles.IMPORTANCE A class of unstructured peptide segments termed disordered regions plays crucial roles in the regulation of protein structure and function. Although HIV-1 Gag precursor protein has multiple disordered elements, molecular mechanisms underlying regulation of unstructured state and viral phenotypes largely remain elusive. In this study, by analyzing in association the structural, evolutionary, and virological roles of the disordered N-terminal region of the HIV-1 Gag protein, we show that an amino acid residue at position 9 of the Gag is able to modulate the N-terminal disordered state, and the level of disorder of the Gag N-terminal region is positively correlated with the level of virus particle formation. Our findings gain new insights into molecular mechanisms of regulation of Gag structure and highlight the importance of a previously unappreciated survival strategy of HIV-1-namely, preservation of the Gag N-terminal disorder.
proteins with high plasticity. The disordered segment lacks a defined three-dimensional structure, adopts versatile conformations under physiological conditions, and yet plays roles in protein functions, which have given rise to a new disorder-function paradigm in protein science (1)(2)(3)(4)(5)(6)(7) as a counterpart to the classical structure-function paradigm. Studies using genome sequence databases indicate that over 40% of the proteins in any eukaryotic proteome include disordered regions (8)(9)(10). The abundance of these unstructured components in proteins would enlarge the conformational landscape of proteins and contribute to the plasticity in protein interaction networks and the fitness of organisms. Although the disordered segments are implicated in some human diseases (11), it remains largely elusive how these segments regulate the fitness of an organism in nature.
Viral proteins are often rich in intrinsically disordered regions. Indeed, in some viruses, disordered residues account for as much as 77.3% of individual viral proteins (12). Accumulating evidence indicates that the disordered regions play a critical role in viral replication in the cells by inducing liquid-liquid phase separation for the effective protein interactions (13). The disordered segments of viral proteins may also play a role in the conformational diversity of the viral proteins. Such a role would be particularly relevant to the highly mutable RNA viruses that exhibit exceptionally quick replication and high adaptability using only small amounts of genetic information.
The Gag precursor protein of HIV-1, the major epidemic variant type of HIV, is a multifunctional structural protein that plays a central role in the late stages of viral replication in infected cells (14). Gag is composed of four domains: matrix (MA), capsid (CA), nucleocapsid (NC), and p6. Each domain of Gag is connected with the disordered linkers and has multiple functions to generate infectious viral particles (14). Unstructured segments also exist within individual Gag domains, such as MA, CA, and NC, as well as at the N and C termini of Gag. A disordered segment in CA functions to allosterically regulate the CA structure for effective assembly of Gag and production of infectious virus particles (15,16). Thus, like many viral proteins, the HIV-1 Gag protein is rich in the disordered regions that are thought to play a regulatory role for viral replication fitness in cells.
The HIV-1 Gag MA domain functions both in the cytoplasm and plasma membrane for the successful production of infectious viral particles (14). The first critical step involving the MA domain is the specific targeting of the newly translated Gag precursor to the plasma membrane. This process is achieved by the interaction between the Gag MA domain and cellular tRNA; tRNA temporally binds to the highly basic region (HBR) near the N-terminus of MA in the cells and is thought to prevent mis-trapping of the Gag precursor in the intracellular vesicle membranes (17)(18)(19)(20). The second critical step involving the MA domain is the formation of an assembly platform to promote Gag-mediated interactions of viral components for producing infectious virus particles in the plasma membrane. This step starts with the anchoring of the Gag precursor in the plasma membrane via a myristic acid moiety covalently attached to the N-terminal glycine residue (21)(22)(23)(24) and HBR (25)(26)(27)(28)(29) of the MA domain. The Gag anchoring to the plasma membrane accelerates Gag-Gag interactions, leading to the formation of the Gag lattice and subsequent membrane curvature (14). The N-terminus of the MA domain again plays a role in inter-Gag interactions and conformational shift during maturation of nascent virus particles (30). In addition, amino acid residues near the MA N-terminus also participate in the effective incorporation of HIV-1 envelope glycoproteins (Env) into the virions (21,26,31). Thus, the N-terminus of MA and its neighboring region fulfill multiple functions at distinct steps during the generation of infectious virus particles in the cells. However, how such multiple functions are regulated remains elusive.
In this study, we revisited the role of the N-terminal segment of HIV-1 Gag pro tein from the viewpoint of the disorder-function paradigm. During our structural and sequence studies on the Gag protein, we noticed the structural, functional, and evolutionary importance of the amino acid residue at position 9 of the N-terminal region of the Gag protein, termed MA-9. We explored the impact of single substitutions at the MA-9 residue on the MA structure and progeny virus particle formation. The obtained data illustrate a hitherto unappreciated mechanism adopted by HIV-1 to optimize progeny virus production in the cells-namely, preservation of the Gag N-terminal disorder.
## RESULTS
## Molecular modeling and structural characterization of the full-length Gag precursor protein of HIV-1
The HIV-1 Gag precursor protein is a multifunctional structural protein that interacts with viral and cellular factors to produce infectious virus particles. The Gag protein has multiple unstructured peptide segments with high disorder scores throughout its structure (Fig. 1A; see Fig. S1 for the disorder scores of the N-terminal end, a region focused in this study), and thereby the overall structure of the full-length Gag protein remains unavailable to date. To gain insights into the overall conformation and the motional dynamics of the Gag protein in solution, we constructed a full-length Gag monomer of the HIV-1 NL4-3 infectious molecular clone as described in Materials and Methods and then subjected this model to molecular dynamics (MD) simulations (Fig. 1B). The root-mean-square deviation (RMSD) (32) between the initial model structure and the structures at given time points of the MD simulation increased rapidly at first and then reached a near plateau with relatively extensive structural fluctuations after 100 ns of the MD simulations (Fig. 1C). Notably, the RMSDs of individual subdomains of Gag did not show such extensive fluctuations after 100 ns of the MD simulations (Fig. S2; MA, CA, SP1, NC, spacer peptide 2 [SP2], and p6). We further examined the RMSDs of multi-domains of Gag monomer. Interestingly, RMSDs of MA-CA and MA-CA-SP1 regions reached a plateau within 200 ns of MD simulations, whereas those of MA-CA-SP1-NC and MA-CA-SP1-NC-SP2 regions were continuously increasing (Fig. S2; MA-CA, MA-CA-SP, MA-CA-SP1-NC, and MA-CA-SP1-NC-SP2). These results suggest that the full-length Gag monomer has an intrinsic property to fluctuate in solution, and the Gag C-terminal region containing NC-SP2 region is critical in causing the continuous fluctuation of HIV-1 Gag monomer in solution.
During the MD simulations, the Gag protein had adopted a more compact conforma tion as compared with the initial model before the simulation; the MA domain bended to reside in close proximity to the CA domain, and the SP1, NC, SP2, and p6 domains were compactly gathered together (Fig. 1D). Consistent with the general principles of the protein folding in aqueous solution (39,40), the areas of water-accessible surfa ces decreased after MD simulations in all of the Gag domains (Fig. 1E). Furthermore, distances of N-terminal-end-to-C-terminal-end were also reduced in the full-length Gag and all of the Gag domains examined, providing quantitative evidence for the compac tion of the Gag monomer during folding in aqueous solution (Fig. S3). Moreover, the overall three-dimensional shape of the Gag model at 200 ns of MD simulations showed striking consistency with the one obtained by the small-angle neutron scattering and hydrodynamic studies with a Gag protein variant lacking the p6 domain (41). The hydrodynamic study, including size-exclusion chromatography and quasi-elastic light scattering analysis, indicated that the hydrodynamic radius (R h ) value of Gag protein with W184A and M185A mutations was 38 Å in solution (41). The sedimentation velocity analysis also indicated a similar R h of 41 Å for the Gag protein in solution (41). These R h values were consistent with our results: the distance between N-and C-termini of the full-length Gag monomer model at 200 ns of MD simulation (45 Å). Together, these results support the conclusion that HIV-1 Gag monomer folds into compact conforma tions in aqueous solution.
Notably, the Gag protein was folded during the MD simulations so that the known Gag interaction surfaces for the later steps of HIV-1 particle formation tended to be less exposed (Fig. 1F). For example, the HBR of the MA domain, which interacts with tRNA for the specific targeting of Gag to the plasma membrane (17)(18)(19)(20), remained fully exposed after MD simulation. In contrast, interaction surfaces that were expected to function at a later stage of virion production, such as the CA-SP1 and CA NTD for the Gag lattice assembly, the zinc finger domain for the genome RNA packaging, and p6 PTAP for the virion budding, tended to be less exposed after MD simulation. The results imply that, in addition to the requirements of appropriate trans-acting factors around the Gag protein, the initial conformation of the unliganded Gag may play a role in determining the order of interactions. The results shown in Fig. 1C through F were reproducible with 10 independent MD simulations under the same conditions, suggesting an intrinsically flexible conformation of the HIV-1 Gag monomer in aqueous solution. Taken together, these results show that our MD-simulation-based Gag monomer model is reasonable from the viewpoint of protein chemistry and HIV-1 virology.
## Molecular modeling and structural characterization of the HIV-1 Gag dimer
HIV-1 Gag-Gag interaction is one of the starting points of the virion formation. To gain insights into the molecular interactions during the Gag dimerization, we constructed a Gag dimer model as described in Materials and Methods and subjected it to MD simulations (Fig. 2A). The RMSDs between the initial dimer structure and the dimers during the MD simulation increased after the onset of the MD simulations and reached a near plateau after 200 ns (Fig. 2B). In contrast to the Gag monomer, the areas of water-accessible surfaces have not decreased during MD simulation: The surface area of the Gag dimer at 1 ns and 1,000 ns was shown to be 54,551.695 Å 2 and 54,789.055 Å 2 , respectively. These results suggest that significant molecular compaction does not occur during Gag dimerization. Consistent with this data, 5-10 of the hydrogen bonds were rapidly formed between the neighboring monomers and maintained thereafter over 1,000 ns of MD simulations (Fig. 2C). We found three hydrophobic areas that could contribute to dimer formation with hydrophobic interactions along the dimerization interface (Fig. 2D). These three interfaces for the dimerization included the known regions critical for the Gag-Gag interactions, such as the helix-helix interaction between two CA domains (42,43) and between SP1-NC regions (44). These results suggest that the Gag monomer has an intrinsic conformation that forms a dimer to bury the exposed hydrophobic surfaces of the monomer, once the binding interfaces are arranged in position in close proximity to each other. Notably, the full-length Gag dimer model disclosed a previously undescribed interaction in addition to the reported ones-i.e., an attractive interaction between the MA and CA domains (Fig. 2E). Serine 9 at the N-terminal disordered region of MA (MA-S9) and glutamine 287 adjacent to the helices 2 and 7 of CA (CA-Q287) had repeatedly formed a hydrogen bond during the MD simulations. The frequency of h-bond formation between MA-S9 and CA-Q287 was 20.19% during 1,000 ns of MD simulation. The hydrogen bond disappeared during the MD simulation when serine 9 was substituted with phenylalanine, suggesting that the MA9 residue plays a role in the adjacency of the two Gag monomers. The hydrogen bond was formed at a site where steric hindrance would not occur by the myristoyl group at glycine 2 of the Gag MA domain (46) (Fig. S4). These results suggest that the MA-9 residue is able to participate in the Gag dimerization. The above results, discussed for Fig. 2B through E, were reproducible with four independent MD simulations under the same conditions, suggesting an intrinsically stable conformation of the HIV-1 Gag dimer in aqueous solution.
## Variation of the amino acid residue at position 9 in the Gag MA domain
The MA-S9-mediated interaction for the Gag dimerization is unique in that it involved an amino acid in the N-terminal disordered region of MA; thus far, the reported interactions for the Gag dimerization have occurred in residues of the structured helices, i.e., between helix 1 and helix 3 of adjacent MAs, helix 9 and helix 9 of adjacent CA CTD s, and helix 12 and helix 12 of adjacent CA-SP1s (30,42,44,47,48). To clarify the variability of the MA-9 residue in vivo, we calculated the Shannon entropy of the individual residues in the MA N-terminus (positions 1-10) using the public HIV sequence database (https:// www.hiv.lanl.gov/content/index; Fig. 3A). The amino acid residues of the N-terminal simulations. The trajectory files during 1,000 ns of MD simulations were used to calculate the number of hydrogen bonds between the two Gag precursors using the cpptraj module in AmberTools 16 (32). (D) Distribution of hydrophobic patches in the binding interface of the two Gag precursors. Hydrophobic patches with a minimum area of 50 Å 2 for protein-protein interactions were estimated using the Protein Patch Analyzer tool in MOE as described previously (15,38). (E) Visualization of contact sites in the two Gag precursors (red and blue stick models). Panel E depicts the same structure as in panel D but rotated 180°. region of Gag-MA were usually highly conserved, but some variations were observed at positions 7 and 9 (Fig. 3A). Among 20 amino acid residues, serine was most abundant at the MA-9 residue, representing about 69% of the Gag sequences (Fig. 3B). Arginine was the next most abundant, representing about 24%. Threonine and lysine were detected as minor populations, representing about a few percentages of the sequences. The number of Gag sequences having other amino acid residues at position 9 was negligi ble, suggesting induction of severe defects in HIV-1 maintenance in humans. Notably, variants having a hydrophobic amino acid residue at MA-9 were rarely detected in the sequence database, indicating a strong selective disadvantage in human populations. These results indicated that the MA-9 residue is variable, but the hydrophobic amino acid substitution is evolutionarily maladaptive.
## Prediction of effects of single substitutions at MA-9 on the disordered potential of the MA N-terminal region
Intrinsically disordered segments of proteins generally comprise an insufficient proportion of hydrophobic amino acids to prevent peptide folding (6,50). Therefore, it is possible that the hydrophobic amino acid substitutions at the MA-9 residue, which were rarely detected in the HIV-1 sequence database, influence the potential of the unstructured state of the MA N-terminal region of the Gag monomer. To address this issue, we examined the effects of a single amino acid substitution at MA-9 on the disordered state of the Gag MA N-terminus (Fig. 4). The disordered levels of individual amino acid residues were estimated with AIUPred (51), a meta-predictor of intrinsically disordered amino acids. The dashed lines at the Y axis of the figures are the threshold lines for disordered/structured residues. The prediction of disorder levels showed that the disorder scores were near 0.5 in the first nine amino acids, higher than the scores in the region downstream of the MA-9 residue (Fig. 4, wild type [WT]). These results are consistent with those of a previous NMR-based analysis, in which the segment containing the first nine amino acids was found to be disordered (46).
Notably, aromatic substitutions at MA-9 (S9F, S9Y, and S9W) all induced drastic reductions in the disorder scores of the MA N-terminal region (Fig. 4A). Similarly, hydrophobic amino acid substitutions often reduced the disordered state of the MA N-terminus (Fig. 4B). In contrast, changes in the disorder scores were mild with hydro philic amino acid substitutions, including acidic/basic amino acid substitutions (Fig. 4C through E); indeed, with the exception of the substitution to cysteine, these scores were similar to the WT score. The amino acid type-dependent changes in the disorder profiles were reproducible with another predictor of disordered regions, the Predictor of Natural Disordered Protein Regions (PONDR; Fig. S5) (33)(34)(35). These data suggest that the N-terminus of the Gag proteins becomes more structured compared to that of the WT when a hydrophobic amino acid substitution occurs at the MA-9 residue.
## Dynamic properties of the N-terminal regions of Gag-MA and Gag-MA-S9F
The structural dynamics of proteins in solution play key roles in molecular interactions (52)(53)(54). Accordingly, we next used NMR to examine the potential role of MA-9 in the regulation of dynamic properties of the N-terminal region of the MA domain. For this purpose, we compared the signal pattern in the two-dimensional (2D) 1 H-15 N HSQC spectrum of 15 N-labeled Gag MA 6His between HIV-1 NL4-3 MAs from the WT and those from the MA-S9F mutant, which was entirely absent from the HIV-1 Gag sequence database (Fig. 3B) and exhibited a reduction in the disordered state of the Gag MA N-terminus (Fig. 4A).
The signal pattern in the 2D 1 H-15 N HSQC spectrum of 15 N-labeled Gag-MA 6His closely resembled that reported by Massiah et al. (Fig. 5A) (55). The assignment of the 2D 1 H- 15 N HSQC spectrum of 15 N-labeled Gag-MA 6His was performed based on Massiah's report provided at pH 8.0 and a series of 2D 1 H- 15 N HSQC spectra obtained under different pH conditions. Signal assignment was achieved for 58% of the spectrum.
Then, we compared the 2D 1 H- 15 N HSQC spectra of 15 N-labeled Gag-MA 6His and 15 N-labeled Gag-MA-S9F 6His (Fig. 5B). The signals of many residues did not change upon the S9F mutation, indicating that this mutant has a similar, if not identical, structure to the WT. However, the signals from residues S6, G10, G11, E12, L13, D14, K15, W16, I19, L21, E52, E55, L85, C87, H89, R91, and V94 displayed some chemical shift changes. We mapped these affected residues onto the previously reported structure of Gag MA (PDB ID: 2H3F) (46) (Fig. 5C). Residues G10, G11, E12, L13, D14, K15, W16, I19, and L21 are within helix I (S9-E17). Considering that S9 in the WT forms the N-terminal cap in helix I, the S9F substitution might have perturbed the conformation of helix I. In addition, residues E52, E55, L85, C87, H89, R91, and V94 were found to be in close proximity in three-dimensional space to S9 (Fig. 5C). The earlier structure indicated a hydrogen-bond ing network involving S9, E12, and H89 (46). Hence, the S9F mutation might have also affected the conformation of the C-terminal region of helix V, to which H89 belongs.
We deduced backbone 1 H- 15 N heteronuclear NOEs to inspect the dynamic properties of Gag-MA and Gag-MA-S9F (Fig. 5D andE). The values of 1 H- 15 N heteronuclear NOE for the residues in α-helical regions of Gag-MA were ~0.8, indicating structural rigidity in these α-helices. The NOE values of Gag-MA in the regions connecting these α-helices ranged between ~0.6 and 0.8, suggesting fast internal motions in these regions. The NOE values of residues in the N-and C-terminal regions were below 0.6, indicating disorder in these regions.
We compared the NOE values of the N-terminal residues A3-L8 between Gag-MA and Gag-MA-S9F. The NOE values of residues A3, A5, S6, and L8 were found to be larger for Gag-MA-S9F than for Gag-MA. This suggests that the internal motion of the disordered N-terminal region was suppressed upon the S9F substitution to some extent. We hypothesize that the motional suppression in the disordered N-terminal region is a result of the bulkiness of F9, which could lead to a hydrophobic interaction involving F9 and hydrophobic residues on the molecular surface (Fig. 5F). Together, these results suggest that the hydrophobic amino acid substitution at MA-9 can reduce motional dynamics of the MA N-terminus and influence the conformation of the MA domain.
## Virion production is drastically reduced by substitutions of Gag-MA-S9 with amino acid residues having aromatic and hydrophobic properties
To investigate the biological role of the MA-9 residue in the HIV-1 life cycle, we conduc ted site-directed mutagenesis using the HIV-1 NL4-3 proviral clone. Mutations at position 9 in Gag-MA (S9R/P/G/H/A/E/I/L/Y/F/W) were selected based on their frequency in the HIV sequence database and the chemical properties of the amino acids (Fig. 3B). Given that the Gag-MA S9 residue can be important for Gag dimerization, virion production is expected to be influenced by mutations at the site. Two well-known virion productiondeficient mutant clones were generated and used as controls. One is a Gag-MA-G2A mutant clone that lacks Gag plasma membrane-targeting activity via a defect in the N-terminal myristoylation site of Gag. The other is a Gag-CA-WM184/185AA (WMAA) mutant clone lacking Gag dimerization ability due to mutations within the Gag-CA C-terminal domain (CTD) (20,(56)(57)(58).
We first examined the effect of Gag-MA S9 mutations on virion production and viral infectivity. Virion production was monitored by measuring Gag-p24 levels in the supernatants from HEK293T cells transfected with proviral clones. Viral infectivity was assessed by measuring the luciferase activity in TZM-bl cells infected with various mutant viruses. As expected from previous reports (14,59), the control mutants (Gag-MA-G2A and Gag-CA-WMAA) displayed a drastic reduction in both virion production and viral infectivity as compared with the WT NL4-3 (Fig. 6A).
Notably, a majority of the amino acid substitutions at position 9 influenced the ability of the virus to produce virus particles and/or infectious virus particles in a manner that was dependent on the type of amino acid residue substituted (Fig. 6A). Among the S9 mutants tested, the effects of aromatic substitutions at position 9 (Gag-MA-S9Y/F/W) were prominent and were associated with a marked reduction both in virion production and viral infectivity to levels comparable to those of the control mutants (Gag-MA-G2A and Gag-CA-WMAA). In contrast, the effects of S9R substitution were relatively mild, with the levels of both virion production and viral infectivity of the Gag-MA-S9R mutant being comparable with those of the WT. Interestingly, S9T and S9P mutants retained a capacity
## Identification of the positive correlation between the disorder level of the HIV-1 Gag MA domain N-terminus and the level of HIV-1 particle production
We next examined the relation between the levels of the virus particle production and the level of disorder of the Gag MA N-terminus. To assess this issue, the mean disorder score of the first nine amino acid residues of the Gag MA N-terminus, which form the N-terminal unstructured cap upstream of helix I (Fig. 5C), was calculated for individual MA-9 mutants, and the relation between this score and the relative viral particle production was examined. We found that these structural and biological values correlated positively with good reproducibility when the disorder scores were calcula ted with two distinct predictors of the disordered state (Fig. 6B, left and right panels; Pearson's correlation coefficient R = 0.87 and 0.78 for AIUPred [51] and PONDR VL-XT [33][34][35], respectively). The positive correlation was also reproducible when the mean disorder scores were calculated with the first 20 and 30 amino acid residues (R = 0.88 and 0.78 for AIUPred and PONDR VL-XT, respectively). These results suggest that the disordered state of the Gag N-terminus plays a key role in the optimal production of HIV-1 particles. Such a positive correlation was less evident when the viral infectivity was used for the counterpart of the correlation analysis (Fig. 6C; Pearson's correlation coefficient R = 0.48 for both the AIUPred-and PONDR VL-XT-mediated calculations of the disorder scores). We found no correlation between the detection frequency of a particular amino acid residue at the MA-9 site in the sequence database (Fig. 3B) and the Gag MA N-terminal disorder (Fig. 4; Pearson correlation coefficient R of -0.1865 and -0.0979 for AIUPred and PONDR VL-XT prediction, respectively). These results suggest that additional factor(s) are involved and should be considered when establishing the fitness of HIV-1 in vivo.
## Gag-MA-S9T/P mutations prevent the Env incorporation into virions
We next examined the molecular mechanisms by which single substitutions at the MA-9 site impaired the production of infectious virus particles. The S9T and S9P mutants exhibited greater than 60% of the progeny virion production of the WT. Notably, however, the infectivities of the produced virions were markedly impaired to less than 20% of those in the WT (Fig. 6A). To gain insights into the mechanisms underlying these findings, we examined whether the decrease in infectivity of Gag-MA-S9T/P clones is associated with their Env incorporation into virions (Fig. 7). HEK293T cells were transfec ted with several proviral clones, and on day 2 post-transfection, virions were collected by ultracentrifugation for analysis by Western blotting using anti-Gag-p24 and anti-Env antibodies. The other Gag-MA-S9 mutant clones that exhibited more than 50% infectivity to NL4-3 (S9R/A/I/G/E; Fig. 4) were selected for comparative analysis.
As shown in Fig. 7A andB, the Env incorporation into virions for Gag-MA-S9T/P was reduced to around 30% of that for NL4-3. Such a marked reduction was not observed with Gag-MA-S9R/A/I/G/E mutants, which retained levels of Env incorporation of more than 70% relative to NL4-3. The Env incorporation into virions in the Gag-MA mutants tested was well correlated with their infectivity (Fig. 7C; Pearson's correlation coefficient, R = 0.90). These results indicate that the Gag-MA-S9T/P mutations reduced viral infectivity likely by impairing Env incorporation into virions.
HIV-1 Env incorporation into virions is assumed to occur during or after Gag lattice formation in the plasma membrane. MA-S9 is located at the interface between two MA trimers in the Gag lattice (Fig. 7D). In our experiments, MA-S9 residues on one MA monomer interacted with serine 6, serine 9, and glutamic acid 52 on the neighboring MA monomer by forming hydrogen bonds (Fig. 7E). This implies that the MA-9 residue can participate in the intermolecular interactions during self-assembly of Gag. Collectively, the above results suggest that MA-S9 may be involved in the regulation of interactions between Gag and Env or other trans-acting factors for the Env incorporation into virions.
## Gag-MA mutations that dramatically decrease virion production impair the Gag oligomerization ability
Since certain mutations of Gag-MA-S9 drastically reduced the virion production level (Fig. 6A), we next examined the effect of Gag-MA-S9 mutations (R/P/H/A/E/I/W/F) on the Gag oligomerization in cells by the EGS-crosslinking method following a protocol similar to that previously reported (60). HeLa cells were transfected with ΔPro/ΔEnv proviral clones, and then the cells were treated with EGS or PBS. After preparing the cell lysates, we evaluated the Gag-oligomerization state based on the migration position of Gag proteins in the Western blot using an anti-Gag antibody. Gag-MA-G2A and Gag-CA-WMAA ΔPro/ΔEnv clones were used as controls. For PBS treatment without EGS (Fig. 8A, right panel), Gag monomer (Gag 1 ) proteins were mainly observed for all clones tested along with small amounts of Gag-Pol proteins. For EGS-crosslinked samples (Fig. 8A, left panel), WT NL4-3ΔPro/ΔEnv produced Gag dimer (Gag 2 ), Gag trimer (Gag 3 ), Gag tetramer (Gag 4 ), and oligomerized Gag (Gag 5~) products at the predicted Full-Length Text migration positions. As expected, for the Gag-CA-WMAA clone, the Gag monomer was predominantly observed, and no oligomerized Gag (Gag 2 ~Gag 5~) was detected. For the Gag-MA-G2A clone, the production of Gag dimer proteins was significantly reduced relative to that for the WT, and no oligomerized Gag products larger than the Gag dimer were observed. Six of the Gag-MA-S9 mutant ΔPro/ΔEnv clones (S9R/P/H/A/E/I) expressed Gag dimer (Gag 2 ), Gag trimer (Gag 3 ), Gag tetramer (Gag 4 ), and oligomerized Gag (Gag 5~) products in a manner similar to the WT, although the band intensities of these products were different. Importantly, the GA-MA-S9F/W clones, which have drastically reduced virion production abilities, exhibited a similar phenotype to Gag-MA-G2A. Although a decrease in Gag dimer products was observed for all Gag-MA-S9 mutants tested, Gag oligomeriza tion (Gag 3 ~Gag 5~) of GA-MA-S9F/W clones basically did not proceed beyond Gag dimer formation, which was also the case for Gag-MA-G2A. These results suggested that the inhibition of virion production by MA-G2A and MA-S9F/W mutations results from defects in Gag oligomerization. This possibility was supported by the results of another assay using a nano-bio luminescence resonance energy transfer (NanoBRET) system to directly measure the efficiency of Gag-Gag interactions in living cells (61, 62) (Fig. 8B). In this assay, MA-S9F/W mutants showed greater reduction in NanoBRET signal production as compared with the other MA-9 mutants, although the impairments were less pronounced than those in the MA-G2A and MA-WMAA mutants. Consistent with the results of the cross-linking experiments in Fig. 8A, the S9I mutant, which exhibited relatively severe defects in progeny virion production, somehow showed only moderate effects on NanoBRET signal production. Consistent with the results of the virion production in Fig. 6A and the cross-linking experiments in Fig. 8A, the effects of the S9A mutation on the NanoBRET signal were much milder than the effects observed in the mutants MA-S9F and MA-S9W with aromatic substitutions at position 9 of MA.
## Virion production-deficient Gag-MA mutants exhibit a drastic decrease in membrane localization of Gag proteins
HIV-1 Gag proteins multimerize at PM during virion assembly (20,(56)(57)(58). To investigate Gag membrane targeting of Gag-MA mutants, which exhibit reduced capacities for virion production and Gag oligomerization, membrane flotation assays were performed (Fig. 9). Lysates of HeLa cells transfected with proviral ΔPro/ΔEnv clones were pre pared and supplemented with high concentrations of sucrose and then overlaid with lower concentrations of sucrose in ultracentrifugation tubes. In this assay, membrane components float at the top of the sucrose layer in tubes after ultracentrifugation. Gag proteins fractionated from the top (membrane fraction: MF) to the bottom (nonmembrane fraction: non-MF) portions were monitored by Western blotting analysis as previously described (15,63). For the NL4-3 ΔPro/ΔEnv clone, Gag proteins were mainly detected in the MF (Fig. 9A andB). In contrast, Gag proteins of the control Gag-MA-G2A clone were marginally observed in the MF but predominantly observed in the non-MF. Among the Gag-MA-S9 mutants tested, the Gag-MA-S9I/A clones exhibited the membrane targeting of Gag proteins, albeit at levels lower than the corresponding activity in the WT. While a high proportion of Gag proteins in the MF was detected for the WT clone (MF/MF + non-MF ratio: ~70%), the Gag-MA-S9I and S9A mutants displayed MF/MF + non-MF ratios of about 40% and 50%, respectively. The results with the S9A mutant are consistent with those of the previous studies (22,64). For the Gag-MA-S9F/W mutant clones, the levels of Gag proteins in the MF were very low relative to those in the non-MF. Of note, the Gag-MA-S9F clone showed a ratio of Gag proteins in the non-MF to total Gag proteins of around 80%, similar to the ratio for the Gag-MA-G2A clone. The Gag localization in the Gag-MA mutants tested seems to be related to their virion production (Fig. 9C; Pearson's correlation coefficient, R = 0.94). Collectively, these results showed that the membrane targeting of Gag is significantly impaired by Gag-MA-S9 mutations, especially S9F/W, and the deficiency of Gag membrane targeting is well correlated with the remarkable reduction in virion production.
## Gag-MA-S9 mutations affect N-myristoylation of Gag proteins
Our Gag-MA mutants with deficiencies of virion production and Gag oligomerization showed a significant reduction in membrane-targeted Gag protein levels (Fig. 6A, 8 and9). N-myristoylation at the G2 residue of Gag-MA is a prerequisite to PM target ing of Gag proteins (20,(56)(57)(58). It has been reported that N-terminal octapeptide amino acid sequences (G-X-X-X-S/T-X-X-X-, X, any amino acid) are important for the N-myristoylation reaction (65,66). We thus examined whether Gag-MA-S9 mutations affect the N-myristoylation of Gag proteins (Fig. 10). Proviral ΔPro/ΔEnv clones were transfected into HEK293T cells, and any myristoylated proteins in the transfected cells were biotinylated through Click-iT reaction followed by Western blot analysis using streptavidin to visualize myristoylated proteins. As shown in Fig. 10, N-myristoylated Gag proteins were readily detected for NL4-3, whereas no N-myristoylated Gag proteins were observed for Gag-MA-G2A mutants. For Gag-MA-S9I/A mutant clones, N-myristoylated Gag proteins were modestly decreased compared to their levels in the WT. Interestingly, Gag-MA-S9F/W mutant clones produced barely detectable levels of N-myristoylated Gag proteins. These results showed that N-myristoylation of Gag is strongly inhibited by certain mutations at position 9 of Gag-MA. This reduction in myristoylated Gag proteins could explain the drastic reductions in Gag membrane targeting and subsequent virion production for the Gag-MA-S9F/W mutants. Together, our virological study showed that the hydrophobic amino acid substitutions at the MA-9 residue impair the elementary and overall processes of progeny virus particle formation in the cells.
## DISCUSSION
In this study, we revisited the role of the HIV-1 Gag N-terminal region from the viewpoint of its structure. Using in silico and experimental methods, we have obtained evidence that the HIV-1 Gag MA-9 residue is involved in the HIV-1 fitness in vivo and has the ability to modulate the level of disorder and motional dynamics of the Gag N-terminal peptide segment. Our study further demonstrated that the MA-9 residue is able to modulate the efficiency of progeny particle formation of HIV-1. Finally, we found that the level of virus particle formation is positively correlated with the levels of N-terminal disorder. To our knowledge, this is the first study showing the structural role of the MA-9 residue in parallel with the biological role of the Gag N-terminal disordered peptide segment.
Our study provides new insights into the regulation of the structure of the HIV-1 Gag N-terminal region. First, it provides evidence that a single hydrophobic amino acid substitution at the MA-9 site is sufficient to reduce the disorder potential of the MA N-terminal region (Fig. 4; Fig. S5). This finding is consistent with a principle of protein folding that intrinsically disordered segments of proteins comprise an insufficient proportion of hydrophobic amino acids to prevent folding (6,50). Second, our NMR study revealed that a single hydrophobic substitution at the MA-9 site is sufficient to induce a reduction in motional dynamics of the MA N-terminal segment and conformational changes in the MA domain (Fig. 5). These findings disclose that the preservation of the non-hydrophobic residue at the MA-9 site is a prerequisite for maintaining high levels of structural disorder and dynamics of the Gag N-terminal segment.
Our study also provides new insights into the biological importance of the MA-9 residue. The MA-9 residue is placed immediately downstream of the N-myristoyl transferase recognition segment (G-X-X-X-S/T) (67) in the Gag MA N-terminal region (Fig. 3A). Previous studies have suggested that the MA-9 residue participates in biological processes such as plasma membrane targeting of the Gag precursor, virus release, and viral fusion with target cell membranes (22-24, 64, 68, 69). Our biological data (Fig. 6 to 10) are consistent with the previous findings, as well as with our structural data. Because molecular interactions are dependent on the structure and dynamics of the interaction regions, it is possible that changes in the disordered state and motional dynamics of the MA N-terminal region and MA conformation (Fig. 2, 4 and 5) influence early events during Gag translation, such as interaction of the Gag monomer with the N-myristoyl transferase. This, in turn, could impact the subsequent events involving the MA domain for viral particle formation, such as Gag protein binding to the tRNA for the Gag targeting to the plasma membrane (17)(18)(19)(20), Gag-Gag interaction for the Gag oligomerization in the plasma membrane (30,43), and incorporation of Env into the virions (21,26,31).
In addition, our comprehensive mutagenesis revealed an amino acid dependency of the mutational effects that was not disclosed in the previous studies; all of the single hydrophobic amino acid substitutions tested impair the HIV-1 particle formation, albeit to different degrees among the different mutations. In agreement with these findings, the hydrophobic amino acid at the MA-9 site was rarely detected in the public HIV-1 sequence database (Fig. 3B). These findings indicate that preservation of the non-hydrophobic residue at the MA-9 site is a prerequisite for optimizing the viral particle formation in the cells and the fitness of HIV-1 in vivo.
Finally, our study provides a novel insight into the significance of the retention of the unstructured state of the Gag N-terminal region. We found that the level of viral particle production is positively correlated with the level of the disorder potential of the Gag N-terminal segment (Fig. 6B). This implies that the preservation of a high level of disorder in the Gag N-terminal segment is critical for the optimal production of virus particles. Such physiological significance of the disordered region of a protein terminus has also been reported for human UDP-α-D-glucose-6-dehydrogenase, in which the C-terminal disordered segment plays a key role in controlling the structural dynamics of the enzyme to favor inhibitor binding (54).
In summary, the present findings clarify the relation between microscopic and macroscopic aspects of the HIV-1 survival strategy. Our data indicate that the Gag MA-9 residue plays a key role in maintaining the unstructured state and motional dynamics of the Gag N-terminal segment in order to optimize progeny virus production in host cells. This model provides a rational explanation for the continued disorder of the N-terminal component of the HIV-1 Gag precursor protein across the evolutionary history of HIV-1.
## MATERIALS AND METHODS
## Prediction of the disordered segment of the N-terminal region of HIV-1 MA
Disorder scores of the first 30 amino acid residues of the N-terminus of the MA domain were estimated with AIUPred (51) and PONDR VL-XT predictor (33)(34)(35) using the Gag full-length and MA domain sequence from the strain HIV-1 NL4-3 (GenBank accession no. AF324493) (36).
## Molecular modeling of the HIV-1 Gag precursor
A three-dimensional model of the Gag monomer of the HIV-1 NL4-3 strain (GenBank accession no. AF324493) (36) was constructed using the reported structures of the HIV-1 Gag subdomains. Briefly, individual structures of Gag subdomains, such as MA, CA, NC, and p6, were constructed by homology modeling and connected with the overlapped regions by using the Molecular Operating Environment (MOE; Chemical Computing Group, Montreal, Quebec, Canada). The structures of the Gag subdomain were obtained from the Protein Data Bank (the PDB IDs were MA: 2H3I [46]; MA-CA: 1L6N [70]; CA-SP1: 5I4T [71]; CA-SP1-NC: 1U57 [72]; NC: 1A1T [73]; and p6: 2C55 [74]). The structure of SP2, which is composed of 14 amino acids, is unavailable at present; we constructed SP2 using the Protein builder application of MOE. The obtained full-length Gag precursor model was optimized via energy minimization using the AMBER10:Extended Hückel Theory (EHT) force field in MOE, which combines the Amber10-and EHT-bonded parameters for large-scale energy minimization (75). The model was subjected to MD simulations. Briefly, the simulations were performed using the pmemd.cuda.MPI module in the Amber 16 program (32) with the ff14SB force field for protein simulation (76). The Gag precursor model was solvated in a truncated box of TIP3P water molecules with a distance of at least 9 Å around the model (77). The Gag precursor model was neutralized and ionized in 150 mM NaCl using the tleap program of Amber software. The size of the system with Gag precursor, water molecules, and ion molecules was 179.11 Å height, 123.13 Å width, and 138.46 Å length (Fig. S6). A non-bonded cut off of 10 Å was used. Bond lengths involving hydrogen were constrained with SHAKE, a constraint algorithm that satisfied Newtonian motion (78). The trajectory data of all MD simulations were collected at 2 fs intervals. After heating calculations were performed for 20 ps up to 310K using the NVT ensemble, simulations were executed using the NPT ensemble at 310K under 1 atm in 150 mM NaCl for a total of 200 ns. The trajectory files during MD simulations were used to calculate the RMSD. RMSDs between the heavy atoms of the initial complex structure and the structure at given time points during the MD simulation were calculated to monitor the overall structural changes. The molecular surface of each Gag region was calculated using the Linear Combinations of Pairwise Overlaps algorithm (79). Calculations of the RMSDs and molecular surface were done by the cpptraj module in AmberTools 16, a trajectory analysis tool (32).
## Molecular modeling of the HIV-1 Gag dimer
A Gag dimer model of HIV-1 was constructed by the MOE using the Gag monomer obtained at 200 ns of MD. The CA CTD dimer model (PDB ID: 4COP) (45) was used as a template for the Gag dimer interface. First, two Gag models were placed around the CTD-CTD interface on an axis and separated until there was no atomic clash. Next, the two Gag models without Gag-Gag interaction were subjected to MD simulations under the same calculation conditions as described above. The size of the system with Gag precursor, water molecules, and ion molecules was 145.16 Å height, 112.65 Å width, and 85.00 Å length (Fig. S6). The simulations were executed for a total of 1,000 ns. The trajectory files during MD simulations were used to calculate the number of hydrogen bonds between two Gag precursors. The formation of hydrogen bonds was determined using geometric criteria, i.e., the distance and angle between an acceptor heavy atom and a donor heavy atom. Calculations of RMSDs and the number of hydrogen bonds were done by the cpptraj module in AmberTools 16, a trajectory analysis tool (32).
## Analysis of the molecular surface area
Trajectory files during MD simulations were used to calculate the molecular surface area of the Gag precursor subdomains (MA, CA, SP1, NC, SP2, and p6) and functional regions of viral particle production (MA-HBR, MA-H2H3 loop, MA-H5H6 loop, CA-MHR, CA-SP1 junction, CA-NTD NTD interaction sites, Zn-zinc finger, and p6-PTAP) using the linear combinations of pairwise overlaps algorithm (79) in cpptraj operated in AmberTools 16 (32).
## Analysis of the Gag length
The trajectory files during MD simulations were used to calculate distances of N-ter minal-end-to-C-terminal-end of the Gag monomer and Gag subdomains (MA, MA-CA, MA-CA-SP1, MA-CA-SP1-NC, and MA-CA-SP1-NC-SP2) using the distance application in cpptraj operated in AmberTools 16 (32).
## Molecular patch analysis
The interaction-prone areas on the Gag precursors were estimated using the Protein Patch Analyzer tool in MOE. A dimer model of the Gag after 1,000 ns of MD simulation was used for the patch analyses. The MA domain model was obtained from the PDB (PDB ID: 2H3F) (46). Briefly, the Protein Patch Analyzer tool was applied to search for the hydrophobic patches (minimal patch area of 50 Å 2 ) that were potentially involved in the interactions with the hydrophobic moieties of molecules.
## Shannon entropy analysis
The amino acid variation at each position of Gag-MA was analyzed with Shannon entropy as described previously (15,80). Full-length matrix amino acid sequences of HIV-1 (n = 25,222) were obtained from the HIV Sequence Database (https:// www.hiv.lanl.gov/content/index). Shannon entropy was calculated on the basis of Shannon's equation ( 49):
where H(i), p(x i ), and i indicate the amino acid entropy score for an individual position, the probability of occurrence of a given amino acid at that position, and the number of the position, respectively. An H(i) score of zero indicates absolute conservation, whereas a score of 4.4 bits indicates complete randomness.
$$H i = - x i p x i log 2 p x i x i = G, A, I, V, … ,$$
## Plasmid construction and protein preparation for NMR analysis
The DNA sequence encoding Gag MA (residues 1-132) was codon-optimized, synthe sized, and inserted into pET21a by VectorBuilder Inc to obtain the protein with a C-terminal 6× histidine-tag (6His), referred to as Gag-MA 6His. An S9F mutant, Gag-MA-S9F 6His, was constructed using inverse PCR. The resulting constructs were confirmed by DNA sequencing.
Gag-MA 6His was expressed in Escherichia coli BL21(DE3)/pLysS cells. Cells were grown in M9 minimal medium containing 1 g/L 15 NH 4 Cl as the sole nitrogen source for the preparation of uniformly 15 N-labeled protein. Protein purification was performed at 4°C using a Ni Sepharose High Performance column (Cytiva) and HiLoad 16/60 Superdex 200 column (GE Healthcare). The obtained protein was dissolved in 25 mM sodium phosphate, pH 5.5, 200 mM NaCl, and 1 mM 1,4-dithiothreitol (DTT). Gag-MA-S9F 6His was also prepared in the same manner. The resulting protein solutions were concentrated using an Amicon Ultra-15 3 kDa (Merck Millipore). Sample concentrations were determined by measuring UV absorbance at 280 nm, utilizing a molar extinction coefficient of 16,960.
## NMR experiments
All NMR spectra were acquired at 308 K using a Bruker AVANCE III HD 600 MHz spectrom eter (Bruker, Billerica, MA) equipped with a cryogenic probe. NMR data processing was performed using TopSpin 3.5.7. An NMR sample containing 100 µM 15 N-labeled Gag-MA 6His, 25 mM sodium phosphate, pH 5.5, 200 mM NaCl, 1 mM DTT, and 5% (vol/vol) D 2 O was used to record a 2D 1 H-15 N HSQC spectrum. NaOH titration was conducted on the NMR sample, and 2D 1 H-15 N HSQC spectra were acquired under five different pH conditions (pH 6.0, 6.5, 7.0, 7.5, and 8.0) for signal assignment. Subsequently, a 2D 1 H-15 N HSQC spectrum of 100 µM 15 N-labeled Gag-MA-S9F 6His was measured in 25 mM sodium phosphate, pH 5.5, 200 mM NaCl, 1 mM DTT, and 5% (vol/vol) D 2 O. The steady-state { 1 H}- 15 N NOE was measured by recording two interleaved spectra: one with a 1 H presaturation time of 3 s and a relaxation delay of 2 s for the NOE experiment, and the other with a 5 s relaxation delay for the reference experiment. In this experiment, NMR samples containing 300 µM of the respective proteins at pH 5.5 were employed. Proton-nitrogen heteronuclear NOE values were calculated as the ratio between the cross-peak intensities with (I) and without (I 0 ) 1 H presaturation (I /I 0 ). The errors were estimated from the root mean square of the baseline noise in the two spectra (81).
## Plasmid DNAs
The full-length proviral clone pNL4-3, the Pro-and Env-deficient proviral clones pNL4-3ΔPro/ΔEnv (15), HaloTag-fused Gag (62), and NanoLuc-fused Gag (82) have been described previously (15,62,82). Their site-specific Gag-MA and Gag-CA mutants were generated by the standard PCR-based mutagenesis method.
## Cells
Monolayer cell lines HEK293T (ATCC CRL-1573), HeLa (ATCC CCL-2), and HeLa-derived reporter TZM-bl (83) were cultured and maintained in Eagle's minimal essential medium containing 10% heat-inactivated fetal bovine serum.
## Assays for virion production and single-cycle infectivity
HEK293T cells were transfected with a full-length pNL4-3 or its mutant clones by Lipofectamine 2000 (Invitrogen), and the culture supernatants were collected at 24 h post-transfection. Virion-associated reverse transcriptase (RT) activity in the culture supernatants was measured by RT assays as previously described (84,85) and used as an index of the virion production level. Equal amounts of viruses (10,000 RT units) were inoculated into TZM-bl cells (4 × 10 3 ). On day 2 post-infection, cells were lysed and subjected to luciferase assays (Promega) to assess viral single-cycle infectivity.
## Analysis of Env incorporation into virions
Proviral clones were transfected into HEK293T cells by the calcium phosphate copre cipitation method. On day 2 post-transfection, the supernatant was passed through a 0.45 µm filter to remove cell debris, layered on top of 25% sucrose solution, and ultracentrifuged at 25,000 rpm for 2 h at 4°C (Himac CP80NX, P40ST rotor; Eppendorf Himac Technologies, Ibaraki, Japan). After centrifugation, the pellets were dissolved in 1× TNE buffer, and then the amounts of virions were monitored by using an HIV-1 p24 antigen enzyme-linked immunosorbent assay kit (ZeptoMetrix Corporation, Buffalo, NY). The samples were analyzed by the Western blotting method using anti-HIV-1 Gag-p24 (183-H12-5C; NIH AIDS Reagent Program; catalog no. 3537), anti-HIV1 gp120 (ab21179, abcam), and anti-β-actin clone AC-15 (Sigma-Aldrich Co.) antibodies as described previously (86). Signal intensities of Gag proteins detected by immunoblotting were quantitated with Fusion Edge Software (Vilber Lourmat).
## Analysis of Gag oligomerization by crosslinking
The crosslinking experiments to assess Gag oligomerization were performed as previously described (60) with some modifications. HeLa cells were transfected with pNL4-3ΔPro/ΔEnv or its derivatives by Lipofectamine 2000, and at 24 h post-transfection, cells were washed with PBS and then treated with 0.1 mM EGS (ethylene glycol bis [succinimidyl succinate]; FUJIFILM Wako Pure Chemical Corporation) or PBS for 30 min at room temperature. Cells were then treated with 0.1 M Tris-HCl, pH 7.5, to quench the EGS reaction, lysed, and subjected to Western blot analysis as described above. Immunoblotting analysis was performed using anti-HIV-1 p55 + p24 + p17 (ab63917; abcam) and anti-β-actin clone AC-15 (Sigma-Aldrich Co.) antibodies.
## NanoBRET assays
Assays were performed as described previously (15). Briefly, vectors encoding Halo Tag-fused Gag and NanoLuc-fused Gag were cotransfected into HEK293T cells. On day 2 post-transfection, NanoBRET activity was measured by the NanoBRET Nano-Glo detection system (Promega).
## Membrane flotation assays
HeLa cells were transfected with pNL4-3ΔPro/ΔEnv or its derivatives. At 24 h post-trans fection, cells were treated with hypotonic lysis buffer (10 mM Tris-acetate, pH 7.4) for 15 min at 4°C. Cells were then homogenized by 30-40 strokes in a Downs homogenizer. A one-ninth volume of 10× salt buffer (10 mM Tris-acetate, pH 7.4, 500 mM KCl, and 1,000 mM NaCl) was added, and the solution was mixed. After centrifugation at 300 g for 10 min at 4°C, the post-nuclear supernatants were adjusted to 75% sucrose in 1× salt buffer by mixing with 87% sucrose, and 1.5 mL of the mixture was added to a centrifugation tube (13 PA tube; Eppendorf Himac Technologies). On top of the mixture, 6.9 mL of 65% sucrose and 2.4 mL of 10% sucrose were layered, and the tubes were ultracentrifuged at 98,400 × g for 20 h at 4°C using a Himac CP80NX and P40ST rotor (Eppendorf Himac Technologies). After ultracentrifugation, 12 fractions were collected from the top to the bottom and stored at -20°C until analysis. Fractionated samples were examined for Gag proteins by Western blotting as described above. Signal intensities of Gag proteins detected by immunoblotting were quantitated with Fusion Edge Software (Vilber Lourmat).
## Analysis of the N-myristoylation of Gag proteins
HEK293T cells were transfected with pNL4-3ΔPro/ΔEnv or its derivatives by Lipofecta mine 2000 (Invitrogen). The culture medium was changed at 6-8 h post-transfection, and Click-iT Myristic Acid Azide (12-Azidododecanoic Acid; Thermo Fisher Scientific) was added to the culture medium (final concentration 25 µM). At 24 h post-transfection, cells were washed with PBS, lysed in myr-buffer (50 mM Tris-HCl, pH 8.0, 1% SDS, protease inhibitor cocktail [Sigma], and 250 U/mL Benzonase [Sigma]), and incubated for 20 min at 4°C. After vortexing for 5 min, supernatants were collected by centrifugation at 18,000 × g for 5 min at 4°C. The preparations thus obtained (100 µg of total protein for each) were reacted with biotin-alkyne (final concentration 40 µM) using a Click-iT Protein Reaction Buffer Kit (Thermo Fisher Scientific) according to the manufacturer's instruc tions, and free biotin was removed using Zeba Dye and Biotin Removal Spin Columns (Thermo Fisher Scientific). The resulting samples were subjected to Western blot analysis using streptavidin-horseradish peroxidase conjugate (Thermo Fisher Scientific) to detect any biotinylated proteins. The immunoblot was reacted with anti-HIV-1 p55 + p24 + p17 (ab63917; abcam) to confirm the migration position corresponding to Gag proteins. Signal intensities of Gag proteins detected by immunoblotting were quantita ted with Fusion Edge Software (Vilber Lourmat). After obtaining the signal intensities of myristoylated Gag (Myr-Gag), total Gag, and β-actin of each sample, those of Myr-Gag and total Gag were normalized by that of β-actin. The myristoylation ratio of Gag was calculated as Myr-Gag/Gag signal intensities in each sample relative to that in NL4-3.
## Statistics and reproducibility
Differences between mutant groups were evaluated by Welch's t-test. Error bars indicated the SDs of three independent experiments. The R values of the relation among the disorder level, viral production, infectivity, Env incorporation, and Gag localiza tion were calculated by Pearson's correlation coefficient. The number of experimental replicates is indicated in the legends for Fig. 6A, 7B, 8B, 9B, and 10B and in the corre sponding Materials and Methods section.
## References
1. Van Der Lee, Buljan, Lang et al. (2014) "Classification of intrinsically disordered regions and proteins" *Chem Rev*
2. Tompa (2012) "Intrinsically disordered proteins: a 10-year recap" *Trends Biochem Sci*
3. Wright, Dyson (2015) "Intrinsically disordered proteins in cellular signalling and regulation" *Nat Rev Mol Cell Biol*
4. (2025) *Full-Length Text Journal of Virology*
5. Uversky, Dunker (2010) "Understanding protein non-folding" *Biochim Biophys Acta*
6. Oldfield, Dunker (2014) "Intrinsically disordered proteins and intrinsically disordered protein regions" *Annu Rev Biochem*
7. Babu, Kriwacki, Pappu (2012) "Versatility from protein disorder"
8. Latysheva, Flock, Weatheritt et al. (2015) "How do disordered regions achieve comparable functions to structured domains?" *Protein Sci*
9. Oates, Romero, Ishida et al. (2013) "D 2 P 2 : database of disordered protein predictions" *Nucleic Acids Res*
10. Potenza, Domenico, Walsh et al. (2015) "MobiDB 2.0: an improved database of intrinsically disordered and mobile proteins" *Nucleic Acids Res*
11. Ward, Sodhi, Mcguffin et al. (2004) "Prediction and functional analysis of native disorder in proteins from the three kingdoms of life" *J Mol Biol*
12. Babu (2016) "The contribution of intrinsically disordered regions to protein function, cellular complexity, and human disease" *Biochem Soc Trans*
13. Xue, Williams, Oldfield et al. (2010) "Viral disorder or disordered viruses: do viral proteins possess unique features?" *Protein Pept Lett*
14. Saito, Shofa, Ode et al. (2021) "How do flaviviruses hijack host cell functions by phase separation?" *Viruses*
15. Freed (2015) "HIV-1 assembly, release and maturation" *Nat Rev Microbiol*
16. Koma, Kotani, Miyakawa et al. (2019) "Allosteric regulation of HIV-1 capsid structure for Gag assembly, virion production, and viral infectivity by a disordered interdomain linker" *J Virol*
17. Liang, Hu, Whitney et al. (2003) "A structurally disordered region at the C terminus of capsid plays essential roles in multimerization and membrane binding of the gag protein of human immunodeficiency virus type 1" *J Virol*
18. Chukkapalli, Hogue, Boyko et al. (2008) "Interaction between the human immunodeficiency virus type 1 Gag matrix domain and phosphatidylinositol-(4,5)-bisphosphate is essential for efficient gag membrane binding" *J Virol*
19. Chukkapalli, Oh, Ono (2010) "Opposing mechanisms involving RNA and lipids regulate HIV-1 Gag membrane binding through the highly basic region of the matrix domain" *Proc Natl Acad Sci*
20. Bou-Nader, Muecksch, Brown et al. (2021) "HIV-1 matrix-tRNA complex structure reveals basis for host control of Gag localization" *Cell Host Microbe*
21. Sumner, Ono (2022) "Relationship between HIV-1 Gag multimeriza tion and membrane binding" *Viruses*
22. Ono, Huang, Freed (1997) "Characterization of human immuno deficiency virus type 1 matrix revertants: effects on virus assembly, Gag processing, and Env incorporation into virions" *J Virol*
23. Ono, Freed (1999) "Binding of human immunodeficiency virus type 1 Gag to membrane: role of the matrix amino terminus" *J Virol*
24. Göttlinger, Sodroski, Haseltine (1989) "Role of capsid precursor processing and myristoylation in morphogenesis and infectivity of human immunodeficiency virus type 1" *Proc Natl Acad Sci*
25. Bryant, Ratner (1990) "Myristoylation-dependent replication and assembly of human immunodeficiency virus 1" *Proc Natl Acad Sci*
26. Yuan, Yu, Lee et al. (1993) "Mutations in the N-terminal region of human immunodeficiency virus type 1 matrix protein block intracellular transport of the Gag precursor" *J Virol*
27. Freed, Orenstein, Buckler-White et al. (1994) "Single amino acid changes in the human immunodeficiency virus type 1 matrix protein block virus particle production" *J Virol*
28. Hermida-Matsumoto, Resh (2000) "Localization of human immunodeficiency virus type 1 Gag and Env at the plasma membrane by confocal imaging" *J Virol*
29. Ono, Orenstein, Freed (2000) "Role of the Gag matrix domain in targeting human immunodeficiency virus type 1 assembly" *J Virol*
30. Ono, Freed (2004) "Cell-type-dependent targeting of human immunodeficiency virus type 1 assembly to the plasma membrane and the multivesicular body" *J Virol*
31. Qu, Ke, Zila et al. (2021) "Maturation of the matrix and viral membrane of HIV-1" *Science*
32. Freed, Martin (1996) "Domains of the human immunodeficiency virus type 1 matrix and gp41 cytoplasmic tail required for envelope incorporation into virions" *J Virol*
33. Case, Cerutti, Duke et al. (2016)
34. Romero, Obradovic, Li et al. (2001) "Sequence complexity of disordered protein" *Proteins*
35. Li, Romero, Rani et al. (1999) "Predicting protein disorder for N-, C-, and internal regions" *Genome Inform*
36. Romero, Obradovic, Dunker (1997) "Sequence data analysis for long disordered regions prediction in the calcineurin family" *Genome Inform*
37. 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*
38. Matsuda, Hirai-Yuki, Kotani et al. (2024) "Loxapine inhibits replication of hepatitis A virus in vitro and in vivo by targeting viral protein 2C" *PLoS Pathog*
39. Izumida, Kotani, Hayashi et al. (2022) "Unique mode of antiviral action of a marine alkaloid against ebola virus and SARS-CoV-2" *Viruses*
40. Anfinsen (1973) "Principles that govern the folding of protein chains" *Science*
41. Fersht (2008) "From the first protein structures to our current knowledge of protein folding: delights and scepticisms" *Nat Rev Mol Cell Biol*
42. Datta, Curtis, Ratcliff et al. (2007) "Conformation of the HIV-1 Gag protein in solution" *J Mol Biol*
43. Zhao, Perilla, Yufenyuy et al. (2013) "Mature HIV-1 capsid structure by cryo-electron microscopy and all-atom molecular dynamics" *Nature*
44. Schur, Hagen, Rumlová et al. (2015) "Structure of the immature HIV-1 capsid in intact virus particles at 8.8Å resolution" *Nature*
45. Purdy, Shi, Chrustowicz et al. (2018) "MicroED structures of HIV-1 Gag CTD-SP1 reveal binding interactions Full-Length Text Journal of Virology October"
46. "with the maturation inhibitor bevirimat" *Proc Natl Acad Sci*
47. Bharat, Menendez, Hagen et al. (2014) "Cryo-electron microscopy of tubular arrays of HIV-1 Gag resolves structures essential for immature virus assembly" *Proc Natl Acad Sci*
48. Saad, Miller, Tai et al. (2006) "Structural basis for targeting HIV-1 Gag proteins to the plasma membrane for virus assembly" *Proc Natl Acad Sci*
49. Gamble, Yoo, Vajdos et al. (1997) "Structure of the carboxylterminal dimerization domain of the HIV-1 capsid protein" *Science*
50. Ivanov, Tsodikov, Kasanov et al. (2007) "Domain-swapped dimerization of the HIV-1 capsid C-terminal domain" *Proc Natl Acad Sci*
51. Shannon (1963) "The mathematical theory of communication" *MD Comput*
52. Uversky, Gillespie, Fink (2000) "Why are "natively unfolded" proteins unstructured under physiologic conditions?" *Proteins*
53. Erdős, Dosztányi (2024) "AIUPred: combining energy estimation with deep learning for the enhanced prediction of protein disorder" *Nucleic Acids Res*
54. Dodson, Lane, Verma (2008) "Molecular simulations of protein dynamics: new windows on mechanisms in biology" *EMBO Rep*
55. Karplus, Kuriyan (2005) "Molecular dynamics and protein function" *Proc Natl Acad Sci*
56. Keul, Oruganty, Bergman et al. (2018) "The entropic force generated by intrinsically disordered segments tunes protein function" *Nature*
57. Massiah, Starich, Paschall et al. (1994) "Three-dimensional structure of the human immunodeficiency virus type 1 matrix protein" *J Mol Biol*
58. Bell, Lever (2013) "HIV Gag polyprotein: processing and early viral particle assembly" *Trends Microbiol*
59. Sundquist, Kräusslich (2012) "HIV-1 assembly, budding, and maturation. Cold Spring Harb Perspect Med 2:a006924"
60. Lerner, Weaver, Anokhin et al. (2022) "Advances in HIV-1 assembly"
61. Von Schwedler, Stuchell, Müller et al. (2003) "The protein network of HIV budding" *Cell*
62. Kutluay, Bieniasz (2010) "Analysis of the initiating events in HIV-1 particle assembly and genome packaging" *PLoS Pathog*
63. Machleidt, Woodroofe, Schwinn et al. (2015) "NanoBRET--a Novel BRET platform for the analysis of protein-protein interactions" *ACS Chem Biol*
64. Miyakawa, Nishi, Matsunaga et al. (2017) "The tumour suppressor APC promotes HIV-1 assembly via interaction with Gag precursor protein" *Nat Commun*
65. Tanaka, Robinson, Chutiraka et al. (2016) "Mutations of conserved residues in the major homology region arrest assembling HIV-1 Gag as a membrane-targeted intermediate containing genomic RNA and cellular proteins" *J Virol*
66. Saad, Loeliger, Luncsford et al. (2007) "Point mutations in the HIV-1 matrix protein turn off the myristyl switch" *J Mol Biol*
67. Boutin (1997) *Myristoylation. Cell Signal*
68. Wang, Dai, Sun et al. (2021) "Protein N-myristoylation: functions and mechanisms in control of innate immunity" *Cell Mol Immunol*
69. Towler, Gordon, Adams et al. (1988) "The biology and enzymology of eukaryotic protein acylation" *Annu Rev Biochem*
70. Spearman, Wang, Vander Heyden et al. (1994) "Identification of human immunodeficiency virus type 1 Gag protein domains essential to membrane binding and particle assembly" *J Virol*
71. Bhatia, Kaushik, Campbell et al. (2009) "Mutation of critical serine residues in HIV-1 matrix result in an envelope incorporation defect which can be rescued by truncation of the gp41 cytoplasmic tail" *Virology (Auckl)*
72. Tang, Ndassa, Summers (2002) "Structure of the N-terminal 283residue fragment of the immature HIV-1 Gag polyprotein" *Nat Struct Biol*
73. Wagner, Zadrozny, Chrustowicz et al. (2016) "Crystal structure of an HIV assembly and maturation switch. Elife 5:e17063"
74. Morellet, Déméné, Teilleux et al. (1998) "Structure of the complex between the HIV-1 nucleocapsid protein NCp7 and the single-stranded pentanucleotide d(ACGCC)" *J Mol Biol*
75. De Guzman, Wu, Stalling et al. (1998) "Structure of the HIV-1 nucleocapsid protein bound to the SL3 psi-RNA recognition element" *Science*
76. Fossen, Wray, Bruns et al. (2005) "Solution structure of the human immuno deficiency virus type 1 p6 protein" *J Biol Chem*
77. Gerber, Müller (1995) "MAB, a generally applicable molecular force field for structure modelling in medicinal chemistry" *J Comput Aided Mol Des*
78. Maier, Martinez, Kasavajhala et al. (2015) "ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB" *J Chem Theory Comput*
79. Jorgensen, Chandrasekhar, Madura et al. (1983) "Comparison of simple potential functions for simulating liquid water" *J Chem Phys*
80. Ryckaert, Ciccotti, Berendsen (1977) "Numerical integration of the cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes" *J Comput Phys*
81. Weiser, Shenkin, Still (1999) "Approximate atomic surfaces from linear combinations of pairwise overlaps (LCPO)" *J Comput Chem*
82. Motomura, Oka, Yokoyama et al. (2008) "Identification of monomorphic and divergent haplotypes in the 2006-2007 norovirus GII/4 epidemic population by genomewide tracing of evolutionary history" *J Virol*
83. Farrow, Muhandiram, Singer et al. (1994) "Backbone dynamics of a free and phosphopeptide-complexed Src homology 2 domain studied by 15N NMR relaxation" *Biochemistry*
84. Adachi, Ono, Sakai et al. (1991) "Generation and characterization of the human immuno deficiency virus type 1 mutants" *Arch Virol*
85. (2025) *Full-Length Text Journal of Virology*
86. Platt, Bilska, Kozak et al. (2009) "Evidence that ecotropic murine leukemia virus contamination in TZM-bl cells does not affect the outcome of neutralizing antibody assays with human immunodeficiency virus type 1" *J Virol*
87. Willey, Smith, Lasky et al. (1988) "In vitro mutagenesis identifies a region within the envelope gene of the human immunodeficiency virus that is critical for infectivity" *J Virol*
88. Nomaguchi, Doi, Sakai et al. (2016) "Natural single-nucleotide variations in the HIV-1 genomic SA1prox region can alter viral replication ability by regulating Vif expression levels" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548450&blobtype=pdf | # Functional and structural insights into HCMV terminase accessory proteins pUL77 and pUL93
C Gourin, F Meo, C Delmon, S Alain, S Hantz
## Abstract
Human cytomegalovirus (HCMV) is a significant cause of morbidity and mortality in immunocompromised and congenitally infected patients. Current antivi ral therapies, primarily targeting the viral DNA polymerase, are limited by toxicity and resistance, underscoring the need for novel therapeutic strategies. Letermovir (LTV), which targets the viral terminase complex (pUL56-pUL89-pUL51), has improved prophylaxis in transplant recipients, but resistance mutations have already emerged. This complex is responsible for cleaving and packaging the viral genome into neo-formed capsids in close interaction with several other proteins. This study focuses on two essential terminase-associated proteins, pUL77 and pUL93, investigating their natural polymorphism, mutations arising under LTV prophylaxis, and the functional relevance of their putative nuclear localization signals (NLS). Sequence alignments across 18 herpesviruses revealed conserved and variable regions in both proteins. New muta tions emerging in patients treated with LTV did not confer resistance, but the pUL77 R43C mutation, when present with pUL56 C325F, significantly reduces LTV resistance compared to C325F alone. In addition, nuclear localization motifs were assessed in the viral replication context and in autonomous cell production after plasmid transfection and revealed motifs essential for viral replication: 43 RVRKRYLRQ 55 and 225 PRWKRV 231 in pUL77 and 505 RDRRGRLRR 513 in pUL93. Finally, AI-based modeling provided some insights into the functions of these motifs for both proteins, notably in interaction with other viral proteins. Collectively, these data provide a wealth of information about putative functional sites in pUL77 and pUL93 that could be new targets for anti-HCMV strategies such as inhibitory peptides.IMPORTANCE Human cytomegalovirus (HCMV) remains a significant health concern, particularly for immunocompromised individuals and in cases of congenital infection. The emergence of resistance to current antivirals targeting viral polymerase has necessitated the development of novel therapeutic approaches. Letermovir, which targets the HCMV terminase complex, represents a promising alternative. However, a deeper understanding of the virus's structural components and their interactions is crucial for identifying new potential drug targets. Here, we focused on two proteins, pUL77 and pUL93, which are associated with the viral capsid and interact with the terminase complex. By analyzing their polymorphism, structure, and functional motifs, we identified essential nuclear localization motifs and conserved regions in these proteins, providing valuable insights for the development of innovative anti-HCMV strategies, such as inhibitory peptides, which could complement existing treatments and address the ongoing challenge of antiviral resistance.
H uman cytomegalovirus (HCMV) is responsible for higher morbidity and mortality rates in immunocompromised patients, such as transplant recipients and AIDS patients (1). HCMV infection is also the main cause of congenital viral infection, responsible for severe malformations (2). Most anti-CMV drugs targeted only the viral DNA polymerase pUL54. Although their established efficacy, their use remains limited not only owing to dose-limiting toxicity but also to the emergence of resistance leading to therapeutic failure. Targeting a new stage in the viral replication cycle, letermovir (LTV) was demonstrated to be effective as a prophylactic treatment against HCMV in hematopoietic stem cell recipients and approved by the US and European Medicines Agencies (3). More recently, the efficacy in prophylaxis was also demonstrated in kidney recipients (4). LTV was shown to inhibit the viral terminase complex made of three proteins, namely pUL56, pUL89, and pUL51 (5). Nevertheless, in vitro and in vivo resistance mutations have already been identified (6)(7)(8)(9)(10). Therefore, there is a critical need for new therapeutic strategies targeting essential viral proteins.
The terminase complex (pUL56, pUL89, and pUL51) is responsible for cleaving DNA into unit-length genomes and packaging them into neoformed capsids (5). The large pUL56 subunit and the small subunit pUL89 contain the functional sites required for DNA cleavage (11,12). Noteworthy, three additional proteins were shown to be involved in this event, namely, pUL52, pUL77, and pUL93 (13,14). pUL77 and pUL93 are, respectively, encoded by the UL77 and UL93 genes, which are found on the unique long region of the HCMV genome. pUL77 and pUL93 are, respectively, 642 and 593 amino acid proteins with molecular masses of 71 and 68 kDa. They were shown to be structural proteins associated with Herpesviridae capsids (15,16). pUL77 and pUL93 proteins are well conserved along the Herpesvirus family, and their homologs in herpes simplex type 1 (HSV-1) are, respectively, pUL25 and pUL17 (17,18). pUL77 contains a putative pyruvoyl decarboxylase site 613 TLGSSLFN 620 in the C-terminal domain (19). The first hundred amino acids of pUL77 (residues 1 to 100) adopt a spiral conformation (CCM) responsible for the protein homodimeric conforma tion (15). The first 48 amino acids of the CCM were proposed to interact with the so-called major capsid protein (MCP), as well as the DNA during the DNA-packaging step (15). Köppen-Rung et al. suggested the existence of two motifs in pUL77 as nuclear localization signals (NLS) in the N-terminal domain, namely 55 RVRKRYLRQ 63 and 219 YYRLKRG(LYTQ)PRWKRV 231 (16). However, their study explained that these motifs were not responsible for nuclear localization by themselves. Furthermore, pUL77 was shown to also interact with pUL48, which is involved in the tegument structure (20). pUL93 has been identified as a crucial partner for the correct subnuclear localization of pUL77 (13). pUL93 interacts with the so-called nuclear egress complex (NEC) consisting of pUL53, pUL50, and pUL97 (21). Two putative NLSs were also identified in the pUL93 sequence: 177 KRDRQHQLATATNHRRR 193 and 442 RARRQ 447 , but have never been studied (21). Moreover, Borst et al. demonstrated that only empty B-capsids were found in the absence of both pUL77 and pUL93 proteins (22). Interestingly, the three-dimensional structures of pUL77 and pUL93 were partially resolved by means of cryogenic electron microscopy within the vertex-specific component complex (CVSC) (23). A complemen tary study based on artificial intelligence has completed the models of the three-dimen sional structure of the HCMV tegument, including CVSC proteins as pUL77 and pUL93 (24).
In the present study, we investigated pUL77 and pUL93 proteins considering (i) their natural polymorphism and (ii) those triggered under LTV selective pressure. New mutations were described and assessed in the LTV plaque assay alone or in combination with UL56 resistance mutations. Additionally, putative NLSs of pUL77 and pUL93 were identified and monitored to provide robust hints regarding their functional domains. Such findings are supported by using structural approaches and the comparison with their respective HSV-1 homologs, namely pUL25 and pUL17 for pUL77 and pUL93, respectively.
## RESULTS
## Determination of conserved domains and polymorphism analysis
The homologs from 18 herpesviruses (Table S1) were considered for sequence align ments: 16 amino acids were identical, 38 were strongly similar, and 15 were weakly similar. Various gaps were observed in the pUL77 N-terminal region (residues 98 to 181), suggesting the existence of a variable region (VR-I, Fig. 1A andB). Sequence alignments also revealed the existence of five conserved regions, namely region I (residues 184 to 285), region II (residues 337 to 382), region III (residues 402 to 466), region IV (residues 502 to 536), and region V (residues 584 to 637) (Fig. 1A andB). By comparing the sequence of pUL77 with that of 18 homologs, the putative pUL77 NLS1 and NLS2 are not conserved among other herpesviruses. In contrast, another motif from residue 522 to 539 ( 522 DPAVTLSQLFPGLALLAV 539 ) located in region IV is relatively conserved among the 18 herpesviruses (Fig. 1C). This domain was deleted in the HCMV-bacterial artificial chromosome (BAC). Eleven days after transfection in MRC-5 cells, no replication of the recombinant virus was observed (Fig. 2).
Similar analyses were conducted with pUL93 (Table S2): 10 amino acids were identical, five were strongly similar, and six were weakly similar. In contrast with pUL77, numerous regions with gaps were observed, suggesting at least three variable regions, namely VR-I (from residue 34 to 87), VR-II (from residue 179 to 264), and VR-III (from residue 436 to 474). Sequence alignment also revealed seven conserved regions located either in the N-and C-terminal domains (Fig. 3A). Polymorphism analysis of pUL93 is in agreement with these results, most of the mutations being in variable regions (Fig. 3B). The putative pUL93 NLSs were all located in non-conserved regions of the protein. Furthermore, these NLSs were not observed in the other 18 herpesviruses, suggesting specific roles for HCMV (Fig. 3C).
Reference strain sequences of AD169, Towne, Toledo, Merlin, and Davis were identical to those from the GenBank database (accession numbers FJ527563, JX512198, GU937742, FJ1616285.1, and NC_006273) for pUL77 and pUL93. Among the 76 naive strains, we reported 51 and 63 amino acid polymorphisms distributed over pUL77 and pUL93, respectively. Twenty and 18 polymorphisms were also present in reference strains for each protein, respectively. The average identities of the HCMV isolates were 98.6% and 99.2% for pUL77 and pUL93, respectively. Analysis of polymorphism distribution across pUL77 revealed a region with a high number of mutations, as well as the deletion of the repetitive 145 PSDAVAPSDAVA 156 motif, observed in most of the reference and clinical HCMV strains. This is consistent with sequence alignment analyses for which the variable region was annotated as pUL77 VR-I. Interestingly, this motif was not found among other herpesviruses (Fig. 1A andB). Likewise, polymorphism analyses revealed a high sequence variability in VR-I, VR-II, and VR-III of pUL93 (Fig. 3A andB).
Interestingly, while considering the 12 sequences of HCMV strains detected in LTV-treated patients, 26 mutations were found in pUL77; 15 and 9 being, respectively, either shared with reference strains or naive strains. However, we observed two new mutations: R43C and A161T (see red arrows in Fig. 1B). R43C was assigned to the non-conserved region belonging to the coiled-coil motif of the protein, while A161T was located in the pUL77 VR-I. For pUL93, 18 mutations were found in HCMV strains detected in LTV-treated patients, of which 13 were identical to reference strains and three to naive strains. Again, we observed two new mutations located either in pUL93 VR-I or VR-II: E73G and R206H (see red arrows in Fig. 3B). Mutation pUL77 R43C was associated with C325F in strain 5 and was combined with pUL77 A161T and pUL56 C325Y in patient 11. pUL93 E73G and R206H were associated with pUL56 V236M in clinical strain 2. An overview of clinical mutations in pUL56, pUL77, and pUL93 is available in Table S3.
## Assessment of new mutations
Even though amino acids identified from pUL77 and pUL93 are not conserved, we investigated their potential impact by introducing these mutations in HCMV-BAC and transfecting it into MRC-5 human embryonic fibroblasts. Plaque assays with LTV exhibited that standalone mutations do not confer LTV resistance to recombinant viruses as compared with AD169 (Table 1). These mutations were also combined with V236M, C325F, or C325Y mutations in pUL56, known to be associated with LTV resistance, as they were detected by resistance genotyping in the clinical follow-up of these patients. Plaque tests with the LTV of the C325F/R43C mutant appeared to significantly reduce the level of resistance compared with the C325F mutant alone (P = 0.0261). For the other mutants, no significant change in EC 50 of strains with the UL56 mutations was observed, independent of the presence of the other pUL77/pUL93 mutations (Table 1). Finally, cell growth was also investigated considering these mutations by means of green fluorescent protein (GFP) counts, using the AD169 strain as a reference. In line with plaque assays, these new mutations were not associated with a modulation of viral fitness (Fig. 4).
## Study of putative nuclear localization signals alone and in a viral replication context
Putative nuclear localization signals reported in the literature were either mutated or deleted in HCMV-BAC GFP and transfected into MRC-5 cells to define their role. Standalone mutations were carried out, and multiple mutation combinations were also performed as reported in Fig. 5. It is important to note that pUL77 NLS1 was only mutated and not deleted due to its overlap region with the ORF region of the pUL76 protein. Moreover, to ensure that pUL77 NLS2 was a bipartite functional domain as described in the literature, three mutants were constructed: NLS2#, NLS2*, and NLS2. Indeed, the second whole putative NLS of pUL77 (NLS2) cited in the literature was 219 YYRLKRG(LYTQ)PRWKRV 231 and was supposed to be bipartite (16). So we construc ted two additional mutants, NLS2# and NLS2*, harboring deletions of the sequences 216 GIWYYRLKRGLYT 224 and 225 QPRWKRV 231 , respectively. This analysis allowed us to assess the bipartite character of the putative NLS2 (Fig. 5A). Growth fitness assays were carried out using the AD169 strain as a reference. In the fitness analysis, the standalone pUL77 NLS1 mutation had a significantly attenuated growth at D4 and D7 (P values = 0.001) (Fig. 5A andD). Likewise, the standalone deletion of the 225 QPRWKRV 231 motif of pUL77 NLS2 (NLS2*) led to the absence of replication. For pUL77 NLS2# deletion, no inhibition of viral replication was observed (P = 0.93) (Fig. 5A andD). The pUL93 ΔNLS1 mutant showed a significant reduction in fitness at D7 (P value = 0.02), but the pUL93 ΔNLS2 and ΔNLS1/2 mutants did not (P = 0.44 and P = 0.14, respectively) (Fig. 5E). Using Geneious software, we obtained isoelectric point (pI = 12.80) of pUL93 505 RDRRGRLRR 513 suggesting a third putative NLS in pUL93 (UL93 NLS3) located in the C-terminal region (Fig. 1A and5B). No cytopathic effects were observed after transfection of the recombinant HCMV-BAC pUL93 ΔNLS3 in MRC-5 cells (Fig. 5A andC).
## Study of putative nuclear localization signal in HEK293T cells
To assess the cellular localization of wild-type or mutated pUL77 and pUL93, mCherryprotein recombinants were built and monitored by confocal microscopy. mCherry was appended to pUL77 and pUL93 N-terminal domains in order to ensure protein functions (Fig. 6A andB). The same experiment was performed with co-transfection of HCMV-BAC.
In this latest experiment, only cells co-expressing GFP and mCherry were analyzed. HCMV AD169 production was assessed by western blot using IE1/IE2 protein detection (Fig. 6B).
The percentage of mCherry in cell nuclei was determined as a function of fluorescence intensity and number of red pixels (see Fig. 6 and7). Wild-type proteins were all observed in cell nuclei, suggesting the crossing of the nuclear membrane (Fig. 6C). Standalone or combined modifications of pUL77 NLSs did not significantly affect pUL77 entry into cell nuclei. Interestingly, the standalone deletion of pUL93 NLS1 or NLS2 and the combined deletion of NLS1/NLS2 led to a significant decrease of pUL93 signal in the cell nucleus (P = 0.009; P = 0.002; P = 0.02, respectively). The deletion of pUL93 NLS3 alone impacted the nuclear crossing of the protein (P < 0.001). However, the co-transfection of AD169 significantly restored the function of pUL93 to entry into the cell nucleus (P = 0.0012) (Fig. 6C and7A).
Western blot assays were conducted to verify the construction of mCherry-fused clones and allowed us to see that proteins were produced at different levels by normalizing them against the actin control protein. The addition of AD169 greatly increased the production of all clones (Fig. 7B).
## Structural investigation of HCMV pUL77 and pUL93
Taking advantage of the recent resolution of the HCMV portal vertex in several configurations (24), we identified pUL77 and pUL93 motifs within the structure resolved in the virion configuration 1 (PDB ID: 8TEP), representing the post-nuclear translocation state with binding to pUL48 (Fig. 8A). Notably, the portal vertex is a pentameric structure, where each subunit is composed of two pUL77 molecules (pUL77-l and pUL77-u), two pUL48 molecules (pUL48-u and pUL48-l), and a single pUL93 molecule, all associated with other capsid proteins (Fig. 8A) (24). Three motifs within pUL77 were resolved in the cryo-EM structure, with NLS 1 located in the capsid-binding domain (CBD). This domain consists of four α-helices formed by the N-terminal regions of pUL77-l and pUL77-u, in association with the C-terminal regions of pUL48-u and pUL48-l (Fig. 8B), consistent with previous reports (23,24). Remarkably, NLS 1 is rich in cationic residues (e.g., Arg57 and Arg55), which form strong electrostatic interactions with the carboxylate groups of the C-terminal leucine residues of pUL48-u and pUL48-l (Fig. 8C). These residues are also in close proximity to anionic residues in pUL77-u (Asp50 and Asp54), suggesting the formation of a robust salt-bridge network critical for maintaining the CBD structure. Additionally, the carbonyl groups of Gly44 and Gly45 were observed within 5 Å of the coil region of the helix-coil-helix configuration in the N-terminal domain of pUL77-l, indicating a weaker hydrogen bond network in this region. NLS 2 and motif 3 are located on the surface of the "head domains" of both pUL77-u and pUL77-l (Fig. 8B andD), where no interactions with other capsid proteins of the same subunit were detected. Likewise, we identified NLS 2 and 3 in the pUL93 resolved structure (Fig. 8E). While NLS 2 is located in the tail domain of pUL93 with no close contact with other capsid protein of the same subunit, NLS 3 was located at the interface between pUL77-l N-terminal domain and one of the five MCP. pUL93 NLS 3 is also rich in cationic residues (e.g., Arg507, Arg510, and Arg513) that are involved in attractive electrostatic interactions with surrounding electronegative of pUL77-l or MCP residues (e.g., Gln476, Gln543, and Glu547, see Fig. 8F). Since pUL93 NLS 1 was not experimentally resolved, we tried to decipher potential location using the HSV1 pUL17 (i.e., pUL93 ortholog) resolved structure (25), as well as AlphaFold2 predicted structure. It is important to note that the HSV-1 capsid structure determination was performed in a different configuration from that performed for HCMV, so the 3D structure observed may differ significantly depending solely on the protein state. Thus, the comparison of certain domains of the two homologs may be limited. We first aligned the resolved structure of HCMV pUL93 and HSV-1 pUL17 structures, for which relatively small structural differences were observed (root mean square deviation [RMSD] = 3.9 Å, Fig. 9A) in spite of a very poor identity score (9.44%). For instance, the front-barrel and back-barrel structure was conserved in pUL93. Interestingly, the region of pUL93 NLS 2 was not resolved in pUL17 ortholog, suggesting a configuration-dependent role. Unfortunately, the predicted location of pUL93 NLS 1 in the AlphaFold model exhibited a poor confidence score, precluding its use to hypothesize its potential role in, e.g., protein-protein interactions (Fig. 9B andC). However, the corresponding region was resolved in HSV1 pUL17, also adopting a kinked helix structure likely owing to the presence of Pro155. AF2 predicted a continuous α-helix structure containing a motif, even though this should be consid ered carefully owing to the poor pLDDT score (prediction local distance difference test: per-residue measure of local confidence) for this region. Moreover, the absence of a resolved structure from the HCMV cryo-EM structure suggests that such a domain is flexible.
## DISCUSSION
HCMV is a major public health problem by increasing morbidity and mortality in populations such as immunocompromised patients and neonates. The terminase complex was shown as a target of choice for the development of treatment against HCMV infections (5). So far, pUL77 and pUL93 have been mostly described as required structural partner proteins of the terminase complex for the genome encapsidation step (15,26). However, the functional domains of these proteins have not been identified yet. Nevertheless, the HSV-1 homologs of pUL77 and pUL93, respectively, pUL25 and pUL17, have been more studied. It has been shown that pUL17 is required for capsid localization in the DNA replication compartments in the nuclei of infected cells, where cleavage and packaging of the viral genome take place (27,28). Furthermore, pUL17 interacts with pUL25 to form the CVSC found on capsids' vertices (17,29). Currently, we know that HSV-1 CVSC is linked to capsid stability and core exit (30,31). Moreover, Huet et al. demonstrated that pUL17 anchors the terminase complex to the capsid and that its interaction with pUL25 at the portal allows retaining the viral genome (32). In relation to this, pUL77 and its homolog pUL25 form the cap of the portal, which seals the capsid (24,32). The functional motifs described in our study could therefore be involved in the sealing function.
Our results suggest that pUL77 and pUL93 are conserved proteins among HCMV reference strains and clinical isolates with average identities of 98.6% and 99.2%, respectively. We were able to identify the variable and conserved regions of pUL77 and pUL93 thanks to sequence alignment and comparison with other herpesvirus homologs. These results were also confirmed by means of polymorphism analysis for both proteins. Regarding pUL77, the only variable region was identified as a rich polymorphism domain in which the repetitive motif 145 PSDAVAPSDAVA 156 is deleted. This suggests that this motif has no functional role for pUL77 but might be involved in favoring protein flexibility during complex formation. The first one hundred amino acids (1 to 92) of pUL77 have been structurally assigned to a coiled-coil motif, which was described as essential for pUL77 homodimerization and supramolecular interactions with MCP (15). These residues also belong to the capsid-binding domain that binds pUL48-u/l and pUL93, thanks to electrostatic interactions, and contributes to the CVSC helix bundle (24). However, our results surprisingly indicated that this region is not highly conserved, suggesting that the structural conformation of this region is more important than its level of conservation. After deletion of the pUL77 522 DPAVTLSQLFPGLALLAV 539 motif, the HCMV-BAC mutant did not replicate in MRC-5 cells. By 3D modeling, we observed that this motif was forming an alpha helix in the core of the protein. These results suggest that this motif is highly important for protein function and viral replication. This motif could therefore be involved in interactions with other proteins, such as pUL48, which are known to interact with pUL77 (13,20). We can hypothesize that pUL77 (522-539) may also be involved in the protein-protein interactions between portal vertex subunits. For pUL93, three variable regions were identified for their rich amino acid polymorphism, while only seven small regions were described as highly conserved. This suggests that these seven conserved regions (I to VII) located in both N-and C-terminal might have relevant functional roles for pUL93 activity. For each protein, two new amino acid mutations were identified in HCMV strains emerging in patients receiving letermovir prophylaxis: R43C and A161T for pUL77; E73G and R206H for pUL93. We built the recombinant mutants in vitro, for which no effect on viral growth nor LTV resistance was observed while considering standalone mutations, suggesting that they are simple polymorphisms, not related to the use of antivirals. In addition, clinical records showed that pUL56 mutations with a high level of resistance to LTV in patients are associated with these pUL77 and pUL93 mutations. Recombinant viruses with combined mutations were thus produced, and LTV EC 50 was assessed. The results reported that the combination of these mutations with pUL56 V236M and C325Y mutations does not modify the LTV EC 50 s, similar to those obtained by Chou et al. in 2015 for pUL56 mutants (7,8). However, the addition of the UL77 R43C mutation to the UL56 C325F mutation significantly attenuated the level of letermovir resistance. Absence of resistance mutations in pUL77 and pUL93 is supplementary proof that allows us to exclude interaction of LTV with the pUL77/pUL93 complex. Nevertheless, more patients with refractory infection under LTV prophylaxis must be explored to confirm our hypothesis. However, our results show that at least pUL77 would be in close contact with pUL56 and could modulate the mechanism of action of LTV.
We also investigated potential functional regions of pUL77 and pUL93 from the literature. Two putative NLSs have been reported for each protein, but those for pUL93 have never been studied (16,21). We here showed that pUL77 NLS2* ( 225 QPRWKRV 231 ) is essential for HCMV viral replication but not for nuclear entry of pUL77. All these observations are in line with structural investigation since NLS1 is essential to (i) maintain C-terminal helices of pUL48-u/l thanks to strong electrostatic interactions, (ii) favor the helix-coil-helix configuration of at least pUL77-u but also likely pUL77-l, even though the flexible domain bridging N-terminal helices of pUL77-l has not been resolved yet. pUL77 NLS2* may play a role in capsid formation since NLS2* is located at the surface of the pUL77 head domain and might interact with the pUL77 protein of other subu nits. Interestingly, evaluation of the pUL77 cellular localization by confocal microscopy showed that the deletion of these NLSs has no impact on nuclear translocation, in agreement with observations made by Köppen-Rung et al. (16). Altogether, these regions cannot be described as NLS domains, but they remain of particular importance for viral replication, suggesting a central functional role. This observation is in line with the AI-based structural study of the specific capsid top component published in 2024, in which pUL77 43 RVRKRYLRQ 55 is described as essential for anchoring the protein in the capsid (24).
Likewise, reported potential NLSs of pUL93 from the literature remain disappointing. Deletion of NLS1 ( 177 KRDRQHLATTTNHRRR 193 ) was associated with a slightly weaker nuclear translocation and a non-significant reduction in viral fitness. Deletion of pUL93 NLS2 ( 442 RARRQ 447 ) significantly impacted nuclear membrane crossing events. For these mutants, the decrease in nuclear localization could be impacted by a new tertiary conformation of the protein, which makes it more difficult to switch to the cell nuclei. However, we here proposed a new putative NLS motif, the so-called NLS3 ( 505 RDRRGRLRR 513 ), whose deletion drastically impaired viral replication in MRC-5 cells and clearly impacted nuclear crossing events in HEK293T cells. Deletion of pUL93 NLS3 may also impact the protein-protein interactions with MCP since NLS3 is located at their interface.
To take our analysis a step further, we co-transfected the HCMV BAC into HEK293T cells to allow the production of all HCMV AD169 proteins. In this context, the nuclear crossover events of all our constructs were enhanced or rescued. As with pUL51, which is required for correct nuclear crossing of pUL56 and pUL89 (33), our results led us to hypothesize that pUL77 and pUL93 are dependent on other proteins linked to the terminase complex or capsid. What's more, we know that pUL77 and pUL93 are dimeric proteins, so their heterodimers may have an impact on their nuclear crossover and may rescue the deletion of different NLS, as has been demonstrated for the ribosomal protein RpS3 and its chaperone Yar1 (34). Further analysis is therefore required to better understand these mechanisms.
Finally, in our study, the 3D structure provides precise information on the conforma tion and atomic interactions of the protein, which is essential for understanding its biological function. However, sequence analysis allows the identification of conserved domains between different homologous proteins, which can reveal crucial functional regions, active sites, or interaction interfaces that are not always evident from the 3D structure alone. Furthermore, sequence-identified conserved domains can guide the functional interpretation of structure, help annotate poorly resolved or flexible regions, and facilitate evolutionary comparison between related proteins. The combination of the two approaches-structural and sequential-offers a more complete understanding of the function, evolution, and therapeutic potential of viral proteins.
According to the presented data, our study provides a better understanding of pUL77 and pUL93 structure and function. Our results suggested that the tertiary structure of these proteins is more important than their amino acid composition for their function.
In addition, we have identified a new functional motif in pUL93 that is significantly linked to the nuclear localization of the protein and provided new data on the other two found by bioinformatics analysis in a previous study (21). Moreover, we demonstrated that nuclear crossing events of the two proteins, pUL77 and pUL93, were also linked to the presence of other viral components. This study also suggests that at least pUL77 could be an additional target for letermovir. Indeed, pUL77 CCM could be involved in modulating resistance to LTV by interacting with pUL56 and pUL89, as it was shown (16). Moreover, pUL77 and pUL93 could be of interest for the development of future therapies by inhibiting their functional domains that we described here through the development of antivirals or inhibitory peptides, as previously demonstrated for pUL51 and the NEC proteins (35,36).
## MATERIALS AND METHODS
## Determination of conserved domains
For both proteins pUL77 and pUL93, the amino acid sequences of reference strain AD169 were aligned with the sequences of 17 homologous proteins from other herpesviruses, as described in Tables S1 andS2. Alignments were performed with the Clustal Omega (Ω) multiple sequence alignment tool, provided by the EMBL-EBI bioinformatics web tools and programmatic tools framework (37)(38)(39). To illustrate the degree of amino acid conservation, "*" was used to indicate a highly conserved amino acid; ":" was used to indicate a site belonging to a group exhibiting strong similarity; and ". " was used to indicate a site belonging to a group exhibiting weak similarity.
## Structural investigations and prediction of unresolved regions of pUL93 and pUL77
FASTA amino acid sequences of pUL93 and pUL77 were submitted to AlphaFold artificial intelligence software to predict protein structure (see Tables S1 andS2). Sequences were also submitted to the Expasy PROSITE Database (SIB Swiss Institute of Bioinformat ics) to find relevant motifs in pUL77, pUL93, and their respective homologs in HSV-1, pUL25, and pUL17. Structural investigations were also conducted using the cryo-EM resolved structure of HCMV portal vertex adopting the virion configuration 1 (PDB ID: 8TEP) (24). In order to compare, the cryo-EM resolved structure of the HSV-1 capsid associated with tegument protein was also used (PDB ID: 6CGR) (25). Structural analyses and visualization, as well as rendering, were performed using the VMD 1.9.4α (Visual Molecular Dynamics) software (25). Cells, bacterial strains, and HCMV strains MRC-5 human fibroblasts (bioMérieux, Craponne, France) were grown in minimal essential medium (MEM) containing 10% fetal bovine serum (FBS) with antimicrobials. HEK-293T (kindly provided by Gaëtan Ligat, Infinity Lab, Toulouse) were harvested in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FBS, 2 mM of L-glutamine, and antimicrobials. All cells were placed at 37°C in 5% CO 2 . For BAC mutagenesis, Escherichia coli strain GS1783 was used (40). The HCMV-BAC (bacterial artificial chromosome containing the genome of the CMV laboratory strain AD169) contained a GFP gene in the unique short region and was derived from the parental strain pHB5, the BAC-cloned genome of HCMV laboratory strain AD169 (41). For the viral strains, five reference strains-AD169 (ATCC VR-538), Davis (ATCC VR-807), Towne (ATCC VR-977), Merlin, and Toledo-45 HCMV strain sequences from GenBank and 43 HCMV clinical isolate sequences from samples collected in various hospitals in France for the National Reference Center for Herpesviruses were studied. Clinical samples came from congenitally HCMV-infected neonates or HCMV-infected transplant patients. Twelve of the 43 patients were HSC recipients who had received LTV prophylaxis.
## Antiviral compound
Letermovir was purchased from MedChemExpress (HY-15233), reconstituted in 10 mM dimethyl sulfoxide (DMSO), and stored at -80°C. For antiviral assays, LMV was diluted in cell medium to a final concentration of 40 nM before making ½ serial dilutions down to 1.25 nM. For resistant strains, additional antiviral dilutions were made, ranging from 40 µM to 10 nM with a final DMSO concentration below 0.1%.
## Amplification and sequencing of the UL77 and the UL93 genes from reference strains and isolates
Complete genes were amplified after DNA extraction (E-Mag, bioMérieux, Craponne, France) from whole blood samples or from isolates by nested PCR using external and internal primers (Table S4) as described (42). Purified internal PCR products were sequenced (Applied Biosystems, Villebon-sur-Yvette, France) with sequencing primers designed using Geneious 9.1.8 software. Primers' specificity was assessed by alignment with GenBank strains. Sequencing results were analyzed using Geneious 9.1.8 software for comparison with AD169 gene sequences.
## Heatmap construction
The heat maps provide a data matrix where staining gives an overview of mutational differences between strains. They were generated using Excel spreadsheets (Microsoft, Washington, USA). Column width was set at 7 and row height at 0.17. Binary heat maps were plotted to show mutations in blue and the absence of mutations in white.
## Bacterial artificial chromosome mutagenesis
As described (40), to identify crucial amino acids involved in the identified putative motif, highly conserved residues were replaced by an alanine by "en passant" mutagenesis, using a two-step markerless red recombination system for BAC mutagenesis in E. coli strain GS1783 (40). Single mutations were introduced into an EGFP-expressing HCMV BAC (AD169 backbone), producing several mutants. Used primers are described in Table S5. For the pUL77 mutNLS1 construct, the amino acids were mutated in such a way as not to modify the pUL76 sequence. For other mutants, all regions of interest were deleted. For the pUL77 NLS2, three mutants (NLS2#, NLS2*, and NLS2) were construc ted to evaluate the binary character of the region as described (16). Mutations were combined to assess a potential synergistic effect (Fig. 5B). SANGER sequencing prior to transfection for each recombinant virus confirmed the presence of the mutation/dele tion. To ensure that the BAC backbone did not contain any other mutations that could have a negative impact on viral replication, we conducted NGS sequencing on both the original BAC and the mutants. The mutations/deletions were found in all mutant BAC sequences, while other SNPs were detected in genes that aren't essential for viral replication and represent less than 30% of the sequences, both in the original BAC and the mutants (11).
## Transient transfection
HCMV recombinant BACs were purified using the NucleoBond Xtra Midi system (Macherey-Nagel, Düren, Germany) following the manufacturer's instructions. To reconstitute virus mutants, purified recombinant BACs were transfected into MRC-5 cells (bioMérieux, France) with Transfast liposomal reagent (Promega, USA) following the manufacturer's instructions (43). The presence of mutations in the UL77 and UL93 genes of each recombinant virus in culture was confirmed by sequencing after extraction of viral DNA from each strain according to Hirt's procedure (44).
## Plaque assays and growth curve experiments
Assessment of the impact of each mutation on viral fitness was carried out as described previously. We inoculated recombinant and HCMV-BAC AD169 WT strains into 48-well MRC-5 culture with a multiplicity of infection of 0.01. From day 1 to day 7 after inocu lation, the number of fluorescent foci with cytopathic effect was counted to establish viral growth curves for each recombinant. The curves are the mean of three independ ent replicates, and the same for plaque assay results. For plaque assays, EC 50 s were calculated with the free EC 50 calculator tool (https://www.aatbio.com/tools/ic50-calcula tor) (45). For statistical analysis, the Mann-Whitney test was applied (*P < 0.05, **P < 0.01, ***P < 0.001) using GraphPad Prism 8.4.3.
## Cloning of NLS mutants and confocal microscopy
Oligonucleotide primers used for vector construction are listed in Table S6. NLS mutants of UL77 and UL93 were PCR amplified from HCMV-BAC construction with primers mCherry-UL77 forward/XbaI-UL77 reverse and mCherry-UL93 forward/XbaI-UL93 reverse for N conformation of the mCherry gene, respectively, to not inhibit their function as seen in Fig. 5A. The mCherry gene was PCR amplified from the strain MG1655 rpoS-mCherry (46) with the primers EcoRI-mCherry forward/77-mCherry reverse and NheI-mCherry forward/93-mCherry reverse according to the desired conformation.
Assembling the PCR mix contained 100 ng of each obtained fragments. The PCR program was as follows: first denaturation at 98°C for 5 min, followed by 35 cycles of 10 s of denaturation at 98°C, 45 s of hybridization at 58°C, and 10 s of elongation with a final elongation at 72°C for 5 min. Digestion was performed with EcoRI/XbaI and NheI/XbaI for inserts and PCI-neo plasmid (Promega, USA) containing UL77 and UL93, respectively. The mCherry gene was fused with the N-terminal part of the UL77 and UL93 genes. Ligation of inserts and vectors was done with T4 ligase (New England Biolabs) at 16°C overnight in a 3:1 ratio. Ligation products were transformed in the chemically competent DH5-α strain by heat shock (Mix and Go, Ozyme). Transformed bacteria were cultivated on LB agar medium containing kanamycin (25 µg/µL) and ampicillin (25 µg/ µL). Colonies were screened with the following primers pCI-neo F/pCI-neo R. Plasmids were extracted with the Xtra-Midi kit (Macherey-Nagel, Düren, Germany) according to the manufacturer's instructions, and 4 µg of plasmid DNA were transfected into HEK293T cells in 6-well plates using Viafect reagent at a 4:1 ratio according to the manufactur er's instructions (Promega, USA). After 72 h, cells were transferred into Labteck II cell culture chambers and were incubated for an additional 24 h. Cells were washed once with phosphate-buffered saline and fixed with a glacial 7:3 ethanol/acetone mix. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue fluorescence), and actin filaments were stained with AlexaFluor 488 phalloidin (ThermoFisher Scientific, USA) (green fluorescence). After staining, samples were mounted in ExPert mounting medium (CellPath, StatLab, USA) and examined by fluorescence confocal microscopy with Zeiss LSM880 confocal laser scanning microscope (Carl Zeiss, Germany). The same experiment was done with GFP-HCMV-BAC co-transfection. Wheat Germ Agglutinin Alexa Fluor 647 Conjugate was used to stain cell membranes (ThermoFisher Scientific, USA) (far-red fluorescence).
## Western blot analysis
Transfected HEK293T cells were lysed using RIPA buffer, and 50 µg of protein extracts were denatured with 2% β-mercaptoethanol mix and reduced at 70°C for 10 min. Proteins were characterized by SDS-PAGE and StainFree analysis (BioRad). Immunos taining was performed with a rabbit recombinant anti-mCherry monoclonal antibody (1:1,000) (#EPR20579, Abcam, USA) with an HRP goat anti-rabbit polyclonal antibody (1:5,000) (#ab97200, Abcam, USA) and an anti-cytomegalovirus IE1/IE2 mouse antibody (1:1,000) (#ab53495, Abcam, USA) with an HRP polyclonal goat anti-mouse IgG (1:1,000) (#ab6728, Abcam, USA). HRP anti-actin antibody was used to mark actin as a control (1:1,000) (#EPR16769, Abcam, USA). After each hybridization, the membrane was washed three times with TBS-T. PVDF membrane was revealed using FUSION FX Spectra (Vilber Lourmat, France).
## Statistical analysis of microscopy and western blot
To assess the reproducibility and truthfulness of our results, a minimum of three independent replicates of transfection for microscopy were done at different times. For each replicate, a minimum of 15 photos were taken at magnification ×63 to assess the percentage of mCherry expression in cell nuclei. Images were analyzed with the QuPath 0.5.0 platform (47) and ImageJ software (48). For all these data, a statistical t-test was applied using GraphPad Prism 8.4.3: *P < 0.05, **P < 0.01, ***P < 0.001.
## References
1. Torres-Madriz, Boucher (2008) "Immunocompromised hosts: perspectives in the treatment and prophylaxis of cytomegalovirus disease in solid-organ transplant recipients" *Clin Infect Dis*
2. Leruez-Ville, Ville (2017) "Fetal cytomegalovirus infection" *Best Pract Res Clin Obstet Gynaecol*
3. Marty, Ljungman, Chemaly et al. (2017) "Letermovir prophylaxis for cytomegalovirus in hematopoietic-cell transplantation" *N Engl J Med*
4. Limaye, Budde, Humar et al. (2023) "Letermovir vs valganciclovir for prophylaxis of cytomegalovirus in highrisk kidney transplant recipients: a randomized clinical trial" *JAMA*
5. Ligat, Cazal, Hantz (2018) "The human cytomegalovirus terminase complex as an antiviral target: a close-up view" *FEMS Microbiol Rev*
6. Chou, Kleiboeker (2022) "Relative frequency of cytomegalovirus UL56 gene mutations detected in genotypic letermovir resistance testing" *Antiviral Res*
7. Chou (2015) "Rapid in vitro evolution of human cytomegalovirus UL56 mutations that confer letermovir resistance" *Antimicrob Agents Chemother*
8. Chou, Satterwhite, Ercolani (2018) "New locus of drug resistance in the human cytomegalovirus UL56 gene revealed by in vitro exposure to letermovir and ganciclovir" *Antimicrob Agents Chemother*
9. Cherrier, Nasar, Goodlet et al. (2018) "Emergence of letermovir resistance in a lung transplant recipient with ganciclovir-resistant cytomegalovirus infection" *Am J Transplant*
10. Muller, Tilloy, Frobert et al. (2022) "First clinical description of letermovir resistance mutation in cytomegalovirus UL51 gene and potential impact on the terminase complex structure" *Antiviral Res*
11. Ligat, Jacquet, Chou et al. (2017) "Identification of a short sequence in the HCMV terminase pUL56 essential for interaction with pUL89 subunit" *Sci Rep*
12. (2025) *Full-Length Text Journal of Virology*
13. Ligat, Couvreux, Cazal et al. (2019) "Highlighting of a LAGLIDADG and a zing finger motifs located in the pUL56 sequence crucial for HCMV replication" *Viruses*
14. Borst, Bauerfeind, Binz et al. (2016) "The essential human cytomegalovirus proteins pUL77 and pUL93 are structural components necessary for viral genome encapsidation" *J Virol*
15. Borst, Wagner, Binz et al. (2008) "The essential human cytomegalovirus gene UL52 is required for cleavage-packaging of the viral genome" *J Virol*
16. Meissner, Köppen-Rung, Dittmer et al. (2011) "A "coiled-coil" motif is important for oligomerization and DNA binding properties of human cytomegalovirus protein UL77" *PLoS One*
17. Köppen-Rung, Dittmer, Bogner (2016) "Intracellular distribution of capsid-associated pUL77 of human cytomegalovirus and interactions with packaging proteins and pUL93" *J Virol*
18. Thurlow, Murphy, Stow et al. (2006) "Herpes simplex virus type 1 DNA-packaging protein UL17 is required for efficient binding of UL25 to capsids" *J Virol*
19. Bowman, Welschhans, Jayaram et al. (2006) "Structural characterization of the UL25 DNA-packaging protein from herpes simplex virus type 1" *J Virol*
21. Yoakum (1993) "Mapping a putative pyruvoyl decarboxylase active site to human cytomegalovirus open reading frame UL77" *Biochem Biophys Res Commun*
22. Kim, Oh, Kwon et al. (2016) "Involvement of the Nterminal deubiquitinating protease domain of human cytomegalovirus UL48 tegument protein in autoubiquitination, virion stability, and virus entry" *J Virol*
23. Derussy, Boland, Tandon (2016) "Human cytomegalovirus pUL93 links nucleocapsid maturation and nuclear egress" *J Virol*
24. Borst, Harmening, Sanders et al. (2022) "A unique role of the human cytomegalovirus small capsid protein in capsid assembly" *mBio*
25. Li, Pang, Dong et al. (2021) "Structural basis for genome packaging, retention, and ejection in human cytomegalovirus" *Nat Commun*
26. Jih, Liu, Liu et al. (2024) "The incredible bulk: human cytomegalovirus tegument architectures uncovered by AI-empowered cryo-EM" *Sci Adv*
27. Dai, Zhou (2018) "Structure of the herpes simplex virus 1 capsid with associated tegument protein complexes" *Science*
28. Derussy, Tandon (2015) "Human cytomegalovirus pUL93 is required for viral genome cleavage and packaging" *J Virol*
29. Taus, Salmon, Baines (1998) "The herpes simplex virus 1 UL17 gene is required for localization of capsids and major and minor capsid proteins to intranuclear sites where viral DNA is cleaved and packaged" *Virology (Auckl)*
30. Scholtes, Baines (2009) "Effects of major capsid proteins, capsid assembly, and DNA cleavage/packaging on the pUL17/pUL25 complex of herpes simplex virus 1" *J Virol*
31. Newcomb, Homa, Brown (2006) "Herpes simplex virus capsid structure: DNA packaging protein UL25 is located on the external surface of the capsid near the vertices" *J Virol*
32. Toropova, Huffman, Homa et al. (2011) "The herpes simplex virus 1 UL17 protein is the second constituent of the capsid vertexspecific component required for DNA packaging and retention" *J Virol*
33. Klupp, Granzow, Keil et al. (2006) "The capsidassociated UL25 protein of the alphaherpesvirus pseudorabies virus is nonessential for cleavage and encapsidation of genomic DNA but is required for nuclear egress of capsids" *J Virol*
34. Huet, Huffman, Conway et al. (2020) "Role of the herpes simplex virus CVSC proteins at the capsid portal vertex" *J Virol*
35. Wang, Zhu, Mcvoy et al. (2012) "Changes in subcellular localization reveal interactions between human cytomegalovirus terminase subunits" *Virol J*
36. Mitterer, Gantenbein, Birner-Gruenberger et al. (2016) "Nuclear import of dimerized ribosomal protein Rps3 in complex with its chaperone Yar1" *Sci Rep*
37. Muller, Hantz (2023) "Identification of a leucine-zipper motif in pUL51 essential for HCMV replication and potential target for antiviral development" *Antiviral Res*
38. Alkhashrom, Kicuntod, Stillger et al. (2022) "A peptide inhibitor of the human cytomegalovirus core nuclear egress complex" *Pharmaceuticals (Basel)*
39. Sievers, Wilm, Dineen et al. (2011) "Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega" *Mol Syst Biol*
40. Mcwilliam, Li, Uludag et al. (2013) "Analysis tool web services from the EMBL-EBI" *Nucleic Acids Res*
41. Li, Cowley, Uludag et al. (2015) "The EMBL-EBI bioinformatics web and program matic tools framework" *Nucleic Acids Res*
42. Borst, Hahn, Koszinowski et al. (1999) "Cloning of the human cytomegalovirus (HCMV) genome as an infectious bacterial artificial chromosome in Escherichia coli: a new approach for construc tion of HCMV mutants" *J Virol*
43. Chee, Bankier, Beck et al. (1990) "Analysis of the protein-coding content of the sequence of human cytomegalovirus strain AD169" *Curr Top Microbiol Immunol*
44. Muller, Gourin, Baumert et al. (2021) "New insights into human cytomegalovirus pUL52 structure" *Viruses*
45. Hantz, Cotin, Borst et al. (2013) "Novel DNA polymerase mutations conferring cytomegalovirus resistance: input of BACrecombinant phenotyping and 3D model" *Antiviral Res*
46. Hirt (1967) "Selective extraction of polyoma DNA from infected mouse cell cultures" *J Mol Biol*
47. (2025) "IC50 calculator"
48. Maisonneuve, Castro-Camargo, Gerdes (2013) "RETRACTED: (p)ppGpp controls bacterial persistence by stochastic induction of toxinantitoxin activity" *Cell*
49. Bankhead, Loughrey, Fernández et al. (2017) "QuPath: open source software for digital pathology image analysis" *Sci Rep*
50. Schneider, Rasband, Eliceiri (2012) "NIH image to ImageJ: 25 years of image analysis" *Nat Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12708658&blobtype=pdf | # Exploring the association between the lineages of human papillomavirus type 16 and viral physical status in the development of cervical cancer in Iran
Hassan Karami, Rahim Soleimani-Jelodar, Zohreh Khezeli, Zabihollah Shoja, Arash Arashkia, Sajad Varmazyar, Somayeh Jalilvand
## Abstract
Human papillomaviruses (HPVs) are well-established etiological causes of invasive cervical cancer (ICC), with type 16 being the main contributor. This virus is often found to be integrated into the host genome in precancerous lesions and advanced cervical cancers. It is suggested that distinct lineages of HPV-16 differ in their persistence, carcinogenic potential, and geographic distribution. Given that genomic integration is one of the most critical steps in the development of cervical cancer, and lineage D has a higher propensity to integrate the viral genome into the host genome, it is necessary to conduct a study to investigate a possible relationship between HPV-16 lineages and physical status in Iran. In this study, a total of 125 laboratory-confirmed HPV-16-positive cervical samples (comprising 30 normal, 39 precancerous, and 56 cancerous specimens) were analyzed. The DNA level of E2 and E6 viral genes was measured using the quantitative Real-Time PCR method, and the E2/E6 ratio was used to calculate the physical status of the HPV-16 DNA in the studied samples. The full sequence of the E6 gene was also sequenced to determine HPV-16 lineages. Three lineages, A (32.8%), C (0.8%), and D (66.4%) of HPV-16, were found in this study. HPV-16 exhibits different integration profiles, with viral DNA detected in three forms: episomal, mixed, and integrated. The physical status of the genome was statistically different based on histology and age. The integrated form was more prevalent in ICC patients than in CIN I-III and normal groups (P=0.000048). Also, the integrated form of the genome was found in higher amounts in the age group >40 years in comparison to <40 years (P= 0.00117). Regarding lineages, no statistically significant differences were identified between HPV-16 lineages and the integration status. However, when the samples were stratified by histology status, an association between lineage D and integrated form was observed, while no association was found for lineage A. In conclusion, lineages A and D were found to be dominant in the Iranian population. Moreover, an association was found between lineage and integration status, as lineage D has a higher propensity to integrate than lineage A. However, it is recommended that further studies with larger sample sizes from different regions of Iran be conducted to estimate whether a specific lineage or sublineage has a higher chance of integrating into the host genome, persisting, and causing cancer.
of cutaneous and mucosal lesions, ranging from benign lesions (e.g., anogenital warts [mostly caused by HPV-6 and 11]), to invasive tumors of both genitourinary tract-which include cervical, vulvar, vagina, penile, and anal carcinomas-as well as head and neck cancers. Among them, invasive cervical cancer (ICC) is believed to be the main well-established HPV-related cancer, affecting a significant proportion of non-immunized women, with an annual incidence of nearly 600,000 new cases and 340,000 deaths worldwide, and 1.90 per 100,000 women in Iran [4][5][6] . These viruses tend to infect the cervix transformation zone, leading to the development of premalignant cervical lesions classified as cervical intraepithelial neoplasia (CIN) grades I, II, and III, which, if left untreated, can progress to carcinoma 4 . HPV-16 is among the most frequently detected viral types in precancerous cervical lesions and ICC [~70%] 7 .
HPV-16 comprises four distinct lineages (A to D) and sixteen sub-lineages designated as A1-4, B1-4, C1-4, and D1-4, characterized if there were 1-10% and 0.5-1% differences in the nucleotide sequences, respectively 8 . The different biological activities of distinct HPV-16 lineages were shown in the world. Indeed, it was indicated that lineage D had an eight-fold increased risk of progression to cervical cancer in comparison to A1-3 variants 9 .
While most HPVs, including cancer-causing types, resolve within two years after initial detection-60% within one year and 90% within two years 4 -concerns remain regarding the virus's potential to cause a persistent infection in a subset of infections by integrating the genetic material into the genome of the infected cells. This integration may occur as either a single copy or several concatemeric copies harboring full-length genomes, resulting in altered E6 and E7 expression levels by complete or partial E2 ORF disruption with a consequence of functional inactivation 10 . The timing and mechanism underlying HR-HPV DNA integration is yet to be well determined; however, the available information is controversial, with some data reporting the viral episomal forms exclusively in early stages of the disease, while others show integrated forms in cases of normal cytology, suggesting integration as an early carcinogenetic event 11 . Moreover, the current understanding of the integration status across different HPV-16 lineages in malignant lesions is limited, and its prognostic significance for the risk of tumor progression in precancerous lesions remains largely undetermined. However, some studies have shown that lineage D has a higher carcinogenic potential and a greater tendency to integrate the viral genome than the European variants, including A1-A3 9,12,13 . According to previous studies in Iran, it has been indicated that lineage D of HPV-16 was more prevalent in ICC patients than in normal individuals [14][15][16] . Given that genomic integration is one of the most critical steps in the development of cervical cancer, and lineage D had a higher propensity to integrate the viral genome into the host genome, it is necessary to conduct a study to find a possible relationship between HPV-16 lineages and the physical status of the viral genome in Iran.
## Materials and methods
## Study population and sampling
This study involved a total of 129 fresh cervical tissue samples previously confirmed to be HPV16-positive. The inclusion criteria included samples that were diagnosed as HPV16 using L1 sequencing or COBAS assays, and the results of histopathology were determined.
Gynecological and histological examinations classified the cervical tissues according to the cervical intraepithelial neoplasia (CIN) classification system into three groups: 30 cases (24%) with normal histology, 39 cases (31.2%) with CINI-III, and 56 cases (44.8%) with ICC, including 47 cases (83.9%) diagnosed as squamous cell carcinoma (SCC) and 9 cases (16%) as adenocarcinoma (AdC). The median age was 40 years. Patients were referred to the women's clinic at Imam Khomeini Hospital or Yas Hospital in Tehran, Iran, between 2022 and 2023. Informed consent was obtained from all participants after a verbal explanation of the study's aims and importance, and the study was approved by the ethics committee of Tehran University of Medical Sciences (TUMS) (IR.TUMS.SPH.REC.1402.126) following the Helsinki Declaration. Demographic data and histopathological diagnoses were collected from participants' medical records.
## DNA extraction and the investigation of lineages of HPV-16
HPV-16 DNA was isolated from tissue specimens by phenol-chloroform assay based on the previously performed procedure 17 and stored at -20 °C until use. To identify lineages of HPV-16, the entire E6 region was amplified and sequenced according to a previously published procedure 14 . As E6 and E7 regions encode oncoproteins of HPV, and the length of the E6 gene was longer than the E7 gene, and the number of mutations was greater, the E6 gene was selected to investigate.
In brief, the nucleotide sequences of HPV 16 E6 (nucleotide 83-559) were investigated by PCR with the following primer pair: 5
## '-C C G A A A C C G G T T A G T A T A A A A G C A-3' and 5'-C A G T T G T C T C T G G T T G C
A A A T C T 3' to amplify a 571 bp amplicon. The PCR reactions and the thermal cycle conditions were done according to our previous study 14 . The PCR reaction was done in a 50 μl reaction mixture containing 100-200 ng of DNA template, 10 pmol of each primer, 2.5 mM MgCl2, 50 μM of each dNTP, and 2 U of Taq DNA polymerase. PCR amplification cycles included an initial 5-minute denaturation at 95°C, followed by 45 cycles of 95°C for 40s, 55°C for 50s, and 72°C for 50s, and a final elongation at 72°C for 5 min. A reaction mixture lacking template DNA, as a negative control, was included in every set of PCR runs.
To investigate the HPV 16 E6 gene variations, all the PCR products were subjected to sequence using bidirectional sequencing with BigDye® Terminator v3.1 Cycle Sequencing Kit and a 3130 Genetic Analyzer Automated Sequencer as specified by Applied Biosystems manuals (Foster City, CA). Obtained sequences were edited by Bioedit software and converted to FASTA format. Then, our sequences were aligned with reference sequences (A1-4, B1-4, C1, C3, C4, and D1-4) that obtained from Home -Nucleotide -NCBI with the following accession numbers: K02718, AF536179, HQ644236, AF534061, AF536180, HQ644298, KU053915, KU053914, AF472509, KU053920, KU053925, HQ644257, AY686579, AF402678, and KU053931 to characterize the (sub) lineages in Bioedit software.
## The physical status of the HPV-16 genome measurement
The DNA level of two viral genes-E2 and E6-were quantified by the method of absolute quantitative Real-Time PCR (qRT-PCR) using the specific
## primers (E2 [F: A C A C A G A C G A C T A T C C A G C G and R: C C G T C C T T T G T G T G A G C T G T] and E6 [F: A A T G T T T C A G G A C C C A C A G G and R: G T T G C T T G C A G T A C A C A C A T T C]
) to investigate the HPV integration status distinguishing the episomal, integrated, and mixed forms following the E2/E6 ratio calculation 18 . Each reaction consisted of 12.5 µL of SYBR® Premix Ex Taq™, 1 µL of each primer (10 pmol), and 1.5 µL of purified DNA in a total volume of 25 µL. In detail, the Real-Time PCR cycling conditions for HPV-16 E2 and E6 were as follows: initial denaturation (3 min at 95°C), followed by 45 cycles of denaturation (10 s at 95°C), annealing (40 s at 55°C), and extension at 60°C for 20 s. All samples were tested in duplicate, and a reaction mixture lacking a template as a negative control was used in each run. A melting curve analysis was performed to verify single gene-specific peaks by heating samples from 72°C to 98°C at the end of the amplification cycles. To ensure assay standardization, the recombinant plasmid pUC57 harboring the E2 and E6 genes of HPV-16 was employed as a quantitative reference for determining HPV E2 and E6 copy number. The assays were evaluated for linearity, sensitivity, specificity, and reproducibility to achieve optimal performance. A standard curve was generated using 10-fold serial dilutions of the recombinant plasmid ranging from 1.15×10⁶ to 1.15×10 1 copies, and the mean cycle threshold (Ct) values of replicate wells were plotted against the corresponding DNA copy numbers. The limit of detection (LoD) was defined as the lowest concentration at which the assay maintained linearity. Intra-assay variability was assessed using three independently prepared dilution series within a single experimental run, whereas inter-assay variability was determined using identical dilution series across three separate runs. The repeatability and reproducibility of the assays were expressed as the percentage of total variance. After quantification, the E2:E6 ratio was calculated. The ratios 0, 1, and <1 were considered for integrated, episomal, and mixed (a mixture of integrated and episomal forms) viral genomes, respectively 18 .
## Statistical analysis
The Mantel-Haenszel χ2 or Fisher exact tests (two-sided) using Epi Info 7; Statistical Analysis System Software was used for data analysis. The P values <0.05 were regarded as statistically significant.
## Results
One hundred twenty-five samples were successfully sequenced, while 4 samples failed to sequence. Our findings revealed that three lineages A, C, and D were found in our samples. The most common lineage was D (66.4%), followed by A (32.8%) and C (0.8%) lineages. All samples of the lineage D belonged to sublineage D1/4 (these two lineages cannot be distinguished as solely D1 or D4 based on the E6 sequence alone). Of the samples belonging to the A lineage, 21 samples (51.2%) were classified in sublineage A1 and 20 samples (48.8%) in sublineage A2. The only detected sample of the C lineage belonged to sublineage C1. Sequence analysis showed that 14 nucleotide substitutions occurred in the entire E6 gene (Table 1). Particular variants, including patterns 2, 3, 5, 8, and 9, were re-sequenced and confirmed that they were not the errors of sequencing. Of these 14 substitutions, 7 substitutions at positions G145T, A131G, A162G, G176A/C, C315G, C335T, T350G, resulted in amino acid changes at positions Q14H, R10G/I, Q20R, D25H/N, S71C, H78Y, L83V of the E6 protein, respectively. At least one amino acid change was observed in most samples (82.4%). Among these 7 amino acid changes, the most frequent change was L83V, which was observed in 103 samples (82.4%), followed by two changes, Q14H and H78Y, which were found in 84 samples (67.2%). The change R10G/I was observed in 4 samples (3.2%), and the three changes C71S, D25H/N, and Q20R were observed in one sample each.
To determine the genomic integration status of 125 studied HPV-16-positive cervical tissues, the expression levels of the E2 and E6 genes were analyzed, and the ratios were calculated. In total, 12 (9.6%), 79 (63.2%), and 34 (27.2%) samples were found to be positive for episomal (E2/E6 ratio=1), mixed (E2/E6 ratio <1), and integrated (E2/E6 ratio=0) HPV-16 DNA, respectively. The results of our investigation revealed that the E2/ E6 ratio was significantly different in the tumor stages, with the most integration being detected in advanced cervical dysplasia (P=0.000048) (Table 2). In other words, the episomal form of DNA was observed in 33.3%, 7.7%, and 3.6% of normal, CIN I-III, and ICC cases, respectively, with frequency decreasing as histological severity increased. Conversely, the integrated form was detected in none of the normal group, 23.1% of the CIN I-III group, and 44.6% of the ICC group.
The results also reveal a statistically significant association between the age of patients and the viral genomic integration status (P=0.00117), with a notable increase in the integration rate among middle-aged and older patients. In our study, most cases harboring an integrated viral genome were diagnosed in women above 40 years (40.6%). Conversely, the episomal and mixed forms were more frequently detected in younger women (aged under 40 years). A comparison of the frequency of HPV-16 genomic integration based on the type of cervical cancer was made. The result showed that the frequency of absolute integration in AdC samples (55.6%) was found to be higher than in the SCC samples (42.5%). However, this difference was not statistically significant (P=0.259).
As shown in Table 2, no significant association was identified between HPV-16 lineages and the integration status (P=0.85). As indicated in Table 3, the HPV-16 lineages and the integration status were investigated with regard to histology status. In the A lineage group, the episomal form was detected in 18.2%, 7.2%, and 6.3% of normal, CIN I-III, and ICC cases, and no statistically significant differences were observed (P=0.116). However, in the D lineage group, the episomal form was detected at a lower frequency in ICC cases (2.5%) than 22.2% and 8% among normal and CIN I-III samples, respectively. This difference reached a statistically significant level (P=0.0028).
In the analysis of HPV-16 lineages and integration status, which was stratified by the type of ICC samples (SCC or AdC), the increased integrated form was found in lineage D than lineage A in the AdC group. However, Table 1. Comparison of nucleotide polymorphic patterns of the papillomavirus 16 E6 gene in cervical samples against reference sequences this difference did not reach a statistically significant level (P=0.489). Also, among SCC patients, no statistically significant differences were found in this regard (P=0.845) (Fig. 1).
As shown in Table 4, the frequency of genomic integration status in terms of mutation at position 350 of the E6 gene of HPV-16 was also investigated. Although the integrated form was higher among samples with G mutation (30.1%) compared to the wild type nucleotide (T) (13.1%) at position 350 of the E6 gene, this difference was not statistically significant (P=0.265).
## Discussion
In this study, three lineages, A, C, and D, were identified in 32.8%, 0.8%, and 66.4% of samples, respectively. Our finding is in accordance with previous studies in Iran, which reported lineage D as the dominant lineage which followed by the A lineage 15,16,19 . It is suggested that the distribution of distinct HPV-16 lineages is populationdependent, and their geographical spreading can vary due to evolution related to the host population's ethnicity [20][21][22] . In a global study, it was found that the A1 and A2 sublineages were most prevalent in Europe, South/Central America, North America, South Asia, and Oceania, while the A3 and A4 were most common in East Asia. Lineages B and C were found only in African samples. Lineage D was more prevalent in South/Central America and North Africa 23 .
The present study found three forms of HPV-16 DNA: episomal, mixed, and integrated in cervical samples. Although a large proportion of HPV-related cancers harbor integrated viral DNA, this is not always the case, as these cancers can also contain either extrachromosomal viral DNA (episomal) or a mix of episomal and integrated forms 24 . This implies that the dysregulation of E6 and E7 gene expression can be observed without DNA integration 25 .
Our findings showed that there is a statistically significant difference in the physical state of HPV-16 DNA across the various stages of cervical lesions, as the integrated viral DNA form was highly prevalent in CIN I-III and cancer lesions compared to controls. This finding is in line with previous studies, which show that a high copy of integrated HPV-16 DNA can be detected in high-grade cervical lesions and is associated with a poor disease prognosis [26][27][28] . Therefore, the examination of HPV-16 physical status is reported to be a promising test providing
CIN I-III 1 ( insight into CC risk 29 . However, in 76.7% of normal samples, a mixed form of HPV-16 genome integration was detected. It is worth mentioning that the rate of integration was lower than 30% in the normal group, while the rate of integration was more than 30% in the CIN I-III and malignant groups. Consistent with Kulmala's study 30 , we identified HPV DNA in mixed form, the most commonly reported physical state in women with normal cervical histology. However, the integrated form was absent among the normal group in the present study. The mixed form of DNA may be a common phenomenon in HPV-16 infection, which could be observed not only in high-grade lesions and ICC but also in low-grade lesions and normal samples infected with the virus 18,27,30 , suggesting that HPV-16 integration may occur in the early stages of cervical neoplastic transformation 27,30 . In our study, the absolute episomal form was also detected in CIN I-III and malignant samples. This finding is in agreement with previous studies reporting the presence of episomal virus in cervical tumors 4 .
The age of patients is also regarded as a risk factor impacting the progression of cervical cancer. In contrast to the results reported by Karbalaie Niya et al. 31 , we found a statistically significant association between age and the frequency of viral genome integration. Compared to HPV-16-infected women with episomal status, those harboring the pure integrated viral DNA tend to be older, showing similarity with the cases in the literature 32 . Although the prevalence of the integrated form of HPV-16 DNA was higher in AdC than in SCC, no statistically significant differences were observed.
Considering genetic differences among HPV-16 lineages, there is a gap in our knowledge regarding whether such small-scale genetic variations influence the frequency of integration. To explore this, we assessed the integration status based on distinct HPV-16 lineages. In total, no integration differences were found between A and D lineages. However, when stratification concerning histology was done, our results indicated that a statistically significant difference was observed for D lineage, as this lineage had a greater tendency to integrate than the A lineage. A study using a three-dimensional organotypic model that supports the natural cycle of the virus showed that the Asian-American variant (lineage D) integrated into the host genome, but the European variant did not. The results of this study showed that lineage D has a greater predisposition to integrate into the host genome 13 . Some studies extensively investigated the genetic variability of HPV-16 by examining the sequence of the E6 oncogene, aiming to discover nucleotide variations and amino acid substitutions impacting the oncogenicity of the virus and subsequently the initiation and progression of ICC 33,34 . The polymorphic mutation most frequently detected in non-European variants is the T350G mutation, which changes leucine to proline (L83V) 35 . In the present study, the frequency of genomic integration status in terms of mutation at position 350 of the E6 gene was evaluated, and the result indicated that variants with a G mutation had a higher integrated form than the wild-type nucleotide (T) at this position. However, no statistically significant differences were found, which may be due to the low sample size in this study.
The most important limitations of this study were the moderately sample size and the lack of differentiation D1 and D4 lineages solely based on the E6 sequence alone.
In conclusion, the results of this study indicate that two lineages, A and D of HPV-16, are common in the Iranian population. Also, our findings reaffirm the crucial role of integration as a key event in HPV-16 carcinogenesis, with integration showing a stronger association with ICC development, and confirm viral integration as a hallmark of ICC development. Regarding lineages, no statistically significant differences were identified between HPV-16 lineages and the integration status. However, when the samples were stratified by histology status, an association between lineage D and integrated form was observed, while no association was found for lineage A. It is recommended that further studies with larger sample sizes from different regions of Iran be conducted to estimate whether a specific lineage or sublineage has a higher chance of integrating into the host genome, persisting, and causing cancer.
## References
1. Mlynarczyk-Bonikowska, Rudnicka (2024) "HPV infections-classification, pathogenesis, and potential new therapies" *Int. J. Mol. Sci*
2. Lagström (2021) "HPV16 and HPV18 type-specific APOBEC3 and integration profiles in different diagnostic categories of cervical samples" *Tumour Virus Res*
3. Burd (2003) "Human papillomavirus and cervical cancer" *Clin. Microbiol. Rev*
4. Vallejo-Ruiz, Gutierrez-Xicotencatl, Medina-Contreras et al. (2024) "Molecular aspects of cervical cancer: a pathogenesis update" *Front. Oncol*
5. Anusha, Brahman, Sesharamsingh et al. (2025) "Electrochemical detection of cervical cancer biomarkers" *Clin. Chim. Acta*
6. Anjam Majoumerd (2024) "Epidemiology of cervical cancer in Iran in 2016: A nationwide study of incidence and regional variation" *Cancer Rep*
7. Van Den Borst, Bell, Van Camp et al. (1940) "Lineages and sublineages of high-risk human papillomavirus types associated with cervical cancer and precancer: A systematic review and meta-analysis" *J. Nat. Cancer Inst. -Bethesda. Md*
8. Burk, Harari, Chen (2013) "Human papillomavirus genome variants" *Virology*
9. Berumen (2001) "Asian-American variants of human papillomavirus 16 and risk for cervical cancer: A case-control study" *J. Natl. Cancer Inst*
10. Myers (2019) "Detecting episomal or integrated human papillomavirus 16 DNA using an exonuclease V-qPCR-based assay" *Virology*
11. Gudleviciene, Kanopiene, Stumbryte (2014) "Integration of human papillomavirus type 16 in cervical cancer cells" *Open Med*
12. Casas (1999) "Asian-american variants of human papillomavirus type 16 have extensive mutations in the E2 gene and are highly amplified in cervical carcinomas" *Int. J. Cancer*
13. Jackson (2016) "Functional variants of human papillomavirus type 16 demonstrate host genome integration and transcriptional alterations corresponding to their unique cancer epidemiology" *BMC Genom*
14. Khezeli, Shoja, Kaffashian et al. (2025) "The lineage and sublineage investigation of human papillomavirus type 16 in Tehran, Iran, During 2022-2023: A Cross-sectional study" *Health Sci. Rep*
15. Vaezi (2017) "Human papillomavirus type 16 lineage analysis based on E6 region in cervical samples of Iranian women Infection, genetics and evolution" *J. Mol. Epidemiol. Evolut. Genet. Infect. Dis*
16. Salehi-Vaziri (2023) "Lineages and sublineages of human papillomavirus type 16 in cervical samples of Iranian women" *Futur. Virol*
17. Jalilvand (2012) "Molecular epidemiology of human herpesvirus 8 variants in Kaposi's sarcoma from Iranian patients" *Virus Res*
18. Peitsaro, Johansson, Syrjänen (2002) "Integrated human papillomavirus type 16 is frequently found in cervical cancer precursors as demonstrated by a novel quantitative real-time PCR technique" *J. Clin. Microbiol*
19. Farhadi (2023) "Type distribution of human papillomaviruses in ThinPrep cytology samples and HPV16/18 E6 gene variations in FFPE cervical cancer specimens in Fars province" *Iran. Cancer Cell. Int*
20. Cornet (2013) "Human papillomavirus type 16 E6 variants in France and risk of viral persistence" *Infect. Ag. Cancer*
21. Cornet (2012) "Human papillomavirus type 16 genetic variants: Phylogeny and classification based on E6 and LCR" *J. Virol*
22. Pimenoff, De Oliveira, Bravo (2017) "Transmission between archaic and modern human ancestors during the evolution of the oncogenic human papillomavirus 16" *Mol. Biol. Evol*
23. Clifford (2019) "Human papillomavirus 16 sub-lineage dispersal and cervical cancer risk worldwide: Whole viral genome sequences from 7116 HPV16-positive women" *Papillomavirus Res*
24. Kristiansen, Jenkins, Holm (1994) "Coexistence of episomal and integrated HPV16 DNA in squamous cell carcinoma of the cervix" *J. Clin. Pathol*
25. Darwich (2011) "Human papillomavirus genotype distribution and human papillomavirus 16 and human papillomavirus 18 genomic integration in invasive and in situ cervical carcinoma in human immunodeficiency virus-infected women" *Int. J. Gynecol. Cancer*
26. Zheng, Peng, Lou et al. (2007) "Human papillomavirus 16 physical status detection in preinvasive and invasive cervical carcinoma by multiplex real-time polymerase chain reaction. Chin.-Ger" *J. Clin. Oncol*
27. Huang, Chao, Lee (2008) "Integration of human papillomavirus type-16 and type-18 is a very early event in cervical carcinogenesis" *J. Clin. Pathol*
28. Dutta (2015) "Physical and methylation status of human papillomavirus 16 in asymptomatic cervical infections changes with malignant transformation" *J. Clin. Pathol*
29. Gallo (2003) "Study of viral integration of HPV-16 in young patients with LSIL" *J. Clin. Pathol*
30. Kulmala (2006) "Early integration of high copy HPV16 detectable in women with normal and low grade cervical cytology and histology" *J. Clin. Pathol*
31. Karbalaie Niya (2018) "Human papillomavirus type 16 integration analysis by real-time PCR assay in associated cancers" *Transl. Oncol*
32. Kulmala (2006) "Early integration of high copy HPV16 detectable in women with normal and low grade cervical cytology and histology" *J. Clin. Pathol*
33. Bletsa (2021) "Genetic variability of the HPV16 early genes and LCR. Present and future perspectives" *Expert. Rev. Mol. Med*
34. Ortiz-Ortiz (2015) "Association of human papillomavirus 16 E6 variants with cervical carcinoma and precursor lesions in women from Southern Mexico" *Virol. j*
35. Huertas-Salgado (2011) "E6 molecular variants of human papillomavirus (HPV) type 16: an updated and unified criterion for clustering and nomenclature" *Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12709833&blobtype=pdf | # BMC Infectious Diseases
Ulugbek Mirzaev, Malokhat Palvannazirovna Oltieva, Dilfuza Alieva, Sitorabonu Bahodir Kizi Abdusattorova, Iroda Pulatovna, Jasur Juraev, Tohirbek Shokir, Shodiev, Erkin Musabaev, Mirzarakhim Baynazarov, Khasanjon Kobiljon Ugli Odilov, Nurmuhammad Jaloliddin, Ugli Jamoliddinov, Ulugbek Khudayberdievich Mirzaev
## Abstract
Background Cervical cancer imposes a substantial public health burden in Uzbekistan, yet comprehensive HPV epidemiological data are lacking. This study characterized high-risk HPV prevalence, genotype distribution, risk factors, and screening coverage in three key regions.
MethodsWe analyzed data from 44,497 women aged ≥ 20 years attending gynecological services in Tashkent, Andijan, and Samarkand (2021-2023). Liquid-based cytology and HPV DNA testing for 12 high-risk genotypes were performed. Sociodemographic, reproductive, behavioral, and clinical variables were collected via standardized questionnaires and examination. Multivariable logistic regression identified associated factors of high-risk HPV infection.Results High-risk HPV was detected in 3,779 women (8.5%), with regional prevalence of 9.2% in Andijan, 8.8% in Tashkent, and 6.5% in Samarkand. Mixed genotype infections comprised 52.8% of cases; HPV-16/18 accounted for 41.1%. Among HPV-positive women undergoing cytology (n = 2,275), 64.1% had negative findings, 14.7% LSIL, 8.4% ASC-US, 6.9% HSIL, 5.1% ASC-H, 0.5% atypical glandular cells, and 0.3% carcinoma. Independent predictors of HPV infection included younger age, single marital status, lower parity, abortion history, and hormonal contraception (all p < 0.05). Foamy vaginal discharge was a significant clinical predictor (adjusted OR 1.53; p < 0.01).
ConclusionsUzbekistan exhibits an intermediate high-risk HPV burden, dominated by vaccine-preventable genotypes, and low screening coverage. Targeted outreach to identified high-risk groups-by age, reproductive history, and clinical features-will optimize resource allocation. These data support accelerated progress toward WHO elimination targets in Uzbekistan.
## Introduction
Human papillomavirus (HPV) infection represents a critical global public health challenge, with persistent infection by high-risk genotypes recognized as the primary etiologic agent in cervical carcinogenesis. Globally, cervical cancer remains the fourth most common cancer among women, accounting for approximately 660,000 new cases and 350,000 deaths annually, with the highest burden concentrated in low-and middle-income countries. The disproportionate impact on resource-limited settings reflects significant disparities in access to HPV vaccination, cervical screening programs, and comprehensive treatment services [1][2][3]. In Uzbekistan, cervical cancer represents a significant public health challenge, ranking as the second most common cancer among women (11.9% of female cancers) and accounting for 5.1% of all cancers. The age-standardized incidence rate is 11.3 per 100,000 women, with approximately 1,887 new cases and 1,103 deaths annually [4].
The epidemiological landscape of HPV infection exhibits considerable geographical variation, with prevalence rates ranging from less than 3% in developed regions such as Australia and the United States to 26% in sub-Saharan Africa. In Central Asia, limited surveillance data suggests HPV prevalence rates of 43-56% among screened populations, significantly exceeding global averages and highlighting the urgent need for comprehensive epidemiological assessment in this region. Kazakhstan, as a neighboring country to Uzbekistan, reports cervical cancer incidence rates of 16.3% in northern regions and 12.7% in western areas, substantially higher than the less than 9% observed in developed European and North American countries [1,5,6].
The complexity of HPV epidemiology extends beyond simple prevalence measurements to encompass genotype-specific distributions and their differential oncogenic potential. While HPV types 16 and 18 are globally responsible for approximately 71% of cervical cancers, regional variations in genotype distribution have significant implications for prevention strategies and clinical management. Mixed genotype infections, which can comprise up to 50% of HPV-positive cases in some populations, present additional clinical challenges, as studies have demonstrated associations between multiple HPV infections and increased disease recurrence and mortality [7][8][9]. The epidemiological complexity of HPV infection is further influenced by HIV co-infection, which significantly increases HPV acquisition, persistence, and progression to high-grade lesions globally. While HIV prevalence data for Uzbekistan remain limited, the potential interaction between HIV and HPV infections represents an important consideration in cervical cancer prevention strategies [10].
Contemporary screening approaches increasingly recognize the superior sensitivity of HPV DNA testing compared to conventional cytological methods, leading to evolving management paradigms that stratify women based on HPV status and cytological findings. The clinical significance of cytological abnormalities among HPV-positive women varies substantially, with negative for intraepithelial lesion or malignancy (NILM) results comprising 60-70% of HPV-positive cases, while atypical squamous cells of undetermined significance (ASC-US), low-grade squamous intraepithelial lesions (LSIL), and high-grade squamous intraepithelial lesions (HSIL) represent progressively increasing cancer risk categories [11][12][13][14]. Beyond squamous cell abnormalities, cytological assessment also encompasses atypical glandular cells (AGC), which, though less common, represent an important category requiring clinical follow-up due to potential association with adenocarcinoma [11][12][13][14].
The identification of sociodemographic and behavioral risk factors associated with HPV infection has consistently demonstrated the importance of age, sexual behavior patterns, reproductive history, and socioeconomic determinants. Younger age represents the most robust predictor of HPV positivity, with infection rates typically peaking in women under 30 years and declining with increasing age. Reproductive factors, including parity, contraceptive use, and pregnancy history, exhibit complex associations with HPV acquisition and persistence, with nulliparity and specific contraceptive methods showing variable risk profiles across different populations [15][16][17].
In the context of Uzbekistan's evolving cervical cancer prevention landscape, recent policy initiatives have established comprehensive screening programs utilizing HPV DNA testing and liquid-based cytology for women aged 30-50 years. The country's HPV vaccination program was implemented in October 2019, initially targeting girls aged 9 years through school-based delivery using a 2-dose schedule of quadrivalent vaccine. The successful implementation achieved 90% coverage among target populations by 2020, providing a foundation for integrated prevention strategies. However, the absence of comprehensive epidemiological data characterizing HPV prevalence, genotype distribution, and associated risk factors among Uzbek women represents a critical knowledge gap that limits evidence-based policy development and clinical decision-making [18][19][20].
A previous epidemiological study conducted by Sharipova et al. examined high-risk HPV prevalence and genotype distribution in a large cohort of women across three regions of Uzbekistan between 2021 and 2023 [9]. The present investigation represents an expanded secondary analysis of this dataset, incorporating comprehensive cytological assessments and employing multivariable statistical modeling to identify independent risk factors for HPV infection and associated cytological abnormalities. This enhanced analytical approach enables the exploration of complex associations between viral, host, and environmental factors that were not fully examined in the original descriptive analysis.
Understanding the multifaceted epidemiology of HPV infection through advanced analytical methods is essential for optimizing screening protocols, developing targeted interventions for high-risk populations, and establishing benchmarks for monitoring prevention program effectiveness. The complex interplay between viral factors, host characteristics, and healthcare system variables necessitates comprehensive multivariable analysis to inform public health strategies and clinical management guidelines tailored to the specific epidemiological context of Central Asian populations.
## Materials and methods
## Study design and setting
This cross-sectional study represents an expanded secondary analysis of a comprehensive dataset originally collected through a collaborative initiative between the Research Institute of Virology of Uzbekistan and the Korea Foundation for International Healthcare [9].
## Data enhancement and preprocessing
Data enhancement and preprocessing were conducted by the research team at the Research Institute of Virology in collaboration with regional partners. The refinement involved systematic identification and correction of inconsistencies, duplicate record removal, standardization of variable formats, and validation of data integrity across all collected parameters. Additional data were sourced from original clinical records and laboratory databases to optimize completeness for multivariable analysis. Further, data enrichment techniques were applied to optimize the dataset for advanced statistical modeling and multivariable analysis. These preprocessing steps resulted in an enhanced dataset comprising 44,497 women, representing a net increase from the original cohort due to improved data recovery and validation processes.
## Study population and sampling
This study utilized specimens and data collected during a comprehensive population-based cervical cancer screening program conducted between 2021 and 2023 in Tashkent, Andijan, and Samarkand regions of Uzbekistan. The detailed methodology regarding participant recruitment, inclusion and exclusion criteria, and baseline demographic data collection has been previously described [9].
## Clinical assessment and gynecological examination
All women underwent a gynecological examination by trained medical professionals at healthcare facilities. The clinical assessment included detailed evaluation of external genitalia, speculum examination of the vagina and cervix, bimanual pelvic examination, and vaginal discharge characteristics [Supplementary 1].
## Sample collection and processing
Cervical specimens were collected using Novaprep® Vial Test liquid-based cytology (Novacyt, Vélizy-Villacoublay, France). Trained medical professionals obtained samples from the cervical transformation zone with sterile cytobrushes, immediately suspending cellular material in transport medium with mucolytic agents (TCM-Ampli-Sens, Moscow, Russia). The samples transportation and storage were described in previous study [9].
## DNA extraction
DNA extraction was performed using the AmpliSens DNA-sorb-AM nucleic acid extraction kit (InterLab Service Ltd., Moscow, Russia) following standardized protocols. The procedure involved lysis of cellular material in specialized buffer solutions, followed by DNA purification using magnetic bead-based separation technology to ensure high-quality nucleic acid recovery.
High-risk HPV detection and genotyping were conducted using real-time polymerase chain reaction (PCR) methodology with the AmpliSens HPV HCR screen-14-titre-FL and AmpliSens HPV HCR genotype-titre-FL kits (InterLab Service Ltd.). The assay targeted 12 highrisk HPV genotypes (HPV 16,18,31,33,35,39,45,51,52,56,58,59) based on established oncogenic potential classifications.
Thermocycling was performed using Rotor-Gene Q systems (Qiagen, USA) in Tashkent and Samarkand facilities, while Andijan utilized the DLAB RT-PCR System Accurate 96 (DLAB SCIENTIFIC CO., LTD., Beijing, China). Due to reagent supply constraints during the COVID-19 pandemic period in Andijan, some testing periods excluded HPV types 66 and 68 from analysis, maintaining consistency with 12 high-risk genotypes across all regions.
## Cytological assessment
Liquid-based cytology slides were prepared using standardized monolayer techniques and interpreted by certified cytotechnologists and pathologists according to Bethesda System classification (NILM, ASC-US, LSIL, HSIL, or malignant cells). Cytological assessment was not performed for all HPV-positive women due to healthcare system capacity constraints, patient factors, and logistical considerations during the study period. Priority was given to women with clinical symptoms or abnormal cervical findings during gynecological examination.
## Data collection and quality control
A structured questionnaire captured sociodemographic, behavioral, and clinical data including age, marital status, reproductive history, sexual behavior, smoking status, and cervical screening history [Supplementary 2]. Quality control procedures included standardized training, regular equipment calibration, positive and negative controls in each PCR run, systematic data validation, and inter-laboratory concordance monitoring.
## Statistical analysis framework
The analytical approach employed both descriptive and inferential statistical methods to address the study objectives. Univariate analyses examined associations between individual risk factors and HPV positivity using chi-square tests for categorical variables. Variables with p-value less than 0.25 in univariate analysis were included in multivariable analysis. Multivariable logistic regression models were constructed to identify independent predictors of high-risk HPV infection, with adjustment for potential confounding variables and assessment of interaction effects. Statistical significance was determined at p < 0.05, with all analyses conducted using JMP version 18.0 (SAS Institute, Inc., Cary, NC, USA).
## Results
This cross-sectional study examined high-risk HPV infection among 44,497 women in Uzbekistan (Table 1). High-risk HPV genotypes were detected in 3,780 women (8.5%). Significant geographic variation was observed, with Andijan exhibiting the highest prevalence (9.2%), followed by Tashkent (8.8%) and Samarkand (6.5%). Among HPV-positive cases, mixed genotype infections predominated (52.8%) over single genotype infections (47.2%). HPV types 16/18 were detected in 41.1% of positive cases, while other high-risk genotypes accounted for 59.0% of infections (Table 2). Among the high-risk genotypes detected beyond HPV 16/18, the most prevalent individual types were 31 (9.8%), 51 (8.8%), 52 (8.6%), etc [9].
Cytological evaluation was performed in 2,275 (60.2%) of HPV-positive women. Normal cytology (NILM) was observed in 64.1% of tested cases. Cytological abnormalities were distributed as follows: LSIL (14.7%), ASC-US (8.4%), HSIL (6.9%), ASC-H (5.1%), atypical glandular cells (0.5%), and squamous cell carcinoma (0.3%). Overall, 35.9% of HPV-positive women who underwent cytological screening demonstrated abnormal findings. Although uncommon, AGC represents a clinically significant finding due to its association with cervical adenocarcinoma and adenocarcinoma in situ (AIS). The detection of glandular abnormalities in our cohort, though infrequent, highlights the importance of comprehensive cytological evaluation beyond squamous cell changes (Table 2).
HPV prevalence was highest among women aged 20-29 years (13.0%). All older age groups demonstrated significantly reduced HPV risk, with the most pronounced reduction in women aged 40-49 years (AOR 0.62, p < 0.01).
Married women demonstrated significantly lower HPV risk compared to single women (AOR 0.60, p < 0.01), while divorced women showed elevated risk (AOR 1.67, p < 0.01). Age of sexual debut showed no significant associations.
Increasing parity conferred protective effects, with significant risk reduction observed in women with 3-4 births (AOR 0.65, p = 0.02) and ≥ 5 births (AOR 0.57, p < 0.01). Abortion history increased HPV risk, with women having 1-2 abortions (AOR 1.15, p < 0.01) and 3-5 abortions (AOR 1.38, p < 0.01) showing significant associations.
No occupational category showed significant association with HPV infection. Smoking demonstrated elevated but non-significant risk. Among contraceptive methods, hormonal contraception was associated with significantly increased HPV risk (AOR 1.26, p = 0.04).
Foamy vaginal discharge showed the strongest association with HPV infection (AOR 1.53, p < 0.01), while colorless discharge demonstrated marginal protective effects (AOR 0.92, p = 0.04). Post-traumatic cervical changes showed borderline significance (AOR 1.09, p = 0.05) (Table 3).
## Discussion
This comprehensive analysis of 44,447 women represent the largest population-based investigation of high-risk HPV prevalence and associated factors in Uzbekistan to date. The overall HPV prevalence of 8.5% positions Uzbekistan within the intermediate global burden range, consistent with epidemiological patterns observed in middle-income countries undergoing demographic transition.
Our findings align with global HPV prevalence patterns, where intermediate prevalence rates are characteristic of transitioning societies. The 8.5% prevalence observed in Uzbekistan is considerably higher than conservative estimates from neighboring Iran (3%) but substantially lower than rates reported in Kazakhstan (39%) [1,5,6]. Recent studies from other regions demonstrate similar intermediate prevalence levels, with Bangladesh reporting 2.6% [21], Saudi Arabia 4.7% [22], and Greece 8.8% [23]. These regional variations reflect complex interactions of cultural, socioeconomic, and behavioral factors influencing HPV transmission dynamics across populations [24]. The pronounced regional differences within Uzbekistan, with Tashkent exhibiting the highest prevalence (8.8%), followed by Andijan (9.2%) and Samarkand (6.5%), underscore the importance of geographic stratification in screening program implementation. Similar intra-country variations have been documented in China, where HPV prevalence varies significantly across provinces, and in other developing nations where urbanization patterns influence transmission dynamics.
The age-related HPV distribution pattern observed in our study demonstrates classic epidemiological characteristics, with peak prevalence among women in their twenties (13.0%). These finding parallels global patterns, where HPV acquisition typically occurs shortly after sexual debut. Recent studies from diverse populations consistently show similar age-related trends, with highest infection rates in younger women followed by gradual decline through reproductive years [16,21,23,25].
The substantial HPV burden documented among women aged 30-49 years has critical implications for cervical cancer prevention, as this population represents the primary target for screening programs. Given the established 10-20-year interval between HPV infection and potential cervical cancer development, current prevalence patterns indicate continued cervical cancer risk without appropriate preventive interventions [25,26].
Beyond age-related patterns, our analysis of HPV genotype distribution in single versus multiple infections reveals clinically significant findings. The pattern of increasing high-risk genotype representation in multiple infections represents a clinically significant finding, with HPV 16 prevalence rising from 32.5% in single infections to 59.3% in multiple infections. This observation aligns with emerging evidence suggesting that multiple HPV infections may be associated with increased viral persistence and accelerated progression to high-grade lesions [25,26].
The cytological findings in our study provide complementary insights into disease severity and screening needs. While squamous abnormalities predominated, the detection of atypical glandular cells (AGC) in 0.5% of cytologically assessed samples warrants particular attention. Although AGC represents a relatively rare finding (typically 0.2-0.7% of cervical specimens), it constitutes a high-risk cytological category with significant cancer potential. The clinical significance of AGC is particularly pronounced when associated with HPV 16/18 infection, which confers up to 17% risk of invasive cervical cancer. Cervical adenocarcinoma, predominantly linked to HPV 16 and 18 (78% of cases), presents unique screening challenges due to the lower sensitivity of cytology for glandular versus squamous lesions. The detection of glandular abnormalities in our population emphasizes the value of combined HPV testing and cytological assessment, as HPV-based screening demonstrates superior effectiveness for preventing cervical adenocarcinoma. These findings support comprehensive follow-up protocols for women with AGC, including colposcopy and endocervical sampling, as glandular lesions may extend beyond visible colposcopic examination [27,28].
The documentation that 85.9% of study participants had never undergone cervical cancer screening represents a critical finding highlighting substantial healthcare access gaps. This coverage rate is significantly below WHO targets of 70% screening coverage by 2030 and contrasts sharply with developed countries where screening coverage exceeds 60-85% [29,30].
Recent systematic reviews demonstrate that screening uptake in least developed countries ranges from 4 to 21%, substantially lower than high-income countries. Our findings confirm that Uzbekistan faces similar challenges to other developing nations in implementing effective population-based screening programs, requiring comprehensive healthcare system strengthening initiatives [30,31].
The 6.9% HSIL rate among HPV-positive women in our study substantially exceeds the 0.4-1% prevalence typically observed in well-screened populations, likely reflecting the consequences of inadequate screening coverage. This elevated rate is characteristic of underscreened populations, where irregular or absent screening allows persistent HPV infections to progress to high-grade disease. Studies demonstrate that adherence to screening recommendations significantly reduces HSIL risk, while women with five or more years between screens show substantially higher rates of advanced lesions. These findings underscore the critical importance of expanding organized cervical screening in Uzbekistan, particularly for unvaccinated women and those infected prior to vaccine availability. Evidence demonstrates that regular screening can prevent 64-70% of cervical cancers in unvaccinated populations, approaching the cancer prevention achieved through vaccination alone. Given that approximately 31% of untreated HSIL cases progress to invasive cervical cancer within 10 years, timely detection and treatment of precancerous lesions is essential.
The implementation of systematic screening programs with coverage exceeding 70-80% has proven highly effective in reducing cervical cancer burden globally. In Uzbekistan, where national HPV vaccination achieved 90% coverage since 2019, complementary screening efforts are essential to address disease burden in older unvaccinated cohorts. The transition to primary HPVbased screening offers particular advantages, demonstrating 94% sensitivity for detecting CIN2 + lesions compared to 72% for cytology alone. Our findings highlight that even in the vaccination era, screening remains indispensable for women infected prior to vaccination, emphasizing the need to strengthen screening infrastructure and improve population coverage to accelerate progress toward WHO cervical cancer elimination targets.
The sociodemographic risk factors identified in our analysis align with established global patterns. The protective effect of marriage and increasing parity parallels findings from diverse populations, likely reflecting behavioral and biological mechanisms including partner stability and pregnancy-associated immune modifications. The elevated risk among healthcare workers warrants targeted investigation, as occupational exposure to HPV in healthcare settings has been increasingly recognized [30,32,33]. The association between hormonal contraception and increased HPV risk observed in our study (AOR 1.26) is consistent with meta-analyses demonstrating that longterm hormonal contraceptive use may facilitate HPV acquisition and persistence through local immune suppression mechanisms. The protective effect of condom use emphasizes the importance of comprehensive sexual health education in prevention programs [34].
The strong association between abnormal vaginal discharge characteristics and HPV positivity provides valuable clinical guidance for resource-limited settings. The finding that foamy discharge was significantly associated with HPV infection (AOR 1.53) suggests that clinical assessment may serve as a useful screening adjunct, particularly where molecular testing capacity is limited [35,36].
Several limitations warrant consideration when interpreting our findings. Selection bias represents the most significant limitation, as our study population comprised women seeking gynecological care at tertiary facilities, potentially overrepresenting those with clinical concerns compared to the general population. This may influence the generalizability of our prevalence estimates and risk factor findings.
The cross-sectional design precludes assessment of HPV persistence, clearance, and progression dynamics. Without longitudinal follow-up, we cannot determine clinical outcomes of detected infections or evaluate the effectiveness of current screening protocols. Our testing protocol focused exclusively on high-risk HPV genotypes, excluding low-risk types 6 and 11 that cause anogenital warts. Since anogenital warts typically decrease within years of vaccination implementation and serve as early indicators of program effectiveness, their exclusion limits comprehensive assessment of vaccination impact. Future surveillance should incorporate low-risk HPV detection. Additionally, lack of information regarding participants' vaccination status prevented direct assessment of vaccine effectiveness. Despite these limitations, our study provides valuable baseline data on high-risk HPV prevalence, genotype distribution, and associated risk factors in Uzbekistan.
## Conclusion
With an 8.5% high-risk HPV prevalence driven largely by vaccine-covered genotypes and over 85% of women unscreened, Uzbekistan has a clear path to dramatically reduce cervical cancer. Expanding HPV vaccination to catch-up cohorts and instituting organized screening could prevent most future cases. Targeted outreach to high-risk groups-identified by age, reproductive history, and clinical indicators-will optimize resource use. These data provide the evidence base to accelerate progress toward WHO elimination targets and secure a healthier future for Uzbek women.
## References
1. De Martel, Plummer, Vignat et al. "Worldwide burden of cancer attributable to HPV by site, country and HPV type" *Int J Cancer*
2. Bray, Laversanne, Sung et al. (2024) "Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries" *CA Cancer J Clin*
3. Hpv, Centre (2025)
4. Kongrtay, Kadrlodinova, Sultankulova et al. (2018) "Prevalence of high-grade HPV types among women in Astana" *J Cancer Metastasis Treat*
5. Aimagambetova, Azizan (2018) "Epidemiology of HPV infection and HPV-related cancers in Kazakhstan: a review" *Asian Pac J Cancer Prev*
6. Mendoza, Haidary, Gabutan et al. (2021) "Mixed and nonvaccine high risk HPV types are associated with higher mortality in Black women with cervical cancer. Sci Rep [Internet]"
7. Ye, Jones, Wang et al. (2024) "Comprehensive overview of genotype distribution and prevalence of human papillomavirus in cervical lesions. Gynecology and Obstetrics Clinical Medicine [Internet]"
8. Sharipova, Musabaev, Sadirova et al. (2025) "Prevalence of high-risk human papillomavirus genotypes among women in Uzbekistan, 2021-2023" *J Gynecol Oncol*
9. Gilles, Konopnicki, Rozenberg (2023) "The recent natural history of human papillomavirus cervical infection in women living with HIV: a scoping review of meta-analyses and systematic reviews and the construction of a hypothetical model" *HIV Med*
10. Wentzensen, Schiffman, Palmer et al. (2015) "Triage of HPV positive women in cervical cancer screening" *J Clin Virol*
11. Hpv, Test Results ; Alrajjal, Pansare et al. (2025) "Squamous intraepithelial lesions (SIL: LSIL, HSIL, ASCUS, ASC-H, LSIL-H) of Uterine Cervix and Bethesda System"
12. Liu, Yin, Zhang et al. (2022) "Diagnostic management of oncogenic HPV cervical infections: the field experience in Wuxi, China" *Front Med (Lausanne)*
13. Roman, Andrade, Hernández et al. (2023) "Biological, demographic, and health factors associated with HPV infection in Ecuadorian women"
14. Roik, Sharashova, Kharkova et al. (2018) "Sociodemographic characteristics, sexual behaviour and knowledge about cervical cancer prevention as risk factors for high-risk human papillomavirus infection in Arkhangelsk, North-West Russia" *Int J Circumpolar Health*
15. Sharipova, Mirzaev, Kasimova et al. (2024) "Optimizing Human Papillomavirus (HPV) screening: urine sample analysis and associated factors in Uzbekistan. Cureus [Internet]"
16. Davies, Aluloski, Arifdjanova et al. "HPV vaccination and cervical cancer screening policies and practices in 18 countries, territories and entities across Eastern Europe and Central Asia" *Asian Pacific Journal of Cancer Prevention*
17. (2007) "Uzbekistan. achieves high HPV vaccination coverage against cervical cancer"
18. Chakraborty, Ferdous, Rahman et al. (2024) "Prevalence and genotypic distribution of high-risk human papillomavirus (HPV) among ever-married women in coastal regions of Bangladesh" *PLoS One*
20. Sait, Anfinan, Sait et al. (2024) "Human papillomavirus prevalence and dynamics: insights from a 5-year population-based study in Jeddah, Kingdom of Saudi Arabia" *Saudi Med J*
21. Tsakogiannis, Zografos, Tzioga et al. "Prevalence and genotype distribution of high-risk hpv genotypes among women in Greece: a retrospective analysis of 3500 women" *Cancers (Basel)*
22. Sultana, Khatun (2025) "Cervical cancer in Asian countries: epidemiology, risk factors and challenges" *Int J Reprod Contracept Obstet Gynecol*
23. Na, Li, Wang et al. "The correlation between multiple HPV infections and the occurrence, development, and prognosis of cervical cancer"
24. Su, Ma, Yu et al. "Clinical significance of extended high-risk human papillomavirus genotyping and viral load in cervical cancer and precancerous lesions" *Gynecology and Obstetrics Clinical Medicine*
25. Qin, Deng, Ling et al. (2024) "Our experience diagnosing 225 patients with cervical glandular lesions: current technologies, lessons learned, and areas for improvement" *Diagn Pathol*
26. "22. Available from: h t t p s"
27. Abbas, De Jonge, Bettendorf (2023) "Distribution and incidence of atypical glandular lesions in cervical cytology focusing on the association with highrisk human papillomavirus subtypes" *Oncol Lett*
28. (2025) "Global partners cheer. progress towards eliminating cervical cancer and underline challenges"
29. Rana, Chan, Law et al. (2025) "Determinants of cervical cancer screening utilisation among women in the least developed countries: A systematic review and meta-analysis" *PLoS One*
30. Gopalkrishnan, Karim "Addressing global disparities in cervical cancer burden: a narrative review of emerging strategies"
31. Han, Huang, Ye et al. (2025) "HPV prevalence and genotype distribution in 2,306 patients with cervical squamous cell carcinoma in central and eastern China"
32. Wei, Zhang, Mei et al. (2015) "Prevalence and genotype distribution of HPV6/11/16/18 infections among 180,276 outpatient females from a Women's and Children's Central Hospital"
33. Aimagambetova, Chan, Ukybassova et al. (2021) "Cervical cancer screening and prevention in Kazakhstan and Central Asia" *J Med Screen*
34. Okunade, Adejimi, So et al. (2015) "An overview of HPV screening tests to improve access to cervical cancer screening amongst underserved populations: from development to implementation" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12807482&blobtype=pdf | # The Journal of Clinical Investigation
C O M M E N T A R Y, Elise Armstrong, Joseph Mizgerd
## Alveolar macrophages and pneumonia
Alveolar macrophages (AMs) are constituent cells of the lung that are essential to health and homeostasis (1, 2). Humans or mice lacking AMs due to interrupted granulocyte macrophage-colony-stimulating factor (GM-CSF) signaling, which can occur as a result of genetic deficiencies or autoantibodies, have pulmonary alveolar proteinosis and are susceptible to lower respiratory infections, implicating AM roles in host defense. The defensive role of AMs is borne out by experiments in which the targeted deletion of AMs makes respiratory infections more severe. AMs have many roles that together help protect the lungs during infection. They patrol the alveoli to sound the alarm upon recognizing pathogens, they ingest and kill microbes, and they enhance the resolution and repair of injured lungs via efferocytosis of dying cells and nurturing of the epithelium.
The pool of AMs in the lungs is dynamic. In early life, AMs are products of the embryonic hematopoietic system and are long lived and capable of selfrenewal. With age, growing fractions of the AM pool are instead derived from postnatal monocyte precursors. Severe lung infections rapidly accelerate this process by causing widespread death of AMs as well as by promoting the recruitment of monocytes, which differentiate into macrophages (from which a portion survive past the infection as AMs). The AMs in lungs that recover from infection differ from the AMs of naive lungs in many ways, with altered surface markers, metabolomes, transcriptomes, and responses to subsequent infections (3-7). The resting states and responses of AMs are highly influenced by prior exposures.
Respiratory viral infections cause a transient susceptibility to bacterial superinfection. Because AMs have essential roles in host defense against bacterial pneumonia, including phagocytosis, sentry, resolution, and repair activities, Malainou et al. considered the possibility that the transient loss of AMs during viral pneumonia may contribute to bacterial superinfection severity (8). They endeavored to define mechanisms underlying the AM loss, manipulate the pathways responsible, and test for functional significance.
## TNF superfamily member TNFSF14 signaling mediates AM death
After infecting mice with influenza A virus, Malainou et al. (8) observed that the peak loss of AM correlated temporally with susceptibility to pneumococcal infection. AMs in the influenza-infected mouse lung increased programmed cell death-related transcripts, as well as caspase activation and annexin V positive staining. In vitro and in vivo inhibition of caspase 8 significantly mitigated AM cell death, implicating apoptosis. In the extrinsic pathway of apoptosis, death receptors like members of the TNF receptor superfamily (TNFRSF) lead to signal transduction and caspase activation. Multiple TNFRSF transcripts were upregulated in AMs from infected lungs, including TNFRSF14 most strongly. TNFRSF14 (also known as HVEM) together with lymphotoxin β receptor (LTβR, also known as TNFRSF3) are receptors for TNF superfamily member 14 (TNFSF14, also known as LIGHT), leading to investigation of a TNFSF14-driven signaling axis in AM death.
During influenza infection, most of the apoptotic AMs stained positive for TNFRSF14 and/or LTβR. TNFRSF14 Related Article: https://doi.org/10.1172/JCI185390 Alveolar macrophages (AMs) help defend the lungs against infection, but during pneumonia many alveolar macrophages die. In this issue of the JCI, Malainou et al. explored the mechanism underpinning AM death during viral pneumonia and its effect on the outcomes of bacterial superinfection, a secondary infection that occurs before the first infection is cleared. In mouse models of influenza A infection, recruited neutrophils secreted TNF superfamily member 14 (TNFSF14), and AMs increased expression of the TNFSF14 receptors TNFSFR14 and type I transmembrane lymphotoxin β receptor (LTβR). TNFSF14 signaling via the LTβR was sufficient to cause AM apoptosis. TNFSF14 deficiency or blockade preserved AMs during influenza infection and diminished bacterial burdens and mouse mortality during pneumococcal superinfection. The adoptive transfer of AMs decreased the severity of pneumococcal superinfections, if those AMs lacked the LTβR. Thus, preserving AMs by interrupting TNFRSF14-LTβR interactions can make virus-infected lungs less susceptible to severe bacterial superinfection.
pneumococcal infection of the influenzainfected lung, TNFSF14-deficient mice displayed significantly increased AM abundance, decreased bacterial burdens, mitigated weight loss, and improved survival. Neutralization with anti-TNFSF14 antibodies before bacterial infection was similarly protective. Most impressively, the adoptive transfer of LTβR-deficient AMs during bacterial superinfection improved the survival of superinfected mice, while the adoptive transfers of WT or TNFRS14-deficient AMs did not. Thus, AMs that could not receive LTβR signaling were protective during superinfection.
Altogether, the data from Malainou et al. (8) compellingly reveal a mechanism shared by mice and humans in which and both TNFSF14 and LTβR were necessary for the AM death observed during influenza infection.
Single-cell RNA-Seq suggested neutrophils as sources of TNFSF14 in the infected mouse lung, and airspace neutrophils from patients with influenza ARDS had increased TNFSF14 expression. Depletion of neutrophils reduced soluble TNFSF14 and rescued the AM population in influenza-infected mice, implicating neutrophil-derived TNFSF14 as triggering AM cell death.
After identifying the TNFSF14/ LTβR pathway as mediating AM death during influenza infection, the authors asked whether the pathway contributes to bacterial superinfection severity. During secondary deficiency did not affect AM numbers, but LTβR deficiency prevented AM loss during influenza infection. The LTβR ligand TNFSF14 was increased in the bronchoalveolar lavage (BAL) fluid and was present throughout the lung parenchyma in influenza-infected mice, and patients with influenza or COVID-19 acute respiratory distress syndrome (ARDS) had elevated levels of TNFSF14 in BAL fluids compared with healthy controls. Recombinant TNFSF14 induced caspase activation and apoptosis in both mouse and human AMs. Deletion or blockade of TNFSF14 in mice decreased caspase activation and led to maintenance of the AM population during influenza infection. Thus, TNFSF14 was sufficient for AM death, 8). These AM-depleted, influenza-infected lungs were susceptible to severe pneumococcal superinfection. However, deficiency of TNFSF14 during influenza infection (or its blockade prior to pneumococcal infection) made bacterial superinfection less severe. Although the adoptive transfer of WT or TNFRSF14-deficient AMs did not alter superinfection severity, the adoptive transfer of AMs lacking the LTβR (hence resistant to TNSF14-induced apoptosis) was sufficient to minimize the severity of superinfection. BMDM, bone marrow-derived macrophage.
adults (19). Therefore, LTβR or TNFSF14 blockade during a respiratory infection may compromise future adaptive immune defenses against related respiratory pathogens. Additionally, apoptosis of AMs is a method of killing microbes and is essential for effective defense against some infections, including pneumococcal infections (9,10). The beneficial functions of AM apoptosis may be lost with TNFSF14 or LTβR blockade.
## Summary and outlook
Malainou et al. identified neutrophil-derived TNFSF14 signaling through LTβR on AMs as a trigger for AM apoptosis that was responsible for AM loss during influenza pneumonia, and they further demonstrated that interrupting TNFSF14/ LTβR signaling can protect mice during pneumococcal superinfection. Analyses of patient samples and cultured human AMs support the translational relevance of these findings. Further studies may help determine whether and when TNFSF14 or LTβR blockade might have clinical utility for respiratory viral infections.
## Funding support
This work is the result of NIH funding, in whole or in part, and is subject to the NIH Public Access Policy. Through acceptance of this federal funding, the NIH has been given a right to make the work publicly available in PubMed Central. to COVID-19 (13), the group of patients treated with anti-TNFSF14 had significantly decreased respiratory failure and mortality. However, whether any respiratory failure or mortality in this study may have involved bacterial superinfection was not reported. TNFSF14 or LTβR blockade might also have benefits for post-acute sequelae. For mice exposed to chronic cigarette smoke, a soluble LTβR protein that inhibits TNFSF14-LTβR interaction ameliorates tertiary lymphoid structures, damage-associated transitional epithelial cells, fibrosis, and emphysema ( 14), all of which can be post-acute sequelae of pneumonia. Cytokines from chronically activated CD8 + T cells are responsible for stimulating the monocyte-derived macrophages and aberrant epithelial cells that drive lung fibrosis after viral pneumonia (15,16), so the roles of TNFSF14 in limiting contraction of activated CD8 + T lymphocytes and enhancing their lung residency (17) might mean that these adverse pulmonary sequelae would be mitigated by TNFSF14 blockade during infection. Along with evidence of roles for TNFSF14 and LTβR in safeguarding AMs (8), there is enthusiasm for further considering TNFSF14 or LTβR blockade in viral pneumonias.
Skepticism about potential therapeutic approaches targeting TNFSF14 or the LTβR is also warranted. In studies of mice that recovered from prior infections, which model the lung-resident immunity that helps protect the lungs of healthy young adult humans, AMs display very different phenotypes compared with the AMs of naive mice (3-7), and infectionexperienced mice can be less reliant on AMs for eliminating pneumococci compared with infection-naive mice (4); thus, the roles of AMs in protecting naive mice against superinfection (8) may not directly translate to more experienced lungs, a context more relevant to most humans. Also, immune functions that may be compromised by acute inhibition of TNFSF14 or the LTβR bear consideration. In humans, biallelic loss-of-function LTβR mutations cause a primary immunodeficiency including hypogammaglobulinemia with diminished memory B cells, Tregs, and follicular Th cells (18). TNFSF14 helps program CD8 + T cells into memory and lung-resident states (17), which are pivotal to immune defense in healthy young neutrophil-derived TNFSF14 signals through the LTβR on AMs to trigger cell death during influenza infection, compromising host defense against secondary pneumococcal infection (Figure 1). The observations that interrupting TNFSF14 or delivering LTβR-deficient AMs can be lifesaving during pneumococcal superinfection of influenza-infected mice suggest that AMs do something in these superinfected lungs that is profoundly helpful.
## Implications and arising questions
While these data (8) demonstrate that the loss of AMs in influenza-infected lungs makes pneumococcal superinfection more severe, the AM roles responsible for protection against superinfection are uncertain. The ability of AMs to ingest and kill bacteria is plausibly an important role that is missing after TNFSF14-induced apoptosis, but many other phagocytes are also present in these influenza-infected lungs. Other activities of AMs, like their orchestration of the local immune response and of local resolution and repair (1, 2), may be as important or more important for the outcomes of bacterial superinfection. Unique required roles for AMs in curbing superinfection remain to be identified.
AM death is characteristic of pneumonias in which it has been studied and can occur via multiple mechanisms. For example, during pneumococcal pneumonia, AM apoptosis can be induced by high numbers of bacteria (9) or, in a manner similar to the findings by Malainou et al. (8), by neutrophil-derived TNFSF10 (also known as TRAIL) binding to TNFRSF10B (also known as DR5) on AMs (10). Many different viruses predispose to many types of superinfections. Circumventing AM death may depend on knowing the pathways responsible in a specific infectious setting. TNFSF14 associates with viral infections other than influenza (11,12), so interrupting TNFSF14/LTβR signaling may have broader applicability for preventing bacterial superinfections secondary to respiratory viruses.
Blocking TNFSF14 or the LTβR during severe viral pneumonia is intriguing as a potential therapeutic approach. In a phase II clinical trial that tested the effects of a TNFSF14-blocking antibody in 62 patients with mild-to-moderate ARDS due
## References
1. (2020) "grams the alveolar macrophage pool" *J Clin Invest*
2. Arafa (2022) "Recruitment and training of alveolar macrophages after pneumococcal pneumonia" *JCI Insight*
3. Aegerter (2020) "Influenza-induced monocyte-derived alveolar macrophages confer prolonged antibacterial protection" *Nat Immunol*
4. Roquilly (2020) "Alveolar macrophages are epigenetically altered after inflammation, leading to long-term lung immunoparalysis" *Nat Immunol*
5. Malainou (2026) "TNF superfamily member 14 drives post-influenza depletion of alveolar macrophages, enabling secondary pneumococcal pneumonia" *J Clin Invest*
6. Aberdein (2013) "Alveolar macrophages in pulmonary host defence the unrecognized role of apoptosis as a mechanism of intra-cellular bacterial killing" *Clin Exp Immunol*
7. Steinwede (2012) "TNF-related apoptosis-inducing ligand (TRAIL) exerts therapeutic efficacy for the treatment of pneumococcal pneumonia in mice" *J Exp Med*
8. Fan (2020) "Plasma TNFSF13B and TNFSF14 function as inflammatory indicators of severe adenovirus pneumonia in pediatric patients" *Front Immunol*
9. Perlin (2020) "Levels of the TNF-related cytokine LIGHT increase in hospitalized COVID-19 patients with cytokine release syndrome and ARDS. mSphere"
10. Perlin (2022) "Randomized, double-blind, controlled trial of human anti-LIGHT monoclonal antibody in COVID-19 acute respiratory distress syndrome" *J Clin Invest*
11. Conlon (2020) "Inhibition of LTβR signalling activates WNT-induced regeneration in lung" *Nature*
12. Narasimhan (2024) "An aberrant immune-epithelial progenitor niche drives viral lung sequelae" *Nature*
13. Lin (2024) "Viral infection induces inflammatory signals that coordinate YAP regulation of dysplastic cells in lung alveoli" *J Clin Invest*
14. Desai (2018) "The TNF superfamily molecule LIGHT promotes the generation of circulating and lung-resident memory CD8 T cells following an acute respiratory virus infection" *J Immunol*
15. Ransmayr (2024) "LTβR deficiency causes lymph node aplasia and impaired B cell differentiation" *Sci Immunol*
16. Traber, Mizgerd (2025) "The integrated pulmonary immune response to pneumonia" *Annu Rev Immunol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671154&blobtype=pdf | # Evaluating the susceptibility of various common cell lines and assessing inactivation conditions to Mpox virus
Shu-Chen Hsu, Ping-Cheng Liu, Shan-Ko Tsai, An-Yu Chen, Hui-Ping Tsai, Jun-Ren Sun, Ti-Yu Li, Pei-Yu Hsieh, Jyh-Yuan Yang, Tein-Yao Chang
## Abstract
Over the past decade, numerous infectious diseases have emerged. In 2022, the World Health Organization declared the Mpox outbreak a global public health emergency, as the virus spread rapidly to over 107 countries. As of May 2023, Taiwan continued to observe sporadic instances of the disease. Understanding host cell susceptibility and viral inactivation conditions is crucial for elucidating viral mechanisms and developing effective antiviral therapies or vaccines. In this study, we evaluated the susceptibility of commonly utilized laboratory cell lines to the Mpox virus by serial passage and assessed various conditions for virus inactivation, such as heat treatment or reacting with multiple reagents. Our results revealed that the Mpox virus could infect multiple cell lines, with the BSC-1 cell line being the most susceptible. Heating at 56°C for 10 min or longer rendered the virus non-infectious, indicating its thermosensitivity. Furthermore, widely used reagents, such as TRIzol, alcohol, Micro-Chem Plus, bleach, and formalin, completely inactivated the virus at recommended concentrations. However, radioimmunoprecipitation assay buffer without SDS should be used with caution, as it may not fully eliminate infectious particles. Our results provide pivotal reference data for future studies and standardization efforts in Mpox virus research, enhancing our understanding and management of this emerging pathogen.
IMPORTANCEThe global resurgence of Mpox highlights the urgent need for robust diagnostic, therapeutic, and biosafety strategies. However, critical gaps remain regarding its replication across various cell types and the effectiveness of disinfection methods. This study systematically evaluates the susceptibility of commonly used laboratory and clinical cell lines to Mpox virus, providing key insights to optimize viral isolation in research and diagnostic settings. Additionally, by assessing the effectiveness of standard disinfectants against Mpox, this work strengthens biosafety protocols for healthcare and high-containment laboratories. These findings have direct implications for public health preparedness, guiding both laboratory practices and biosafety measures. KEYWORDS Monkeypox, Mpox, cell lines susceptibilities, disinfectants, thermosensitive T he World Health Organization declared the Mpox (Monkeypox) outbreak in Europe in May 2022 a global public health emergency. The outbreak was first identified in the United Kingdom and rapidly spread to over a hundred countries within less than 6 months (1). The first imported case of Mpox was reported in July 2022 in Taiwan. To date, over 107 countries have reported Mpox outbreaks, with more than 87,000 confirmed cases at the time of writing this article (2).The Mpox, cowpox, and variola viruses belong to the Orthopoxvirus genus within the Poxviridae family. Compared to other viruses, the Mpox virus has a large size and a linear, double-stranded DNA genome of approximately 197 kbp (3). The virus is classified into two clades: clade I (Congo Basin) has a mortality rate approaching 10%. In contrast,
clade IIa (West Africa) shares over 95% genomic similarity with clade I but has a mild mortality rate of less than 1% (4,5). A comparison of clade I and clade II reveals that although most of their genomes exhibit similarity, clade II demonstrates variations attributed to mutations or deletions in several virulence-associated genes, such as D10L, B10R, or B14R. As a result, clade II Mpox exhibits a lower mortality rate than clade I (6). African animal importation caused a localized outbreak of Mpox, with a lineage belonging to clade IIa, in the United States in 2003 (7). However, the 2022 Mpox outbreak was classified as clade IIb because of its similarity to the clade IIa genome sequences, despite its lower mortality rate (6)(7)(8)(9). Clades I and IIa cause zoonotic diseases, whereas clade IIb shows broad human-to-human transmission. The respiratory transmission of the Mpox virus was inefficient in the laboratory, with direct animal contact and close human interactions critically regulating human transmission (10)(11)(12). The incubation period for the disease ranges from 3 to 17 days. Individuals with Mpox may present with a characteristic rash that can manifest on various body parts, including the hands, feet, chest, face, mouth, or near the genital region (13). Statistical analysis indicated that the R 0 , which denotes the number of a single infected individual that can cause spread to another, of the 2022 Mpox outbreak was estimated to vary from 1.56 to 3.80 by country (14).
Understanding cell susceptibility to viral infections and identifying effective inactivation conditions for emerging pathogens are important in virology research and public health. The former provides valuable insights into viral pathogenesis, host-virus interactions, and the development of antiviral strategies (14)(15)(16)(17). Neverthe less, most previous studies on the Mpox virus chose cell lines based on references to other poxviruses (18), without conducting specific assessments of the susceptibility of conventional laboratory cell lines to the Mpox virus. This lacuna in the evaluation may have introduced bias in choosing particular cell lines. Establishing optimal inactivation conditions is crucial for ensuring laboratory safety, preventing viral transmission, and preserving sample integrity.
In this study, we used various commonly employed laboratory cell lines to evaluate their susceptibility to the Mpox virus. In addition, we examined the ability of disinfec tants and cell lysis reagents widely used in research environments and experimental settings to inactivate viruses effectively. Our findings establish a comprehensive set of experimental conditions that provide valuable reference data for future studies and standardization efforts on the Mpox virus. Moreover, the findings from this study would help determine the effectiveness of different disinfectants in inactivating the Mpox virus for controlling the global epidemic.
## MATERIALS AND METHODS
## Cells and virus
Cell lines from the American Type Culture Collection, including BSC-1, HeLa, LLC-MK2, MRC-5, RD, Vero E6, BHK-21, HEp-2, Huh7, MDCK, CHO, HEK-293, and MM55.K, were cultured in Dulbecco's Modified Eagle Medium (DMEM; Cat# SH30243.02, HyClone, Logan, USA), supplemented with 10% fetal bovine serum (FBS; Cat# 26140079, Gibco, Grand Island, USA) and 1% antibiotic-antimycotic (Cat# 15240-062, Gibco, Grand Island, USA) at 37°C with 5% CO 2 . The Mpox virus strain, hMpxv/Tai wan/CVDCDC-110-231642/2022 (GISAID Accession ID: EPI_ISL_13632071), was isolated from a Taiwanese patient, a man in his 20s with a documented history of travel to Germany. The BSC-1 cells were used to propagate the virus. Following adsorption, the inoculum was removed and replaced with DMEM supplemented with 2% FBS (E-2 medium). A viral cytopathic effect (CPE) was observed, and virus harvesting involved subjecting the cells to three freeze-thaw cycles and collecting the supernatant and cell debris. This study was approved by our Biosafety Committee (reference number B1-112-0006). All live-virus experiments were conducted in a Biosafety Level 4 (BSL-4) laboratory.
## Different cell lines' susceptibility test
To evaluate Mpox virus susceptibility, each cell line was seeded at 4 × 10 5 cells per well in 12-well plates (Cat# PC312-0050, GeneDireX, New Taipei City, Taiwan) in duplicate the night before the experiment. Upon removal of the growth medium, the cells were infected with 300 µL of Mpox at a multiplicity of infection (MOI) of 0.001. Following an hour of incubation for viral adsorption at 37°C, the virus was discarded and replaced with 2 mL/well of E-2 medium. Incubating these cells at 37°C in 5% CO 2 for 4 days was the first passage of infection. Next, we placed the 12-well plate in a -80°C freezer overnight to ensure complete freezing, then transferred it to a 37°C incubator for 1 h to thaw. This constituted one freeze-thaw cycle. Based on previous studies, three freeze-thaw cycles were performed to lyse the cells and ensure the complete release of viral particles. The supernatant was mixed with cell debris, and 300 µL was inoculated into fresh cells. Following an hour of virus adsorption, we replaced the culture medium with E-2 medium and incubated the cells at 37°C in 5% CO 2 for 4 days, which we define as the second passage (P-2) (15,19). All samples were collected and stored at -80°C until real-time quantitative PCR analysis or plaque assay.
## Real-time qPCR
The Mpox virus G2R gene copies were quantified using qPCR with specific primers. The specific primers used were G2R-G_F (5′-GGAAAATGTAAAGACAACGAATACAG-3′), G2R-G_R (5′-GCTATCACATAATCTGGAAGCGTA-3′), and G2R_G_probe (5′-FAM-AAGCCGTA ATCTATGTTGTCTATCGTGTCC-3′-BHQ1) (20). All primers and probes were purchased from Integrated DNA Technologies. The total nucleic acids of the samples were extracted using a TANBead Maelstrom 4800 automatic extraction system with a TANBead OptiPure Viral kit (Cat# M665S46, TANBead, Taoyuan, Taiwan). Subsequently, the extracted nucleic acids were analyzed using the Roche LC480 system (Roche, Basel, Switzerland) with automatic threshold and baseline settings. The QuantiNova Probe RT-PCR Kit (Cat# 208252, QIAGEN, Hilden, Germany) was used as the reagent, and the user's manuscript was followed to perform the experiments.
## Plaque assay
For the plaque assay, BSC-1 cells were seeded in 24-well plates (Cat# 3524, Corning, NY, USA) at a density of 8 × 10 4 cells per well and incubated at 37°C in 5% CO 2 overnight. The virus samples were serially diluted 10-fold by E-2 medium, and 200 µL of the dilutions were added to the cells. The cells were incubated for 1 h at 37°C with shaking at 15 min intervals to facilitate viral adsorption. Next, the virus was removed, and 1 mL of E-2 medium was added to the cells, which were incubated for 3 days at 37°C with 5% CO 2 . The supernatant was discarded, and the monolayer was fixed with 1 mL of 10% formalin (Cat# HT501320, Sigma-Aldrich, St. Louis, USA) for at least 1 h, at room temperature. The monolayer was stained with 0.4% crystal violet (Cat# C0775, Sigma-Aldrich, St. Louis, USA) in 20% methanol (Cat# 1.06009.1000, Merck, Darmstadt, Germany) and washed with tap water to visualize plaques. Plaques were quantified and recorded as PFU per milliliter.
## The median tissue culture infectious dose
The assay was performed to evaluate viral infectivity. Briefly, BSC-1 cells at a concentra tion of 8 × 10 3 cells/well were seeded in a 96-well plate (Cat# PC396-0100, GeneDireX, New Taipei City, Taiwan) and cultured overnight. Viral supernatants containing cell debris were subjected to 10-fold serial dilutions in E-2 medium. BSC-1 cells were washed twice with 200 µL of phosphate-buffered saline (PBS) per well. Subsequently, 100 µL of the diluted samples was inoculated into the 96-well plates pre-seeded with BSC-1 cells. Each dilution was performed in sextuplicate. Following incubation at 37°C with 5% CO 2 for 5 days to allow CPE development, the plates were fixed with 200 µL of 10% formalin and subsequently stained with crystal violet to visualize the CPE.
## Viral growing curve
To characterize Mpox virus replication kinetics in various cell lines, a multi-step growthcurve analysis was performed. A total of 8 × 10 4 cells per well were seeded into 24-well plates and incubated overnight. Cells were then infected with Mpox virus at an MOI of 0.001. After 1 h of adsorption at 37°C, the inoculum was removed and replaced with 500 µL of fresh E-2 medium. Cultures were maintained at 37°C, and samples were collected at the indicated time points. Each sample underwent three freeze-thaw cycles to release intracellular virus, after which viral titers were quantified by qPCR and plaque assay. All conditions were tested in quadruplicate.
## Plaque assay conditions test
Two cell lines, BSC-1 and Vero E6, were evaluated in overlapping media to compare the optimized plaque assay conditions. Briefly, 8 × 10 4 cells were seeded into each well of a 24-well plate and incubated overnight. Virus samples were then serially diluted and added to the cells at a volume of 200 µL, followed by a 1 h incubation at 37°C to allow virus adsorption. The plate was gently rocked every 15 min during adsorption. After removing the inoculum, cells were overlaid with 1 mL of either methylcellulose (Cat# M7140-1KG, Sigma-Aldrich, St. Louis, USA) in E-2 medium at concentrations of 1.55%, 0.78%, and 0.4%, or microcrystalline cellulose (Avicel RC-591, DuPont, Wilmington, USA) in E-2 medium at concentrations of 1.25% and 0.63%. The control group was overlaid with E-2 medium only. Each group performed in quadruplicate. The plate was then incubated at 37°C with 5% CO 2 for 3-6 days. After fixation with 10% formalin, plaques were visualized by staining with crystal violet, and the experimental results were recorded.
## Viral inactivation
For the heat-inactivation assay, 100 µL of Mpox virus (8 × 10 4 PFU) was dispensed into eight-strip tubes (Cat# MB-P08-A, Gunster, Taiwan). Samples were divided into the following groups: negative control, unheated control, and heat treatments at 56°C, 65°C, and 95°C. All groups were processed in octuplicate. Heating was performed in a PCR thermocycler (TCLT9610, Blue-Ray Biotech, Taiwan) to ensure precise temperature and experimental consistency. The samples were harvested at the time points of 1, 3, 5, 10, 15, and 30 min. A total of 100 µL of E-2 medium was added to each tube to obtain a final volume of 200 µL. The entire volume was inoculated onto BSC-1 cells seeded in 24-well plates and allowed to adsorb for 1 h at 37°C. Following adsorption, the viral supernatant was removed, fresh E-2 medium was added, and the cultures were incubated for 3 days. Finally, cells were fixed with 10% formalin and stained with crystal violet to visualize plaque formation (21).
For evaluating the effect of lysis buffers and disinfectants, 8 × 10 4 PFU of Mpox virus was mixed with the respective solutions listed in Table 3 and incubated for 10 min at room temperature. The reaction volume was then adjusted to 4 mL with PBS and passed through an Amicon Ultra-4 100 kDa centrifugal filter (Millipore, Ireland) at 3,500 rpm by centrifuge (DSC-N158A, Digisystem Laboratory Instrument, Taiwan) until ~250 µL remained. The retentate was brought back to 4 mL with PBS, and the wash was repeated twice more to remove residual chemicals. The retentate remaining on the filter was brought to a final volume of 500 µL with E-2 medium. Each experimental condition was tested in quadruplicate. Then, the viruses were inoculated into BSC-1 cells in a 24-well plate and incubated for 3 days. The cells were fixed and stained with crystal violet to visualize the viral plaques.
To validate the effect of formalin and paraformaldehyde on inactivating the virus. BSC-1 cells were grown to full confluence in 24-well plates. After PBS washes twice, the cells were infected with the virus at an MOI of 0.1 and incubated for 24 h. The medium was then removed, and 1 mL of a chosen fixative was added to each well: 10%, 5%, or 1% formalin, or 4% or 1% paraformaldehyde. Mock-infected controls were subjected to the same fixation procedures in parallel. Viral-control wells received 1 mL PBS instead of fixative. All conditions were set up in four replicates. Fixation lasted 1 h at room temperature. The fixatives were discarded, and the monolayers were rinsed three times with 1 mL PBS to remove chemical residues. Each well was then loaded with 500 µL of PBS, and the plate underwent three freeze-thaw cycles to release virus from the cells. The whole suspension was collected, and 200 µL of it was used to infect fresh confluent BSC-1 monolayers. After 1 h of adsorption, the inoculum was removed, 0.5 mL E-2 medium was added, and the cultures were incubated for 72 h. Finally, cells were fixed with 10% formalin for 1 h and stained with crystal violet to visualize the plaques.
## Viral freeze-thaw cycles stability test
To evaluate Mpox virus stability, 100 µL aliquots containing 8 × 10 4 PFU were dis pensed into microcentrifuge tubes (Cat# 72.692.005, Sarstedt, Nümbrecht, Germany) and subjected to one to six successive freeze-thaw cycles; each condition was tested in triplicate. After the designated cycles, viral titers were quantified by plaque assay and qPCR as described above.
## Statistical analysis
The number of independent replicates for each experiment is indicated in the cor responding Materials and Methods subsections. Data are expressed as mean ± SD. Descriptive statistics and hypothesis testing were performed in GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA) or Microsoft Excel 2019. Differences between the two groups were evaluated with Student's t-test, and P < 0.05 was considered statistically significant.
## RESULTS
## The Mpox virus propagates in various conventional cell lines
The P1-0 h represents the time point immediately after 1 h of viral absorption, serving as a baseline indicating the initial viral load for Mpox virus replication in various cell lines, ranging from 3.84 × 10 2 to 3.68 × 10 4 copies/mL (Fig. 1A). All cells showed viral titers ranging from 8.49 × 10 3 to 1.52 × 10 7 copies/mL during P1. The fold change in viral replication was calculated for each cell type, yielding log-transformed values from 0.98 to 4.06 (Fig. 1A; Table 1). The fold-change analysis classified the cells into four groups based on their viral replication capacity. Replication fold changes exceeding 1,000-fold were observed in BSC-1, MRC-5, Vero E6, and 293T cell lines, whereas MK-2, RD, and MM55.K cells exhibited replication fold changes above 100-fold. BHK-21, MDCK, and CHO cells demonstrated a replication fold change above 10-fold, and Hep-2 and Huh-7 cells had the least of less than 10-fold (Table 1). In contrast, during the viral cultures' P-2, viral titers ranged from 4.43 × 10 3 to 2.72 × 10 7 copies/mL (Fig. 1A). BHK-21 and Hep-2 cells showed increased viral replication, while MDCK cells exhibited decreased replication (Table 1).
Six cell lines were selected to analyze the Mpox virus's growth curve. After virus adsorption at an MOI of 0.001, the virus and cell debris were collected at various time points, and CPE was observed. The results categorized the cell lines into three groups. Vero E6, MRC-5, RD, and BHK-21 cells exhibited viral replication from 6 to 24 h, reaching a peak at 48 h (~10 5 PFU/mL). BSC-1 cells showed rapid viral replication, starting at 24 h and reaching a peak at 72 h (~10 6.8 PFU/mL). MM55.K cells exhibited slow viral replication, with viral levels plateauing at 72 h (Fig. 1B). Similar results were observed for viral genome copies. The highest viral copy number was observed in BSC-1 cells at 96 h post-infection. However, viral replication was observed as early as 6 h post-infection and peaked at 48 h in most cell lines, except BHK-21 and MM55.K, with delayed viral growth at 24 h (Fig. 1C). CPEs were observed in BSC-1, RD, BHK-21, and MRC-5 cells 48 h post-infection, whereas Vero E6 and MM55.K cells showed CPEs 72 h post-infection (Fig. 1D).
Our results indicated that the Mpox virus could propagate in various cell types, and within 48 h of post-infection, most cells exhibited similar viral titers. However, we chose BSC-1 cells for viral cultivation to achieve higher viral yields.
## The Mpox virus efficiently forms viral plaques in BSC-1 cells
We tested the plaque assay conditions for the Mpox virus using BSC-1 and Vero E6 cell lines and compared the effects of different overlapping media. Mpox virus formed relatively smaller plaques in Vero E6 cells, which were clearly visible to the naked eye on day 6. In contrast, clear plaques were observed as early as day 3 in the BSC-1 cells. The overlapping media used in this experiment did not significantly affect the number of viral plaques, except in the 1.55% methylcellulose group, which produced smaller plaques than the other groups (Fig. 2A). The viral plaques formed effortlessly, even when the E-2 culture medium was used, reflecting the intercellular spreading characteristic of the virus. Our experimentation with various overlay media formulations demonstrated the consistent efficiency of the Mpox virus plaque formation in BSC-1 cells. However, it is noteworthy that comet-shaped plaques might form when using the E-2 medium alone.
Furthermore, we compared the conditions for titrating the Mpox virus using median tissue culture infectious dose (TCID 50 ) in BSC-1 cells and tested the results for 3-6 days (data not shown). The TCID 50 collected on day 6 demonstrated either the presence or absence of CPEs (Fig. 2B). Additionally, we summarized the viral loads obtained from the plaque assay, TCID 50 , and qPCR for the same specimen tube (Table 2).
## Most reagents and disinfectants effectively inhibited Mpox virus activity at conventional concentrations
Our findings indicate that heat treatment of the Mpox virus at 56°C for 5 min can reduce plaque counts while extending the treatment to 10 min or longer achieves complete viral inactivation (Fig. 3A). Heat treatment is one of the most convenient methods for pathogen inactivation. Current pathogen detection methods rely on nucleic acid detection (such as real-time qPCR) as the primary diagnostic test in the laboratory because of its high sensitivity, accuracy, and reliability (22,23). Therefore, we further evaluated how heat treatment at different temperatures might damage the viral genome by heat to assess its impact on the sensitivity of nucleic acid testing. This investigation did not yield significantly different Ct values among the virus samples subjected to differential heat treatment. Interestingly, our findings revealed that the average Ct value of the virus treated at 56°C was slightly higher than that of those treated at 65°C and 95°C (Fig. 3B), suggesting that 56°C treatment for inactivation might lead to a lower viral genome copy number (we will further discuss later).
We also assessed the effects of commercial lysis buffers and disinfectants on Mpox virus infectivity. Briefly, 8 × 10 4 PFU of Mpox virus was incubated in each reagent for 10 min, residual chemicals were removed with Amicon Ultra-4 100 kDa centrifugal filters, and infectious virus was quantified by plaque assay. Complete loss of plaque formation was considered evidence of inactivation. Under these conditions, most reagents abolished Mpox infectivity at their standard working concentrations (Fig. 4A). The item details and statistics results from multiple repeats are shown in Table 3. Because the chemical resistance of the filter might influence the outcome, we evalu ated the compatibility of each formulation. Based on the manufacturer's user guide and safety data sheets, the maximal concentrations of active ingredients that reached the membrane during our wash protocol were below the recommended limits for all reagents except Trizol. To confirm that Trizol did not compromise filter integrity, we performed an additional validation experiment (Fig. S1). The results showed identical virus-retention efficiencies for Trizol-treated and untreated filters with no virus detected in the flow-through, demonstrating that the filter remained fully functional. Unexpect edly, a few viral plaques persisted after repeated treatments with commercial radioim munoprecipitation assay (RIPA) buffer, indicating that a 10 min exposure was insufficient for complete inactivation. Extending the reaction time to 5-30 min progressively reduced viral activity, yet infectious particles were still detectable after 30 min (Fig. 4B).
In contrast, 10% formalin and 4% paraformaldehyde showed complete inactivation. However, it is noteworthy that one of the four repeated experiments using 1% parafor maldehyde did not entirely inhibit viral activity (Table 4).
Our findings demonstrate that the Mpox virus is sensitive to heat treatment, with a minimum of 10 min at temperatures above 56°C, effectively inhibiting its activity. Commonly used laboratory lysis buffers and disinfectants also efficiently deactivated the virus at the recommended concentrations. However, the composition of RIPA buffer varies depending on its purpose and brand, resulting in varying outcomes. Additionally, paraformaldehyde at concentrations lower than 1% might leave viral residues. These two aspects will be further discussed.
## The Mpox virus exhibits high stability after multiple freeze-thaw cycles
Next, we evaluated the stability of infectious Mpox virus by subjecting it to freeze-thaw cycles. Virus aliquots were subjected to 1-6 freeze-thaw cycles. Plaque assays were used to determine the viral survival rate and to calculate the infectivity titer (Fig. 5A). Samples thawed once as a reference to calculate the relative percentage of viral activity after multiple freeze-thaw cycles. Interestingly, we observed that the viral titer increased during the first three freeze-thaw cycles, remained stable after the third cycle, and did not show any decrease in infectivity even at the end of the experiment (Fig. 5B). Furthermore, we examined whether freeze-thaw cycles degraded viral nucleic acids and affected qPCR results. While a slight decrease in the Ct value was observed after one or more freeze-thaw cycles, there was no significant difference across all six freeze-thaw cycle samples (Fig. 5C).
The Mpox virus exhibited remarkable stability during multiple freeze-thaw cycles, as evidenced by the absence of a reduction in viral titer following six cycles. Moreover, qPCR analysis revealed no observable impairment to viral nucleic acid integrity, indicating the resistance of the viral genome to damage under these conditions.
## DISCUSSION
Cell susceptibility is a critical determinant of viral infection, which depends on host cells for replication and propagation. The ability of a virus to infect a specific cell type is determined by the presence of specific cell-surface receptors that facilitate binding and entry, such as the sialic acid, which served as a crucial receptor for the Influenza virus (24,25). Previous studies have elucidated the unique cellular entry mechanisms of the vaccinia virus, which may involve membrane fusion or endocytosis (26). It has been reported that the Mpox virus shares similar receptors, such as heparan sulfate, chondroitin sulfate, and glycosaminoglycan (27), with the vaccinia virus. However, the precise receptors involved in these interactions remain unknown.
In the susceptibility study, we removed non-adsorbed viruses and evaluated the number of viruses attached to the cells as a baseline for calculating the virus replication rate. This approach better reflects the actual viral replication fold change than the total number of viruses added as a reference. The baseline viral load revealed that different cell types have different affinities for the Mpox virus. This suggests that cells with stronger binding affinities and their surface molecules could be further investigated to understand the diversity of virus receptors on host cells. Furthermore, we observed that the number of copies per milliliter in the viral growth curve and susceptibility test was similar for nearly all cell lines at 96 h post-infection (Fig. 1B). However, the growth curves of actual infectious particles revealed that only BSC-1 cells maintained high viral sensitivity, with a higher fold change in virus replication. Analysis of copies-to-PFU ratios revealed that only the BSC-1 and Vero E6 demonstrated better virus packaging efficiency, as evidenced by the average log numbers ranging from 0.74 to 1.16. In contrast, the remaining cell lines averaged between 1.98 and 2.19 (Table 5) (28,29). In this section, the replication of the Mpox virus can be observed in multiple cell types. However, when generating virus stocks, the ratio of viral copies per PFU might be of critical considera tion. Cells with a lower copies-to-PFU ratio, such as BSC-1, exhibit better viral replication and packaging efficiency, concurrently yielding fewer defective particles. Several cell lines, such as BSC-1, MRC-5, VeroE6, and 293T, maintained similar viral titers between P1 and P2 (Table 1). This observation likely reflects their high permissive ness to Mpox virus, as these cells can efficiently support viral replication without the need for adaptation. The lack of further increase in P2 may indicate that virus replication had already reached a plateau. In contrast, BHK-21 and HEp-2 cells showed an increase in P2 titers. These cells initially supported lower replication in P1, but improved yields in P2 suggest that Mpox virus may have begun adapting to these less-permissive environ ments. Similar patterns of virus adaptation through serial passage have been reported in other poxviruses (30). Conversely, viral titers declined in P2 cultures of RD, Huh-7, MDCK, and CHO cells, implying that these cell lines are only semi-permissive for Mpox virus. The reduced yield is most likely due to an intracellular block rather than inefficient viral entry. For instance, previous studies demonstrated that in CHO cells, vaccinia virus and related orthopoxviruses abort at the stage of intermediate protein synthesis, resulting in markedly lower progeny production. A similar restriction mechanism may therefore explain the replication we observed in these four cell lines (31). These patterns are evident not only from the copies to PFU ratios in Table 1 but also from the CPE observed at 96 h post-infection in P1 and P2 cultures. In highly permissive cell lines such as BSC-1 and VeroE6, extensive CPE was already visible at P1-96h, and by P2-96h, the monolayers were nearly destroyed. In moderately permissive cell lines such as BHK-21 and HEp-2, the degree of CPE at 96 h increased from P1 to P2, suggesting progressive viral adaptation. In contrast, cell lines that likely restrict Mpox replication, including MDCK and CHO, showed little to no CPE throughout the experiment (Fig. S2). Together, these results highlight distinct patterns of host-cell interactions with Mpox virus, including stable support, adaptation, and restriction. These differences are important considerations when selecting cell lines for virus propagation or antiviral screening. The plaque assay is a fundamental technique for determining the viral titers and remains the gold standard for accurately measuring infectious virus particles (32). Traditionally, agarose or methylcellulose has been used in solid or semisolid overlays to prevent viral spread through liquid convection. However, a simplified and safer approach is required in laboratories with high biosafety levels, such as BSL-3 facilities. Therefore, agarose was phased out as an overlay medium for the plaque assays. We compared the conditions and effectiveness of methylcellulose and microcrystalline cellulose as overlay media in plaque assays. Mpox virus is predominantly transmitted via cell-to-cell contact. The plaque assay using the E-2 medium alone may generate numerous comet-shaped plaques, which are smaller and can lead to misinterpretation of viral titer calculations. Furthermore, the virus titers calculated from the E-2 medium may be overestimated. Nevertheless, this approach can be used if the plates are incubated for only 2-3 days, given the significant size difference between the comet-shaped and primary virus plaques. Adding at least 0.4% methylcellulose or 0.63% microcrystalline cellulose effectively mitigated the appearance of comet-shaped plaques in plaque assays. Due to the absence of previous records of Mpox cases in Taiwan, no live virus was available to test the efficacy of various inactivation conditions. As a result, inactivation standards for the Mpox virus were previously established based on literature references and data from related viruses, such as the vaccinia virus (33). This study investigated the inactivation conditions and stability of Mpox virus from four different perspectives to provide a foundation for future guidelines in Taiwan.
The first aspect explored the inhibitory effect of heat treatment on the Mpox virus, which revealed 10 min at 56°C effectively inactivated the virus. Our findings were largely consistent with those of Kaplan et al. and Christophe et al., although the latter study applied 56°C for 30 min, which was insufficient for complete inactivation. However, the main difference between our study and Christophe's study lies in the viral load used for the reaction, with Christophe's study using approximately 7 × 10 6 PFU of the virus. In contrast, we used 8 × 10 4 PFU of the virus. Additionally, they used a water bath for heating, whereas we used a more precise PCR machine for temperature control, which may account for the inconsistencies in the data between the two studies (34,35). However, the outcomes of both studies indicate the Mpox virus to be thermosensitive. We observed that virus inactivation at 56°C resulted in slightly higher Ct values than at higher temperatures. This indicated that nucleic acid-based detection following 56°C treatment may yield lower viral genome or copy numbers than treatments at higher temperatures. We hypothesize that temperatures below 65°C may preserve nuclease activities, leading to degradation of the viral genome despite viral inactivation (36,37). These findings highlight the importance of considering this phenomenon in future nucleic acid-based viral detection strategies.
Second, understanding the efficacy of various lysis buffers and disinfectants in suppressing Mpox virus is essential for effective pathogen control in public health settings and for enhancing laboratory safety practices. Our testing of various reagents could completely inhibit the Mpox virus at standard working concentrations, consis tent with results reported in other publications (38)(39)(40). RIPA buffer is a widely used reagent for protein extraction and an essential component of many protein experiments. Therefore, verifying the conditions under which the RIPA buffer can effectively inactivate the Mpox virus is critical for future protein-related experiments. Surprisingly, our findings indicated that even after 10 min of incubation with the Mpox virus at the standard working concentration, viral plaques were formed despite a 99.94% inhibition rate (Fig. 4A; Table 3). Subsequently, we conducted a time-course experiment with RIPA buffer and discovered that complete viral inactivation was not achieved even after 30 min of exposure (Fig. 4B). After examining the components of the commercial RIPA buffer used in this study, we found that it lacked 0.1% SDS, unlike the conventional recipe (41). We then procured a different brand of RIPA buffer containing 0.1% SDS and supplemented the original RIPA buffer used in this experiment with 0.1% SDS. The virus inhibition experiment was repeated, and the results demonstrated complete viral inactivation (data not shown).
Third, we investigated the inhibitory effects of commonly used fixatives on the Mpox virus. Formalin or paraformaldehyde was widely used as a fixative for immunofluorescence staining, formalin-fixed paraffin embedding, and in situ hybridization (42). The effectiveness of several concentrations commonly used for cell fixation in suppressing the activity of the Mpox virus was evaluated in our study. Our results revealed that standard working concentrations of formalin effectively inhibited viral activity for over 30 min. Notably, incomplete inactivation was observed in one out of four replicates with 1% paraformaldehyde, as evidenced by CPE observations (Table 3). These findings are consistent with a previous study by Ellen et al., which reported that 1.5% paraformal dehyde treatment for 30 min could suppress approximately 10 7 PFU of the vaccinia virus, whereas 0.3% can suppress 10 2 PFU of the virus (43). Our study used a virus concentration approaching 10 5 PFU, which may explain the observed CPE under 1% paraformaldehyde conditions, consistent with similar studies despite the differences between viruses.
During freezing, the formation of ice crystals within cells can puncture the cell membrane (44). This phenomenon can also occur in enveloped viruses such as influenza, whose outer layer consists of a phospholipid bilayer (45). Although the Mpox virus is also enveloped, its viral genome may be damaged upon exposure to multiple freezethaw cycles (46). Therefore, testing the tolerance of the Mpox virus to freeze-thaw cycles is highly significant for specimen preservation, transportation, and experimental operations. Our results demonstrate that Mpox virus exhibits a considerable degree of tolerance to repeated freeze-thaw cycles, which may be related to its life cycle. Most virus particles accumulate inside cells, where the microenvironment during freezing provides a relatively high protein concentration that serves as a protective mechanism.
In this study, several limitations should be considered. First, using a single virus strain of virus limits the generalizability of our findings. Mpox virus includes clade I and clades IIa and IIb. Due to the absence of local cases in Taiwan, only the IIb lineage of the virus was available for validation. Additionally, a comparative analysis with other viruses or orthopoxviruses was not performed. Previous literature demonstrates extensive research on the vaccinia virus, thus guiding our primary focus toward studying the Mpox virus. Second, the range of cell lines used in this study was relatively limited. The chosen cell lines primarily represent those commonly employed in laboratory settings and do not comprehensively simulate infection across diverse organ-derived cell lines. To address this limitation, we have initiated plans to expand our cell repository to facilitate broader susceptibility testing.
In conclusion, our study provides evidence of the susceptibility of several widely used cell lines to Mpox virus. These findings have important implications for future research, particularly in selecting appropriate expression systems. Moreover, we validated the efficacy of inactivation protocols for the Mpox virus using a range of disinfectants and lysis reagents commonly employed in laboratory and environmental disinfection. The insights obtained from this research have potential applications in experimental design, laboratory sterilization, and environmental sanitation practices.
## References
1. Thornhill, Barkati, Walmsley et al. (2022) "Monkeypox virus infection in humans across 16 countries"
2. Who, Outbreak (2022)
3. Kaler, Hussain, Flores et al. (2022) "Monkeypox: a comprehensive review of transmission, pathogenesis, and manifesta tion" *Cureus*
4. Likos, Sammons, Olson et al. (2005) "A tale of two clades: monkeypox viruses" *J Gen Virol*
5. Shchelkunov, Totmenin, Safronov et al. (2002) "Analysis of the monkeypox virus genome" *Virology (Auckl)*
6. Chen, Li, Liszewski et al. (2005) "Virulence differences between monkeypox virus isolates from West Africa and the Congo basin" *Virology (Auckl)*
7. Ligon (2004) "Monkeypox: a review of the history and emergence in the Western hemisphere" *Semin Pediatr Infect Dis*
8. Gigante, Korber, Seabolt et al. (2022) "Multiple lineages of monkeypox virus detected in the United States, 2021-2022" *Science*
9. Ulaeto, Agafonov, Burchfield et al. (2023) "New nomenclature for mpox (monkeypox) and monkeypox virus clades" *Lancet Infect Dis*
10. Hutson, Gallardo-Romero, Carroll et al. (2013) "Transmissibility of the monkeypox virus clades via respiratory transmission: investigation using the prairie dog-monkeypox virus challenge system" *PLoS One*
11. Fleischauer, Kile, Davidson et al. (2005) "Evaluation of human-to-human Methods and Protocols Microbiology Spectrum December"
12. "transmission of monkeypox from infected patients to health care workers" *Clin Infect Dis*
13. Americo, Earl, Moss (2023) "Virulence differences of mpox (monkeypox) virus clades I, IIa, and IIb.1 in a small animal model" *Proc Natl Acad Sci*
14. Cdc (2023) "Mpox in the U.S"
15. Branda, Pierini, Mazzoli (2023) "Monkeypox: early estimation of basic reproduction number R 0 in Europe" *J Med Virol*
16. Dighe, Sarkale, Patil et al. (1962) "Differential cell line susceptibility to the SARS-CoV-2 Omicron BA.1.1 variant of concern" *Vaccines (Basel)*
17. Madani, Abuelzein, Bell-Sakyi et al. (2013) "Susceptibility of tick cell lines to infection with alkhumra haemorrhagic fever virus" *Trans R Soc Trop Med Hyg*
18. 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*
19. Rosa, Ferreira De Castro, Vieira Da Silva et al. (2023)
20. Dai, Wu, Wu et al. (2021) "Differential cell line susceptibility to Crimean-Congo hemorrhagic fever virus" *Front Cell Infect Microbiol*
21. Li, Zhao, Wilkins et al. (2010) "Real-time PCR assays for the specific detection of monkeypox virus West African and Congo basin strain DNA" *J Virol Methods*
22. Jureka, Silvas, Basler (2020) "Propagation, inactivation, and safety testing of SARS-CoV-2" *Viruses*
23. Yang, Rothman (2004) "PCR-based diagnostics for infectious diseases: uses, limitations, and future applications in acute-care settings" *Lancet Infect Dis*
24. Dutta, Naiyer, Mansuri et al. (2022) "COVID-19 diagnosis: a comprehensive review of the RT-qPCR method for detection of SARS-CoV-2" *Diagnostics (Basel)*
25. Maginnis (2018) "Virus-receptor interactions: the key to cellular invasion" *J Mol Biol*
26. (2018) "Host range, host-virus interactions, and virus transmis sion" *Viruses*
27. Roberts, Smith (2008) "Vaccinia virus morphogenesis and dissemination" *Trends Microbiol*
28. Shi, He, Song et al. (2022) "Kinetic and structural aspects of glycosaminoglycan-monkeypox virus protein a29 interactions using surface plasmon resonance" *Molecules*
29. Americo, Earl, Moss (2017) "Droplet digital PCR for rapid enumeration of viral genomes and particles from cells and animals infected with orthopoxviruses" *Virology (Auckl)*
30. Weidmann, Sall, Manuguerra et al. (2011) "Quantitative analysis of particles, genomes and infectious particles in supernatants of haemorrhagic fever virus cell cultures" *Virol J*
31. Elde, Child, Eickbush et al. (2012) "Poxviruses deploy genomic accordions to adapt rapidly against host antiviral defenses" *Cell*
32. Ramsey-Ewing, Moss (1995) "Restriction of vaccinia virus replication in CHO cells occurs at the stage of viral intermediate protein synthesis" *Virology (Auckl)*
33. Baer, Kehn-Hall (2014) "Viral concentration determination through plaque assays: using traditional and novel overlay systems" *J Vis Exp*
34. Cutchins, Warren (1958) "Comparative susceptibility of cell cultures to vaccinia virus: application to the standardization of smallpox vaccine" *Proc Soc Exp Biol Med*
35. Kaplan (1958) "The heat inactivation of vaccinia virus" *J Gen Microbiol*
36. Batéjat, Grassin, Feher et al. (2022) "Heat inactivation of monkeypox virus" *J Biosaf Biosecur*
37. Hilbert, Hayes, Stone et al. (2017) "The large terminase DNA packaging motor grips DNA with its ATPase domain for cleavage by the flexible nuclease domain" *Nucleic Acids Res*
38. Landry, Vu, Levin (2014) "Purification of an inducible DNase from a thermophilic fungus" *Int J Mol Sci*
39. Meister, Tao, Brüggemann et al. (2023) "Efficient inactivation of monkeypox virus by world health organization-recommended hand rub formulations and alcohols" *Emerg Infect Dis*
40. De Oliveira, Rehfeld, Guedes et al. (2011) "Susceptibility of vaccinia virus to chemical disinfec tants" *Am J Trop Med Hyg*
41. Fischer, Gallogly, Schulz et al. (2022) "Evaluation of five buffers for inactivation of monkeypox virus and feasibility of virus detection using the panther fusion ® open access system" *Viruses*
42. Ngoka (2008) "Sample prep for proteomics of breast cancer: proteomics and gene ontology reveal dramatic differences in protein solubilization preferences of radioimmunoprecipitation assay and urea lysis buffers" *Proteome Sci*
43. Liu, Babka, Kearney et al. (2020) "Molecular detection of SARS-CoV-2 in formalin-fixed, paraffinembedded specimens" *bioRxiv*
44. Hulskotte, Dings, Norley et al. (1997) "Chemical inactivation of recombinant vaccinia viruses and the effects on antigenicity and immunogenicity of recombinant simian immunodeficiency virus envelope glycoproteins" *Vaccine (Auckl)*
45. Reite (1966) "Mechanical forces as a cause of cellular damage by freezing and thawing" *Biol Bull*
46. Greiff, Blumenthal, Chiga et al. (1954) "The effects on biological materials of freezing and drying by vacuum sublimation. II. Effect ON influenza virus" *J Exp Med*
47. Holohan, Hanrahan, Feely et al. (2021) "Influence of viral transport media and freeze-thaw cycling on the sensitivity of qRT-PCR detection of SARS-CoV-2 nucleic acids" *Nanoscale* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12513004&blobtype=pdf | # Evolution of split-protein technologies in virology: From mechanistic discovery and diagnostics to therapeutic promise
Kai Zhou, Yingshou Lei, Yaoqi Zhou, Jian Zhan
## Abstract
The persistent challenge posed by viruses such as HIV, HBV, and SARS-CoV-2 necessitates the continuous evolution of molecular tools for their study and for advancing therapeutic research. Split-protein complementation assays (PCAs), where a reporter protein is divided into two inactive fragments, have evolved from simple reporters of biological events into an increasingly important tool in modern virology. This review traces the evolutionary trajectory of split-protein systems. We begin with their foundational use in mechanistic discovery, where they first visualized viral-host interactions in living cells. We then explore their translation into practical applications, such as high-throughput drug screening and rapid point-of-care diagnostics. A step in this evolution was the development of systematic engineering platforms, dramatically accelerating the creation of novel biosensors. Finally, we discuss the latest frontier: engineering therapeutically active "split effectors." By integrating principles from synthetic biology, these advanced systems can function as programmable logic gates that respond to specific viral signatures. While therapeutic translation remains preclinical, split-protein platforms are emerging as tangible tools for advanced research and potential therapeutic development.
## 1. Introduction：The Need for New Antiviral Weapons
## 1.1. The unfinished fight against chronic and emerging viruses
An ongoing battle against viral pathogens has marked the 21st century. Despite historic achievements such as eradicating smallpox, humanity remains in an arms race with a diverse and adaptable virosphere. Chronic infections caused by pathogens like Human Immunodeficiency Virus (HIV), which establishes a latent reservoir that persists despite current antiretroviral therapy, and Hepatitis B and C viruses (HBV/ HCV), leading to progressive liver damage and cancer, affect hundreds of millions of people globally. 1,2 These viruses highlight a crucial unmet need for tools to study and eradicate persistent infections. At the same time, emerging and re-emerging viruses continually threaten human health. The recent impact of SARS-CoV-2 and the ongoing risk of zoonotic spillovers from influenza and Ebola viruses underscore the urgent need for platform technologies that can rapidly adapt to new pathogenic threats. 3,4 Moreover, the rapid evolution of viral escape mutants requires flexible therapeutic strategies based on a thorough understanding of essential viral biology.
## 1.2. The importance of molecular-resolution readouts for mechanistic understanding
Achieving durable control of chronic viral infections such as HIV, and developing countermeasures against emerging pathogens, requires a multifaceted approach. This approach requires a comprehensive, mechanistic understanding of the entire viral life cycle, encompassing cellular entry and uncoating, genome replication, protein synthesis, virion assembly, and egress. Many of these processes are coordinated by complex and often transient interactions between viral and host proteins within specific subcellular compartments. 5,6 Although traditional biochemical techniques, such as co-immunoprecipitation, are powerful, they usually rely on cell lysates that average signals across large cell populations, thereby removing these events' spatial and temporal context. Consequently, it becomes challenging to determine the precise location and timing of molecular interactions within a single living cell using these methods. Therefore, live-cell imaging and interaction assays are crucial for identifying vulnerable points in the viral life cycle, 7 and this is where split-protein technologies may provide a transformative advantage.
## 1.3. Split proteins as an evolving platform for modern virology
Split-protein complementation assays (PCAs) have become a highly versatile tool for virologists. The basic idea involves dividing a reporter protein (such as a fluorescent protein or a luciferase) into two inactive constituents. These two constituents have a weak natural affinity for each other but cannot efficiently self-assemble. However, when attached to two interacting proteins, the interaction brings the fragments close together, increasing their local concentration. This proximity encourages their refolding and reassembly, restoring the reporter's function and producing a detectable optical signal. 7,8 This approach enables the conversion of a complex molecular event-such as a protein-protein interaction, a conformational change, or a cleavage event-into an easily detectable optical signal, often in real-time and within living cells or entire organisms. Over the past two decades, this simple yet powerful principle has served as the foundation for a continuously evolving platform. This review traces this evolutionary journey-from foundational discovery tools to platforms with therapeutic promise-and highlights how split-protein technologies are becoming valuable components of the modern virologist's toolkit.
## 2. The split-protein toolkit: fundamental principles and important variants
## 2.1. The core mechanism: from BiFC to split enzymes
This idea was initially demonstrated convincingly using split Green Fluorescent Protein (GFP) in what is now known as Bimolecular Fluorescence Complementation (BiFC). 5,7 In a typical BiFC assay, the N-and C-terminal halves of GFP, or its brighter, faster-folding variant Venus, are fused to two potentially interacting proteins. Their association promotes the irreversible self-assembly of the GFP beta-barrel structure, which then enables the internal chromophore to mature, producing a fluorescent signal that permanently marks the location of the interaction. 9 This principle was quickly applied to luciferases, which enable lightproducing chemical reactions. Split luciferases, such as Renilla (RLuc) and Firefly (FLuc), offer a luminescent readout with a higher signal-tobackground ratio and a much broader dynamic range compared to fluorescence. 8,10 A significant breakthrough was the development of NanoLuc, a small (19 kDa) and highly bright luciferase. Its optimized split version, NanoBiT, includes a large fragment (LgBiT; 18 kDa) and a minimal 11-amino-acid peptide (SmBiT, now called HiBiT). 10 The affinity between LgBiT and HiBiT is low; however, when fused to interacting partners, they reassemble into a highly bright luciferase. Importantly, this interaction is reversible, enabling the measurement of both protein association and dissociation kinetics in real time, a task that is challenging to accomplish with BiFC. In addition to optical reporters, this principle has been applied to enzymes such as dihydrofolate reductase (DHFR), resulting in reporters that can confer antibiotic resistance and are helpful for genetic selection screens 11 (Fig. 1).
The assay is based on dividing a reporter protein into two nonfunctional fragments. These fragments are then fused to two potentially interacting proteins (Protein A and Protein B). Left panel (OFF state): In the absence of an interaction between Protein A and B, the reporter fragments remain separated, and no signal is produced. Right panel (ON state): An interaction between Protein A and B brings the fragments into close proximity, driving their reassembly into a functional reporter and generating a detectable signal. This versatile principle can be applied to various reporters, including fluorescent proteins (e.g., GFP in BiFC), luciferases (e.g., NanoBiT), and enzymes for selection (e.g., DHFR). The assay is based on dividing a reporter protein into two non-functional fragments. These fragments are then fused to two potentially interacting proteins (Protein A and Protein B). Left panel (OFF state): In the absence of an interaction between Protein A and B, the reporter fragments remain separated, and no signal is produced. Right panel (ON state): An interaction between Protein A and B brings the fragments into close proximity, driving their reassembly into a functional reporter and generating a detectable signal. This versatile principle can be applied to various reporters, including fluorescent proteins (e.g., GFP in BiFC), luciferases (e.g., NanoBiT), and enzymes for selection (e.g., DHFR).
## 2.2. Performance metrics: selecting the appropriate tool for the job
Choosing a PCA system involves careful evaluation of various performance factors. These include:
• Signal-to-Background Ratio: Measures the strength of a specific signal compared to noise from spontaneous fragment reassembly. Luciferases, especially NanoBiT, typically provide the highest signalto-background ratio, enabling the detection of low-level expression or weak interactions. reporters is a key goal to enhance tissue penetration. 12 A side-by-side comparison of the most commonly used split-protein systems is summarized in Table 1.
## 2.3. Challenges, artifacts, and best practices
While powerful, the application of PCAs requires careful experimental design to mitigate potential artifacts and ensure robust data interpretation. A primary concern is false positives, which often arise when fusion proteins are overexpressed. High cellular concentrations can promote non-specific reassembly of the reporter fragments, masking the genuine biological interaction. Conversely, false negatives can occur if the fused reporter fragments sterically hinder the natural interaction site or disrupt protein folding.
For false positives, solutions include using inducible promoters to control expression or, ideally, creating stable cell lines with nearendogenous expression levels. 29 Moreover, testing both N-and C-terminal fusions, as well as optimizing the length and flexibility of peptide linkers, can often resolve the issue. Because any fusion tag can potentially alter protein function, key findings should always be validated by orthogonal methods, such as co-immunoprecipitation of the native proteins or proximity-labeling proteomics. 30 One key advance in overcoming overexpression artifacts is the synergy between CRISPR/Cas9-mediated genome editing and AI-driven structural prediction. CRISPR/Cas9 allows for the precise knock-in of split-protein tags into an endogenous gene, while tools like AlphaFold2 are crucial for identifying an optimal insertion site that does not disrupt the protein's native function. This powerful combination was effectively demonstrated by Xu et al., who used AlphaFold2 predictions to guide the CRISPR-mediated insertion of a split-GFP tag into endogenous tubulin, enabling artifact-free visualization of the cytoskeleton. 31 In addition to these potential experimental artifacts, selecting an appropriate PCA system requires considering its intrinsic biophysical properties. The irreversible nature of BiFC, for instance, is advantageous for trapping weak or transient interactions, but it precludes the study of interaction dynamics. In contrast, reversible systems like NanoBiT have become the gold standard for quantitative applications, such as measuring binding kinetics or screening for inhibitors, due to their high signal-to-background ratio and real-time response. 10 Nevertheless, even the most sensitive systems have limits; extremely transient or low-affinity interactions (e.g., in the high micromolar Kd range) may not be sufficient to drive reporter assembly above background noise. Therefore, the choice of a PCA must be carefully matched to the anticipated strength and dynamics of the biological question being investigated.
## 3. Stage 1: mechanistic discovery -illuminating the viral lifecycle
Armed with this foundational toolkit and an awareness of its best practices, researchers first applied split-protein technologies to address one of the most fundamental questions in virology: how, where, and when do viruses interact with their hosts? This marked the first major stage in the evolution of these assays: their utility in mechanistic discovery.
Understanding where and when a virus utilizes host machinery is crucial for identifying treatment targets. By attaching LgBiT to viral proteins and HiBiT to host proteins, interaction networks can be mapped within living cells with precise spatial and temporal detail. 6 This approach has highlighted key processes across various viral systems. For HIV-1, split-luciferase assays initially mapped the interaction between viral integrase and the host factor LEDGF/p75. High-throughput screens later identified small-molecule LEDGINs that block this interface and inhibit late-stage replication. 32 Split-GFP imaging subsequently revealed dynamic HIV-1 Gag assembly at plasma membrane budding sites. 33 In HBV research, split-luciferase complementation revealed targetable interactions between the HBV core protein and nuclear transport receptors. 34 At the same time, BiFC assays elucidated the anti-apoptotic mechanisms of the HBV X protein through its interactions with mitochondrial proteins. 35 For SARS-CoV-2, these assays rapidly identified essential host factors, 36 demonstrating the platform's Unstable (flash) 28 Stable a Fragment sizes are approximate and depend on the exact split junction and tag/linker sequences.
adaptability to emerging pathogens.
## 4. Stage 2: application -from discovery to diagnostics and drug screening
The detailed mechanistic insights gained from Stage 1 studies naturally paved the way for the next evolutionary step: translating these biological discoveries into practical tools. This second stage focuses on the application of split-protein systems for high-throughput drug screening and rapid diagnostics, directly leveraging the specific viralhost interactions that have been previously identified. The versatility of this approach has enabled a wide range of practical applications across the antiviral pipeline, from large-scale screening to point-of-care solutions (Fig. 2).
The platform's versatility enables a range of use cases across the antiviral pipeline. (A) Target identification: mapping virus-host interactions in living cells to uncover targets. 5,6 (B) High-throughput screening (HTS): cell-based assays for inhibitors of viral processes (e. g., entry). 34 (C) Rapid diagnostics: a surrogate neutralization format using split NanoLuc (LgBiT/SmBiT) fused to RBD and ACE2; neutralizing antibodies block RBD-ACE2 complementation and reduce luminescence in patient serum. 37,38 (D) Replication monitoring: real-time biosensors reporting viral protease activity (example: 3CLpro), with a schematic readout shown as relative luminescence (a.u.). 36 Conceptual schematic; not to scale; no unpublished data. Abbreviations: ACE2, angiotensin-converting enzyme 2; RBD, receptor-binding domain;
LgBiT/SmBiT, Large/Small NanoLuc fragments; HTS, high-throughput screening; a.u., arbitrary units.
## 4.1. Screening for antivirals and host factors
The optical readout of split-protein assays is ideally suited for highthroughput screening (HTS) campaigns. Engineered cell lines expressing split reporters for key viral-host interactions enable rapid compound screening, as the disruption of these interactions causes measurable signal loss, immediately flagging potential inhibitors. 39 During the COVID-19 pandemic, Wang et al. developed a split-NanoLuc syncytium-formation assay (ACE2-LgBiT/Spike-HiBiT) that enabled rapid screening of entry inhibitors and neutralizing sera. 40 At the same time, Rawson et al. developed a 3CLpro-cleavable split-NanoLuc biosensor, which contributed to the identification of nirmatrelvir analogs. 36 Beyond small molecules, genome-wide CRISPR screens coupled to PCAs have systematically identified host dependency and restriction factors. For SARS-CoV-2, this approach revealed unexpected cellular pathways essential for viral replication, including previously uncharacterized membrane trafficking components. 36 The versatility of these platforms extends to variant surveillance, with split-protein assays adapted to rapidly distinguish neutralizing antibody responses against different viral strains within 30 min, 41 demonstrating their potential for both drug discovery and diagnostic applications.
## Fig. 2. Key applications of split-protein technologies in virology (conceptual).
The platform's versatility enables a range of use cases across the antiviral pipeline. (A) Target identification: mapping virus-host interactions in living cells to uncover targets [5,6]. (B) High-throughput screening (HTS): cell-based assays for inhibitors of viral processes (e.g., entry) [34]. (C) Rapid diagnostics: a surrogate neutralization format using split NanoLuc (LgBiT/SmBiT) fused to RBD and ACE2; neutralizing antibodies block RBD-ACE2 complementation and reduce luminescence in patient serum [37,38]. (D) Replication monitoring: real-time biosensors reporting viral protease activity (example: 3CLpro), with a schematic readout shown as relative luminescence (a.u.) [36]. Conceptual schematic; not to scale; no unpublished data. Abbreviations: ACE2, angiotensin-converting enzyme 2; RBD, receptorbinding domain; LgBiT/SmBiT, Large/Small NanoLuc fragments; HTS, high-throughput screening; a.u., arbitrary units.
## 4.2. Engineering viral protease sensors as a window into replication
Many viruses rely on a viral protease to cleave polyprotein precursors. Split reporters bridged by protease recognition sequences yield direct, real-time readouts of replication and facilitate HTS for protease inhibitors. 36,42
## 4.3. Point-of-care diagnostics
Split-luciferase platforms underpin rapid tests, such as LUCAS, which detect anti-SARS-CoV-2 antibodies in under an hour with ELISA-like sensitivity. 37,38 Adaptability suggests similar assays for diverse pathogens.
## 5. Accelerating the evolution: systematic engineering and AIguided design
While the applications described in Stage 2 were powerful, their development was often constrained by a significant bottleneck: the slow, manual process of designing new split-protein sensors for each new target. The need to overcome this limitation spurred the development of more systematic approaches. These high-throughput discovery platforms have dramatically accelerated the evolution of the entire field by shifting sensor design from an art to a data-driven science.
## 5.1. The "split site" problem and the manual era
A key challenge in developing a new PCA has been finding the best "split site" where the protein can be cut without losing its ability to refold. For years, this was a slow, trial-and-error process, precise knockins guided by structural intuition, involving manual testing of dozens of individual constructs. This handcrafted approach significantly slowed the development of new sensors. 43
## 5.2. Systematic discovery of split sites
The development of high-throughput discovery platforms accelerates biosensor design by transforming it from a manual, trial-and-error process into a systematic, data-driven workflow. A representative example is the High-Throughput Split-protein Profiling (HiTS) platform, whose core involves three key steps, with the functional selection principle illustrated in Fig. 3.
The diagram illustrates a HiTS system based on antibiotic resistance, using the rapamycin-inducible FRB-FKBP interaction. In the presence of rapamycin (interaction ON), functional complementation of the antibiotic resistance gene (ARG) fragments allows cells to survive on a selective medium. In the absence of rapamycin (interaction OFF), no complementation occurs, and cells are eliminated. This conditional survival enables the enrichment of clones with permissive split sites. Adapted from . 44 Conceptual schematic; not to scale; no unpublished Fig. 3. The principle of functional selection in a HiTS screen (conceptual). The diagram illustrates a HiTS system based on antibiotic resistance, using the rapamycin-inducible FRB-FKBP interaction. In the presence of rapamycin (interaction ON), functional complementation of the antibiotic resistance gene (ARG) fragments allows cells to survive on a selective medium. In the absence of rapamycin (interaction OFF), no complementation occurs, and cells are eliminated. This conditional survival enables the enrichment of clones with permissive split sites. Adapted from [44]. Conceptual schematic; not to scale; no unpublished data. data.
1) Library Generation: First, a DNA transposon is used to randomly fragment the gene of a reporter protein, such as an antibiotic resistance enzyme (ARG in the figure), creating a vast library of potential split sites. 2) Functional Selection: This is the critical step where functional variants are identified. As depicted in the figure, the library of reporter fragments is fused to a pair of inducible interacting partners, FRB and FKBP. The interaction between FRB and FKBP is tightly controlled by the small molecule rapamycin. When this library is expressed in cells, a selection pressure is applied. In the presence of rapamycin, FRB and FKBP interact, forcing the complementation of the antibiotic resistance enzyme fragments. Only cells containing a "permissive" split site-one that allows the fragments to reassemble into a functional enzyme-will gain phenotypic activity and survive on a medium containing the corresponding antibiotic. In the absence of rapamycin, no interaction occurs, and the cells remain sensitive to the antibiotic. This conditional survival allows for a highly stringent selection of optimal split sites. 3) Deep Sequencing: Finally, the enriched population of surviving cells is collected, and their plasmids are analyzed by next-generation sequencing. The sequencing reads are then mapped back to the reporter gene, generating a comprehensive "split-site atlas" that reveals all permissible fragmentation points. This process greatly streamlines the discovery of optimal split sites, providing a rich dataset for the rapid customization of new biosensors. 44
## 5.3. The synergy with AI and machine learning
The impact of such high-throughput approaches extends beyond simply accelerating discovery; the massive datasets they generate are ideal for training machine learning models. By analyzing thousands of permissive and non-permissive split sites, AI-guided tools could be developed to predict optimal split points for any protein in silico, potentially bypassing the need for experimental screening altogether. This synergy between high-throughput experimentation and computational prediction represents the next wave of innovation, promising to make the design of bespoke biosensors nearly instantaneous and driving the entire field towards a fully automated Design-Build-Test-Learn (DBTL) cycle.
## 6. The therapeutic frontier: engineering split-effectors towards programmable therapeutics
With the ability to rapidly engineer and optimize biosensors, the scientific community began to ask a more ambitious question: Can we repurpose this technology not just to see the virus, but to destroy it? This inquiry marks the beginning of the most exciting and challenging stage in this evolutionary journey-the push towards the therapeutic frontier and the engineering of programmable effectors for selective antiviral action and future therapeutic applications. This entire evolutionary journey, from passive reporters to controllable therapeutic agents, is conceptually summarized in Fig. 4. Fig. 4. Evolution of split-protein technologies from reporters to therapeutics (conceptual) (A) Passive reporters (e.g., BiFC). for live-cell visualization of molecular events. 5,6 (B) Advanced NIR reporters (e.g., Akaluc). optimized for in-vivo imaging via systematic engineering. 12 (C) Controllable therapeutic effectors: multi-input logic, thresholding, and temporal controls reconstitute an effector only under defined viral contexts, enabling programmable therapeutic responses. 45 Enabling technologies (machine learning, DNA nanotechnology, gene editing, high-throughput engineering) accelerate this trajectory. Conceptual schematic; not to scale; no unpublished data.
## 6.1. The core logic: from reporter to programmable effector
The conceptual leap from a split reporter to a therapeutic effector involves redesigning the system as a modular intracellular sensoractuator device. Instead of reassembling a passive reporter, the system reconstitutes or activates a potent, cell-killing or therapeutic protein-an "effector"-exclusively within infected cells. The key to this strategy lies in leveraging a unique molecular signature of the virus as a trigger. Viral proteases, which are essential for the viral life cycle, represent ideal triggers due to their high specificity. For instance, the papain-like protease (PLpro) of SARS-CoV-2 is a well-validated antiviral target, making it a perfect candidate for activating such a system. 46 This creates a highly specific "AND-gate" logic: the therapeutic action is unleashed only when (A) the cell is infected (viral protease is present) AND (B) the therapeutic construct is present.
## 6.2. Building the effector: protease-controlled activation and delivery
The principle of protease-activated therapy has been elegantly demonstrated through generalizable platforms like RELEASE (Retained Endoplasmic Cleavable Secretion). In this system, a therapeutic protein is engineered to be trapped within the endoplasmic reticulum. Only in the presence of a specific protease-such as one from a virus-is the protein cleaved and released to perform its function, such as inducing apoptosis or signaling to the immune system. 47 This modular design effectively converts a protease signal into a programmable therapeutic output.
Crucially, this concept has been validated in vivo. Researchers have developed a "provector," an adeno-associated virus (AAV) vector that is activated by disease-associated proteases. In a model of myocardial infarction, this provector demonstrated site-specific gene delivery to damaged tissue with high protease activity, while significantly reducing off-target effects in healthy organs like the liver. 48 This study provides powerful proof-of-concept that protease-activated systems can achieve high target specificity in a complex living organism, a critical step towards clinical translation for viral diseases.
## 6.3. Advanced designs: multi-input logic and safety switches
While a single trigger is powerful, the future of therapeutic design lies in more sophisticated logic to enhance safety and precision. An alternative or complementary approach is to sense viral proteins on the surface of an infected cell using engineered receptors like Synthetic Notch(synNotch). 45 This allows for the creation of programmable immune cells that can recognize an infected cell and then, for example, activate a localized therapeutic response.
The paramount concern for any therapeutic effector is safety. The risk of off-target toxicity from even minimal "leaky" activation or the emergence of escape mutants must be rigorously addressed. Here, inspiration comes from advanced biocontainment systems. Designs like the "Deadman" and "Passcode" kill switches utilize complex transcriptional circuits to ensure cell death in the absence of specific survival signals. 45 Furthermore, the "demon and angel" construct provides a genetically stable kill-switch by making a single essential gene responsible for both cell viability and conditional killing, thus reducing the probability of loss-of-function mutations. 49 The potential of this multi-layered safety approach is stunningly illustrated in an engineered Mycobacterium tuberculosis strain containing a triple-kill-switch. This strain was rapidly cleared in animal models upon removal of stabilizing drugs, with an almost undetectable escape mutation rate. 50 Moreover, continuous engineering efforts, such as the development of an enhanced synNotch receptor with significantly reduced ligand-independent activation, demonstrate a clear trajectory towards building safer and more precise therapeutic modules. 51 These examples provide a clear roadmap for building next-generation antiviral effectors with multiple, redundant safety mechanisms, supporting a roadmap toward safer, more precise programmable therapeutics.
## 6.4. Biosafety, ethical, and regulatory considerations
Translating split-protein systems from observational reporters to interventional split-effectors demands structured risk management. While reporter-only work with non-infectious constructs is typically performed under BSL-2 with Institutional Biosafety Committee (IBC) oversight, 52 studies involving infectious agents require agent-appropriate containment. Because effectors are programmable, they are subject to dual-use research of concern (DURC) policies, necessitating measures such as sequence screening and internal risk review. 53 Designs must minimize leakiness and off-target activation using multi-input logic (e.g., AND gating), orthogonal recognition pairs, and conditional degradation, and should be validated by quantitative reporting of basal versus induced activity, dynamic range, kinetics, and reversibility.
A minimal preclinical matrix should include in vitro specificity panels and in vivo studies of dose-response, biodistribution/persistence (and shedding if vectorized), and toxicology under 3Rs/ARRIVE guidelines. 54 The regulatory path diverges by application: diagnostics require analytical and clinical validation within IVD frameworks (e.g., CLSI EP05/EP17), 55,56 whereas therapeutic split-effectors follow gene-therapy guidance, including pre-IND consultation, GLP preclinical studies, and GMP/CMC packages that define identity, purity, stability, and potency-supporting a responsible, staged progression toward programmable therapeutics. 57,58
## 6.5. Current limitations and path-to-translation
It is crucial to distinguish between the maturity of split-protein applications: while reporters are established research tools, 6 interventional split-effectors remain at the preclinical proof-of-concept stage. 47 Their path to the clinic is constrained by significant hurdles, including effective in vivo delivery, ensuring absolute specificity while eliminating off-target activation, 47 potential immunogenicity, 57,58 and the challenges of scalable manufacturing and quality control. 57,58 The path forward requires a staged, evidence-driven approach. It begins with rigorous in vitro validation of a construct's safety and efficacy, followed by comprehensive preclinical in vivo studies to assess pharmacology and toxicology. 57,58 Successful candidates then advance to an IND-enabling package, supported by GLP studies and a full CMC plan, to gain regulatory approval for early-stage clinical trials. This progression is intrinsically linked to the biosafety and regulatory frameworks detailed in 6.4. 52
## 7. Conclusion
Split-protein technologies have evolved from molecular curiosities into a vital and dynamic platform for modern virology. By converting complex molecular events into clear optical signals, they have consistently accelerated discovery at every stage of the antiviral pipeline-from fundamental mechanistic discovery to high-throughput drug screening and rapid diagnostics. This evolution is now entering its most transformative phase, propelled by the convergence of data-driven systematic engineering, 44 AI-guided design, 59 and the powerful toolkits of synthetic biology. 47,50,60 Looking forward, the most exciting frontier lies in translating these systems into programmable therapeutics. While this remains a long-term aspiration, the conceptual and technological blueprints are actively being drawn. By integrating split-protein sensors with multi-input logic gates, deploying them via smart delivery vectors validated in vivo, and fortifying them with genetically encoded, multi-layered safety switches, a new class of antiviral agents will emerge. While formidable hurdles to clinical translation certainly remain, these intelligent therapeutic systems hold the potential to enable selective targeting of infected cells,
## References
1. (2017) "Global prevalence and genotype distribution of hepatitis C virus infection in 2015: a modelling study" *Lancet Gastroenterol Hepatol*
2. (2021) "Report on HIV, Viral Hepatitis and Sexually Transmitted Infections, 2021: Accountability for the Global Health Sector Strategies 2016-2021: Actions for Impact"
3. Morens, Fauci (2020) "Emerging pandemic diseases: how we got to COVID-19" *Cell*
4. Jones, Patel, Levy (2008) "Global trends in emerging infectious diseases" *Nature*
5. Hu, Chinenov, Kerppola (2002) "Visualization of interactions among bZIP and Rel family proteins in living cells using bimolecular fluorescence complementation" *Mol Cell*
6. Kerppola (2008) "Bimolecular fluorescence complementation (BiFC) analysis as a probe of protein interactions in living cells" *Annu Rev Biophys*
7. Cabantous, Terwilliger, Waldo (2005) "Protein tagging and detection with engineered self-assembling fragments of green fluorescent protein" *Nat Biotechnol*
8. Paulmurugan, Gambhir (2003) "Monitoring protein-protein interactions using split synthetic renilla luciferase protein-fragment-assisted complementation" *Anal Chem*
9. Nagai, Ibata, Park et al. (2002) "A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications" *Nat Biotechnol*
10. Dixon, Schwinn, Hall (2016) "NanoLuc complementation reporter optimized for accurate measurement of protein interactions in cells" *ACS Chem Biol*
11. Pelletier, Campbell-Valois, Michnick (1998) "Oligomerization domain-directed reassembly of active dihydrofolate reductase from rationally designed fragments" *Proc Natl Acad Sci U S A*
12. Li, Kawamura, Dong et al. (2025) "Near-infrared bioluminescence assays for protein-protein interactions and cellular membrane fusion in deep tissues using split Akaluc reconstitution" *Anal Chem*
13. Kodama, Hu (2010) "An improved bimolecular fluorescence complementation assay with a high signal-to-noise ratio" *Biotechniques*
14. Jester, Cox, Gaj et al. (2010) "A coiled-coil enabled splitluciferase three-hybrid system: applied toward profiling inhibitors of protein kinases" *J Am Chem Soc*
15. Michnick, Levy, Landry et al. (2016) "The dihydrofolate reductase protein-fragment complementation assay: a survival-selection assay for large-scale analysis of protein-protein interactions" *Cold Spring Harb Protoc*
16. Luker, Smith, Luker et al. (2004) "Piwnica-Worms H, Piwnica-Worms D. Kinetics of regulated protein-protein interactions revealed with firefly luciferase complementation imaging in cells and living animals" *Proc Natl Acad Sci U S A*
17. Ohmuro-Matsuyama, Chung, Ueda (2013) "Demonstration of protein-fragment complementation assay using purified firefly luciferase fragments" *BMC Biotechnol*
18. Villalobos, Naik, Bruinsma (2010) "Dual-color click beetle luciferase heteroprotein fragment complementation assays" *Chem Biol*
19. Remy, Michnick (1999) "Clonal selection and in vivo quantitation of protein interactions with protein-fragment complementation assays" *Proc Natl Acad Sci U S A*
20. Romei, Boxer (2019) "Split green fluorescent proteins: scope, limitations, and outlook" *Annu Rev Biophys*
21. England, Ehlerding, Cai (2016) "NanoLuc: a small luciferase is brightening up the field of bioluminescence" *Bioconjug Chem*
22. Inouye (2009) "Firefly luciferase: an adenylate-forming enzyme for multicatalytic functions" *Cell Mol Life Sci CMLS*
23. Wan, Bennett, Wilson (2014) "Toward resolving the catalytic mechanism of dihydrofolate reductase using neutron and ultrahigh-resolution X-ray crystallography" *Proc Natl Acad Sci U S A*
24. Miller, Kim, Huh et al. (2015) "Bimolecular fluorescence complementation (BiFC) analysis: advances and recent applications for genome-wide interaction studies" *J Mol Biol*
25. Filonov, Verkhusha (2013) "A near-infrared BiFC reporter for in vivo imaging of protein-protein interactions" *Chem Biol*
26. Schwinn, Machleidt, Zimmerman (2018) "CRISPR-mediated tagging of endogenous proteins with a luminescent peptide" *ACS Chem Biol*
27. Lai, Chiang (2013) "Bimolecular fluorescence complementation (BiFC) assay for direct visualization of protein-protein interaction <em>in vivo</em>" *Bio-Protoc*
28. (2024) "Luciferase Assay System Technical Bulletin TB281"
29. Loew, Heinz, Hampf et al. (2010) "Improved Tet-responsive promoters with minimized background expression" *BMC Biotechnol*
30. Roux, Kim, Raida et al. (2012) "A promiscuous biotin ligase fusion protein identifies proximal and interacting proteins in mammalian cells" *J Cell Biol*
31. Xu, Li, Mao (2024) "AlphaFold2-guided engineering of split-GFP technology enables labeling of endogenous tubulins across species while preserving function" *PLoS Biol*
32. Desimmie, Schrijvers, Demeulemeester (2013) "LEDGINs inhibit late stage HIV-1 replication by modulating integrase multimerization in the virions" *Retrovirology*
33. Ivanchenko, Godinez, Lampe (2009) "Dynamics of HIV-1 assembly and release" *PLoS Pathog*
34. Wei, Gan, Cui (2018) "Identification of compounds targeting hepatitis B virus core protein dimerization through a split luciferase complementation assay" *Antimicrob Agents Chemother*
35. Wang, Oishi, Shimakami (2017) "Hepatitis B virus X protein induces hepatic stem cell-like features in hepatocellular carcinoma by activating KDM5B" *World J Gastroenterol*
36. Rawson, Duchon, Nikolaitchik et al. (2021) "Development of a cell-based luciferase complementation assay for identification of SARS-CoV-2 3CL (pro) inhibitors" *Viruses*
37. Yao, Drecun, Aboualizadeh (2021) "A homogeneous split-luciferase assay for rapid and sensitive detection of anti-SARS CoV-2 antibodies" *Nat Commun*
38. Kim, Cho, Lee (2025) "Ultrasensitive and long-lasting bioluminescence immunoassay for point-of-care viral antigen detection" *Nat Biomed Eng*
39. Ma, Sacco, Xia (2021) "Discovery of SARS-CoV-2 papain-like protease inhibitors through a combination of high-throughput screening and a FlipGFP-based reporter assay" *ACS Cent Sci*
40. Wang, Guo, Yang (2022) "Rapid quantitative monitoring of SARS-CoV-2 spike protein-mediated syncytia formation using split NanoLuc" *J Med Virol*
41. Zhang, Davis, Chen et al. (2021) "Emergence of a novel SARS-CoV-2 variant in southern California" *JAMA*
42. Brecher, Li, Liu (2017) "A conformational switch high-throughput screening assay and allosteric inhibition of the flavivirus NS2B-NS3 protease" *PLoS Pathog*
43. Ozawa, Kaihara, Sato et al. (2001) "Split luciferase as an optical probe for detecting protein-protein interactions in mammalian cells based on protein splicing" *Anal Chem*
44. Zhou, Litfin, Solayman et al. (2022) "High-throughput splitprotein profiling by combining transposon mutagenesis and regulated proteinprotein interactions with deep sequencing" *Int J Biol Macromol*
45. Chan, Lee, Cameron et al. (2016) "Deadman" and "Passcode" microbial kill switches for bacterial containment" *Nat Chem Biol*
46. Ershov, Yablokov, Mezentsev (2022) "SARS-COV-2 coronavirus papain-like protease PLpro as an antiviral target for inhibitors of active site and protein-protein interactions" *Biophysics*
47. Vlahos, Kang, Aldrete (2022) "Protease-controlled secretion and display of intercellular signals" *Nat Commun*
48. Guenther, Brun, Bennett (2019) "Protease-activatable adeno-associated virus vector for gene delivery to damaged heart tissue" *Mol Ther J Am Soc Gene Ther*
49. Kato, Mori (2024) "Genetically stable kill-switch using "demon and angel" expression construct of essential genes" *Front Bioeng Biotechnol*
50. Wang, Su, Wallach (2025) "Engineered Mycobacterium tuberculosis triple-killswitch strain provides controlled tuberculosis infection in animal models" *Nat Microbiol*
51. Yang, Yu, Cai et al. (2020) "Engineering of an enhanced synthetic Notch receptor by reducing ligand-independent activation" *Commun Biol*
52. (2020) "Biosafety in Microbiological and Biomedical Laboratories (BMBL)"
53. (2014) "United States government policy for institutional oversight of life sciences dual use research of concern"
54. Du Sert, Hurst, Ahluwalia (2020) "The ARRIVE guidelines 2.0: updated guidelines for reporting animal research" *PLoS Biol*
55. (2012) "Evaluation of Detection Capability for Clinical Laboratory Measurement Procedures" *EP*
56. (2014) "EP05: Evaluation of Precision of Quantitative Measurement Procedures"
57. (2020) "Information for Human Gene Therapy Investigational New Drug Applications (INDs): Guidance for Industry" *Chemistry*
58. Schmitz, Ballestin, Liang (2024) "Int&in: a machine learning-based web server for active split site identification in inteins" *Protein Sci Publ Protein Soc*
59. Morsut, Roybal, Xiong (2016) "Engineering customized cell sensing and response behaviors using synthetic notch receptors" *Cell* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12537335&blobtype=pdf | # Retraction notice to "Understanding the implications of SARS-CoV-2 re-infections on immune response milieu, laboratory tests and control measures against COVID-19" [Heliyon 7 (2021) e05951]
Jelili Mustapha, Idris Nasir Abdullahi, Odunayo Ajagbe, Anthony Emeribe, Samuel Fasogbon, Solomon Oloche Onoja, Charles Egede Ugwu, Modesta Umeozuru, Folake Ajayi, Natasha Tanko, Pius Omosigho, Abdulmumuni Aliyu, Halima Shuwa, Justin Nwofe, Amos Dangana, Ovye Alaba, Peter Ghamba, Yakubu Ibrahim, Dorcas Aliyu, Sunday Olawale, Animasaun, Nkechi Blessing Ugboaja, Mala Alhaji, Baba Mallam, Sharafudeen Dahiru Abubakar, Maijidda Aminu, Hadiza Yahaya, Silifat Oyewusi |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12614448&blobtype=pdf | # An Era Ended, But the Legacy Lingers On: A Personal Reflection on Dr. David Baltimore
Parameswaran Ramakrishnan
## Abstract
Dr. David Baltimore's contributions to modern biology span more than six decades and continue to shape the fields of virology, immunology, biochemistry, and molecular biology. Beyond his landmark discoveries-such as reverse transcriptase and NF-κB, as well as the Baltimore classification of viruses-his influence endures through his mentorship, leadership, and the generations of scientists he trained and inspired. In this essay, I recount my journey as his postdoctoral trainee at the California Institute of Technology, offering a personal glimpse into the mind, character, and legacy of a scientist whose approach to thinking, teaching, and living science remains timeless.
## KEYWORDS
Reverse Transcriptase; NF-kappaB; Baltimore Classification of Viruses; HIV; Retroviruses; Recombination Activating Gene; Mentorship; Science Policy; Nobel Prize THE POWER OF A SINGLE WORD When I think of Dr. David Baltimore, the first word that comes to mind is "think. " That single word, so deceptively simple, holds the key to unlocking the magic and mystery of science. For David, thinking was not merely a cognitive process-it was a philosophy, a lifestyle, and the very foundation of his scientific creativity. In my view, "think" to David was what "imagine" was to Walt Disney: both words had the power to transport minds into wonderlands of discovery or where the impossible becomes possible.
I was privileged to spend five and a half years in David's "wonderland, " his laboratory at the California Institute of Technology (Caltech), as his postdoctoral trainee. My initial admiration for him and profound interest in his work originated much earlier, at the time I was first introduced to his name through his Molecular Cell Biology textbook during my master's studies. At that time, the seed of an almost impossible dream was planted in my mind-learning directly from Dr. Baltimore-one that, remarkably, would later come true. My heart was racing-part nervousness, part exhilaration-as I realized I was about to sit beside my scientific idol and present my original work, which was a surreal experience.
At the outset, he mentioned he had limited time since he was preparing a keynote lecture for the AACR meeting the following morning. But, as anyone who has interacted with David knows, his curiosity and scientific engagement are boundless, and nothing can stop David's scientific interest. What was meant to be a relatively brief meeting evolved into an intense discussion lasting two hours and forty-five minutes, deeply immersed in biochemistry and molecular biology. Only later did I learn how deeply David valued biochemistry, the discipline that had provided the foundation for his scientific career, back when molecular biology was still in its infancy.
At the end of that meeting, he gently patted my shoulder and said, "I think you will do well in my lab. We need some biochemistry, and I'm happy to offer you a postdoctoral position. Think about it and let me know whether you' d like to accept. " For a few moments, I was speechless. When I finally absorbed what had just happened and managed to respond, I said, "I still can't believe I'm sitting next to you and being offered a position in your lab. " He smiled and replied, "Relax. I'm just another person like you who likes to do science. That's what brings us together. "
That humility, from a man whose discoveries fill chapters of textbooks, whose career spans over six decades with more than 700 publications, and whose influence has shaped the entire field of biology, remains one of the most defining aspects of his character, that allowed him to carry himself with disarming simplicity.
## PATHOGENS AND IMMUNITY-A THEMATIC UMBRELLA
By mere coincidence, specifically for the readers of this journal, most of David's work could be summarized under the theme "Pathogens and Immunity. " His groundbreaking discoveries in virology and immunology connected the molecular workings of viruses to the complex choreography of immune responses. His identification of the transcription factor NF-κB-the nuclear factor that binds to the kappa light chain enhancer in B lymphoid cells-elegantly revealed the interface between pathogens and immunity. It deciphered complex molecular and cellular mechanisms regulating infections and the interplay between innate and adaptive immune responses during infections.
## THE POWER OF THOUGHT AND THE PURSUIT OF FOUNDATIONAL SCIENCE
David's career embodies the triumph of logic-driven curiosity. His brilliance was often manifested through simple yet profound reasoning. From his logical deduction and the legendary conclusion -"therefore, there is an enzyme"-that led to the Nobel Prize-winning discovery of reverse transcriptase, which reshaped the central dogma of biology, to his insights into antibody formation and the identification of NF-κB-a molecule that became a cornerstone of modern immunology, his work consistently illuminated the hidden architecture of life. The Baltimore classification of viruses, his discovery of recombination-activating gene (RAG) enzymes mediating antibody and T cell diversity, exploration of protein tyrosine phosphorylation, foundational work on human immunodeficiency virus (HIV) biology, elucidation of NF-κB signaling in inflammation and autoimmunity, and studies on microRNAs-all stand as pillars of molecular science, which makes one often wonder why he never received a second Nobel Prize.
Even as a high school student, his curiosity was evident-he often mentioned he was intrigued by seeing the obese mice at Jackson Laboratories. That early curiosity evolved into a lifelong passion for in vivo experimentation and mouse models of disease. Throughout his career, he emphasized that deep biological insights arise from observing living systems in their natural complexity.
## MENTORSHIP: TEACHING INDEPENDENCE THROUGH THINKING
David's mentorship style was both challenging and empowering. The very day he offered me a postdoctoral position, he asked me to draft my proposed research on NF-κB in the format of an NIH R01 grant application. At that time-my second day ever in the United States-I had absolutely no idea what an R01 grant was. Yet, in his own subtle way, David was teaching me to think independently, to plan like a principal investigator. Eventually, my aims were incorporated into a larger NF-κB-focused R01 grant that was funded shortly after I began my work in his lab. I would later learn that I was the last official recruit to join his laboratory, focusing solely on NF-κB research.
On one occasion, he told me something that has stayed with me ever since: "The foundational core content is what endures in science. It doesn't matter what language you think in-what matters is how deeply you think. Others can edit your words, but only you can think your thoughts. "
That was quintessential David-an unwavering belief in the power of thought over form. That advice has become a compass for the careers of many of his trainees. David's lectures and writings embodied the same principle-clarity of thought above all. In his lectures, he could distill complex ideas into messages that resonated long after the talk ended. At times, he also entertained his audience impromptu. For example, those who attended his April 15, 2007, AACR talk on Sunday morning will remember the unique way he started his talk and paused after a short while, telling us "that's the Sunday Sermon! Now science, " eliciting laughter from a captivated audience.
David's mastery extended to writing as well. He had an uncanny ability to convey intricate ideas with simplicity and precision in his articles. He always tried to provide critical suggestions to his trainees to improve their research papers, while encouraging them to explore their full potential in scientific writing. While editing one of my manuscripts, he once said, "It's perfectly fine to start a sentence with 'because'-it prepares the reader for what follows. " On another occasion, he praised my use of the word thereby, remarking, "It's a powerful word to express signal transduction-underused in cell signaling papers. " His attention to each word revealed the same meticulous care that defined his experiments. Every word mattered to him-a reflection of his belief that precision in language mirrors precision in thought.
A SCIENTIST WHO DREW ENERGY FROM THE WORLD David drew inspiration and energy for his science from the fascination he had with the world around him and his energy seemed inexhaustible. During his 2008 visit to India, he delivered four distinguished lectures in as many days. When I later asked him about the trip, he smiled and said, "It was hectic, but exhilarating. Seeing thousands of students smiling, waving, and attending my lectures reminded me how deeply this country respects science. Their enthusiasm recharged me to do more. " On another occasion, after flying in from London, he came straight to a lab meeting. Seeing him sort through medications, we asked if he was unwell. He laughed, "Not at all-These are just to fix my circadian clock. These drugs are good to do that job. " The next sentence he said was this. "These medicines have years of research behind them, and they work like a charm. I must read before bed what is new on them. " That was quintessential David-endlessly curious, scientifically disciplined, completely unpretentious, blended humor, and grace.
Like life comes out of pictures in fairy tales, David's guidance and aura had the power to make life out of scientific data and images. His guidance often turned chance observations into discoveries. On one occasion, I serendipitously found that a protein previously reported as a loading control in Western blotting showed unexpected changes upon cell stimulation. When I mentioned it to David, he encouraged me to explore it further, saying, "It may be a control for someone else, but perhaps not for you. If your data are solid, follow the lead-see where it takes you. " That open-mindedness defined his lab's spirit-follow the science, not the convention. This opennessto question assumptions and follow evidence wherever it led-defined his lab's culture and was key to its constant evolution.
Over the years, the Baltimore Lab evolved-from virology to molecular biology to immunologywith each transition reflecting David's fearless intellectual curiosity and marking a new chapter in modern biology. His discovery of reverse transcriptase led to the naming of RNA tumor viruses as retroviruses. His subsequent studies extended toward retroviruses and cancer biology, which then extended to HIV, NF-κB signaling in the immune and non-immune cells and expanding frontiers of immunology research. Through every shift, David emphasized the same truth: foundational science is the wellspring of translational breakthroughs.
This philosophy guided his co-founding of the Whitehead Institute for Biomedical Research affiliated with Massachusetts Institute of Technology (MIT) in 1982, with entrepreneur Edwin C. Jack Whitehead. The institute's mission-to pursue fundamental biological discovery-mirrored David's lifelong philosophy. Beyond research, David left an indelible imprint on scientific leadership, serving as president of Rockefeller University, Caltech, and the American Association for the Advancement of Science (AAAS). He helped shape policies on recombinant DNA technology, HIV/AIDS, and human genome editing-always advocating for curiosity-driven, ethical science.
## THE SCIENTIST, THE HUMAN, THE HUMORIST
What set David apart was not just his intellect, but his humanity. He treated everyone with equal respect, from first-year undergraduates to fellow Nobel laureates. He possessed a rare ability to adapt to the knowledge level of the person he was speaking with and elevate the conversation to a higher plane of thought. He could tailor his explanations to anyone's level, then lift them to new heights of understanding. Many of his trainees have remarked that he saw their potential before they themselves did, guiding them toward success with gentle precision.
At a lab birthday celebration, he once joked, "One reason I love my profession is that I never feel older-the average age across from me stays the same, and that keeps me young!" Indeed, his energy was boundless, not only for science but also for radiating joy around him, not just in science but in life. He personally hosted the annual lab Christmas party at Caltech's Athenaeum and took great delight in moderating the Yankee Swap gift exchange. He shared his personal joys freely-from the private plane ride piloted by his wife, Dr. Alice Huang, to his fishing adventures in Montana (where he proudly posed for a photo holding a 20-pound trout that he caught), to showcasing his journey through China with a dynamic, slide-driven narrative, and even demonstrating the "Sport" mode of his new Audi on busy U.S. Route 101.
His humor was sharp and spontaneous. Once, the lab arranged for a look-alike from Caltech to surprise him at a weekly lab meeting near his birthday. When David entered and saw his double, he chuckled and said, "Ah, now we can do twice as much science and train twice as many people!" At a dinner, when someone remarked that the sushi was "excellent and reasonably priced, " David dryly replied, "Good sushi and economy don't usually go together. " On another occasion, a visiting faculty member asked David what he felt was the main difference moving from Boston to Southern California. As usual, David's reply was swift, "Aah! (in his unique tone) When I walked out in Boston, people recognized me, and I could not go more than 50 feet without stopping to say hello to someone. But here in California, no one knows me, probably because I have not starred in a movie yet!" Here, one could easily see David's clever evasion of the comparison between the two leading institutes.
Even outside the lab, he engaged deeply with the world. He visited Case Western Reserve University in 2015 to deliver a keynote lecture at the 38th Annual Biomedical Graduate Student Symposium (BGSS). While driving through Cleveland Heights with its large mansions, he remarked, "If a Clevelander ever gets the chance, they should live in one of these storybook houses-they carry history. " Then he astonished me by time traveling to the 1900's and demonstrating his depth of knowledge about local architecture and Cleveland history; he could converse meaningfully with experts in any field-be it science, geography, history, policy, or cuisine. I also recall a lab reunion dinner he hosted for his mentees and lab alumni in Santa Fe in 2012, following a Keystone Symposium on NF-κB. As the waiter proudly started to describe a Bordeaux wine in elaborate detail, David gently took over the conversation, recounting its entire lineage, vineyard origins, the vintners' philosophy, and stories of small French wineries he had personally visited. Everyone at the table watched, smiling-it was classic David: erudite, effortless, endlessly engaging.
## FINAL ENCOUNTERS AND ENDURING WISDOM
My last in-person meeting with David was on December 26, 2022, at his home in Pasadena. We spent two unforgettable hours together. For half of that time, we stood near his koi pond, talking about science, life, and curiosity. He listened intently as I described my independent research and then offered a final piece of advice that remains etched in my heart: "If you believe in something, take it to completion-whether it gives what you expected or something entirely different. Both outcomes are knowledge gained. Leaving something halfway is no different from not doing it at all. " As we talked near the koi pond, I was reminded of a moment years earlier when he had delivered an impromptu lecture on ichthyology to my middle-schoolaged, fish enthusiast son, with the same zest he brought to Nobel-level discussions. That was David-a man who approached everything with pure, unfiltered passion, whether it involved fishkeeping or tackling challenges in fundamental biology.
## LEGACY AND CONTINUITY
During my time in his lab-and likely long before-the Baltimore Lab functioned like a miniature research institute unto itself. David often remarked that Caltech's Division of Biology had the perfect name, because "real life science should transcend disciplinary boundaries. " My time in his lab was surrounded by projects spanning HIV, microRNAs, T-cell engineering, B-cell biology, myeloid cell differentiation and function, cancer, skin pathology, pregnancy complications, DNA recombination, portable PCR design, and more. Lab meetings felt like multidisciplinary symposia, attended by budding faculty, physician-scientists, and students alike.
Despite his enormous responsibilities, David prioritized his trainees and never distanced himself from daily research life. He attended nearly every lab meeting, and sometimes he would casually walk to a trainee's bench, saying, "We should talk. " Those spontaneous discussions often redirected entire projects, and invariably for the better.
Perhaps his greatest legacy as a mentor lies in the remarkable number of independent scientists he trained-more than 200 faculty members and principal investigators around the world, a feat unlikely to be surpassed. He always talked highly about his mentors at leading institutes in the U.S. and recognized the key role of mentors in shaping the career of trainees. He paid attention to continue the legacy that he received from his mentors to advance his career for his trainees as well. He allowed each postdoc to carry their projects forward into independent careers, ensuring continuity of ideas and expansion of the scientific frontier. In doing so, he passed on not only knowledge but a tradition of curiosity, integrity, and fearless inquiry. This fostered independence and continuity. Through his direct trainees, their students, and the generations beyond-his "scientific grandchildren, " as he fondly called them-David's influence and vision will continue to shape the future of modern biology.
## EPILOGUE: THE ETERNAL LEGEND
To borrow from The Christmas Song, "Although it's been said many times, many ways, " the legend of Dr. David Baltimore will remain eternal-not only in the annals of science but in the hearts and minds of those privileged to think, learn, and dream in his world. His legacy reminds us that great science begins not with equipment or data, but with the simple act of thinking deeply and daring to wonder. For those interested in hearing more about science from David's perspective, you can read or watch an August 2021 interview conducted by senior editors of Pathogens and Immunity. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607702&blobtype=pdf | # Host cell state: an overlooked factor impacting the production of influenza A deletion-containing viral genomes and noninfectious particles
| Virology, | Minireview, Ilechukwu Agu, Samuel Díaz-Muñoz
## Abstract
Influenza A virus remains a global public health threat, prompting the need for novel, broad-spectrum therapeutics. Deletion-containing viral genomes (DelVGs) produced during influenza replication have shown broad-spectrum therapeutic potential via defective interference, where DelVG accumulation depletes the relative abundance of standard viral genomes, diminishing the viral yield needed to sustain pathogenesis. Decades of research have focused on the viral factors affecting the production and maintenance of DelVGs in influenza infections. Surprisingly, the study of host factors that affect the emergence of DelVGs has been neglected. Uncovering host factors that affect DelVG production could help predict infection outcomes based on host state; facilitate the manipulation of host metabolism to increase DelVG production, potentially leading to milder clinical outcomes; and enhance biomanufacturing. The therapeutic potential of increasing in situ production of DelVGs is evident, but barriers to progress have persisted for decades, such as a lack of tailored methodologies to reliably quan tify defective interference and early research findings that dismissed host cell involve ment in the production of DelVGs and the defective interfering particles that carry them. Thus, the molecular mechanism of de novo DelVG production remains unknown in spite of evidence implicating host involvement. This review summarizes (i) newly discovered associations between host cell state and influenza DelVGs, (ii) the extensive host-virus metabolic signaling crosstalk that refocused the host as a potential contributor to DelVG/non-infectious particle production, and (iii) the methodological innovations that facilitated these recent discoveries. We conclude by providing an outlook on new avenues in DelVG basic and applied research.
W ithin the infected host, influenza A virus has a bipartite existence as intracellu lar viral genomic segments undergoing replication and assembly into progeny particles and extracellular progeny particles in search of new cells to infect. Owing to the low fidelity of the viral polymerase, there is a striking diversity of mutants among the genomic segments that in turn carry over to the particle level as mutant genomes are packaged into combinatorially variant particles. This within-host pool of closely related yet distinct "individuals" possesses variable fitness upon which natural selection-in the form of host immune responses, antiviral treatments, and virus-virus interactions among the genome segments and particles-acts, favoring the variants better adapted to prevailing conditions (1)(2)(3)(4)(5)(6)(7)(8). Fortunately for the sieged host, there is a periodic emergence of nonstandard viral genomes (9); specifically, in the case of influenza, many of these are deletion-containing viral genomes (DelVGs). These DelVGs fail to express the standard, full-length protein, while simultaneously outcompeting standard genomes in their representation within progeny viral particles, either through replication or packaging competition (10)(11)(12)(13)(14)(15). Because particles that contain one or more DelVGs do not express the full complement of full-length proteins, they are not capable of sustaining a complete infectious cycle independently and require coinfection by a virion expressing the missing full-length proteins.
DelVGs can also undermine the capacity of standard viral genomes and fully infectious particles to sustain propagation, leading to the term defective interfering particles (DIPs) for those virions that harbor one or more DelVGs. Notably, these truncated segments can encode proteins that can interfere with replication machinery (16). Moreover, DelVGs and DIPs also accelerate the host innate immune response (17), leading to the self-limiting pathogenesis and mild disease severity that characterizes defective interference (18,19). In fact, host immune modulation by nonstandard viral genomes has been known since the 1970s, in studies that correlated early production of high levels of nonstandard viral genomes with the establishment of persistently infected cells, increased survival, and decreased pathogenicity (20,21). These initial findings have been corroborated by a body of research detailing how nonstandard viral genomes modulate the host immune system (22)(23)(24). The therapeutic potential of on-demand defective interference induction remains unrealized because the mechanism(s) of de novo DelVG production during influenza A infection remains unsolved (18,(25)(26)(27). This mechanism could be elucidated by identifying more factors that influence DelVG production, as groups of these factors likely share common processes involved in DelVG formation. Host cell metabolism exemplifies one such shared process, as evidenced by our studies that recently discovered the impact of metabolic drugs on DelVG production (28,29).
The host cell's metabolic signaling state is the most recent of the few known in situ modulators of DelVG and DIP production, joining the likes of infection multiplicity (12,30), virus polymerase gene mutations (19,31), virus non-structural gene mutations (32,33), and virus matrix gene mutations (34), as also found for lymphocytic chorio meningitis virus (LCMV) matrix protein (35). In an initial study, we showed that phar macological disruption of virus-host metabolic crosstalk with an inhibitor of growth metabolic signaling in the host cell modulated in situ hallmarks of defective interference early in infection and to varying degrees depending on flu virus strain (28). Pre-expo sure of MDCK cells to a PI3Kα inhibitor-alpelisib-prior to A/Texas/50/2012(H3N2) infection significantly increased the relative abundance of non-infectious progeny viral particles at a wide dosage range (2.5, 5, 10, and 20 µM alpelisib), increasing non-infec tious particles by ~12% at an early infection stage (18 h post-infection). In contrast, the A/California/07/2009(H1N1pdm) strain did not show changes in the proportion of non-infectious particles but variable changes in the total number of particles produced (28). However, at the genomic level, both strains showed an increasing trend of DelVG production with higher concentrations of alpelisib, with A/California/07/2009(H1N1pdm) infections pre-treated with 20 µM of alpelisib increasing polymerase complex segment DelVGs from 3.11%-4.40% (controls) to 24.67%-40.48%. These findings provided proof of concept that altering the metabolic state of the cell can change non-infectious particle production and DelVG production.
Armed with this proof of concept, we then set out to find other novel metabolic signaling drugs that act through the host to modulate in situ hallmarks of DI early in infection by examining the relative production of DelVGs specifically, e.g., controlling for impacts on full-length replication (29). We found that adenosine was a potent and consistent amplifier of DelVG production. Adenosine increased DelVG production in all three polymerase complex segments by 35.00%-80.61% in both tested H1N1 and H3N2 strains and also increased DelVGs in other segments (29). Insulin showed strainspecific effects on polymerase complex DelVGs, increasing DelVG production in the H3N2 subtype (29). Tricarboxylic acid (TCA) cycle inhibitors 4-OI and UK5099 significantly boosted total viral genome production across multiple segments (29), mimicking the Warburg effect observed in tumor cells, where cells forgo oxidation of nutrients to favor increased biomass. These metabolic signaling molecules collectively link host metabo lism to DelVG production through their shared impact on altering metabolic signaling pathways within the host cell. In addition to these drugs, a low dose of the mutagenic nucleoside analog precursor Favipiravir increased total viral genome production across H1N1 and H3N2 subtypes, while slightly reducing DelVG proportions in the H3N2 subtype, indicating that nucleotide misincorporation may not be essential for DelVG production (29). Lastly, cyanobacterial extracts selectively and almost completely shut down the production of antigenic segments (hemagglutinin and neuraminidase) in the H3N2 subtype, highlighting the potential of natural products in modulating segmentspecific forces that act during replication and capsid assembly (29,36).
Of the few known modulators of DelVG and DIP production, the host cell metabolic signaling state is unique for being the most therapeutically actionable, i.e., it can be monitored readily and provides targets that are potentially easily druggable. This discovery was made possible in part by the extensively characterized crosstalk between host metabolism and influenza infection outcomes.
## HOST-INFLUENZA INTERACTIONS: A HISTORY OF METABOLIC CROSSTALK
At the cellular level, influenza A induces various metabolic changes within infected cells that favor productive pathogenesis. In a particularly striking example, A/Puerto Rico/8/34(H1N1) infection disrupted proteasomal degradation of hypoxia-inducible-fac tor-1α (HIF-1α) in the mitochondria of human lung cells (A549) and mouse lung tissue, allowing for accumulation and translocation of this transcription factor (HIF-1α) to the nucleus where it facilitates expression of pro-glycolytic enzymes (37). This aberrant reprogramming of host cell glucose catabolism disrupts the normoxic oxidation of pyruvate by upregulating hexokinase (HK2), pyruvate kinase M2 (PKM2), and pyruvate dehydrogenase kinase (PDK3) (38). This state, similar to that of tumor cells, is character ized by enhanced glycolysis and the redirection of pyruvate from complete breakdown into CO 2 gas, preserving the reduced-carbon biomass needed to feed anabolic pathways that drive the proliferation of tumor cells (39,40) or viruses.
Influenza A virus also upregulates host biosynthetic pathways that support viral proliferation via direct binding of the viral NS1 effector protein to the regulatory subunit (p85β) of host class 1a phosphoinositide-3-kinase (PI3Kα) (41). This interaction relea ses the catalytic p110α subunit to initiate the PI3Kα signaling cascade (42)(43)(44)(45)(46)(47), with strain-dependent intensity (28,44), even in the absence of typical growth factors like insulin. Consequently, NS1's actions shift host metabolism toward a state marked by increased pools of precursor metabolites (48)(49)(50) vital for uninterrupted replication of virion components.
Conversely, host metabolism can equally influence the course of influenza infection. In a straightforward example at the cellular level, growth media supplementation with the Tricarboxylic Acid Cycle inhibitor malonate drove dose-dependent decreases in the total particle yield of viral progeny (51). At the organism level, other metabolic factors and host states such as obesity, diabetes, nutritional status, and pregnancy have been shown to impact susceptibility to influenza infection and follow-on disease severity. For instance, extreme nutritional states of either diet-induced obesity or caloric restriction impaired immune function and increased the risk of complications from influenza infection in mice (52). Additionally, obese patient groups shed higher viral loads, which also contained more virulent mutants relative to the non-obese cohorts in human infections (53,54). Pregnancy has long been known to be a state associated with more severe infection. More recent research has shown that this is due to the host's changed immune landscape, which impairs host defenses and facilitates the emergence of more pathogenic viral variants (55,56).
There is a steadily expanding body of work on how DelVGs affect the host, including the molecular mechanisms underlying host cell detection of the Z-conformation RNA of DelVGs (17); DelVG-induced variation of the host transcriptional program (47); and interferon-independent protection of co-infected influenza defective interfering virus on types I and III interferon-deficient mice (57). Strangely, the reverse is not the case, as evidenced by a striking absence of research into the effects of host cells on DelVG production. In truth, host cell involvement in DI virus production was wrongly dismissed well over half a century ago, and the extensive host-influenza metabolic crosstalk characterized in the intervening decades served to refocus the host as a key contributing factor.
## THE "NOT" STAR: EARLY INQUIRIES WRONGLY DISMISSED HOST CELL INVOLVEMENT IN INFLUENZA A DI VIRUS PRODUCTION
Shortly after the discovery of non-infectious influenza A particles and their antiviral potential (11,12), scientists speculated about the role of host cells in their productionno doubt in response to mounting evidence of viral sensitivity to perturbation in host cell functions like the TCA cycle (51), glycolysis (58), and vitamin A metabolic signaling (59). However, initial attempts to establish a connection between host parameters and DIP production found only a meandering correlation with virally induced cell damage (60). These early probes also had some limitations, like overlooking the effect of multiplicity of infection (MOI) (61,62) and the misattribution of control variables (61,63). As such, these studies were unsuccessful in disentangling host effects from MOI and consistently found MOI to be the primary determinant of DIP production, while the host had no effect. These findings and subsequent support for MOI as the main determinant of DIP production (12,30) appear to have diverted the collective pursuit for inducers of DIP production away from the host cell. This diversion is evidenced by the abrupt drop-off in research, which failed to resurge despite the emergence of supporting evidence over the subsequent decades.
The intervening decades following the dismissal of host cell involvement in DIP production saw several missed opportunities to refocus the host cell as a potential controlling factor. A recent review (27) observed that the per-segment DelVG profile of flu strains may differ with the cell type infected, citing the disparate DelVG outcomes of A/Puerto Rico/8/34 infection in two independent studies that used different cell types, embryonated eggs (64) and MDCK cells (65). In another case, researchers discovered that the fatty acid and phospholipid profile of the A/Puerto Rico/8/34 envelope differed significantly between infectious and non-infectious particles (66,67). This discrepancy suggests that lipid-driven changes in host cell membrane rigidity might differentially affect the efficiency with which nascent DelVG and standard genomes are packed into progeny particles. Lastly, influenza A polymerase replication fidelity suffers under low concentration of its ribonucleotide triphosphate (rNTP) substrate in vitro (68). Given that internal deletions of DelVGs have the appearance of a replication error product, their de novo production may also be driven by low rNTP concentration or other environmental determinants of polymerase physiochemistry and fidelity, such as pH, temperature, rNTP pool balance, choice of metal ion cofactor (Mg 2+ or Mn 2+ ), crowding, and other factors (68,69) that are regulated by host signaling networks. Moreover, altering the physio chemistry of influenza polymerase through sequence mutations directly affected DelVG accumulation (19,31), indicating the potential for physiochemical changes induced by various factors, including those originating from the host, to yield similar effects.
The above-mentioned examples are just a glimpse into a broader pool of uncura ted findings implicating host involvement in DIP production, which has been slow to rekindle interest in this once-dismissed area (60)(61)(62)(63). A persistent barrier to discovering more factors shaping DIP production is the absence of methodologies capable of not only quantifying DI phenomena at sufficient resolution but also doing so with precision and high throughput. In fact, assays enabling precise quantification of DelVGs (28,70,71) and non-infectious viral particles (28,72,73) have only recently been developed. We now turn to review past and current methods of quantifying DelVGs and DIPs, with an emphasis on methodological shortfalls and recent innovations that promise to revitalize research in this field.
## METHODOLOGICAL APPROACHES TO QUANTIFYING DEFECTIVE INTERFERING PHENOMENA, PAST AND PRESENT
This review focuses on DelVGs and DIPs. From a methodological perspective, we focus on non-infectious particles as opposed to DIPs. Non-infectious particles, which cannot complete a full cycle of replication independently, can lose their infectious ability for a number of reasons, including DelVGs, gene-lethal mutations, or failure to express a segment (semi-infectious particles sensu [73]), among others (reviewed in reference 74). Fundamentally, there are no methods to distinguish the different potential types of non-infectious particles at the virion level; thus, we use the term non-infectious particle to precisely convey what the methods measure. Standalone titration of fully infectious particles (FIP) or non-infectious particles (NIP) provides an incomplete representation of disease state due to their entangled antagonism, which shapes influenza A pathogen esis. The same rings true at the genome level, where standalone counts of SVGs or DelVGs overlook the entangled effects of both segment types on disease progression and outcomes. This is why interference is best quantified in terms of the ratios or proportions of different viral sub-groups relative to each other-both at the particle and genome level. Options to quantify interference at the particle level include NIP:FIP ratio or NIP relative abundance. In the same vein, genome-level interference can be represented via the DelVG:SVG ratio or DelVG relative abundance.
## Particle-level DI: quantifying non-infectious particles
Particle-level interference during influenza A virus infection was first quantified as the ratio of FIPs to total particles (60). FIPs were titrated via plaque assay (75) and reported as the number of plaque forming units (PFU), while total particles were titrated via hemagglutination assay (76) and reported as the number of hemagglutination units (HAU). However, the PFU:HAU ratio had low precision because HAU is only an approxima tion of total particles, not an actual count. Additionally, the PFU:HAU ratio did not directly report on interference but rather productive infectivity, meaning that observed changes in the metric may or may not be due to interference.
The infectious center reduction assay (77) was the breakthrough assay that first quantified an interference metric from influenza A virus infection, albeit indirectly and with imprecision. Co-infecting a viral sample (sample A) of known PFU with a viral sample (sample B) of unknown titer and then measuring the reduction in titer of sample A allowed researchers to quantify sample B's interference as defective interfering units (DIU/mL) (62,77,78). The DIU:PFU ratio could now be derived to report on interference in an influenza A infection. However, the DIU metric is imprecise because it does not directly measure interfering particles but derives interference from another metric. A similar assay, the plaque reduction assay (79), has been used more recently to quantify the activity of influenza DIPs. An alternate approach developed for respiratory syncytial virus (RSV) was the colorimetric assay that stained cells that were protected by DIPs (80).
We developed the cluster-forming assay to titrate influenza A viral particles (28). This assay builds on the method pioneered by Brooke et al. (73), which uses single-cell immunofluorescence to reproducibly titrate non-infectious and infectious influenza A particles directly, in a physiologically relevant adherent cell monolayer model. The methodological innovation of the cluster-forming assay lies in (i) the replacement of the plaque assay's solid agar overlay with a semi-solid overlay that is aspirated postassay, which facilitates high-throughput immunofluorescence staining and imaging of the monolayer; and (ii) the automated computational image analysis pipeline. In the resulting immunofluorescence image, infectious and non-infectious particles are clearly resolvable based on whether an infection event has propagated to adjacent cells (fully infectious) or remains confined to a single cell (non-infectious) (28). Fully infectious and non-infectious particles are then summed to yield the total particles, which is used to divide the number of non-infectious particles to derive the relative abundance of non-infectious particles; a precise metric of defective interference based on the actual count of infective and non-infective particles (28).
## Genome-level DI: quantifying deletion-containing viral genomes
Progress in particle-level defective interference quantitation initially outpaced DelVG quantitation, with the infectivity-hemagglutination ratio (PFU:HAU) (60) entering use a full 20+ years ahead of the discovery of influenza A DelVGs. Initially termed "subgenomic RNAs, " DelVGs were discovered via PAGE of phenol-extracted viral RNA (77,81). They were quantified either qualitatively by the presence or absence of a gel band (81) or quantitatively by determining the molar ratios of standard and DelVG segments relative to a reference segment (77). This quantitative method involved creating an autoradio graph from a PAGE gel of radioactively labeled RNA segments, where band intensities on the autoradiograph correlated with the amount of RNA and were analyzed using densitometry to measure the counts per minute (CPM) for each band. These CPM values were then compared to the CPM of a reference segment to determine the relative abundance of each RNA segment (77). Although pioneering, the molar ratio method was imprecise because CPM is only an approximation of total genomes per segment, not an actual count.
Quantitative PCR (qPCR) allows for influenza A virus DelVG detection via the use of internally binding primer sets that target regions flanking the known deletion sites in the viral genome (82). This allows the amplification of both full-length and deletion-contain ing genomes, but detection in this manner is limited to DelVGs of known deletion sites. As such, a forward approach to detect all possible deletion junctions in any given viral sample will require a vast amount of custom primer sets, which will significantly reduce throughput and prove technically difficult. Moreover, the specific DelVGs that qPCR manages to detect are subject to imprecise quantitation due to (i) PCR amplification bias and (ii) the inference of genomic cDNA production from the probe fluorescence instead of being directly counted.
More recent workflows to detect internal deletions in influenza A genomic segments pair next-or third-generation, long-read sequencing with downstream bioinformatics. Long-read sequencing in particular allows investigators to classify sequenced reads on the basis of internal deletions or other recombination events they harbor, but these recombination events must first be flagged via the alignment of sequenced reads to the matching reference influenza A genome. DelVG-tailored sequence alignment tools like ViReMa (83)(84)(85), DI-tector (86), and VODKA2 (87), as well as other aligners like TopHat2 (88), STAR Aligner (89), and HMMER (90) have been successfully used to detect deletion-containing reads in short-and long-read sequencing data sets (15,16,19,28,70,71,91). However, both short-and long-read sequencing platforms currently boast different efficiencies and capabilities with regard to detecting and quantifying influenza A DelVGs.
The highest throughput and accuracy (nearly 100.0% per base) for influenza A virus sequencing is achieved with the next-generation sequencing (NGS) short-read sequencing Illumina platform (92,93). However, the need for sequencing library fragmentation in NGS makes it impossible to distinguish fragmented DelVG and full-length segments. Consequently, DelVG deletion junction mapping is the current limit of the Roche/454 (15) and Illumina (94) NGS platforms regarding DelVG identi fication. Experimental and computational artifacts from physical DNA fragmentation can be mitigated by various methods, such as using simulated control data sets to validate deletion breakpoints (94) or employing ClickSeq to avoid physical fragmentation and enzyme-mediated ligation of sequencing adapters. In ClickSeq, sequencing library preparation starts with a reverse transcriptase reaction using semi-random DNA primers, deoxyribonucleotides, and a 3′-modified nucleotide analog that randomly terminates DNA synthesis, producing variably sized 3′-blocked cDNA fragments similar to dideoxy-Sanger sequencing (95). These fragments are then purified and reacted with sequencing adapters bearing a 5′-modified chemical group, which binds both molecules at their 3′ and 5′ ends into ssDNA substrate for PCR amplification to generate a viral cDNA library (95). Despite ClickSeq's advantages, the dependence on library fragmentationwhether physical or non-physical-limits NGS platforms to reporting the location and abundance of deletion breakpoints per segment (70). Fortunately, advancements in long-read sequencing platforms have allowed researchers to overcome the limitations of short-read sequencing data.
End-to-end sequencing of the influenza A virus genomic segments on the Oxford Nanopore Technologies long-read sequencing platform has made it possible to detect the deletions of all possible lengths in any given viral segment and, therefore, classify and count the number of standard and deletion-containing genomes in a given sample (28)-a feat as yet not achieved with short read genomic data of influenza A. Addition ally, the current generation Oxford Nanopore hardware and software have a modal per-base accuracy of 97.21% for influenza sequencing, a 1.35% point improvement from the previous generation (96). This trend of improvement puts Oxford Nanopore on track to rival Illumina in per-base accuracy in the coming years and outperform Illumina if the short-read platform is unable to expand its capabilities to include long-read sequencing.
In recent influenza A DelVG investigations, cDNA synthesis followed by PCR amplification with universal influenza primers has become the preferred method of sequencing library preparation (15,16,19,28,71). However, PCR amplification introdu ces a risk of bias (97,98), affecting the accuracy of segment and deletion junction counts. To enhance precision, researchers now use DNA primers with unique molecular identifiers (UMIs) during the reverse transcriptase reaction, incorporating UMIs into viral cDNA before amplification. In the bioinformatics process, amplicons with identical UMIs are collapsed into a single representative read, accurately reflecting the true count of viral cDNA, and can also be used to increase the per-base accuracy through consen sus of multiple reads sharing the same UMI (99). Finally, advancements in direct RNA sequencing on third-generation sequencing platforms (100) promise further innova tion by eliminating the need for PCR and UMI deduplication by quantifying native RNA molecules (16). In lieu of the mainstream adoption of direct RNA sequencing in the Influenza A DelVG research, which is hampered by a lack of officially supported multiplexing (101), UMI-deduplicated amplicons are classified as DelVGs and SVGs and divided to obtain a precise relative abundance of DelVGs and an accurate count of mapped deletion junctions based on the actual count of genomic segments (28).
## CONCLUSION AND OUTLOOK
The mechanism(s) of de novo DelVG production during influenza A infection remains unsolved. The discovery of related factors that impact the production of DelVGs and non-infectious viral particles holds the promise of informing the chain of events that lead to the de novo emergence of DelVGs. The potential for the host cell to modulate influenza A DelVG production is a logical target of inquiry because of the strong dependence of viral pathogenesis on cell machinery. Host effects were pursued for a time (60)(61)(62)(63) but abandoned in the wake of findings that increasingly pointed to the multiplicity of DIPs in the inoculum as the main contributor of DelVG production (30,62,63). There has also been an absence of a concerted effort to uncover the causes and mechanisms behind DelVG de novo emergence in the subfield of influenza A DelVG research-likely due to the absence of methodologies capable of quantifying DelVGs and non-infectious particles with the requisite precision, at meaningful resolutions, and with sufficient throughput. Fortunately, methodological innovations-such as the cluster-forming assay to titer non-infectious particles (28,72) and the combination of long-read genome sequencing with unique molecular identifier deduplication to titer DelVGs directly (28,70)-have led to the discovery of associations between the host cell state, particularly metabolism and metabolic signaling, and DelVG production.
The renewed evidence that host cell metabolism can affect DelVG production opens up several avenues for basic and applied research. There is a new class of mechanisms within the host to investigate DelVG production, which should assist in the search to elucidate how exactly DelVGs form in the host cell in the first place. Furthermore, these findings can inform the dynamics of virus-virus interactions, as the host context can alter the composition of the DelVG population (2, 7), which can alter the course of the infection and has potential implications for viral host jumps. Host involvement in DelVG production also presents many possibilities for diagnostics, therapeutics, and their manufacturing. An exciting possibility is to link host cell states that are particularly susceptible to severe influenza infection with existing metabolite screens as a diagnostic of infection severity to prioritize medical resources. The induction of host cell states that steer flu pathogenesis toward milder outcomes using metabolic drugs is a novel therapeutic approach that could take advantage of existing, approved drugs. However, this approach comes with an important caveat: timing of nonstandard viral genome production is crucial, and late administration can have neutral effects or even exacerbate disease (22). Thus, more research is needed to determine the feasibility of inducing host states that can alter pathogenesis. Finally, as therapeutic interfering particles are identified and manufactured, compounds that can boost the cell production of DelVGs could be a very useful tool to scale production of DelVG-based therapeutics. Hopefully, these new methodologies and discoveries will multiply into breakthroughs that finally allow mapping of mechanisms that underlie de novo DelVG production and harnessing this information to stem the ongoing burden of influenza epidemics and pandemics.
## References
1. Díaz-Muñoz, Boddy, Dantas et al. (2016) "Contextual organismality: beyond pattern to process in the emergence of organisms" *Evolution*
2. Díaz-Muñoz, Sanjuán, West (2017) "Sociovirology: conflict, cooperation, and communication among viruses" *Cell Host Microbe*
3. Domingo, Holland (1997) "RNA virus mutations and fitness for survival" *Annu Rev Microbiol*
4. Eigen (1971) "Selforganization of matter and the evolution of biological macromolecules" *Naturwissenschaften*
5. Holland, De, Torre et al. (1992) "RNA virus populations as quasispecies" *Curr Top Microbiol Immunol*
6. Leeks, West, Ghoul (2021) "The evolution of cheating in viruses" *Nat Commun*
7. Leeks, Bono, Ampolini et al. (2023) "Open questions in the social lives of viruses" *J Evol Biol*
8. Sanjuán (2021) "The social life of viruses" *Annu Rev Virol*
9. Aparicio, López (2024) "Selection of nonstandard viral genomes during the evolution of RNA viruses: a virus survival strategy or a pesky inconvenience" *Adv Virus Res*
10. Davis, Hiti, Nayak (1980) "Influenza defective interfering viral RNA is formed by internal deletion of genomic RNA" *Proc Natl Acad Sci*
11. Henle, Henle (1943) "Interference of inactive virus with the propagation of virus of influenza" *Science*
12. Von Magnus (1954) "Incomplete forms of influenza virus" *Adv Virus Res*
13. Nayak, Sivasubramanian, Davis et al. (1982) "Complete sequence analyses show that two defective interfering influenza viral RNAs contain a single internal deletion of a polymerase gene" *Proc Natl Acad Sci*
14. Mendes, Russell (2021) "Library-based analysis reveals segment and length dependent characteristics of defective influenza genomes" *PLoS Pathog*
15. Saira, Lin, Depasse et al. (2013) "Sequence analysis of in vivo defective interfering-like RNA of influenza A H1N1 pandemic virus" *J Virol*
16. Ranum, Ledwith, Alnaji et al. (2024) "Cryptic proteins translated from deletion-containing viral genomes dramatically expand the influenza virus proteome" *Nucleic Acids Res*
17. Zhang, Yin, Boyd et al. (2020) "Influenza virus Z-RNAs induce ZBP1-mediated necroptosis" *Cell*
18. Dimmock, Easton (2014) "Defective interfering influenza virus RNAs: time to reevaluate their clinical potential as broad-spectrum antivirals?" *J Virol*
19. Vasilijevic, Zamarreño, Oliveros et al. (2017) "Reduced accumulation of defective viral genomes contributes to severe outcome in influenza virus infected patients" *PLoS Pathog*
20. Lehmann-Grube, Slenczka, Tees (1969) "A persistent and inapparent infection of L cells with the virus of lymphocytic choriome ningitis" *J Gen Virol*
21. Rabinowitz, Huprikar (1979) "The influence of defective-interfering particles of the PR-8 strain of influenza A virus on the pathogenesis of pulmonary infection in mice" *J Infect Dis*
23. Felt, Sun, Jozwik et al. (2021) "Detection of respiratory syncytial virus defective genomes in nasal secretions is associated with distinct clinical outcomes" *Nat Microbiol*
24. Genoyer, López (2019) "The impact of defective viruses on infection and immunity" *Annu Rev Virol*
25. Tapia, Kim, Sun et al. (2013) "Defective viral genomes arising in vivo provide critical danger signals for the triggering of lung antiviral immunity" *PLoS Pathog*
26. Manzoni, López (2018) "Defective (interfering) viral genomes reexplored: impact on antiviral immunity and virus persistence" *Future Virol*
27. Vignuzzi, López (2019) "Defective viral genomes are key drivers of the virus-host interaction" *Nat Microbiol*
29. Wu, Zhou, Sheng et al. (2022) "Defective interfering particles of influenza virus and their characteristics, impacts, and use in vaccines and antiviral strategies: a systematic review" *Viruses*
30. Agu, José, Oberbauer et al. (2024) "Influenza A defective viral genomes and non-infectious particles are increased by host PI3K inhibition via anti-cancer drug alpelisib" *bioRxiv*
31. Agu, José, Sl (2024) "Influenza A defective viral genome production is altered by metabolites, metabolic signaling molecules, and cyanobacteria extracts" *bioRxiv*
32. Bangham, Kirkwood (1990) "Defective interfering particles: effects in modulating virus growth and persistence" *Virology (Auckland)*
33. Fodor, Mingay, Crow et al. (2003) "A single amino acid mutation in the PA subunit of the influenza virus RNA polymerase promotes the generation of defective interfering RNAs" *J Virol*
34. Ngunjiri, Buchek, Mohni et al. (2013) "Influenza virus subpopulations: exchange of lethal H5N1 virus NS for H1N1 virus NS triggers de novo generation of defective-interfering particles and enhances interferon-inducing particle efficiency" *J Interferon Cytokine Res*
35. Odagiri, Tobita (1990) "Mutation in NS2, a nonstructural protein of influenza A virus, extragenically causes aberrant replication and expression of the PA gene and leads to generation of defective interfering particles" *Proc Natl Acad Sci*
36. Pérez-Cidoncha, Killip, Oliveros et al. (2014) "An unbiased genetic screen reveals the polygenic nature of the influenza virus anti-interferon response" *J Virol*
37. Ziegler, Eisenhauer, Bruce et al. (2016) "The lymphocytic choriomeningitis virus matrix protein PPXY late domain drives the production of defective interfering particles" *PLoS Pathog*
38. Silva, Salomon, Hamerski et al. (2018) "Inhibitory effect of microalgae and cyanobacteria extracts on influenza virus replication and neuraminidase activity" *PeerJ*
39. Ren, Zhang, Han et al. (2019) "Influenza A virus (H1N1) triggers a hypoxic response by stabilizing hypoxia-inducible factor-1α via inhibition of proteasome" *Virology (Auckland)*
40. Ren, Zhang, Zhang et al. (2021) "Influenza A virus (H1N1) infection induces glycolysis to facilitate viral replication" *Virol Sin*
41. Liu, Summer (2019) "Cellular metabolism in lung health and disease" *Annu Rev Physiol*
42. Lunt, Heiden (2011) "Aerobic glycolysis: meeting the metabolic requirements of cell proliferation" *Annu Rev Cell Dev Biol*
43. Cho, Zhao, Shi et al. (2020) "Molecular recognition of a host protein by NS1 of pandemic and seasonal influenza A viruses" *Proc Natl Acad Sci*
44. Hale, Jackson, Chen et al. (2006) "Influenza A virus NS1 protein binds p85beta and activates phosphatidylinositol-3kinase signaling" *Proc Natl Acad Sci*
45. Li, Anderson, Liu et al. (2008) "Mechanism of influenza A virus NS1 protein interaction with the p85β, but Not the p85α, subunit of phosphatidylinositol 3-kinase (PI3K) and up-regulation of PI3K activity" *Journal of Biological Chemistry*
46. Ayllon, Hale, García-Sastre (2012) "Strain-specific contribution of NS1-activated phosphoinositide 3-kinase signaling to influenza A virus replication and virulence" *J Virol*
47. Kuss-Duerkop, Wang, Mena et al. (2017) "Influenza virus differentially activates mTORC1 and mTORC2 signaling to maximize late stage replication" *PLoS Pathog*
48. Lopes, Domingues, Zell et al. (2017) "Structure-guided functional annotation of the influenza A virus NS1 protein reveals dynamic evolution of the p85β-binding site during circulation in humans" *J Virol*
49. Wang, Forst, Chou et al. (2020) "Cell-to-cell variation in defective virus expression and effects on host responses during influenza virus infection" *mBio*
50. Al-Saffar, Jackson, Raynaud et al. (2010) "The phosphoinositide 3-kinase inhibitor PI-103 downregulates choline kinase alpha leading to phosphocholine and total choline decrease detected by magnetic resonance spectroscopy" *Cancer Res*
51. Luo, Xu, Li et al. (2018) "Weighing In on mTOR complex 2 signaling: the expanding role in cell metabolism" *Oxid Med Cell Longev*
52. Saha, Connelly, Jiang et al. (2014) "Akt phosphorylation and regulation of transketolase is a nodal point for amino acid control of purine synthesis" *Mol Cell*
53. Ackermann, Klernschmidt (1951) "Concerning the relation of the Krebs cycle to virus propagation" *J Biol Chem*
54. Gardner, Beli, Clinthorne et al. (2011) "Energy intake and response to infection with influenza" *Annu Rev Nutr*
55. Honce, Schultz-Cherry (2019) "Impact of obesity on influenza A virus pathogenesis, immune response, and evolution" *Front Immunol*
56. Honce, Karlsson, Wohlgemuth et al. (2020) "Obesity-related microenvironment promotes emergence of virulent influenza virus strains" *mBio*
57. Engels, Hierweger, Hoffmann et al. (2017) "Pregnancy-related immune adaptation promotes the emergence of highly virulent H1N1 influenza virus strains in allogenically pregnant mice" *Cell Host Microbe*
58. Ghedin, Schultz-Cherry (2017) "Host response: pregnancy impairs flu defences" *Nat Microbiol*
59. Wang, Honce, Salvatore et al. (2023) "Influenza defective interfering virus promotes multiciliated cell differentiation and reduces the inflammatory response in mice" *J Virol*
60. Kilbourne (1959) "Inhibition of influenza virus multiplication with a glucose antimetabolite (2-deoxy-D-glucose)" *Nature*
61. Blough (1963) "The effect of vitamin A alcohol on the morphology of myxoviruses. I. The production and comparison of artificially produced filamentous virus" *Virology (Auckland)*
62. Ginsberg (1954) "Formation of non-infectious influenza virus in mouse lungs: its dependence upon extensive pulmonary consolidation initiated by the viral inoculum" *J Exp Med*
63. Choppin (1969) "Replication of influenza virus in a continuous cell line: High yield of infective virus from cells inoculated at high multiplicity" *Virology (Auckland)*
64. De, Nayak (1980) "Defective interfering influenza viruses and host cells: establishment and maintenance of persistent influenza virus infection in MDBK and HeLa cells" *J Virol*
65. Choppin, Pons (1970) "The RNAs of infective and incomplete influenza virions grown in MDBK and HeLa cells" *Virology (Auckland)*
66. Jennings, Finch, Winter et al. (1983) "Does the higher order structure of the influenza virus ribonucleoprotein guide sequence rearrangements in influenza viral RNA?" *Cell*
67. Frensing, Pflugmacher, Bachmann et al. (2014) "Impact of defective interfering particles on virus replication and antiviral host response in cell culture-based influenza vaccine production" *Appl Microbiol Biotechnol*
69. Blough, Merlie, Tiffany (1969) "The fatty acid composition of incomplete influenza virus" *Biochem Biophys Res Commun*
70. Blough, Merlie (1970) "The lipids of incomplete influenza virus" *Virology (Auckland)*
71. Aggarwal, Bradel-Tretheway, Takimoto et al. (2010) "Biochemical characterization of enzyme fidelity of influenza A virus RNA polymerase complex" *PLoS One*
72. Ganai, Johansson (2016) "DNA replication-a matter of fidelity" *Mol Cell*
73. Jaworski, Routh (2017) "Parallel ClickSeq and Nanopore sequencing elucidates the rapid evolution of defective-interfering RNAs in Flock House virus" *PLoS Pathog*
74. Velthuis, Long, Bauer et al. (2018) "Mini viral RNAs act as innate immune agonists during influenza virus infection" *Nat Microbiol*
75. Amarilla, Modhiran, Setoh et al. (2021) "An optimized high-throughput immunoplaque assay for SARS-CoV-2" *Front Microbiol*
76. Christopher, Ince, Wrammert et al. (2013) "Most influenza A virions fail to express at least one essential viral protein" *J Virol*
77. Brooke (2014) "Biological activities of "noninfectious" influenza A virus particles" *Future Virol*
78. Cooper (1961) "The plaque assay of animal viruses" *Adv Virus Res*
79. Hirst (1942) "The quantitative determination of influenza virus and antibodies by means of red cell agglutination" *J Exp Med*
80. (2025) "Minireview mBio November"
81. Nayak, Tobita, Janda et al. (1978) "Homologous interference mediated by defective interfering influenza virus derived from a temperature-sensitive mutant of influenza virus" *J Virol*
82. Janda, Davis, Nayak et al. (1979) "Diversity and generation of defective interfering influenza virus particles" *Virology (Auckl)*
83. Marcus, Ngunjiri, Sekellick (2009) "Dynamics of biologically active subpopulations of influenza virus: plaque-forming, noninfectious cell-killing, and defective interfering particles" *J Virol*
84. Treuhaft (1983) "A colorimetric assay for quantification of defective interfering particles of respiratory syncytial virus" *J Gen Virol*
85. Crumpton, Dimmock, Minor et al. (1978) "The RNAs of defective-interfering influenza virus" *Virology (Auckland)*
86. Fonville, Marshall, Tao et al. (2015) "Influenza virus reassortment is enhanced by semi-infectious particles but can be suppressed by defective interfering particles" *PLoS Pathog*
87. Routh, Johnson (2014) "Discovery of functional genomic motifs in viruses with ViReMa-a virus recombination mapper-for analysis of nextgeneration sequencing data" *Nucleic Acids Res*
89. Sotcheff, Zhou, Yeung et al. (2023) "ViReMa: a virus recombination mapper of next-generation sequencing data characterizes diverse recombinant viral nucleic acids"
90. Yeung, Routh (2022) "ViReMaShiny: an interactive application for analysis of viral recombination data" *Bioinformatics*
91. Beauclair, Mura, Combredet et al. (2018) "DI-tector: defective interfering viral genomes' detector for nextgeneration sequencing data" *RNA*
92. Achouri, Felt, Hackbart et al. (2023) "VODKA2: a fast and accurate method to detect non-standard viral genomes from large RNA-seq data sets" *RNA*
94. Kim, Pertea, Trapnell et al. (2013) "TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions" *Genome Biol*
95. Dobin, Davis, Schlesinger et al. (2013) "STAR: ultrafast universal RNA-seq aligner" *Bioinformatics*
96. Eddy (2011) "Accelerated profile HMM searches" *PLoS Comput Biol*
97. Alnaji, Reiser, Rivera-Cardona et al. (2021) "Influenza A virus defective viral genomes are inefficiently packaged into virions relative to wild-type genomic RNAs" *mBio*
98. Cane, Sanderson, Barnett et al. (2024) "Nanopore sequencing of influenza A and B in Oxfordshire and the United Kingdom"
99. Cheng, Fei (2023) "Methods to improve the accuracy of nextgeneration sequencing" *Front Bioeng Biotechnol*
100. Alnaji, Holmes, Rendon et al. (2019) "Sequencing framework for the sensitive detection and precise mapping of defective interfering particle-associated deletions across influenza A and B viruses" *J Virol*
101. Andrew, Head, Ordoukhanian et al. (2015) "ClickSeq: fragmentation-free next-generation sequencing via click ligation of adaptors to stochastically terminated 3′-azido cDNAs" *J Mol Biol*
102. Ratcliff, Merritt, Gooden et al. (2024) "Improved resolution of avian influenza virus using Oxford Nanopore R10 sequencing chemistry" *Microbiol Spectr*
103. Head, Komori, Lamere et al. (2014) "Library construction for nextgeneration sequencing: overviews and challenges" *BioTechniques*
104. Ma, Shao, Tian et al. (2019) "Analysis of error profiles in deep next-generation sequencing data" *Genome Biol*
105. Karst, Ziels, Kirkegaard et al. (2021) "High-accuracy long-read amplicon sequences using unique molecular identifiers with Nanopore or PacBio sequencing" *Nat Methods*
106. Keller, Rambo-Martin, Wilson et al. (2018) "Direct RNA sequencing of the coding complete influenza A virus genome" *Sci Rep*
107. Van Der Toorn, Bohn, Liu-Wei et al. (2025) "Demultiplexing and barcode-specific adaptive sampling for nanopore direct RNA sequencing" *Nat Commun* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12549545&blobtype=pdf | # Editorial: Advancing immunotherapy in the elderly: overcoming metabolic and inflammatory barriers
Zechen Wang, Chao Liu, Guang-Liang Chen, Peter Brossart
## Abstract
Editorial on the Research TopicAdvancing immunotherapy in the elderly: overcoming metabolic and inflammatory barriersAs populations age, cancer care is being reshaped by a growing burden of malignancy in older adults (1). Aging is the dominant risk factor for cancer and brings a distinct host biology, including immunosenescence, inflammaging, and dysregulated immunometabolism, often compounded by comorbidity and frailty (2-4). Together, these forces alter treatment selection, tolerance, and outcomes. This Research Topic, Advancing Immunotherapy in the Elderly: Overcoming Metabolic and Inflammatory Barriers, assembles studies that directly address these challenges. The contributions move beyond treating age as a covariate to interrogate the aging host biology and offer strategies to sharpen diagnostics, improve therapeutic efficacy, and mitigate debilitating systemic symptoms.
Redefining the efficacy of immunotherapy in the aging hostA prevailing concern in geriatric oncology has been that immunosenescence may render immune checkpoint inhibitors (ICIs) ineffective in older patients. The work presented in this topic provides a powerful rebuttal to this dogma. A multicenter observational analysis by Drobniak et al. evaluated the combination of nivolumab and ipilimumab in older adults with metastatic renal cell carcinoma (mRCC). Their findings provide reassuring real-world evidence: patients aged ≥65 years demonstrated superior outcomes compared to younger patients (<65 years), with longer progression-free survival, higher ORR/DCR, and acceptable immune-related toxicity profiles. This challenges the conventional wisdom that age alone should be a limiting factor for immunotherapy, Frontiers in Immunology frontiersin.org 01
proving that an "aged" immune system remains remarkably capable of being harnessed against cancer.
on optimization of the host-tumor interface, Zhan et al. explored a multimodal strategy in a real-world cohort with locally advanced gastric cancer (LAGC). By integrating ICIs into a neoadjuvant regimen with chemotherapy and anti-angiogenic therapy, they achieved impressive rates of pathological complete response (23.7%) and major pathological response (47.4%), with R0 resection 97.4% and encouraging 1-year overall survival (OS) 100%/ disease-free survival (DFS) 94.7%. Median OS and DFS had not been reached at the time of reporting. Importantly, while not elderly-specific, the cohort's median age was 65 and 63.2% were >60 years old, supporting applicability to many older candidates for surgery. These data illustrate a practical multimodal path to enhance surgical readiness and early outcomes; longer follow-up is needed to confirm durability. When considered alongside realworld mRCC results showing at least comparable disease control and PFS in patients ≥65 on nivolumab plus ipilimumab, the case for expanding ICI use in well-selected older adults, and for combination strategies that optimize the host-tumor interface, grows stronger.
## Targeting the inflammatory milieu to improve patient outcomes
Beyond direct anti-tumor activity, a critical frontier in geriatric oncology is managing the systemic consequences of cancer, which are often exacerbated by inflammaging. Cancer cachexia, a debilitating syndrome of weight loss and muscle wasting driven by systemic inflammation, severely impacts quality of life and treatment tolerance. The work of Chen et al. provides a targeted therapeutic approach. They investigated the use of tocilizumab, an IL-6 receptor inhibitor, in combination with corticosteroids to manage cancer cachexia. Their results show that this immunomodulatory strategy was associated with improved inflammatory and nutritional indices, weight gain, increased strength, and higher rates of restarting anticancer therapy, without excess infections or ICU admissions. These findings support targeted anti-inflammatory approaches that directly block a key cytokine pathway as an effective bridge to maintain or resume cancer treatment.
## Innovating diagnostics for a new era of precision medicine
Effective treatment begins with accurate and early detection; a need that is particularly acute in frail elderly patients for whom invasive procedures carry higher risk. In this Topic, the study on exosome-derived lncRNA PITPNA-AS1 in pleural effusions by Chen et al. shows how a liquid-biopsy signal can aid differential diagnosis across lung cancer subtypes and track disease burden. Their research shows that exosomal lncRNA PITPNA-AS1 distinguishes malignant from benign lung conditions with high sensitivity and specificity across several lung-cancer subtypes. Mechanistically, PITPNA-AS1 binds FMR1, blocks its ubiquitination, and thereby stabilizes the protein, an interaction that may drive oncogenesis. Consequently, the study moves beyond diagnostics: by implicating PITPNA-AS1 in tumorigenesis, it reveals a potential therapeutic target at the nexus of cellular metabolism and cancer signaling. Overall, PITPNA-AS1 emerges as a promising, non-invasive biomarker; future work should assess its prognostic value and longitudinal performance.
## Conclusion and future outlook
Collectively, the studies in this Topic show that we can expand the appropriate use of ICIs in older adults, modulate inflammatory and angiogenic barriers to improve perioperative efficacy, and deploy non-invasive diagnostics that ease the burden on frail patients. The path forward is not to withhold therapy based on chronological age, but to personalize care by integrating immunosenescence, inflammaging, and immunometabolic considerations, so that more older patients can both receive and benefit from cancer immunotherapy.
## References
1. Zhu, Yang, Bao et al. (2025) "Global, regional, and national burden of cancer in the elderly population, 1990-2021: analysis of data from the global burden of disease study 2021" *Med Res*
2. Chen, Su, Xue (2025) "Targeting T-cell aging to remodel the aging immune system and revitalize geriatric immunotherapy" *Aging Dis*
3. Loṕez-Otıń, Blasco, Partridge et al. (2023) "Hallmarks of aging: An expanding universe" *Cell*
4. Li, Chen, Zhou et al. (2021) "Increased risk of lymphoma in men or the elderly infected with tuberculosis" *Mediterr J Hematol Infect Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282154&blobtype=pdf | # Correction for Lv et al., "Differential antigenic imprinting effects between influenza H1N1 hemagglutinin and neuraminidase in a mouse model"
Huibin Lv, Wen Qi, Chang-Chun Teo, Weiwen Lee, Danbi Liang, Kevin Choi, Madison Mao, Akshita Ardagh, Arjun Gopal, Matt Mehta, Roberto Szlembarski, Ian Bruzzone, Nicholas Wilson, Chris Wu, Mok, Nicholas Wu, Chris Mok |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772252&blobtype=pdf | # A post-implementation evaluation of BioFire FilmArray Meningitis/Encephalitis panel for pathogen detection in cerebrospinal fluid with a special focus on clinical significance of HHV-6
Manoshi Perera, Hemalatha Varadhan, Aileen Oon
## Abstract
Syndromic molecular testing for the diagnosis of infections has been advantageous given high sensitivity and rapid diagnostics; however, challenges remain with appropriate interpretation. A post-implementation evaluation of the cerebrospi nal fluid (CSF) BioMerieux BioFire FilmArray Meningitis/Encephalitis panel in an adult and pediatric population was performed. The aims were to assess the prevalence and epidemiology of all targets and to correlate laboratory parameters with results from human herpesvirus 6 (HHV-6) detection in CSF to identify opportunities to apply diagnostic stewardship. A retrospective observational study was conducted on CSF Nucleic Acid Amplification Test (NAAT) results performed at a referral laboratory in Australia between 19 March 2022 and 31 July 2023. Laboratory data were extracted, with additional clinical data obtained for the HHV-6 positive subcohort. Of 1,490 CSF specimens identified, 287 (19.3%) had a positive NAAT. Non-neonate CSF specimens with white cell count (WCC) above the normal range had a higher proportion of NAAT-positive rates (35.3%) compared to those with WCC < 5 × 10 6 cells/L (6.5%) (P < 0.001). A positive NAAT result was detected in 90 (39.1%) specimens in neonates, with enterovirus and parechovirus detected most frequently. Concordance of bacterial and fungal targets and culture results was seen in 16 (33.3%) specimens. Forty-eight CSF specimens had HHV-6 detected, with a median age of 24.3 years (interquartile range: 0.5-58.7) at collection and 10 specimens with a documented immunocompromised status. This study highlights the caution required when utilizing syndromic diagnostic assays for the detection of CSF pathogens without associated diagnostic stewardship strategies and the potential utility of CSF WCC as a criterion for further review. IMPORTANCE Utilization of molecular testing methods enables rapid and high-sensitiv ity diagnostics, which can have significant impacts upon clinical outcomes in meningitis and encephalitis. However, challenges remain with the interpretation of results when syndromic panels are used, such as with the detection of human herpesvirus 6 (HHV-6) in cerebrospinal fluid (CSF). This study aimed to evaluate the BioMerieux BioFire FilmArray Meningitis/Encephalitis panel for the detection of pathogens in CSF in an adult and pediatric population. Data demonstrated that performing CSF nucleic acid amplifica tion testing without diagnostic stewardship strategies in place can be associated with increased detection of potential bystander pathogens, including HHV-6.
T he introduction of syndromic molecular testing methods for the diagnosis of central nervous system (CNS) infections has been advantageous, given high sensitivity and subsequent implications of rapid diagnostics leading to prompt antimicrobial rationaliza tion, optimal clinical outcomes, and an impact upon financial burden (1).
Global epidemiology studies have identified a comparatively low rate of CNS infection in Australia, with an incidence of encephalitis of 2.08 per 100,000 person-years in 2019 (2), and an incidence of meningitis in 2016 of 0.5 per 100,000 persons (3). A large prospective cohort study identified enteroviruses, human parechovirus, herpes simplex virus (HSV), and Streptococcus species as the highest causes of meningoencephalitis among Australian children (4). Similarly, in a study over an 18-year period, HSV and varicella zoster virus (VZV) were identified as the most common causes of hospitalization due to encephalitis among adults. The etiology of CNS infections has likely also been impacted by the implementation of nationwide targeted immunization schedules for Streptococcus pneumoniae, Haemophilus influenzae, and Neisseria meningitidis (5).
Human herpesvirus 6 (HHV-6) is a ubiquitous herpesvirus with a wide spectrum of disease from benign, self-limiting illness in children to encephalitis within an immuno compromised host (6). The seroprevalence of HHV-6 in the adult population globally is estimated to be >90%, with the majority of primary infections occurring in childhood (7). HHV-6 has a unique ability among other members of the family Herpesviridae to integrate into the subtelomeric regions of human chromosomes in every host cell (6). Described as chromosomally integrated HHV-6 (ciHHV-6), it has an estimated prevalence of 1% in humans (8). Hence, detection of HHV-6 nucleic acid in cerebrospinal fluid (CSF) may represent primary infection, latency, secondary reactivation, or ciHHV-6 present with host cells within CSF (1). Subsequently, challenges arise in the interpretation of results from a highly sensitive assay like the BioMerieux BioFire FilmArray Meningitis/Encephali tis (FAME) assay and necessitate appropriate correlation with clinical findings (9).
No standardized treatment guidelines are available for the management of HHV-6; however, literature has reported positive outcomes with the use of ganciclovir, foscar net, and less frequently, cidofovir (10,11). These antiviral therapies are associated with significant adverse effects, including anemia, electrolyte disturbance, and nephrotoxicity, and hence require close monitoring during use (10). Subsequently, the implications of reporting and interpretation of results need to be carefully considered by laboratories that perform diagnostic testing.
In March 2022, our laboratory implemented the BioMerieux BioFire FAME assay as part of CSF multiplex Nucleic Acid Amplification Test (NAAT) testing in order to facilitate more rapid diagnostics for a large, heterogeneous adult and pediatric patient popula tion. Prior to the implementation of the assay in March 2022, CSF specimens were referred for testing of targeted pathogens as specifically requested by the clinician.
## Aim
Evaluation of the diagnostic performance of the BioMerieux BioFire FAME panel (from this point forward referred to as CSF NAAT), on CSF in a population of adult and pediatric patients was undertaken with two main aims. The first is to assess the prevalence and epidemiology of all NAAT targets, and the second is to correlate laboratory parameters with results from HHV-6 detection on CSF to identify opportunities for the laboratory to apply diagnostic stewardship.
## MATERIALS AND METHODS
A retrospective observational study was conducted of CSF NAAT results, performed at a single referral laboratory in Newcastle (New South Wales), Australia between 19 March 2022 and 31 July 2023. The laboratory service performs testing for regional and rural health facilities in the region and is co-located with an adult and a specialist children's hospital, which includes transplant, hematology, and oncology clinical services. Inclusion criteria included pediatric and adult CSF specimens obtained from lumbar puncture, at least 200 µL volume collected, and referred specimens from Hunter New England and Mid-North Coast Local Health Districts. Exclusion criteria included forensic specimens, repeat specimens collected from the same patient during the same admission period, and specimens collected external to the predefined local health districts.
## Laboratory procedures
CSF multiplex NAAT testing is ordered at clinician discretion, and the BioMerieux BioFire FAME assay is performed as per the manufacturer's instructions for use. The CSF NAAT is an automated, real-time assay from BioMerieux BioFire Diagnostics which enables the isolation, amplification, and detection of nucleic acid from seven viral targets-HSV 1 and 2, VZV, HHV-6, human parechovirus, enterovirus, and cytomegalovirus (CMV)-as well as Cryptococcus neoformans/gattii and six bacterial targets-Listeria monocytogenes, Neisseria meningitidis, Streptococcus pneumoniae, Streptococcus agalactiae, Escherichia coli K1, and Haemophilus influenzae. Analysis occurs simultaneously from a single 200 µL CSF specimen, producing a qualitative result within approximately 1 h. Detection of a pathogen is considered a critical result and is verbally notified to the treating clinician.
CSF microscopy and bacterial culture are performed onsite as per local laboratory standard procedures. Microscopy, including cell count and Gram stain, is performed manually, and culture plates are incubated for a minimum of 5 days. Organism identifica tion is performed using MALDI-ToF (Bruker Biotyper).
HHV-6 serum quantitative NAAT and serology were referred to an external reference laboratory. No other targeted NAAT testing was performed on CSF specimens.
## Data collection
Laboratory data were extracted from the laboratory information system (AUSLAB). This included results of CSF NAAT, CSF culture, and CSF microscopy (white cell count, WCC), in addition to patient demographics of age and gender. Additional laboratory data were obtained for the CSF HHV-6 positive subcohort, including CSF microscopy (red blood cell, RBC; WCC differential), CSF biochemistry (glucose, lactate, and protein), peripheral WCC and differential, serum HHV-6 viral load, and HHV-6 serology. Additional clinical details, including immunocompromised status, documented clinical diagnosis, and antiviral therapy received, were obtained for the CSF HHV-6 positive subcohort.
Normal CSF WCC in adults and children (aged between 1 month and 18 years) was defined as <5 × 10 6 cells/L and RBC as <5 × 10 6 cells/L (12,13). Normal CSF WCC in neonates (≤1 month) was defined as <20 × 10 6 cells/L (14,15). Normal ranges for CSF biochemistry were defined as follows: glucose 2.8-4.5 mmol/L with lower limit of normal (LLN) < 2.8 mmol/L, protein 0.2-0.7g/L with upper limit of normal (ULN) > 0.7 g/L, and lactate < 2.8 mmol/L with ULN ≥ 2.8 mmol/L (13). Normal CSF biochemistry in neonates included glucose ≥ 2.0 mmol/L and protein < 1.0 g/L (12).
## Statistical analysis
Standard descriptive statistics were used to analyze participant demographics, microbiological data, and clinical data. Qualitative/categorical variables were described as counts and proportions and compared with the chi-square test. Statistical significance was accepted as P < 0.05. Microsoft Excel version 16.0 and IBM SPSS version 29.0 were used for statistical analysis.
## RESULTS
## Total cohort
## Overview
During the study period, 1,490 CSF specimens were identified from 1,457 different patients. Thirty-three patients had more than one CSF NAAT performed in the study period during different admission episodes, with the median number of days between collection being 25 (interquartile range [IQR]: 17.0-65.0). There was a male preponder ance of 52% with the median age at CSF collection of 34.9 years (range 0.3-61.7 years). The cohort comprised 958 (64.3%) specimens from adults, 302 (20.3%) from children, and 230 (15.4%) from neonates.
## CSF NAAT
A positive CSF NAAT was identified in 287 specimens (19.3%), with five specimens (0.3%) having more than one organism detected. This included one specimen with VZV and Streptococcus pneumoniae detected, two specimens with HHV-6 and enterovi rus, one specimen with HHV-6 and Streptococcus agalactiae, and one specimen with HHV-6 and VZV (Table 1). Enterovirus was detected in 130 specimens (8.7%), and HHV-6 was detected in 48 specimens (3.2%). There was a variable distribution of pathogens detected between the neonate, children, and adult subcohorts (Fig. 1).
## CSF culture
There was no growth from culture in 1,421 (95.4%) specimens, and culture was not performed for 15 (1.0%) specimens. Of samples which isolated an organism from culture, 38 (70.4%) were not target pathogens within the CSF NAAT panel-these included coagulase-negative staphylococci (20, 37%), Candida dubliniensis (1, 1.9%), Staphylococ cus aureus (1, 1.9%), other Gram-positive organisms (13, 24.1%), and other Gram-negative organisms (3, 5.6%). Concordance of bacterial and fungal NAAT targets and culture results was seen in 16 (33.3%) specimens, with concordance seen in all Cryptococcus cases (Table 2). Four CSF specimens with a virus detected on NAAT isolated bacteria on CSF culture, including coagulase-negative staphylococci and other Gram-positive organisms.
## CSF microscopy
The median WCC on CSF microscopy was 3 × 10 most frequently detected (Fig. 2). In contrast, there were 347 CSF specimens from non-neonates with elevated WCC (≥5 × 10 6 cells/L) and negative CSF culture, of which 115 (33.1%) had a positive NAAT. Overall, non-neonate CSF specimens with a WCC above normal range had a higher proportion of NAAT-positive rates (35.3%) compared to those with a WCC < 5 × 10 6 cells/L (6.5%) (P < 0.001). NAAT-positive specimens in adults, there were also more frequent bacterial/fungal detections compared to the pediatric cohort. This was supported by 38 positive CSF cultures, including five Cryptococcus isolates, five Streptococcus pneumoniae, and three Haemophilus influenzae.
## Age-based subcohorts
## HHV-6 positive subcohort
Forty-eight (48/1,490) CSF specimens had HHV-6 detected via NAAT, with a male preponderance (58.3%) and a median age of 24.3 years (IQR: 0.5-58.7) at CSF collection. An elevated CSF RBC above the normal range was noted in 41.7% of specimens (Table 4).
An immunocompromised status was documented with 10 CSF specimens, of which two underwent HHV-6 quantitative NAAT and serology.
## DISCUSSION
This study outlines the epidemiology of CSF NAAT positivity in a heterogeneous adult and pediatric population. Furthermore, it highlights parameters for the application of diagnostic stewardship, as exemplified by the frequency of detection of HHV-6.
The advantages of utilizing a CSF multiplex NAAT are often recognized in cases of aseptic meningitis-CSF specimens with elevated WCC and negative culture (16). Our study had positive identification in 33.1% of these specimens via NAAT, including bacterial pathogens. Enterovirus was the most commonly detected pathogen on NAAT in our study, and this was similarly reported in a recent study as the most common cause of aseptic meningitis in a cohort of pediatric and adult patients (16). The over all concordance between culture and bacterial/fungal targets was lower in our study (33.3%) than previously described by Myint et al., though comparable H. influenzae and N. meningitidis targets (17). Interestingly, our study had 100% concordance between CSF NAAT and culture results for Cryptococcus gattii and Cryptococcus neoformans. This contrasts with previous studies which have highlighted a high false-negative rate for the CSF NAAT, particularly among patients on anti-fungal therapy and/or who have a low disease burden (18). Other instances of false negatives have been described in association with HSV 1/2 (19). Hence, recommendations made by our laboratory to repeat CSF collection if clinical suspicion for HSV infection remains. Instances of NAAT bacteria detection with sterile CSF culture were noted; however, an attributable cause was not directly assessed-a limitation of our study. Previously ascribed causes include antimicrobial treatment prior to specimen collection in hospital or community, fastidiousness of organism, and organism concentration at assay limit of detection (16,20). The majority (80.7%) of CSF specimens in our study had a negative NAAT panel, which is similar to rates reported by Precit et al., who had a comparable patient cohort and size (20). This raises concerns about overutilization of the assay (20), emphasizing the need to rationalize testing algorithms and employ diagnostic stewardship practices (21). A lenient testing criterion can be detrimental in placing an added burden on clinicians with regard to the interpretation of results. Restriction criteria previously explored include CSF WCC within the normal range; however, studies have described that the absence of leukocytosis is not always reliable (10,22). Important exceptions to consider include specimens from neonates and detection of parechovirus, which is commonly associated with the absence of reactive pleocytosis (15,23). This was similarly seen in our study with detection of bacterial pathogens, including S. agalactiae and E. coli K1 target via NAAT in neonates with CSF WCC within normal range. Within our cohort, there was a statistically significant difference in NAAT positivity between non-neonate specimens with normal and elevated CSF WCC. There was only 6.5% NAAT positivity among non-neonates with CSF WCC within normal range, with a minority of targets requiring directed therapy and positive cultures with growth of what were deemed likely contaminants. Most pathogens detected in this subcohort included viruses associated with reactivation but not clinical CNS disease, such as HHV-6 and VZV (1). Of the two positive NAAT bacterial detection, S. pneumoniae was deemed clinically unlikely and a possible false positive due to contamination, which has previously been reported by Leber et al. (16). While, detection of H. influenzae on NAAT with a sterile culture, may be attributable to antimicrobial treatment prior to specimen collection (17). Thereby supporting the use of this WCC criterion among non-neonate specimens as a prompt for clinician discussion prior to proceeding with CSF NAAT.
Another consideration with the use of multiplex NAAT testing methods is the identification of targets which may not correlate with clinical pathogenicity. This is particularly important with viruses given the possibility of subclinical reactivation, latency, and genomic integration (1). The incidence of reported HHV-6 associated meningoencephalitis or transverse myelitis in the HHV-6 positive cohort (22.9%) was similar to that reported by Pandey et al. (24). However, compared to Pandey et al., which was only of pediatric patients, our study cohort had a median age of 24.3 years, with only 35.7% of CSF specimens from patients aged <18 years (24). This is an important consideration, given HHV-6 meningoencephalitis among immunocompetent adults is reportedly rare, and our study cohort included a majority (79.2%) without an immu nocompromised status documented (25). Furthermore, the predominant peripheral neutrophilia among the immunocompetent patients may suggest a non-viral etiology. In contrast, HHV-6 associated with severe clinical outcomes has been described among immunocompromised individuals, with manifestations of encephalitis, hemophagocytic lymphohistiocytosis, long-term neurological sequelae, and mortality (26). Although HHV-6 reactivation can occur in up to two-thirds of hematopoietic stem cell transplant recipients, the prevalence of encephalitis among this cohort remains small, as low as 1.4% (27). Further to the low prevalence of HHV-6 meningoencephalitis, the clinical symptoms remain nonspecific, including headache, delirium, and seizures (28), further impacting upon clinician confidence in diagnosing infection. This may have contributed to the diagnosis of two immunocompetent patients within the study cohort with HHV-6 CNS infection who received targeted anti-viral therapy with ganciclovir. Other factors which can further complicate interpretation are in CSF specimens with elevated RBC, suggestive of a traumatic lumbar puncture and possible blood contamination. Thereby, the presence may represent serum levels, rather than be representative of CNS infections (25). This may have contributed to CSF detection in 41.7% of specimens in the HHV-6 cohort in our study, which had CSF RBC levels above the ULN.
Supplementary diagnostic aids which can be utilized for HHV-6 meningoencephali tis include the presence of lymphocytic pleocytosis and elevated protein in CSF (27), although not consistently seen (25), as was similar to our study cohort with a median mononuclear cell count of 82% and elevated protein in only 25% (12/48) of speci mens. Incorporation of serology into HHV-6 infection diagnostics has been infrequently reported in the literature, due to limited utility in distinguishing acute infection or reactivation from previous exposure (19). Our study had 12.5% of the HHV-6 subcohort undergo serological testing, although the diagnostic utility of one neonatal specimen is questionable unless compared to maternal serology. Furthermore, two patients with serology consistent with past infection received targeted antiviral therapy, although this may reflect the prolonged turnaround time in receiving serology results.
Serum HHV-6 quantitation can assist in differentiating active infection or reactiva tion; however, a high level does not consistently occur with CNS infection and can be isolated from otherwise healthy individuals (28). In contrast, digital droplet NAAT enables absolute quantitation of ciHHV-6 by fragmenting the specimen and perform ing multiple parallel amplification reactions concurrently (14). Although this was not performed in our study, an approximation can be made via quantitative viral load of whole blood or serum. There is a consensus in the literature that a whole blood viral load >5.5log 10 copies/mL or serum viral load >3.5log 10 copies/mL is suggestive of chromosomal integration (14,24,29,30). However, pre-analytic factors also need to be considered prior to interpretation of viral load, including WCC count, which can significantly elevate levels (9). Of note, 50% (3/6) of patients who had serum HHV-6 quantitative NAAT performed had a viral load of >3.5log 10 copies/mL suggestive of ciHHV-6. Overall, our study had underutilization of these supplementary tests, particu larly among the subcohort of immunocompromised patients with HHV-6 detected; only two patients proceeded to have further testing, including HHV-6 quantitative NAAT. This renders interpretation challenging, given the likelihood of subclinical reactivation among immunocompromised patients and in those with severe acute illness due to another etiology (19). Hence, this emphasizes the importance of clinicians having a high pre-test probability, with concordant clinical presentation and risk factors. Pursuing supplemen tary testing such as quantitative HHV-6 NAAT on CSF and serum can further help clarify this, although it was not accessible in our laboratory testing network (31). This may be further prompted by the inclusion of an interpretative comment alongside HHV-6 results, suggesting detection may represent primary infection, latency, or secondary reactivation and for consideration of quantitative NAAT testing if clinically indicated.
Antiviral therapy such as ganciclovir, which targets HHV-6 DNA polymerase to inhibit viral DNA synthesis, is utilized in HHV-6 management; however, it can be associated with adverse effects such as renal impairment and bone marrow suppression (24). Other considerations including the significant financial burden of antiviral therapy and potential increased length of stay for administration of intravenous therapy. The efficacy is also poorly understood in immunocompetent pediatric populations (31), which can likely be extrapolated to immunocompetent adult patients. These concerns are similarly raised in our study, where among the 14.6% (7/48) of patients who received anti-viral therapy, three were immunocompetent hosts, including one case in a pediatric patient.
Strength of the study was the inclusion of a heterogeneous patient population, suggesting generalizability of study findings. This included analysis of CSF specimens from both pediatric and adult patients, immunocompromised patients, and inclusion of multiple regional and rural institutions. Limitations of this study include the retrospective nature and reduced sample size in some subcohorts; therefore, no power calculations were performed. Furthermore, full clinical and laboratory data were not available on all patients, in addition to lack of data regarding the prevalence of ciHHV-6 and supportive evidence such as radiological imaging to support diagnoses of HHV-6 encephalitis.
Subsequent review of these study findings provided confidence in implementing a diagnostic stewardship algorithm within our laboratory service. This involved initiating a medical microbiologist review if a non-neonate specimen had CSF WCC within normal range to determine if CSF NAAT should be performed. This included review of medi cal records and discussion with requesting clinician to determine the indication and underlying risk factors, e.g., immunocompromised host. Future studies worth exploring include a prospective trial following implementation of restrictive criteria for perform ing CSF NAAT testing on the basis of WCC and the implications on frequency of HHV-6 detection, alongside other pathogens, as well as laboratory factors, including turnaround time and cost. Development of an algorithm or tool to aid clinicians with interpretation and further investigation of HHV-6 detection on CSF specimens specific to local epidemiology, laboratory workflow practices, and available resources may also be beneficial.
## Conclusions
This study highlights the caution required when utilizing high-sensitivity, rapid diagnostic assays for the detection of CSF pathogens without laboratory stewardship. Syndromic testing is not without its challenges, and this study has highlighted that HHV-6 detection in CSF, in particular, can be prone to misinterpretation. Utilization of CSF WCC as a trigger for further review of non-neonate samples may be a useful criterion to minimize this.
## References
1. Waldrop, Zucker, Boubour et al. (2022) "Clinical significance of positive results of the BioFire cerebrospinal Fluid FilmArray meningitis/encephalitis panel at a tertiary medical center in the United States" *Arch Pathol Lab Med*
2. Wang, Zhao, Wang et al. (2022) "Global magnitude of encephalitis burden and its evolving pattern over the past 30 years" *Journal of Infection*
3. Zunt, Kassebaum, Blake et al. (2016) "Global, regional, and national burden of meningitis, 1990-2016: a systematic analysis for the Global Burden of Disease Study" *Lancet Neurol*
4. Britton, Dale, Blyth et al. (2020) "Causes and clinical features of childhood encephalitis: a multicenter, prospective cohort study" *Clin Infect Dis*
5. Gora, Smith, Wilson et al. (2022) "The epidemiology and outcomes of central nervous system infections in Far North Queensland"
6. Greninger, Naccache, Pannaraj et al. (2019) "The brief case: inherited chromosomally integrated human herpesvirus 6 (HHV-6) in the age of multiplex HHV-6 testing" *J Clin Microbiol*
7. Aimola, Beythien, Aswad et al. (2020) "Current understanding of human herpesvirus 6 (HHV-6) chromosomal integration" *Antiviral Res*
8. Leong, Tuke, Tedder et al. (2007) "The prevalence of chromosomally integrated human herpesvirus 6 genomes in the blood of UK blood donors" *J Med Virol*
9. Green, Pereira, Miko et al. (2018) "Clinical significance of human herpesvirus 6 positivity on the FilmArray meningitis/encephalitis panel" *Clin Infect Dis*
10. Phan, Lautenschlager, Razonable et al. (2018) "HHV-6 in liver transplantation: a literature review" *Liver Int*
11. Pritchett, Naesens, Montoya et al. (2014) "Treating HHV-6 Infections: The laboratory efficacy and clinical use of anti-HHV-6 agents"
12. (2019) "The Royal Children's Hospital Melbourne"
13. (2024) "Cerebrospinal Fluid Examination"
14. Kestenbaum, Ebberson, Zorc et al. (2010) "Defining cerebrospinal fluid white blood cell count reference values in neonates and young infants" *Pediatrics*
15. Ngo Nsoga, Pérez-Rodriguez, Mamin et al. (2023) "Rational use of microbiological tests in the diagnosis of central nervous system infections using restrictive criteria: a retrospective study" *Microbiol Spectr*
16. Leber, Everhart, Balada-Llasat et al. (2016) "Multicenter evaluation of BioFire FilmArray meningitis/encephalitis panel for detection of bacteria, viruses, and yeast in cerebrospinal fluid specimens" *J Clin Microbiol*
17. Myint, Soria, Gao et al. (2025) "Comparison of positive BioFire FilmArray meningitis/encephalitis (ME) panels, CSF cultures, CSF parameters, clinical presentation and in-patient mortality among patients with bacterial and fungal meningitis" *Microbiol Spectr*
18. O'halloran, Lainhart, Burnham et al. (2017) "Pitfalls associated with the use of molecular diagnostic panels in the diagnosis of cryptococcal meningitis" *Open Forum Infect Dis*
19. Tansarli, Chapin (2019) "Diagnostic test accuracy of the BioFire FilmArray meningitis/encephalitis panel: a systematic review and metaanalysis" *Clin Microbiol Infect*
20. Precit, Yee, Pandey et al. (2020) "Cerebrospinal fluid findings are poor predictors of appropriate FilmArray meningitis/encephalitis panel utilization in pediatric patients" *J Clin Microbiol*
21. Patel, Fang (2018) "Diagnostic stewardship: opportunity for a laboratory-infectious diseases partnership" *Clin Infect Dis*
22. Cunha (2006) "Distinguishing bacterial from viral meningitis: the critical importance of the CSF lactic acid levels" *Intensive Care Med*
23. Chakrabarti, Warren, Vincent et al. (2018) "Outcome of routine cerebrospinal fluid screening for enterovirus and human parechovirus infection among infants with sepsis-like illness or meningitis in Cornwall, UK" *Eur J Pediatr*
24. Pandey, Greninger, Levin et al. (2020) "Pathogen or bystander: clinical significance of detecting human herpesvirus 6 in pediatric cerebrospinal fluid" *J Clin Microbiol*
25. Chia, Wong, Loh (2024) "Human herpesvirus-6 infection in a critically ill and immunocompetent patient: a case report" *J Med Case Rep*
26. Raouf, Ouf, Elsorady et al. (2023) "Human herpesvirus-6 in hematopoietic stem cell transplant recipients: a prospective cohort study in Egypt" *Virol J*
27. Li, Zhang, Wang et al. (2024) "Case report: Acute HHV6B encephalitis/myelitis post CAR-T cell therapy in patients with relapsed/refractory aggressive B-cell lymphoma" *Front Neurol*
28. Rebechi, Bork, Riedel (2021) "HHV-6 encephalitis after chimeric antigen receptor T-cell therapy (CAR-T): 2 case reports and a brief review of the literature" *Open Forum Infect Dis*
29. Pellett, Ablashi, Ambros et al. (2012) "Chromosomally integrated human herpesvirus 6: questions and answers" *Rev Med Virol*
30. Ward, Leong, Thiruchelvam et al. (2007) "Human herpesvirus 6 DNA levels in cerebrospinal fluid due to primary infection differ from those due to chromosomal viral integration and have implications for diagnosis of encephalitis" *J Clin Microbiol*
31. Crawford, Kadom, Santi et al. (2007) "Human herpesvirus 6 rhombencephalitis in immunocompetent children" *J Child Neurol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12280948&blobtype=pdf | # Mycophenolic acid treatment drives the emergence of novel SARS-CoV-2 variants
Toni Meis, Maximil Nocke, Natali Heinen, Thomas Burkard, Yannick Brüggemann, Saskia Westhoven, Bettina Trü, Nadine Ebert, L Thomann, P Lubieniecki, Joanna Lubieniecka, Kristina Döring, Maike Herrmann, Sibylle Haid, Thomas Pietschm, Bettina Wiegm, Ronny Ta, Susann Pfefferle, Marylyn Addo, Volker Thiel, Ingo Drexler, Nina Babel, Phuc Nguyen, Richard Brown, Daniel Todt, Eike Steinmann, Stephanie Pfaender
## Abstract
SignificanceMycophenolic acid (MPA), an immunosuppressant widely used in posttransplant regimens, exhibits antiviral activity by depleting cellular guanosine triphosphate, thereby inhibiting viral replication. However, prolonged exposure to MPA can drive the emergence of novel viral variants with enhanced replication capabilities. Here, we identified specific mutations in severe respiratory syndrome coronavirus 2 that conferred altered viral fitness, allowing for faster replication and increased viral titers despite MPA treatment. Importantly, these mutations have been observed in vivo, suggesting a real-world risk of variant evolution under immunosuppressive treatment. However, these mutations have not yet been identified together in a single infected individual. These findings underscore the importance of vigilant monitoring in immunosuppressed patients, as treatment may inadvertently foster viral variants with a competitive advantage.
Viral infections pose a significant global threat, contributing to disease, mortality, and economic losses. A large number of deaths are attributed to infections with viral pathogens globally ( 1 , 2 ). This raises significant concerns, especially given the limited availability of around 90 drugs for treating nine viral species and the approval of only about 15 vaccines for various viral species at present ( 3 ). Recent viral outbreaks including Zika (ZIKV), Ebola (EBOV), and the more recent Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) pandemic highlight the urgent need for novel antiviral strategies.
Approximately 22% of the global population display at least one underlying health condition, which puts them at higher risk of adverse outcomes for infectious diseases ( 4 ). Indeed, the recent SARS-CoV-2 pandemic highlighted the vulnerability of specific pop ulations with comorbidities toward less favorable disease outcomes ( 5 , 6 ). Among those, immunosuppressed patients have become the center of attention. In the United States, an estimated 2.7 to 4% of all adults self-report immunosuppression, which refers to a diverse group of individuals with inherited or acquired immune deficiencies ( 7 ). This group includes individuals with various malignancies and immunologic diseases, as well as patients who have undergone solid organ transplantation ( 8 ). The latter are typically treated with immunosuppressive drugs, which can be classified into different categories, including calcineurin inhibitors (such as Tacrolimus and Cyclosporin A), antiproliferative agents (such as Mycophenolate mofetil and azathioprine), mTOR inhibitors (such as Rapamycin), and steroids (Prednisolone). Mycophenolate mofetil, a prodrug of myco phenolic acid (MPA), is an inhibitor of the inosine monophosphate dehydrogenases (IMPDH) thereby negatively impacting purine biosynthesis and causing cell cycle arrest ( 9 ). In addition to being an immunosuppressive drug, MPA has been shown to effectively inhibit the replication of flaviviruses such as dengue virus (DENV) ( 10 , 11 ) and hepatitis C virus (HCV) ( 12 , 13 ), as well as the Old World arenavirus lymphocytic choriomeningitis virus ( 14 ) and several other viruses including Mpox virus (MPXV) ( 15 -22 ) in vitro. Laboratory experiments have also demonstrated that MPA can impair the replication of several coronaviruses such as Middle East Respiratory Syndrome coronavirus (MERS-CoV), SARS-CoV-2, and endemic coronaviruses (such as human coronavirus HCoV-OC43 and -NL63) ( 14 , 23 , 24 ). Additionally, MPA was recently shown to effectively inhibit OPEN ACCESS 2 of 12 https://doi.org/10.1073/pnas.2500276122 pnas.org bacterial infections including Chlamydia ( 25 ) and also displayed beneficial effects on gastric cancer cells by targeting several KEGG pathways ( 9 ). Intriguingly, in animal models including mice and common marmosets, the administration of MPA resulted in enhanced infection rates for SARS-CoV-2 and MERS-CoV. However, the underlying mechanism and potential for viral adap tation is still to be determined.
Despite ongoing debates regarding the severity of the disease ( 26 -28 ) and mortality rates ( 27 -29 ), there is a consistent body of evidence reporting prolonged or chronic infections with SARS-CoV-2 in immunocompromised individuals ( 30 -34 ). Certain reports even proposed that SARS-CoV-2 Variants of Concern (VOCs), which became globally dominant, including the Alpha variant, are thought to have originated from individuals with compromised immune systems ( 35 -37 ). This phenomenon has been observed in chronic infections with other viruses, such as DENV ( 38 ) and hepatitis E virus (HEV) ( 39 ). Continuous replication of the virus potentially favors the accumulation of mutations in order to adjust to environmental and immunological pressure and therefore presumably facilitates the emergence of new variants with enhanced epidemic or pandemic potential ( 35 , 40 -43 ). Additionally, immunocompromised patients often require pharmacological interventions to clear infections, which may also further drive viral mutagenesis. Direct acting antivirals such as the nucleotide analogues Remdesivir or Ribavirin (RBV) facilitate lethal mutagenesis during viral replication, which leads to defective or replication-incompetent viral genomes and subsequent viral extinction ( 44 , 45 ). However, in some cases the selective pressure drives viral mutagenesis leading to the formation of intrahost pop ulations ( 46 , 47 ). RBV treatment of patients infected with HEV was associated with increased viral heterogeneity, eventually result ing in the emergence of RBV-resistant mutants and/or variants with increased viral fitness, causing therapy failure ( 46 -49 ). Remdesivir has also been shown to further facilitate viral mutagen esis upon administration of low concentrations ( 47 , 50 ). The emergence of resistance mutations has also been reported during treatment with antiretroviral therapy of HIV infection ( 51 ).
The emergence of novel viral variants presents a risk to the gen eral population by affecting various aspects such as transmission dynamics, public health measures, vaccine efficacy, antiviral treat ment, disease severity, and pathogenesis, among others. Therefore, it is crucial to thoroughly investigate potential drivers of mutagen esis and understand the impact of specific mutations. Here, we aimed to explore viral replication under immunosuppressive ther apy to better predict and manage the spread of novel variants.
## Results
## Mycophenolic Acid Inhibits a Wide Range of Viruses. Immuno suppression can have a profound impact on viral infections (52).
To further investigate this, we tested the antiviral activity of MPA, Rapamycin, Cyclosporin A, Tacrolimus and Prednisolone against SARS-CoV-2 infection in vitro (SI Appendix, Fig. S1 and Table S1). To this end, VeroE6 cells were treated with single or combinatorial doses of immunosuppressants at clinically relevant concentrations to mimic long-term immunosuppression, reflecting the common practice of prescribing multiple immunosuppressants to patients (53)(54)(55)(56)(57)(58)(59)(60). Viral replication was determined by an end-point dilution assay to determine infectious particles released into the supernatant (SI Appendix, Fig. S1A) and cell associated viral M-gene expression was quantified by qRT-PCR (SI Appendix, Fig. S1B). Consistent with previous reports (10), MPA treatment led to a significant decrease in infectious viral titers (1.5 logarithmic units [log] TCID 50 / mL), whereas none of the other immunosuppressive drugs exhibited an antiviral effect upon single treatment (SI Appendix, Fig. S1 A andB). None of the drug combinations affected cell viability (SI Appendix, Fig. S1C). Combinations encompassing MPA reduced virus replication, without any synergistic effects with drugs that are frequently coadministered (SI Appendix, Fig. S1 A and B). To validate these findings in a more physiologically relevant system, differentiated primary human airway epithelial cells (hAEC) in air liquid interface (ALI) culture were exposed to single treatments or different combinations of immunosuppressive drugs, and the production of infectious particles (Fig. 1 A-F, Left) and cell viability (Fig. 1 A-F, Right) were monitored over time. In line with previous observations, solely MPA-encompassing treatments reduced the production of infectious progeny, without affecting cell viability (Fig. 1 A-F). These observations confirm an antiviral effect for MPA against SARS-CoV-2 in vitro.
To investigate the breath of antiviral activity, we subsequently examined the susceptibility of unrelated RNA and DNA viruses from various families to MPA inhibition. Consistent with prior findings ( 15 -21 ), we observed a dose-dependent decrease in virus infection upon MPA treatment for the SARS-CoV-2 Omicron variant BA2.86, HCoV-229E, HEV, respiratory syncytial virus (RSV), Influenza A virus (IAV), and MPXV (SI Appendix, Fig. S2 ). These viruses exhibited comparable susceptibility to MPA treat ment, with IC 50 values ranging from 0.07 (IAV) to 1.59 µg/mL (MPXV). IAV and MPXV infections were completely abolished at the highest concentration of MPA, while reduced virus production was detected for HCoV-229E, HEV, and RSV. The IC 50 value for the Omicron variant BA2.86 was similar to what has been observed for the SARS-CoV-2 B.1.1.70 variant. These findings indicate MPA's broad activity against diverse viruses is likely due to its tar geting of cellular processes rather than specific viral proteins.
## Mycophenolic Acid Targets Postentry Stages of the Coronavirus Lifecycle via Depletion of Intracellular Guanosine Triphosphate
(GTP) Pools. Immunosuppressive drugs impact a variety of cellular signaling pathways including JAK/STAT, NF-κB, PI3K/AKT-mTOR, MAPK, and Keap1/Nrf2/ARE pathway (61,62), thereby modulating host cell immune responses. Although the cellular targets of MPA are well defined, its specific antiviral mechanisms remain elusive. MPA has been shown to affect the expression of cellular structural proteins, fatty acid, and lipid metabolism, as well as nucleotide-dependent processes including the reduction of GTP pools (63). To further study which step of the viral replication cycle was affected, we first focused on virus entry, employing pseudotyped VSV where the native glycoprotein was deleted and replaced with the SARS-CoV-2 Spike protein (VSVΔG + S), encoding a firefly luciferase reporter (64). VeroE6 cells were treated with MPA, Tacrolimus, and/or Prednisolone at different time points relative to infection, including 1 h before infection (pre 1 ), before and during infection (pre-during 2 ), during the entire experiment (pre-during-post 3 ) or only after infection with VSVΔG + S (post 4 ) (SI Appendix, Fig. S3A). Cell viability remained unaffected throughout the experiment (SI Appendix, Fig. S3B). While drug treatment prior to infection or before and during infection did not impact luciferase activity, administering MPA after infection with VSVΔG + S resulted in reduced luciferase activity suggesting that MPA does not play a role during virus entry but affects the expression of the reporter gene (SI Appendix, Fig. S3C). Next, similar time-of-addition experiments were performed using authentic SARS-CoV-2. A549-A/T cells were treated with either the vehicle control (EtOH) or MPA for the specified time periods (Fig. 2A) and viral titers were determined. MPA treatment, as long as administered shortly after infection or for an extended period of time before infection (24 h), led to a significant reduction in SARS-CoV-2 secretion compared to the vehicle control. In contrast, a single pretreatment dose administered 1 h before infection did not alter SARS-CoV-2 titers.
A time-course analysis and subsequent determination of intraand extracellular viral titers ( Fig. 2B ) revealed the formation of infectious viral particles within A549-A/T cells takes approxi mately 6 h post infection (hpi) and was delayed by an additional hour before becoming detectable in the supernatant. When MPA treatment is initiated simultaneously with infection, there is no discernible difference in viral loads until the 10-h time point. We also pretreated the cells 10 h before infection, mimicking immu nosuppressive treatment before infection. In this case, we observed a delay of infectious particle production by another 3 h for both intra-and extracellular virus ( Fig. 2B ). Furthermore, immunoflu orescence staining confirmed a reduction of SARS-CoV-2 N-protein positive A549-A/T cells upon MPA treatment ( Fig. 2C ), suggesting that a reduced number of cells establish productive infection upon MPA treatment. Together, these results demon strate that MPA affects postentry stages of the SARS-CoV-2 rep lication cycle, thereby reducing the amount of infectious progeny produced over time.
Similar to RBV, MPA inhibits IMPDH, an essential enzyme in cellular metabolism that regulates the production of guanine nucle otides necessary for DNA and RNA synthesis ( 63 ). Specifically, IMPDH facilitates the conversion of inosine monophosphate (IMP) to xanthosine monophosphate (XMP), a critical step in the de novo biosynthesis of guanosine nucleotides. Consequently, MPA-mediated inhibition of IMPDH leads to reduced levels of GTP ( 63 ). However, studies have shown that this effect can be reversed by introducing exogenous guanosine (G) or guanosine monophosphate (GMP) during RBV treatment ( 20 , 65 ). To explore this mechanism in the context of infected cells, we treated A659-A/T cells with a single dose of MPA while simultaneously adding different concentrations of G and/or GMP ( Fig. 3 ). Increasing concentrations of G and/or GMP resulted in restored viral titers, thus counteracting the antiviral effects of MPA ( Fig. 3 C and F ), without affecting cell viability ( Fig. 3 A andD ). Of note, individual addition of either GMP, G or GMP and G together did not affect virus titers ( Fig. 3 B andE ). Addition of the monophos phates AMP, CMP, UMP, or TMP during MPA treatment did not lead to a rescue of viral infectivity (SI Appendix, Fig. S4 ). These findings suggest that MPA's broad antiviral effect is likely mediated by the depletion of cellular GTP and can be reversed by the exog enous substitution of G and/or GMP, but not other monophosphates.
We hypothesized that combinational treatment of MPA with direct acing antivirals could enhance the antiviral effect, resulting in complete virus inhibition. Currently approved treatment options for COVID-19 include the direct acting antivirals Nirmatrelvir und Ritonavir (Paxlovid®, Pax), as well as RBV ( 66 ). In line with previous reports, single treatment of either compound significantly inhibited SARS-CoV-2 replication (SI Appendix, Fig. S5A ) without any effect on cell viability (SI Appendix, Fig. S5B ). Strikingly, com bination treatments with MPA significantly enhanced the antiviral activity, confirming additive effects and superior activity compared to single treatment (SI Appendix, Fig. S5 C andD ). These observa tions underline the potential superiority of combinational treatment approaches targeting both viral and host processes. Antiviral activity of immunosuppressive drugs in hAECs over time. hAECs were treated with MPA (2.5 µg/mL), Rapamycin (6 ng/mL), Cyclosporin A (150 ng/mL), Tacrolimus (6 ng/mL), and/or Prednisolone (20 ng/mL) in indicated combinations (A-F). After 1 h the cells were infected with SARS-CoV-2 (25,000 PFU) for 2 h and subsequently washed thrice with HBSS. Directly after infection as well as 24, 48, 72, and 96 h post infection, infectious progeny virus was collected from the apical site and subjected to an end-point dilution assay to determine TCID 50 /mL (left half of each graph). Viral titers were normalized to viral titers derived from the untreated cells (dotted line). Simultaneously, cell culture medium from the basal site was obtained to evaluate cell viability by an LDH assay (right half of each graph). All experiments were performed in three different donors (mean ± SD). MPA-mycophenolic acid. ns-not significant.
## SARS-CoV-2 Can Rapidly
driving replication beyond the error threshold, leading to viral extinction, or facilitating the emergence of variants with enhanced viral fitness (46)(47)(48)(49). Medications directed at cellular proteins typically pose a lower risk of promoting antiviral resistance. However, in the case of MPA, the identified mode of action could still play a role in viral mutagenesis. To investigate this, SARS-CoV-2 was serially passaged in A549-A/T cells under different treatment regimens encompassing MPA (M); MPA and Tacrolimus (MT); or MPA, Tacrolimus, and Prednisolone (MTP). Viral titers were monitored throughout the passaging experiments (Fig. 4A). Consistent with our initial findings (Fig. 1 and SI Appendix, Fig. S1), viral titers decreased upon MPA treatment compared to virus propagated on untreated cells (UTC) until passage 5. However, within 8 passages, viral titers recovered to comparable levels as the UTC, pointing toward adaptation to the treatment. No further changes in viral titers were observed during the remaining passaging time (up to 20 passages, p20). The different combinational treatment regimens displayed no discernible differences. Alongside recovered infectivity, phenotypic changes resulting in a significantly increased plaque size, that was not observed in the UTC, became apparent (Fig. 4 B andC). These findings indicate viral adaptation, potentially involving changes in replication fitness and/or resistance to treatment.
To exclude a potential gain-of-function, passaged viruses were tested for their potential to evade neutralizing antibodies and anti viral treatment approaches. Although passaged viruses exposed to MPA exhibited reduced sensitivity (SI Appendix, Fig. S6 A andB ), their responsiveness to clinically relevant drugs such as Molnupiravir, Paxlovid® or Remdesivir and neutralization by neutralizing antibod ies remained unchanged ( Fig. 5 ). Notably, extracellular infectivity was found to be enhanced, particularly for viruses passaged in the presence of MPA, compared to the virus passaged in the UTC cells (SI Appendix, Fig. S6C ). This suggests altered viral fitness noticeable through accelerated replication kinetics and the efficient release of infectious viral particles from A549-A/T cells. Given the increased pressure upon depletion of cellular GTP pools, we hypothesized that virus-host interactions may be modulated as viruses adapt toward differential cellular conditions. Therefore, the ratio of virushost reads and cellular transcriptional responses to infection in UTC and MPA-treated cells were investigated by RNA-sequencing (RNA-seq) ( 67 ) ( Fig. 6 ). In UTC cells, virus read numbers generally remained stable at ~20% of cellular reads across all passages. In contrast, in MPA treated cells, early passage virus-mapped reads (p1, p3, and p5) were demonstrably lower (<10%), followed by a massive increase to over 70% of total cellular reads at later passages (p9, p15, and p20) ( Fig. 6A ). No obvious dysregulation of viral cofactors or lung markers was observed ( Fig. 6B ). Based on these data, comparison of global transcriptional responses to infection in UTC cells (n = 6) were made separately to cells infected with either early passage virus where replication was suppressed, (n = 3) or late passage virus where replication was enhanced (n = 3) ( Fig. 6 C -F ). In early virus passage infected cells, differentially expressed gene (DEG) and Gene Ontology (GO) analysis unveiled a dysregulation of cellular genes ( Fig. 6 C and D ), associated with distinct biological processes ( Fig. 6 D -F ). Specifically, MPA-treated cells showed enhanced protein modification processes as well as response to stress, NFκB signaling, and biosynthetic and metabolic processes ( Fig. 6 E and F ). In contrast, late passage virus-infected cells showed min imal transcriptional differences when compared to UTC ( Fig. 6 C -F ). Together, these data highlight the host genes and cellular processes associated with MPA suppression of SARS-CoV-2. Additionally, the rapid loss of MPA associated inhibition, evidenced by the massive increase of viral replication observed with late passage virus, are indicative of the emergence of breakthrough mutations, conferring altered viral fitness restoring the ability of the virus to replicate efficiently in the context of cellular metabolic changes induced by MPA.
Mutational Changes during Viral Adaptation. The recovery of infectious viral titers of SARS-CoV-2 when passaged in MPAtreated A549-A/T cells, coupled with evident phenotypic changes in plaques, strongly suggests the emergence of variants potentially carrying mutations that alter viral fitness and/or reduce sensitivity to MPA. To investigate this further, viral RNA from each passage (p0-20) was sequenced. The SARS-CoV-2 genome coverage was comparable across different passages of the virus cultured in UTC cells, while the coverage increased in later passages for the virus continuously exposed to MPA (Fig. 7A). Despite the selective pressure from MPA, nucleotide usage across the SARS-CoV-2 genome remained consistent throughout the passaging process (SI Appendix, Fig. S7 A andB). This suggests that while MPA's antiviral activity is mediated by cellular GTP depletion, this metabolic stress did not prompt the virus to alter its nucleotide composition as an adaptive response. When examining polymorphic sites, the vast majority of sites contained variants at substantially less than 1% frequencies, and no enrichment of mutation frequencies at synonymous sites was observed, indicating no obvious mutagenic effects of MPA. We found that only a minority of SNVs occurred at frequencies >1% of the viral population, with only a small proportion of these SNVs becoming fixed in the population and leading to a consensus sequence change (SI Appendix, Fig. S7C). In total, seven mutations were detected with a frequency exceeding 50% in at least one passage (Fig. 7B). Among these mutations, two were also observed in the untreated control (UTC), indicating likely adaptations to cell culture conditions. Notably, three mutations (S P812R, ORF3 Q185H, and E S6L) became dominant and were correlated with "breakthrough" viral load increases from continued exposure to MPA, providing direct evidence of SARS-CoV-2 adaptation to MPA treatment (Fig. 7C). These mutations became dominant in both intracellular and extracellular viruses. To study the impact of the MPA-related mutations, S P812R, ORF3 Q185H, and E S6L were introduced into the wild-type (WT) SARS-CoV-2 genome (with D614G) through reverse genetics, generating three single mutants and a triple mutant harboring all three mutations. Interestingly, the triple variant demonstrated increased production of infectious virus, in contrast to individual mutants, which exhibited comparable titers to WT virus (Fig. 8A).
In line with this finding, time kinetics indicate enhanced replication of the triple variant in comparison to the WT virus and all three single mutants by immunofluorescence staining for the SARS-CoV-2 Nucleoprotein (NP) in relation to the total cell number (Fig. 8B). Furthermore, bright field microscopy and analysis of plaque size suggested enhanced cell death for the triple mutant (Fig. 8 C andD). However, no significant increase in IC 50 values was observed for any of the variants. The mutations in the E and ORF3 proteins instead increased susceptibility to MPA (Fig. 8E) which is presumably offset by the massive increase in replication and virus production. In hAECs, the triple variant showed no clear replication advantage compared to the WT strain; however, the single variants appeared to modestly reduce infectious progeny production, a deficit that seems to be compensated for in the triple variant (SI Appendix, Fig. S8). In conclusion, these findings suggest that MPA-selected mutations confer altered viral fitness rather than MPA resistance.
## Discussion
Immunosuppression, whether induced by disease or medical treat ment, places individuals at an increased risk for severe infections. This vulnerability stems from the compromised state of their immune systems, leaving them less equipped to mount an effective defense against invading pathogens. Numerous studies indicate that immunocompromised individuals, even after vaccination, tend to experience more severe disease upon viral infection ( 30 -33 ). In the context of the recent SARS-CoV-2 pandemic, it has been reported that approximately 40% of hospitalized individuals with SARS-CoV-2 breakthrough infections after vaccination are immu nocompromised ( 31 ). Furthermore, the use of antiproliferative agents has previously been associated with a poor humoral immune response ( 68 ). Next to potentially more severe disease outcomes, immunosuppression has been intensively discussed concerning viral adaptation and the emergence of novel variants ( 35 -37 ).
Especially in the context of viral infections, and specifically due to the ongoing circulation of SARS-CoV-2, it is thus crucial to enhance our understanding on how immunosuppression impacts viral infection. In line with previous observations, we could demonstrate that the immunosuppressive drug MPA broadly reduces virus infection in vitro. Mechanistically, we provide com pelling evidence demonstrating that MPA influences virus repli cation by depleting intracellular GTP pools. This is substantiated by the dose-dependent restoration of viral titers through the ectopic substitution of guanosine and/or GMP. Conversely, viral titers could not be recovered for the alternative monophosphates, suggesting a direct link between GMP availability and MPA treat ment efficacy. Coupled with the observation that individuals undergoing immunosuppression frequently experience prolonged viral replication ( 35 , 40 -43 ), MPA-induced immunosuppression may thus serve as a potential source for the emergence of new variants. Chronic viral infections as caused by HCV ( 69 -71 ) or HEV ( 39 ), but also SARS-CoV-2 ( 34 ), have already demonstrated an increased risk of quasispecies development, characterized by heterogeneous viral swarms or mutant clouds. These viral popu lations, composed of closely related but nonidentical genomes typical for RNA viruses ( 72 , 73 ), play a crucial role in the adap tation of viruses to environmental changes and selective pressures imposed by the host's immune system ( 74 ). We could demonstrate that increased exposure to MPA treatment led to rapid adaptation of SARS-CoV-2, which, next to viral titer recovery, was accom panied by phenotypic changes resulting in increased plaque sizes. We found that the majority of SNVs occurred at low frequencies and only a very small proportion of these SNVs became fixed in the population, suggesting that most mutations were transient and did not contribute to long-term genomic adaptation. We observed a rapid jump in virus production and intracellular viral RNA levels under MPA-treatment, which coincided with the par allel emergence of three nonsynonymous mutations. We con firmed these mutations confer enhanced virus replication fitness and not MPA resistance. On the virus side, no obvious mutagenic effects or base exchange biases were observed in viral genomes mediated by MPA treatment, outside of the detected mutations. On the host side, MPA-mediated transcriptional dysregulation affected a range of cellular processes. Thus, the exact mechanism underlying MPA-driven mutational emergence remains elusive. However, in line with the evolutionary dynamics of rapidly evolv ing RNA virus populations, we propose reduced viral population sizes observed under MPA treatment can facilitate the more rapid fixation of mutations in the population. In these restricted pop ulations, adaptive mutations which confer a replication advantage will rapidly become dominant in the absence of immune pressure. Along these lines, SARS-CoV-2 Alpha VOC was hypothesized to have emerged in an immunosuppressed suppressed patient, where chronic infections due to impaired humoral and cellular responses results in greater chances of accumulating fitness-enhancing muta tions, and subsequent onward transmission ( 75 ). Specifically, deep sequencing revealed the presence of three commonly arising mutations during adaptation, present in the Spike protein (S P812R), the accessory protein ORF3 (ORF3 Q185H) and the Envelope protein (E S6L), respectively. Interestingly, only the combination of all three mutations increased viral fitness in comparison to the WT virus, whereas the single variants did not affect viral fitness suggesting that they may have complementary roles. This hypothesis is supported by the obser vation that these mutations, when arising independently as in the UTC, do not affect viral fitness. The linked emergence of these three mutations in the sequencing analysis further suggests a com pensatory relationship, indicating they likely function collectively rather than independently. Importantly, susceptibility to neutral izing antibodies as well as approved antiviral treatment regimens remained comparable to the WT. This holds particular significance in light of the growing number of viral variants linked to immune escape and/or diverse pathogenesis ( 75 ).
In general, propagation in cell culture correlates with a signifi cant increase in substitutions in the genomic region coding for the Spike protein, which could mediate enhanced infectivity ( 76 , 77 ). Here, we identified S P812R to be of significant interest. Although reported as a cell culture adaptation conferring a selec tive phenotypic advantage ( 78 ), S P812R had also been reported to occur in vivo ( 79 ). An S P812S mutation has been observed in association with upper airway specific evolution in an immuno compromised individual with chronic SARS-CoV-2 infection suggesting a mutational hotspot that also contributes to evasion from neutralizing antibodies through structural rearrangements ( 80 ). In fact, S P812 is located close to the S2' cleavage site. Upon Nucleoside triphosphate biosynthetic process amino acid substitution from proline to arginine the sequence changes from P SKR to R SKR, which corresponds to creation of an additional furin cleavage motif ( 81 ). Introducing a second cleavage site could potentially increase membrane fusion and sub sequently infectivity thereby compensating for suboptimal con ditions ( 82 , 83 ). Interestingly, the proline at position 812 is also present in SARS-CoV whereas an arginine is found in MERS-CoV ( 77 ). Next to the Spike protein, ORF3 has been identified as a driver of pathogenicity affecting virulence, infectivity, ion channel activity, morphogenesis, and virus release ( 84 , 85 ), as well as a potent interferon antagonist ( 86 ). Here, we identified a substitu tion within ORF3 at position 185 (ORF3 Q185H). Notably, this mutation located in the transmembrane domain of ORF3a has been frequently detected in waste water samples in New York City ( 87 ) and is one of the characteristic mutations found in lineage B.1.258.17, which previously circulated in Slovenia, Switzerland, Germany, Sweden, and Austria ( 88 -90 ). It has also been linked with disease progression ( 84 ). Controversially, ORF3 Q185H was additionally found to substantially decrease the stability of the protein ( 91 ), adding ambiguity toward its proposed function. The amino acid substitution in the envelope (E S6L) has been reported to occur in vivo ( 92 ) while others have reported emergence of this variant during in vitro passaging experiments ( 36 , 93 ). The enve lope protein is the smallest of the major structural proteins, and associated with viral assembly, budding, envelope formation, and pathogenesis. Furthermore, the assembly to pentameric channels directly affects virus replication ( 94 , 95 ). Currently, there is no evidence indicating specific functional or phenotypic conse quences associated with this mutation. However, since all reported mutations have also been observed in vivo, albeit separately and never within the same genome, we still believe that each likely holds biological relevance.
$$B I R C 3 C H U K C U L 1 D I C E R 1 M A P 3 K 1 4 N F K B 2 P S M A 1 P S M$$
$$9 P S M E 1 R E L R E L B S K P 1 early late NIK/NFκB signaling A B R A X A S 1 A C T B A D R B 2 A R A R R B 2 A T X N 3 B A B A M 1 B A B A M 2 B R C A 1 C D K 1 D D B 2 E N Y 2 I K B K G I N O 8 0 C J O S D$$
$$P S M E 1 R U V B L 1 S M A D 3 S N X 3 T A F 9 B T G F B R 1 T O M M 2 0 U B A 5 2 U B E 2 D 1 U C H L 1 U C H L 3$$
## Protein deubiquitination early late
Comparable findings regarding the appearance of viral variants with altered viral fitness or characteristic mutations similar to those displayed by VOCs were documented during the passaging of SARS-CoV-2 in the presence of low doses of Remdesivir ( 50 ) and similarly with poliovirus exposed to RBV ( 46 , 49 ). Our findings also suggest that mutations observed in circulating variants can emerge through in vitro passaging. Consequently, this passaging approach holds significant potential as a valuable tool for assessing the development of viral variants under certain conditions. The emergence of new viral variants presents a risk to the general pop ulation by affecting various aspects of virus-host biology such as transmission dynamics, public health measures, vaccine efficacy, antiviral treatment, disease severity, and pathogenesis, among others. Therefore, it is crucial to thoroughly investigate potential drivers of mutagenesis and understand the impact of specific mutations. While immunosuppression has always been a major concern in pathogen-induced diseases, the identification of novel variants raise questions about potential consequences for the long-term manage ment of viral infections in immunocompromised individuals. This is particularly noteworthy not only because SARS-CoV-2 continues to circulate in humans and wild animal reservoirs, posing a constant threat of new variant emergence, but also considering the growing numbers of immunocompromised individuals. Understanding these dynamics becomes crucial in developing more effective and sustain able strategies for treating infections in vulnerable populations. The insights gleaned from our study have the potential to be extrapolated to other virus infections and can contribute to evaluating the pan demic potential of emerging variants in distinct populations. Fully differentiated hAECs were washed with Hank's balanced salt solution (HBSS) apically before infection.
Ethics. The experiments involving the adaptation of SARS-CoV-2 to MPA were approved by the local commission for ethics in security-relevant research (KEF).
Data, Materials, and Software Availability. RNA-seq data is deposited in the NCBI GEO database (GSE295532) (67). All other data are included in the manuscript and/or SI Appendix.
$$A P C A S X L 1 A T X N 7 A T X N 7 L 3 B I R C 3 C D C 2 5 A C Y L D E P 3 0 0 E P O P G A T A 3 H C F C 1 I N O 8 0 I N O 8 0 D M D M 4 M Y C M Y S M 1 N A F 1 N F K B I A O T U D 3 O T U D 4 O T U D 7 B R H O T 1 S M A D 2 S M U R F 2 S P A T A 2 S T A M T N F A I P 3 T R A F 2 T R A F 3 T R A F 6 T R R A P U I M C 1 U S P$$
$$P I P 1 Y O D 1 Z C 3 H 1 2 A Z R A N B 1 Protein deubiquitination early late >2 < -2 -2 -1 0 1 2 log10 RPKM$$
## References
1. Kirk (2015) "World Health Organization estimates of the global and regional disease burden of 22 foodborne bacterial, protozoal, and viral diseases, 2010: A data synthesis" *PLoS Med*
2. Toor (2021) "Lives saved with vaccination for 10 pathogens across 112 countries in a pre-COVID-19 world" *Elife*
3. De Clercq, Li (2016) "Approved antiviral drugs over the past 50 years" *Clin. Microbiol. Rev*
4. Clark (2020) "Global, regional, and national estimates of the population at increased risk of severe COVID-19 due to underlying health conditions in 2020: A modelling study" *Lancet Glob. Health*
5. Burrell, Howard, Murphy (2017) "Epidemiology of viral infections" in Fenner and White's Medical Virology"
6. Jeleff (2022) "Vulnerability and one health assessment approaches for infectious threats from a social science perspective: A systematic scoping review" *Lancet Planet. Health*
7. Harpaz, Dahl, Dooling (2013) "Prevalence of immunosuppression among US adults" *JAMA*
8. (2022) "Altered immunocompetence: General best practice guidelines for immunization: Best practices guidance of the advisory committee on immunization practices (ACIP)"
9. Dun (2013) "Transcriptomic changes induced by mycophenolic acid in gastric cancer cells" *Am. J. Transl. Res*
10. Kato (2019) "Identification of inhibitors of dengue viral replication using replicon cells expressing secretory luciferase" *Antivir. Res*
11. Diamond, Zachariah, Harris (2002) "Mycophenolic acid inhibits dengue virus infection by preventing replication of viral RNA" *Virology*
12. Pan (2012) "Mycophenolic acid augments interferon-stimulated gene expression and inhibits hepatitis C virus infection in vitro and in vivo" *Hepatology*
13. Henry (2006) "Mycophenolic acid inhibits hepatitis C virus replication and acts in synergy with cyclosporin A and interferon-alpha" *Gastroenterology*
14. Wan (2021) "High-throughput screening of an FDA-approved drug library identifies inhibitors against arenaviruses and SARS-CoV-2" *ACS Infect. Dis*
15. Marroquí, Estepa, Perez (2008) "Inhibitory effect of mycophenolic acid on the replication of infectious pancreatic necrosis virus and viral hemorrhagic septicemia virus" *Antivir. Res*
16. Wang (2014) "Calcineurin inhibitors stimulate and mycophenolic acid inhibits replication of hepatitis E virus" *Gastroenterology*
17. Sebastian (2011) "Mycophenolic acid inhibits replication of Japanese encephalitis virus" *Chemotherapy*
18. Robertson, Hermann, Coombs (2004) "Mycophenolic acid inhibits avian reovirus replication" *Antivir. Res*
19. Debing (2014) "Ribavirin inhibits in vitro hepatitis E virus replication through depletion of cellular GTP pools and is moderately synergistic with alpha interferon" *Antimicrob. Agents Chemother*
20. Kosugi (1994) "Antiviral activities of mizoribine and other inosine monophosphate dehydrogenase inhibitors against several ortho-and paramyxoviruses" *Antivir. Chem. Chemother*
21. To (2016) "Mycophenolic acid, an immunomodulator, has potent and broad-spectrum in vitro antiviral activity against pandemic, seasonal and avian influenza viruses affecting humans" *J. Gen. Virol*
22. Hishiki (2023) "Identification of IMP dehydrogenase as a potential target for anti-mpox virus agents" *Microbiol. Spectr*
23. Shen (2019) "High-throughput screening and identification of potent broad-spectrum inhibitors of coronaviruses" *J. Virol*
24. Hart (2014) "Interferonβ and mycophenolic acid are potent inhibitors of Middle East respiratory syndrome coronavirus in cell-based assays" *J. Gen. Virol*
25. Rother (2018) "Combined human genome-wide RNAi and metabolite analyses identify IMPDH as a host-directed target against Chlamydia infection" *Cell Host Microbe*
26. (2020) "Coronaviruses and immunosuppressed patients: The facts during the third epidemic" *Liver Transpl*
27. Minotti (2020) "How is immunosuppressive status affecting children and adults in SARS-CoV-2 infection? A systematic review" *J. Infect*
28. Schoot (2020) "Immunosuppressive drugs and COVID-19: A review" *Front. Pharmacol*
29. Devresse (2022) "Immunosuppression and SARS-CoV-2 infection in kidney transplant recipients"
30. (2022) "People who are immunocompromised: Know how to protect yourself and what to do if you get sick"
31. Brosh-Nissimov (2021) "BNT162b2 vaccine breakthrough: Clinical characteristics of 152 fully vaccinated hospitalized COVID-19 patients in Israel" *Clin. Microbiol. Infect*
32. Belsky (2021) "COVID-19 in immunocompromised patients: A systematic review of cancer, hematopoietic cell and solid organ transplant patients" *J. Infect*
33. Abbasi (2021) "Researchers tie severe immunosuppression to chronic COVID-19 and virus variants" *JAMA*
34. Hettle (2022) "Persistent SARS-CoV-2 infection in immunocompromised patients facilitates rapid viral evolution: Retrospective cohort study and literature review" *Clin. Infect. Pract*
35. Weigang (2021) "Within-host evolution of SARS-CoV-2 in an immunosuppressed COVID-19 patient as a source of immune escape variants" *Nat. Commun*
36. Sonnleitner (2022) "The mutational dynamics of the SARS-CoV-2 virus in serial passages in vitro" *Virol. Sin*
37. Cele (2022) "SARS-CoV-2 prolonged infection during advanced HIV disease evolves extensive immune escape" *Cell Host Microbe*
38. Ng (2019) "Persistent dengue infection in an immunosuppressed patient reveals the roles of humoral and cellular immune responses in virus clearance" *Cell Host Microbe*
39. Lhomme (2014) "Influence of polyproline region and macro domain genetic heterogeneity on HEV persistence in immunocompromised patients" *J. Infect. Dis*
40. Avanzato (1901) "Case study: Prolonged infectious SARS-CoV-2 shedding from an asymptomatic immunocompromised individual with cancer" *Cell*
41. Kemp (2021) "SARS-CoV-2 evolution during treatment of chronic infection" *Nature*
42. Choi (2020) "Persistence and evolution of SARS-CoV-2 in an immunocompromised host" *N. Engl. J. Med*
43. Corey (2021) "SARS-CoV-2 variants in patients with immunosuppression" *N. Engl. J. Med*
44. Contreras (2002) "Viral RNA mutations are region specific and increased by ribavirin in a full-length hepatitis C virus replication system" *J. Virol*
45. Severson (2003) "Ribavirin causes error catastrophe during Hantaan virus replication" *J. Virol*
46. Vignuzzi, Stone, Andino (2005) "Ribavirin and lethal mutagenesis of poliovirus: Molecular mechanisms, resistance and biological implications" *Virus Res*
47. Heyer (2022) "Remdesivir-induced emergence of SARS-CoV2 variants in patients with prolonged infection" *Cell Rep. Med*
48. Todt (2016) "In vivo evidence for ribavirin-induced mutagenesis of the hepatitis E virus genome" *Gut*
49. Pfeiffer, Kirkegaard (2003) "A single mutation in poliovirus RNA-dependent RNA polymerase confers resistance to mutagenic nucleotide analogs via increased fidelity" *Proc. Natl. Acad. Sci. U.S.A*
50. Szemiel (2021) "In vitro selection of Remdesivir resistance suggests evolutionary predictability of SARS-CoV-2" *PLoS Pathog*
51. Clavel, Hance (2004) "HIV drug resistance"
52. Carretero-Ledesma (2024) "Antiviral activity of immunosuppressors alone and in combinations against human adenovirus and cytomegalovirus" *Int. J. Antimicrob. Agents*
53. (2009) "Kidney Disease: Improving Global Outcomes (KDIGO) Transplant Work Group, Kdigo clinical practice guideline for the care of kidney transplant recipients" *Am. J. Transplant*
54. Rose, Yurchak, Jusko (1981) "Dose dependent pharmacokinetics of prednisone and prednisolone in man" *J. Pharmacokinet. Biopharm*
55. Kahan (2000) "Therapeutic drug monitoring of sirolimus: Correlations with efficacy and toxicity" *Clin. Transplant*
56. Akhlaghi (2006) "Pharmacokinetics of mycophenolic acid and metabolites in diabetic kidney transplant recipients" *Ther. Drug Monit*
57. Patel (2007) "Concentrations of mycophenolic acid and glucuronide metabolites under concomitant therapy with cyclosporine or tacrolimus" *Ther. Drug Monit*
58. Tönshoff (2020) "Immunosuppressants in organ transplantation" *Handb. Exp. Pharmacol*
59. Claeys, Vermeire (2019) "Immunosuppressive drugs in organ transplantation to prevent allograft rejection: Mode of action and side effects" *J. Immunol. Sci*
60. Cruse, Lewis, Wang (2004) "Chapter 18-Transplantation immunology"
61. Xu, Chu (2022) "Advances in immunosuppressive agents based on signal pathway" *Front. Pharmacol*
62. Meneghini, Bestard, Grinyo (2021) "Immunosuppressive drugs modes of action" *Best Pract. Res. Clin. Gastroenterol*
63. Heischmann (2017) "The immunosuppressant mycophenolic acid alters nucleotide and lipid metabolism in an intestinal cell model" *Sci. Rep*
64. Zettl (2020) "Rapid quantification of SARS-CoV-2-neutralizing antibodies using propagation-defective vesicular stomatitis virus pseudotypes" *Vaccines*
65. Chung (2007) "Ribavirin reveals a lethal threshold of allowable mutation frequency for Hantaan virus" *J. Virol*
66. Ali (2020) "Treatment options for COVID-19: A review" *Front. Med. (Lausanne)*
67. Meister (2025) "Mycophenolic acid treatment drives the emergence of novel SARS-CoV-2 variants" *Gene Expression Omnibus*
68. Linares (2021) "A propensity score-matched analysis of mortality in solid organ transplant patients with COVID-19 compared to non-solid organ transplant patients" *PLoS One*
69. Gömer (2022) "Intra-host analysis of hepaciviral glycoprotein evolution reveals signatures associated with viral persistence and clearance" *Virus Evol*
70. Weiner (1992) "Evidence for immune selection of hepatitis C virus (HCV) putative envelope glycoprotein variants: Potential role in chronic HCV infections" *Proc. Natl. Acad. Sci. U.S.A*
71. Kurosaki (1993) "Rapid sequence variation of the hypervariable region of hepatitis C virus during the course of chronic infection" *Hepatology*
72. Domingo, Schuster (2016) "What is a quasispecies? Historical origins and current scope" *Curr. Top. Microbiol. Immunol*
73. Domingo, Sheldon, Perales (2012) *Viral quasispecies evolution. Microbiol. Mol. Biol. Rev*
74. Lauring, Andino (2010) "Quasispecies theory and the behavior of RNA viruses" *PLoS Pathog*
75. Carabelli (2023) "SARS-CoV-2 variant biology: Immune escape, transmission and fitness" *Nat. Rev. Microbiol*
76. Foster (2023) "Long-term serial passaging of SARS-CoV-2 reveals signatures of convergent evolution" *J. Virol*
77. Ramirez (2021) "Overcoming culture restriction for SARS-CoV-2 in human cells facilitates the screening of compounds inhibiting viral replication" *Antimicrob. Agents Chemother*
78. Bamford (2022) "Comparison of SARS-CoV-2 evolution in paediatric primary airway epithelial cell cultures compared with vero-derived cell lines" *Viruses*
79. Van Dorp (2020) "No evidence for increased transmissibility from recurrent mutations in SARS-CoV-2" *Nat. Commun*
80. Cheng (2024) "Signatures of omicron-like adaptation in early SARS-CoV-2 variants and chronic infection" *bioRxiv*
81. Hoffmann (2020) "SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor" *Cell*
82. Belouzard, Chu, Whittaker (2009) "Activation of the SARS coronavirus spike protein via sequential proteolytic cleavage at two distinct sites" *Proc. Natl. Acad. Sci. U.S.A*
83. Wolf, Kwan, Kamil (2022) "Structural dynamics and molecular evolution of the SARS-CoV-2 spike protein" *mBio*
84. Hassan, Attrish, Ghosh et al. (2021) "Pathogenic perspective of missense mutations of ORF3a protein of SARS-CoV-2" *Virus Res*
85. Ren (2020) "The ORF3a protein of SARS-CoV-2 induces apoptosis in cells" *Cell. Mol. Immunol*
86. Konno (2020) "SARS-CoV-2 ORF3b is a potent interferon antagonist whose activity is increased by a naturally occurring elongation variant" *Cell Rep*
87. Gregory (2022) "Genetic diversity and evolutionary convergence of cryptic SARS-CoV-2 lineages detected via wastewater sequencing" *PLoS Pathog*
88. Klikovits "eyJ pYX QiO jE2 MDk 0NT kyM DAs ImV 4cC I6N DA3 MDk wOD gwM Cwi dXN lci I6M Cwi Z3J vdX BzI jpb MCw tMV 0sI mZp bGU iOi Jma Wxl YWR taW 4vQ UdF U18 yMD IyL zZf Rk"
89. Franco (2022) "Characterization of genetic variants of SARS-CoV-2 isolated from cadavers in the region of Lisbon: Comparative and validation study between Real-Time PCR and Next Generation Sequencing methodologies"
90. Brejová (2021) "A SARS-CoV-2 mutant from B.1.258 lineage with ∆H69/∆V70 deletion in the Spike protein circulating in Central Europe in the fall 2020" *Virus Genes*
91. Hossain (2021) "Genome-wide identification and prediction of SARS-CoV-2 mutations show an abundance of variants: Integrated study of bioinformatics and deep neural learning" *Inform. Med. Unlocked*
92. Kim (2021) "Genome sequences of two GH clade SARS-CoV-2 strains isolated from patients with COVID-19 in South Korea" *Microbiol. Resour. Announc*
93. Zhou (2022) "Mutations that adapt SARS-CoV-2 to mink or ferret do not increase fitness in the human airway" *Cell Rep*
94. Rahman (2021) "Mutational insights into the envelope protein of SARS-CoV-2. Gene Rep"
95. Hoque (2020) "Genomic diversity and evolution, diagnosis, prevention, and therapeutics of the pandemic COVID-19 disease" *PeerJ*
96. Matsuyama (2020) "Enhanced isolation of SARS-CoV-2 by TMPRSS2-expressing cells" *Proc. Natl. Acad. Sci. U.S.A*
97. Schuhenn (2022) "Differential interferonα subtype induced immune signatures are associated with suppression of SARS-CoV-2 infection" *Proc. Natl. Acad. Sci. U. S. A*
98. Barbet (2024) "Bronchoscopic biopsies-A novel source for primary airway epithelial cells in respiratory research" *Respir. Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12710325&blobtype=pdf | # Giant viruses specific to deep oceans show persistent presence and activity
Wenwen Liu, Komei Nagasaka, Junyi Wu, Hiroki Ban, Ethan Mimick, Lingjie Meng, Russell Neches, Mohammad Moniruzzaman, Takashi Yoshida, Yosuke Nishimura, Hisashi Endo, Yusuke Okazaki, Hiroyuki Ogata
## Abstract
Giant viruses (GVs) of the phyla Nucleocytoviricota and Mirusviricota are large double-stranded DNA viruses that infect diverse eukaryotic hosts and impact biogeochemical cycles. Their diversity and ecological roles have been well studied in the photic layer of the ocean, but less is known about their activity, population dynam ics, and adaptive strategies in the aphotic layers. Here, we conducted eight seasonal time-series samplings of the surface and mesopelagic layers at a coastal site in Muroto, Japan, and integrated 18S metabarcoding, metagenomic, and metatranscriptomic data to investigate mesopelagic GVs and their potential hosts. The analysis identified 48 GV genomes including six that were exclusively detected in the mesopelagic layer. Notably, these mesopelagic-specific GVs showed persistent activity across seasons. To further investigate the distribution and phylogenomic features of GVs at a global scale across broader depths, we compiled 4,473 species-level GV genomes from the OceanDNA MAG project and other resources and analyzed 1,890 marine metagenomes. This revealed 101 deep-sea-specific GVs, distributed across the GV phylogenetic tree, indicating that adaptation to deep-sea environments has occurred in multiple lineages. One clade enriched with deep-sea-specific GVs included a GV genome identified in our Muroto data, which displayed a wide geographic distribution. Seventy-six KEGG orthologs and 74 Pfam domains were specifically enriched in deep-sea-specific GVs, encompassing functions related to the ubiquitin system, energy metabolism, and nitrogen acquisition. These findings support the scenario that distinct GV lineages have adapted to hosts in aphotic marine environments by altering their gene repertoire to thrive in this unique habitat.IMPORTANCE Giant viruses are widespread in the ocean surface and are key in shaping marine ecosystems by infecting phytoplankton and other protists. However, little is known about their activity and adaptive strategies in deep-sea environments. In this study, we performed metagenomic and metatranscriptomic analyses of seawater samples collected from a coastal site in Japan and discovered giant virus genomes showing persistent transcriptional activity across seasons in the mesopelagic water. Using a global marine data set, we further uncovered geographically widespread and vertically extensive groups of deep-sea-specific giant viruses and characterized their distinctive gene repertoire, which likely facilitates adaptation to the limited availability of light and organic compounds in the aphotic zone. These findings expand our under standing of giant virus ecology in the dark ocean. KEYWORDS giant viruses, deep ocean, the Kochi Prefectural Deep Seawater Labora tory, metatranscriptome, adaptation G iant viruses (GVs), encompassing the phylum Nucleocytoviricota and the recently proposed phylum Mirusviricota, are characterized by their large virion and genome
## RESULTS AND DISCUSSION
## Data overview
The pump systems at the study site (i.e., off the coast of Muroto; Fig. S1) enabled us to collect and filter a large volume of water from two depths (0.5 m and 320 m) in a short time (Table S1). Size fractionations were performed to concentrate different microorganisms: pico-fraction (0.2-3.0 µm or 0.2-5.0 µm), total-fraction (0.2-150 µm), and nano/micro-fraction (3.0-150 µm or 5.0-150 µm) (Table S2). From these samples, we successfully extracted adequate quantities (>500 ng) of high-quality DNA and RNA for sequencing (Table S3). We generated 18S rDNA and rRNA metabarcodes (4,981 ASVs; nano/micro-fraction and either pico-or total-fraction) (Table S4), metagenomic data (757 Gbp; pico-or total-fraction), and metatrancriptomic data (598 Gbp; nano/micro-fraction and either pico-or total-fraction).
## Mesopelagic microeukaryote community is distinct from that in the surface water at the sampling site
In this study, 18S rDNA metabarcodes were used to measure the relative abundance of microeukaryotes at the ASV level although the abundance of 18S rDNA is biased by the copy number of rRNA genes, which varies among organisms (41,42). 18S rRNA metabarcodes were used as a proxy for the metabolic activity of organisms (43). The microeukaryote community in the mesopelagic layer substantially differed from that of the surface layer at the study site (Fig. 1a through d), suggesting that the two layers harbor distinct host candidates for GVs. Furthermore, the mesopelagic microeukaryote community characterized by 18S rRNA differed from that of 18S rDNA metabarcodes, suggesting that some microeukaryotes are active in the mesopelagic layer, while others are inactive (i.e., in a dormant state or dead cells settled from the surface). For instance, Polycystinea had the highest abundance but was ranked 11th in activity (Fig. S2). Spirotrichea did not show high abundance but ranked third in activity (Fig. S2). These results suggest that the mesopelagic communities are a mixture of active and inactive organisms.
Non-metric multidimensional scaling (NMDS) ordination confirmed a clear separation between the surface and mesopelagic microeukaryote communities, regardless of whether 18S rDNA (Fig. 1e; R 2 = 0.25, PERMANOVA, P-value < 0.01) or 18S rRNA (Fig. 1f; R 2 = 0.24, PERMANOVA, P-value < 0.01) metabarcodes were used. Furthermore, the mesopelagic microeukaryote communities showed less enhanced seasonal changes than the surface communities. To compare the persistence of ASVs between the layers, we calculated Levins' standardized niche breadth (BA) using only the surface and mesopelagic samples collected on the same day. In both the 18S rDNA and 18S rRNA data sets, the BA values were significantly higher for the mesopelagic than the surface ASVs in both size fractions (Fig. S3, P-value < 0.001), indicating higher persistence for the mesopelagic ASVs. These results suggest that the mesopelagic environment may provide a unique and stable host community for GVs.
## GVs exclusively detected in mesopelagic metagenomes show persistency at the study site
Forty-eight GVMAGs recovered from the Muroto metagenomic data met our quality criteria (Fig. S4; see Materials and Methods), including 43 Nucleocytoviricota MAGs (NCV-MAGs) and 5 Mirusviricota MAGs (MV-MAGs) (Fig. 2a). These GVMAGs exhibited genome sizes ranging from 54 kbp to 824 kbp and GC contents ranging from 24% to 53% (Table S5). To assess the recovery level of GV lineages by GVMAGs, we investigated family-B DNA polymerase (PolB) sequences in the metagenomic data set. Through phylogenetic analysis (Fig. S5), we identified 163 PolB sequences predicted to be of GV origin. Of these PolB sequences, 34 (21%) were found in the Muroto GVMAGs, suggesting that many GVs in the study site were not represented in our GVMAG data set. Phylogenetic analysis of 7 conserved NCV marker genes revealed that the 43 NCV-MAGs belonged to 5 orders of Nucleocytoviricota: 19 imiterviruses, 15 algaviruses, 5 pimascoviruses, 2 pandoraviruses, and 2 asfuviruses (Fig. S6a). Two NCV-MAGs belonging to the order Pandoravirales were closely related to Emiliania huxleyi viruses (average nucleotide identity, ~85%). The five MV-MAGs were classified based on HK97 major capsid protein (MCP) phylogeny, including three in the MR2 marine clade and two in a clade composed of mirusvirus MAGs from a freshwater lake (Fig. S6b) (33). Mapping of the metagenomic reads on the 48 GVMAGs revealed that these MAGs represent 0.028%-1.28% (0.48% on average) of the total reads for the surface data compared with 0.01%-0.028% (0.019% on average) for the mesopelagic data (Fig. S7a). The lower relative abundance of GVs in the mesopelagic samples explains the lower level of GVMAG recovery from these samples. Among the 48 GVMAGs, 17 GVMAGs (15 NCV-MAGs, 2 MV-MAGs) were detected in the mesopelagic metagenomes, including six (NCV01, 13, 26, 27; MV2, 4) that were exclusively found in the mesopelagic samples (Fig. 2b; Table S6). These "meso-exclusive" GVMAGs tended to show persistent presence in the mesopelagic layer across different seasons (Fig. 2b). The other 11 GVMAGs detected in the mesopelagic metagenomic data were also detected in the surface metagenomes (Fig. 2b). In sharp contrast with the six meso-exclusive GVMAGs, many of these GVMAGs showed strong seasonal dynamics (e.g., NCV30, 42) akin to the pattern observed for GVs in a shallow coastal area (23). Endo et al. previously showed that the community structure of GVs in the mesopelagic layer (200-1,000 m) is different from that in the photic layer based on the Tara Oceans expedition data (19). Our results confirmed not only the distinct community structure but also the difference in the community dynam ics between the surface and mesopelagic layers.
## Mesopelagic-specific GVs are active across seasons
Mapping of the metatranscriptomic reads on the GVMAGs revealed a similar trend to the metagenomic data (Fig. 2c; Fig. S7), with the meso-exclusive GVMAGs showing seasonally stable activity and the surface layer community showing strong seasonal dynamics. Seventeen GVMAGs showed transcriptional activity in the mesopelagic layer (Table S6). These included all six meso-exclusive GVMAGs (Fig. 2c). For five of these GVMAGs (NCV01, 26, 29; MV2, 4), transcripts were detected for over 10% of genes encoded in the genomes (Fig. S8), with NCV01, MV2, and MV4 expressing up to 37.3%-63.6% of their genes (Fig. S8). This result suggests ongoing viral replication in the mesopelagic environment. Furthermore, the transcriptional activities of many of these meso-exclusive GVMAGs were relatively constant over time (e.g., NCV01, NCV26, MV2, MV4) (Fig. 2c; Fig. S9). To the best of our knowledge, this is the first report of the transcriptional activity of GVs exclusively detected in an aphotic environment. The stable community structure and activity of meso-exclusive GVs across seasons parallel the stability of the microeukaryotic community in the mesopelagic zone at the study site (Fig. 1). In contrast to the meso-exclusive GVMAGs, several GVMAGs that were detected in both surface and mesopelagic metagenomes showed transcriptional activity only in the surface samples (e.g., NCV11, NCV30). These GVs in the mesopelagic layer may have originated from the surface through sinking processes, as previously suggested (19,30). Overall, our findings support the existence of active and stable GV populations specific to the mesopelagic zone. Virus-host interactions may be less influenced by seasonal fluctuations in the relatively stable ecological conditions in this layer.
The difference in mapping depth for the metatranscriptome data between the surface and mesopelagic layers was less pronounced than the case for the metagenome data, with some sampling days showing comparable (or even higher) relative transcrip tional activity of GVMAGs in the mesopelagic than the surface data. This suggests that even though fewer GVMAGs were reconstructed from the mesopelagic data, they may be highly active. For the total-fraction, the proportions of mapped transcriptomic reads were 0.0021%-0.54% (0.13% on average) for the surface samples and 0.0024%-0.0079% (0.0043% on average) for the mesopelagic samples. For the nano/micro-fraction, the proportions of mapped transcriptomic reads were 0.00028%-0.21% (0.052% on average) for the surface samples and 0.00024%-0.0017% (0.00074% on average) for the mesope lagic samples. The higher proportions of mapped mesopelagic transcriptomic reads in the total-fraction than in the nano/micro-fraction suggest that active infection of GVs in the mesopelagic layer is primarily associated with smaller host cells included in this fraction.
## Mesopelagic GVMAGs encode specific gene functions
The six meso-exclusive GVMAGs recovered in this study contained between 51 and 913 predicted genes per genome (Table S5), with 880 genes (35.5%) being assigned to 594 KO terms (Table S7). The six meso-exclusive GVMAGs showed distinct gene composi tions enriched in functions related to virus-host interactions and signaling, including components of the ubiquitin system, cytoskeletal proteins, and signal transduction regulators.
Among genes in the ubiquitin system, E3 ubiquitin-protein ligase RNF181 was found in 23 GVMAGs. However, other ubiquitin system genes, such as Ariadne-1 (NCV01, NCV26, NCV27), E3 ubiquitin-protein ligase RNF13 (NCV01, NCV26), and NEDD4-binding protein 2 (NCV01, NCV26), were specific to meso-exclusive GVMAGs. Actin, myosin, and flagellar basal-body rod protein FlgG were exclusively present in NCV01 and NCV26 and could benefit viruses by manipulating the localization of viral replication machi nery during infection (14,44,45). NCV01 also encoded ninein-like protein, which is a member of the γ-tubulin complexes binding proteins (46) and may function to modulate cytoskeleton. Additionally, phosphoinositide 3-kinase (PI3K), a regulator of membrane dynamics and endocytic trafficking, was detected in 4 of 6 meso-exclusive GVMAGs. PI3K has been implicated in supporting the formation of replication organelles of enterovi ruses and may similarly support GV replication (47).
Notably, among the top 20 transcriptionally active KO terms in the mesopelagic samples, 7 were related to the ubiquitin system (e.g., Ariadne-1, NEDD4-binding protein 2, ubiquitin B, and E3 ubiquitin-protein ligases RNF181, EDD1, SIAH1, RNFT1) (Fig. S10b), compared with only 1 in the surface samples (Fig. S10a). The sequential action of ubiquitin systems conjugates ubiquitin to proteins and target them to proteasomes for degradation (48). Eukaryotic viruses can hijack the ubiquitin-proteasome system to aid in various stages of viral propagation (49). Since normal protein folding is hindered by low temperature or high pressure (50,51), the ubiquitin-related genes of GVs may function to modulate the host ubiquitin-proteosome system to quality control viral and host proteins in harsh condition in the mesopelagic layer, as previously hypothesized for GVs in cold environments (52). As the ubiquitin-proteasome system is also known to regulate apoptosis (53), the GV ubiquitin-related genes may function to regulate programmed cell death of the infected host cells. In the mesopelagic layer, which is characterized by relatively low host biomass (or encounter rates), the ability to suppress apoptosis might represent an adaptive strategy to prolong the life of the infected host cells, thereby allowing for greater viral production before host lysis and ensuring the continued propagation of the GV in resource-scarce environments.
## Putative hosts of meso-exclusive GVMAGs
To gain insights into the potential hosts of the 48 GVMAGs, we first performed homol ogy searches against the MarFERReT database (Fig. S11). The best eukaryotic matches differed between meso-exclusive and other GVMAGs, with the former displaying fewer matches to green algae (i.e., Mamiellophyceae and Chlorophyceae). Meso-exclusive GVMAGs were enriched in matches to Prymnesiophyceae, Dinophyceae, and Mediophy ceae. Of other GVMAGs, MV5 and NCV18 showed many matches to a single taxon (Fig. S11). MV5 showed four genes with best matches to Cryptophyceae, which was previously predicted as a host group of mirusvirus (6). NCV18, which is closely related to Emiliania huxleyi viruses, showed many matches to Emiliania huxleyi (class Prymnesio phyceae) probably due to the inclusion of viral sequences or homologs in the MarFERReT database.
We conducted phylogenetic analyses of protein sequences from meso-exclusive GVs that showed significant similarities with eukaryotic protein sequences (Fig. S12). We found three cases that supported the predictions by the above MarFERReT-based analysis. The NifU-like N terminal domain (NifU_N) of MV2 was clearly related to eukaryotic homologs with the closest relationship with sequences from dinoflagellates (Kryptoperidinium) and stramenopiles. Two other cases suggested relationships between meso-exclusive GVs and specific eukaryotes. However, interpreting these two trees as a support for specific virus-host relationships is not straightforward as most of other related sequences were bacterial sequences.
We further conducted host inference by examining co-occurrence patterns between meso-exclusive GVs and microeukaryotes. Using metatranscriptome and 18S rRNA data from mesopelagic samples, we calculated Spearman's correlation coefficients for the relative abundance between meso-exclusive GVs and ASVs in pico-or total-fraction. Nine ASVs exhibited statistically significant positive correlations with five meso-exclusive GVs (P-value < 0.05; Fig. S13). The predicted potential hosts were heterotrophs (or potentially mixotrophs), belonging to Cercozoa (3 ASVs; Cercozoa, Cercozoa_X, Filosa-Imbricatea; amoeboids or flagellates of Rhizaria), Picozoa (1 ASV; heterotrophic picoeukaryotes), Dinophyceae (1 ASV), and Syndiniales (4 ASVs; early branching dinoflagellates). These potential hosts were undetected in the 18S rRNA/rDNA data from the surface layer, suggesting mesopelagic or deeper layers as their main habitat. As members of Syndi niales parasitize dinoflagellates, their co-occurrence with GVs may reflect shared hosts rather than host-virus relationships. Cercozoa, Picozoa, and Dinophyceae were reported in the mesopelagic layer, with Cercozoa in particular showing the highest activity among all depths (43). Host-virus associations have been suggested for Cercozoa and Dinoflagellata based on single-cell genomics data from induced algal bloom samples (24).
Although these analyses point to potential hosts of meso-exclusive GVs, the limited representation of mesopelagic protist genomes in current reference databases, the small number of samples in this study, and the low abundance of GVs in the mesopelagic layer mean that these host predictions should be considered preliminary. Other approaches such as single-cell genomics may directly link GVs to their hosts, offering validation beyond sequence homology and co-occurrence (24).
## Deep-sea-specific GVs distribute widely in the global ocean
Our discovery of meso-exclusive GVs at a coastal site in Japan raises questions regard ing their distributions and adaptation across different oceans and depth ranges. To address this issue and to catalog GV genomes specific to aphotic layers of the ocean, we constructed the GVGR database (see Materials and Methods) (Table S8), containing 4,473 GVMAGs and analyzed 1,890 publicly available marine metagenomes collected from depths ranging from 0 m to 10,899 m (Fig. S14).
The GVMAGs were found to have a wide vertical distribution, extending to a depth of 5,601 m (Fig. 3a). Based on the occurrence of GVMAGs across depths, we identified 101 deep-sea-specific GVMAGs that were only or predominantly distributed in the mesopela gic or deeper layers (Fig. 3b; Table S8) (see Materials and Methods). To further investi gate their biogeographic patterns, we analyzed the distribution of deep-sea-specific GVMAGs across eight geographic regions and quantified their geographic ranges. The total number of deep-sea-specific GVMAGs was relatively high in the North Atlantic Ocean, North Pacific Ocean, and Arctic Oceans, ranging from 44 to 53 (Fig. 3c). Only eight were detected in the Southern Ocean, which may be due to limited sampling efforts.
The connectivity map revealed sharing of GVMAGs between multiple oceanic regions (Fig. 3c), suggesting the existence of widely distributed deep-sea-specific GVMAGs. Indeed, 55.4% of the deep-sea-specific GVMAGs were detected in two or more regions, with 10 GVs detected in five or more regions (Fig. 4d). The Arctic Ocean and Southern Ocean exhibited a large proportion of unique deep-sea-specific GVMAGs (36.4% and 37.5%, respectively). These polar oceans, thus, represent a "hotspot" of endemic deepsea-specific GVMAGs, consistent with previous findings (19), and may be explained by limited water mass exchanges and steep environmental gradients that forms strong ecological barriers between high and lower latitude regions.
Phylogenetic analysis revealed that identified deep-sea-specific GVMAGs are scat tered across the tree (Fig. 4a), reminiscent of the phylogenetic distribution of GVs specific to cold environments (20). One clade within the order Imitervirales was enriched with deep-sea-specific GVMAGs (Fig. 4b); two Muroto meso-exclusive GVMAGs (NCV01 and NCV26) were also placed within this clade. The members of this clade showed high relative abundances (Fig. 4c) and wide geographic distributions (Fig. 4d). The Muroto meso-exclusive NCV01 (represented by ERS493705_165 after dereplication) was distributed in four oceanic regions (Fig. 3c and4d).
To investigate potential functional adaptations of deep-sea-specific GVMAGs, we identified significantly enriched KO terms and Pfam domains in deep-sea-specific GVMAGs using Fisher's exact test with FDR correction (P-value < 0.05) (see Materials and Methods; Table S9). Seventy-six KO terms were found to be enriched in deep-sea-specific GVMAGs (Fig. S15 andS16). Consistent with our results from the Muroto data, genes related to the ubiquitin system were found as deep-sea-specific KOs. Ammonium transporters and glutaminase were also deep-sea-specific. A previous study also revealed an enrichment of ammonium transporters in GVs detected below 200 m and suggested that these transporters may enhance nitrogen acquisition by their hosts, possibly in competition with Thaumarchaeota (30). The above mentioned Imitervirales deep-sea clade showed the highest number of deep-sea-specific KO terms (Fig. S16), suggesting lineage-specific specialization. Pfam domain analysis revealed 74 domains enriched in deep-sea-specific GVMAGs (Table S9), including homologs of cytochrome P450, CTP synthases, S-adenosylmethionine decarboxylase, and NifU_N. P450 has been previously reported in many GV genomes and has been proposed to be functionally connected to the 2-oxoglutarate and Fe (II)-dependent dioxygenase (54), which was also enriched in deep-sea-specific GVMAGs. The function of P450 in GVs remains unknown; however, it has been hypothesized that these genes might modulate viral or host lipid pools to aid energy production (55). Additionally, NifU is involved in Fe-S cluster formation, which plays roles in diverse cellular processes. NifU has been previously reported in eukaryotes, although its specific function remains unresolved (56). Two KO terms and 22 Pfam domains were enriched in the non-deep-sea-specific GVMAGs (Table S9). These included domains for DNA damage repair (UV damage endonuclease and 5ʹ-3ʹ exonucleases), rhodopsins, phosphate transport associated proteins (PhoH), and sulfotransferases. The enrichment of PhoH suggests adaptation to phosphate-limited conditions (57,58).
Taken together, the KO and Pfam enrichment patterns revealed that GVs in the deep-sea have evolved to encode several genes of unique functions, potentially enabling them to maintain replicative efficiency by modulating host metabolic programs under environmental constraints such as limited light availability and scarce organic com pounds.
## Summary
Our study provided the first integrative multi-omics evidence for the persistent activity of GVs in a mesopelagic environment. Six GVMAGs from the Muroto metagenomes were exclusively detected in the mesopelagic layer, many showing persistent transcriptional activity across seasons. Some GVMAGs detected in the mesopelagic layer were active in the surface layer but inactive in the mesopelagic layer, suggesting that GV populations detected among mesopelagic metagenomes may partially originate from the surface through vertical transport. Furthermore, we identified 101 deep-sea-specific GVMAGs from a global metagenomic data set. Our results suggest that multiple lineages of GVs have independently adapted to deep-sea environments and are widely distributed across the global ocean. We also revealed clear genomic difference between deep-seaspecific and other GVs, including 76 KO terms and 74 Pfam domains enriched in the deep-sea-specific GVMAGs. Deep-sea-specific GVs exhibited distinct gene content, notably including genes for the ubiquitin system. These findings collectively support the hypothesis that distinct GV lineages may have evolved to adapt to deep-sea environ ments by acquiring genes to manipulate host cellular processes to thrive in the aphotic environment, which largely differs from that of the photic zone.
## MATERIALS AND METHODS
## Sample collection
Eight time-series seawater samples were collected from two depths (surface: 0.5 m, mesopelagic: 320 m) at the Kochi Prefectural Deep Seawater Laboratory, Kochi Prefec ture, Japan (Fig. S1a), from December 2021 to April 2024. The seawater was pumped directly from the ocean and obtained via the deep-sea water intake system at the facility (37). The filtration system with membranes directly connected to the mesopela gic water tap was built to ensure that the water remained in its original condition (Fig. S1b; Table S1). After pre-filtration through a 150-µm-pore-size nylon mesh to remove large organisms, a large volume of seawater, especially for mesopelagic samples (30-1,110 L), was filtered through 3-µm-or 5-µm-pore-size polycarbonate membranes (Merck, Germany) and 0.2-µm-pore-size Sterivex-GP PES filters (Merck). Three groups of size fractionations were performed (Table S2). The first group (3.0-150 µm and 5.0-150 µm) mainly corresponds to nano-and microplankton and is referred to as the "nano/micro-fraction. " The second group (0.2-150 µm) covers pico-, nano-, and microplankton and is referred to as the "total-fraction. " The last group (0.2-3.0 µm and 0.2-5.0 µm) mainly corresponds to picoplankton and is referred to as the "pico-fraction. " We collected pico-fraction samples and nano/micro-fraction samples in early samplings (until September 2022), and total-fraction samples and nano/micro-fraction samples in later samplings. All filtrated membranes were immediately frozen and transported in a cryo-shipper at liquid nitrogen temperatures to the laboratory and stored at -80°C until DNA and RNA were extracted. Filtration and cryo-preservation of each filter were completed within 25 min.
Total DNA and RNA were extracted using the AllPrep RNA/DNA kit (Qiagen, Ger many) following the protocol described by Okazaki et al. (59). The DNA/RNA quantity was assessed using a Qubit fluorometer (Thermo Fisher Scientific, USA). The extracted DNA/RNA was used for 18S rDNA metabarcoding (28 samples), 18S rRNA metabarcod ing (28 samples), metagenomic (13 samples), and metatranscriptomic (25 samples) sequencing (Fig. S1c). Environmental variables, including temperature, salinity, and pH, were recorded continuously by the Kochi Prefectural Deep Seawater Laboratory (Table S10).
## Metabarcoding sequencing and analyses
Microeukaryote amplicon sequencing was performed for both DNA and RNA-derived cDNA (rDNA and rRNA sets, respectively). For RNA, genomic DNA removal and first-strand cDNA synthesis were conducted using the SuperScript IV VILO Master Mix with the ezDNase enzyme kit (Thermo Fisher Scientific) following the manufacturer's protocol. The V4 region of the 18S rRNA gene from both cDNA and DNA was amplified by the KAPA HiFi HotStart ReadyMix (Roche) using the universal eukaryotic primer set E572F/ E1009R (60) attaching Illumina overhang adapters. The PCR conditions were as follows: initial pre-denaturation at 95°C for 2 min, followed by 30 cycles of denaturation at 98°C for 20 s, annealing at 61°C for 15 s, and extension at 72°C for 30 s, with a final extension at 72°C for 2 min. Triplicate PCR products were pooled and then purified using VAHTS DNA Clean Beads (Vazyme). The amplicon libraries were sequenced on the Illumina MiSeq platform to generate paired-end reads (2 × 300 bp).
Raw sequencing reads were processed using QIIME2 v2024.10 (61) with the DADA2 plugin (62). The "dada2 denoise-paired" command was used for adapter removal, primer removal, low-quality read trimming, dereplication, chimera removal, and amplicon sequencing variant (ASV) identification. Rare ASVs that appeared in less than two sequences across all samples were excluded. Taxonomic classification was performed on the remaining ASVs based on a pre-trained naive Bayes classifier trained against the PR2 reference database v5.0.0 using the "feature-classifier classify-sklearn" plugin (63). ASVs classified as non-protist lineages (i.e., metazoa, fungi) were removed. Finally, 1,616,977 reads from 56 samples were grouped into 5,031 ASVs. To enable comparisons between samples, the ASV table was rarefied to the minimum read depth per sample (7,581 reads), resulting in a final data set of 4,981 ASVs and 424,536 reads. Statistical differences among sample groups (e.g., depth and season) were tested by PERMANOVA (9,999 permutations) on the Bray-Curtis dissimilarity calculated from the Hellinger-transformed ASV abundance table, using the Vegan v2.6.10 package in R (64). The standardized Levins index (BA) (65,66), representing the niche breadth, was calculated using the R package "spaa. " For each depth and size fraction, BA was estimated from the relative abundance of ASVs in the surface and mesopelagic samples collected on the same day.
## Metagenomic sequencing and analyses
Metagenomic shotgun sequencing libraries were constructed using the Illumina DNA Prep kit (Illumina) and sequenced on the Illumina NovaSeq 6000 platform, generating 150 bp paired-end reads with an average of 58 Gbp per sample. Raw reads were quality-controlled using fastp v0. 23.4 (67) with default parameters. For each metage nome, trimmed reads were assembled into contigs using MEGAHIT v1.2.9 (68) in "meta-sensitive" mode. Contigs over 2.5 kb were retained for downstream analysis. Trimmed reads from all samples were primarily used for cross-mapping against contigs assembled from individual samples by Bowtie2 v2.4.5 (69). Additionally, a co-assembly was performed by pooling reads from all mesopelagic samples. Contigs were clustered into bins using MetaBAT2 v2.15.15 (70). Metagenomic bins were generated for each of the 13 metagenomes and for the pooled mesopelagic metagenomes. Genes in MAGs were predicted using Prodigal-gv v2.11.0 (71,72) with "meta" mode for marker gene identification and functional annotation.
NCV-MAGs and MV-MAGs were reconstructed separately following previously published methods (23,33). To identify NCV-MAGs, we used a custom pipeline named hedera v.0.0.5 (https://github.com/banhbio/hedera). Briefly, the pipeline employs a gene density index (NCLDV index [23]) calculated using the genome size and the presence of 20 marker genes selected from the Nucleo-Cytoplasmic Virus Orthologous Groups (NCVOG) (73). MAGs with an NCLDV index over 5.75 are identified as potential NCV-MAGs. The NCV verification and decontamination is performed using the results of Viralrecall v2.1 (score > 0) (74), Virsorter2 v2.2.3 (max score group "NCLDV") (75), CAT v5.2.3 ("Nucleocytoviricota") (76), and hidden Markov models (HMMs) built with 149 NCVOGs (E-value <1.0 × 10 -3 ) (23). Bins selected by hedera were further investigated for the possibility of chimeric bins (e.g., multiple viral genomes) by manual inspection using Anvi'o v8 (77). First, we primarily focused on deep branching clades that showed markedly different occurrence patterns or GC contents from other clades. If such clades showed the seven core marker genes of GVs (PolB, A32 packaging enzyme, superfamily II helicase, VLTF3 transcriptional factor, topoisomerase family II, transcription factor IIB, and RNA polymerase second largest subunit [RNAPS]), they were retained as a different bin; otherwise, they were discarded. Second, for the remaining clades, if we identified clearly different patterns of occurrence or GC content, we split the bins into two or more sub-bins. Finally, the modified bins were re-assessed for the NCLDV index and only those with an NCLDV index over 5.75 were retained. For MV-MAGs, bins were identified as a mirusvirus if the HK97 MCP gene was detected using HMMER v3.4 (E-value <1.0 × 10 -3 , bit score >100) (78). After manual inspection with Anvi'o, MAGs containing a contig encoding HK97-MCP were retained. Decontamination was performed using CheckV v1.0.1 (database v1.4) (79) by concatenating all contigs of each MAG and using "end_to_end" mode to remove prokaryotic contigs and provirus segments (33). Dereplication was conducted using dRep v3.5.0 (80) at an average nucleotide identify (ANI) of 95% with the parameters "--ignoreGenomeQuality --S_algorithm ANImf -sa 0.95 --clusterAg single -sizeW 1". The resulting non-redundant Muroto GVMAGs (NCV-MAGs and MV-MAGs) were species-level representatives. GVMAG names were assigned as follows: "NCV" or "MV, " followed by a serial number indicating the rank of genome size from large to small (e.g., NCV10, MV5).
To determine the relative abundances of Muroto GVMAGs, the coverage and transcripts (or reads) per million (TPM) (81) of each MAG were calculated by mapping all quality-controlled trimmed reads to Muroto GVMAGs using CoverM v0.6.1 (82).
## Metatranscriptomic sequencing and analyses
RNA samples were subjected to metatranscriptomic sequencing (Fig. S1c). Poly-A strand-specific RNA sequencing libraries were constructed using the NEBNext Poly(A) mRNA Magnetic Isolation Module and NEBNext Ultra ll Directional RNA Library Prep Kit (New England Biolabs, USA). Libraries were sequenced on the Illumina NovaSeq 6000 platform and 150 bp paired-end reads were generated. An average of 24 Gbp of sequence per sample was obtained. Raw reads were quality-controlled using fastp v0. 23.4 with the "-l 100" parameter to remove low-quality reads and reads shorter than 100 bp. SortMeRNA v4. 3.6 (83) was used to filter out rRNA reads against the Silva v138.1 LSU NR99, Silva v138.1 SSU NR99, RFAM 5s, and RFAM 5.8s databases using default settings. Filtered mRNA reads were mapped against the Muroto GVMAGs using Salmon v1.10.1 (84) with the parameters "--meta --minScoreFraction 0.95 --validateMappings". Transcript abundance was normalized as TPM. The results were used to quantify the relative expression level of each gene and MAG.
## Compiling the GVGR database and identifying deep-sea-specific GVMAGs
To investigate the global deep-sea GV distribution and adaptation, we compiled a comprehensive giant virus genome reference, named GVGR, by integrating the Muroto GVMAGs with publicly available GV metagenomic data sets (Fig. S4). We first extrac ted NCV-MAGs from metagenomic bins generated using MetaBAT2 by the OceanDNA MAG project (39). Metagenomic bins with an NCLDV index over 5.75 were consid ered as potential NCV-MAGs. These MAGs were decontaminated through retaining contigs identified as "NCLDV" by Virsorter2 v2.2.4 or those with a ViralRecall v2.1 score greater than 0. Host-derived sequences and proviral segments were removed using CheckV v1.0.1, as mentioned above. For MV-MAG classification, the initial screening was consistent with the process described for Muroto MV-MAGs. Contigs classified as "chromosome" or "plasmid" using geNomad v1.8.0 (72) were excluded. MAGs with a size between 50 kbp-3 Mbp in length were retained as GVMAGs.
Additionally, we incorporated publicly available GV genomes from multiple sources, including GOEVdb (2), 1,065 GVMAGs from Uranouchi Inlet, Japan (23), 293 LBGVMAGs from Lake Biwa (33), and 4 circular endogenous Mirusviricota (6,40). All collected MAGs were pooled and dereplicated using dRep v3.5 at an ANI of 95% as previously men tioned. Within the ANI-based cluster, the MAGs with the highest N50 value among the top 30% largest sized genomes were selected as representative.
To assess MAG quality, a phylogeny-informed MAG assessment (PIMA) based on a phylogenetic tree and orthologous groups (OGs) was performed on the above-men tioned data set and Muroto GVMAGs (23). For phylogenetic tree reconstruction, seven conserved marker genes were identified using the script "ncldv_markersearch.py" (57). For MV-MAGs, four marker genes (HK97-MCP, RNAPS, RNA polymerase largest subunit [RNAPL], and PolB) were identified using HMMER v3.4 (78) with pairwise alignment against the reference HMM profiles. These marker genes were aligned and concaten ated using Clustal Omega v1.2.4 (85), trimmed with trimAl v1.2.1 (-gt 0.1) (86), and a phylogenetic tree was built with FastTree v2.1.11 (87) using default parameters, serving as an input for PIMA of NCV-MAGs and MV-MAGs, separately. OGs were classified using GVOG HMM profiles (88) using HMMER v3.4 (E-value < 1.0 × 10 -10 ) (78). If multiple OGs were assigned to a single sequence, the OG with the lowest E-value was retained.
PIMA was applied at a relative evolutionary divergence threshold of 0.65, correspond ing to genus-or family-level classification (23). MAG consistency was evaluated based on the proportion of core genes present, while redundancy was calculated as the propor tion of duplicated core genes exceeding the mode copy number observed across all MAGs within each lineage. MAGs exhibiting redundancy over 50% were removed. A second round of dRep dereplication was subsequently performed to integrate Muroto GVMAGs, producing the final GVGR data set. The taxonomy of GVMAGs in the GVGR data set was validated using TIGTOG (89). Gene prediction and abundance analyses in all 1,890 global samples (39) were performed using methods identical to those described above.
Each of these GVMAGs was categorized as "deep-sea-specific" or "non-deep-sea-spe cific" We considered genomes showing overrepresentation in deep-sea samples (depth > 200 m) as deep-sea-specific GVMAGs. The overrepresentation was ascertained using Mann-Whitney U tests with Benjamini-Hochberg (BH) corrected P values (P-value < 0.05) (90). We also considered genomes with signals only in deep-sea samples (TPM > 0) as deep-sea-specific GVMAGs. Other GVMAGs, not assigned to deep-sea-specific category, were referred to as "non-deep-sea-specific. "
## Protein prediction and annotation
Proteins predicted from Muroto GVMAGs and the GVGR database were annotated with KEGG orthology (KO) using KofamScan v1.3.0 (91). KO annotations with the lowest E-values were retained. We further generated Pfam annotations of the genomes using AnnoMazing (https://github.com/BenMinch/AnnoMazing), a pipeline to annotate proteins based on HMM profiles. The annotations were performed using the Pfam database (92) with an E-value cut-off of 1.0 × 10 -5 .
To identify KO terms and Pfam domains significantly enriched in deep-sea-specific GVMAGs, Fisher's exact test was conducted to compare the frequency of each annotated KO term in deep-sea-specific GVMAGs and non-deep-sea specific GVMAGs. KO terms with a total frequency ≤ 2 were excluded from further analysis.
To infer the potential eukaryotic hosts of GVs, we performed protein homology searches using BLASTP in DIAMOND v2.1.10 (E-value < 1.0 × 10 -5 and identity > 50%) (93) against the MarFERReT ver1.1.1 database (94). Hits with the lowest E-values were retained and used to assign candidate host taxonomy. For each retained protein, up to 100 homologs were further retrieved from a combined MarFERReT and NR database using DIAMOND BLASTP (E-value < 1.0 × 10 -5 , identity > 50%). These homologous sequence sets were subsequently included in the phylogenetic analyses described below to assess putative host associations suggested by horizontal gene transfer.
## Analyses of phylogenetic diversity and clade delineation
A phylogenetic tree was first constructed based on PolB sequences (>500 amino acids) extracted from Muroto contigs (>1 kbp). The GOEV database and a wide range of eukaryotic and additional viral lineages (2,95) were used as a reference set of PolB sequences to evaluate the completeness of GV recovery in Muroto GVMAGs. Another tree for Mirusviricota was generated using HK97-MCP sequences from Muroto MV-MAGs, combined with previously reported Mirusviricota genomes from the GOEV database, LBGVMAGs, and circular endogenous MV-MAGs. For Nucleocytoviricota, a concatenated phylogenetic tree was constructed using the 7 core marker genes (88) for Muroto NCV-MAGs, along with 220 reference genomes from culture (2). Another concatenated phylogenetic tree was generated using the same 7 core marker genes, incorporating all deep-sea-specific MAGs, Muroto meso-exclusive MAGs, and the GVGR data set redundant by dRep v3.5.0 at an ANI of 85%. The marker gene search, alignment, concatenation, and trimming methods were the same as those described for the PIMA above. The tree was inferred using IQ-TREE v2.2.0 (96) with the ModelFinder Plus option to determine the best-fitting model. Support values were inferred using 1,000 ultrafast bootstraps (97). Tree visualization and rooting were carried out using iTOL v7 (98).
## FUNDING Funder
Grant(s)
## References
1. Schulz, Abergel, Woyke (2022) "Giant virus biology and diversity in the era of genome-resolved metagenomics" *Nat Rev Microbiol*
2. Gaïa, Meng, Pelletier et al. (2023) "Mirusvi ruses link herpesviruses to giant viruses" *Nature*
3. Archibald, Chung, Brask et al. (2020) "Persistent mirusvirus infection in the marine protist Aurantiochytrium"
4. Meng, Endo, Blanc-Mathieu et al. (2021) "Quantitative assessment of nucleocytoplasmic large DNA virus and host interactions predicted by co-occurrence analyses"
5. Zhao, Meng, Hikida et al. (2024) "Eukaryotic genomic data uncover an extensive host range of mirusviruses" *Curr Biol*
6. (2025)
7. Wang, Li, Lin (2023) "Active viral infection during blooms of a dinoflagellate indicates dinoflagellate-viral co-adaptation" *Appl Environ Microbiol*
8. Hevroni, Vincent, Ku et al. (2023) "Daily turnover of active giant virus infection during algal blooms revealed by single-cell transcriptomics" *Sci Adv*
9. Funaoka, Hiromoto, Morimoto et al. (2023) "Diversity in infection specificity between the bloom-forming microalga Heterosigma akashiwo and its dsDNA virus, Heterosigma akashiwo virus" *Microbes Environ*
10. Rosenwasser, Ziv, Creveld S Van et al. (2016) "Virocell metabolism: metabolic innovations during host-virus interactions in the ocean" *Trends Microbiol*
11. Monier, Pagarete, De Vargas et al. (2009) "Horizontal gene transfer of an entire metabolic pathway between a eukaryotic alga and its DNA virus" *Genome Res*
12. Kaneko, Blanc-Mathieu, Endo et al. (2021) "Eukaryotic virus composition can predict the efficiency of carbon export in the global ocean"
13. Hingamp, Grimsley, Acinas et al. (2013) "Exploring nucleo-cytoplasmic large DNA viruses in Tara Oceans microbial metagenomes" *ISME J*
14. Ha, Moniruzzaman, Aylward (2021) "High transcriptional activity and diverse functional repertoires of hundreds of giant viruses in a coastal marine system"
15. Carradec, Pelletier, Silva et al. (2018) "A global ocean atlas of eukaryotic genes" *Nat Commun*
16. Li, Hingamp, Watai et al. (2018) "Degenerate PCR primers to reveal the diversity of giant viruses in coastal waters" *Viruses*
17. Mihara, Koyano, Hingamp et al. (2018) "Taxon richness of "Megaviridae" exceeds those of bacteria and archaea in the ocean" *Microbes Environ*
18. Schulz, Roux, Paez-Espino et al. (2020) "Giant virus diversity and host interactions through global metagenomics" *Nature*
19. Endo, Blanc-Mathieu, Li et al. (2020) "Biogeography of marine giant viruses reveals their interplay with eukaryotes and ecological functions" *Nat Ecol Evol*
20. Meng, Delmont, Gaïa et al. (2023) "Genomic adaptation of giant viruses in polar oceans" *Nat Commun*
21. Bäckström, Yutin, Jørgensen et al. (2019) "Virus genomes from deep sea sediments expand the ocean megavirome and support independent origins of viral gigantism" *mBio*
22. Prodinger, Endo, Takano et al. (2022) "Year-round dynamics of amplicon sequence variant communities differ among eukaryotes, Imitervirales and prokaryotes in a coastal ecosystem" *FEMS Microbiol Ecol*
23. Fang, Meng, Xia et al. (2025) "Genome-resolved year-round dynamics reveal a broad range of giant virus microdiversity. mSystems 10:e0116824"
24. Fromm, Hevroni, Vincent et al. (2024) "Single-cell RNA-seq of the rare virosphere reveals the native hosts of giant viruses in the marine environment" *Nat Microbiol*
25. Xia, Kameyama, Prodinger et al. (2022) "Tight association between microbial eukaryote and giant virus communities in the Arctic Ocean" *Limnol Oceanogr*
26. Danovaro, Corinaldesi, Dell'anno et al. (2017) "The deepsea under global change" *Curr Biol*
27. Abrahão, Silva, Silva et al. (2018) "Tailed giant Tupanvirus possesses the most complete translational apparatus of the known virosphere" *Nat Commun*
28. Mizuno, Ghai, Saghaï et al. (2016) "Genomes of abundant and widespread viruses from the deep ocean" *mBio*
29. Jian, Yi, Wang et al. (2021) "Diversity and distribution of viruses inhabiting the deepest ocean on Earth" *ISME J*
30. Sheam, Luo (2025) "Vertical transport and spatiotemporal dynamics of giant viruses in the North Pacific subtropical gyre" *ISME J*
31. Acinas, Sánchez, Salazar et al. (2021) "Deep ocean metagenomes provide insight into the metabolic architecture of bathypelagic microbial communities" *Commun Biol*
32. Ha, Moniruzzaman, Aylward (2023) "Assessing the biogeogra phy of marine giant viruses in four oceanic transects" *ISME Commun*
33. Zhang, Meng, Fang et al. (2024) "Spatiotemporal dynamics of giant viruses within a deep freshwater lake reveal a distinct dark-water community" *ISME J*
34. Wu, Gao, Zhang et al. (2011) "Optimization of wholetranscriptome amplification from low cell density deep-sea microbial samples for metatranscriptomic analysis" *J Microbiol Methods*
35. Steiner, De Corte, Geijo et al. (2019) "Highly variable mRNA half-life time within marine bacterial taxa and functional genes" *Environ Microbiol*
36. Guo (1991) "The Kuroshio. Part II. Primary productivity and phytoplank ton" *Oceanogr Mar Biol Annu Rev*
37. Yamaguchi, Inoue, Hirakawa et al. (2003) "Deep-sea water suction technology" *Furukawa Review*
38. Kang, Kawamoto, Kaneda et al. (2004) "Characteriza tion of selenium in the deep ocean water pumped up at Muroto" *Japan. Deep Ocean Water Res*
39. Nishimura, Yoshizawa (2022) "The OceanDNA MAG catalog contains over 50,000 prokaryotic genomes originated from various marine environments" *Sci Data*
40. Collier, Rest, Gallot-Lavallée et al. (2023) "The protist Aurantiochytrium has universal subtelomeric rDNAs and is a host for mirusviruses" *Curr Biol*
41. Prokopowich, Gregory, Crease (2003) "The correlation between rDNA copy number and genome size in eukaryotes" *Genome*
42. Zhu, Massana, Not et al. (2005) "Mapping of picoeucar yotes in marine ecosystems with quantitative PCR of the 18S rRNA gene" *FEMS Microbiol Ecol*
43. Giner, Pernice, Balagué et al. (2020) "Marked changes in diversity and relative activity of picoeukaryotes with depth in the world ocean" *ISME J*
44. Cunha, Gaia, Ogata et al. (2022) "Giant viruses encode actin-related proteins" *Mol Biol Evol*
45. Kijima, Hikida, Delmont et al. (2024) "Complex genomes of early nucleocytoviruses revealed by ancient origins of viral aminoacyl-tRNA synthetases" *Mol Biol Evol*
46. Li, Zhan (2011) "The role of centrosomal Nlp in the control of mitotic progression and tumourigenesis" *Br J Cancer*
47. Wu, Fan, Chen et al. (2024) "The class III phosphatidylinosi tol 3-kinase VPS34 supports EV71 replication by promoting viral replication organelle formation" *J Virol*
48. Hershko (2005) "The ubiquitin system for protein degradation and some of its roles in the control of the cell division cycle" *Cell Death Differ*
49. Tang, Wu, Chen et al. (2018) "Pleiotropic roles of the ubiquitinproteasome system during viral propagation" *Life Sci*
50. Ramanujam, Alderson, Pritišanac et al. (2020) "Protein structural changes characterized by high-pressure, pulsed field gradient diffusion NMR spectroscopy" *J Magn Reson*
51. Kozuch, Stillinger, Debenedetti (2019) "Low temperature protein refolding suggested by molecular simulation" *J Chem Phys*
52. Buscaglia, Iriarte, Schulz et al. (2024) "Adaptation strategies of giant viruses to low-temperature marine ecosystems" *ISME J*
53. Abbas, Larisch (2021) "Killing by degradation: regulation of apoptosis by the ubiquitin-proteasome-system" *Cells*
54. Lamb, Follmer, Goldstone et al. (2019) "On the occurrence of cytochrome P450 in viruses" *Proc Natl Acad Sci*
55. Lamb, Goldstone, Belhaouari et al. (2025) "Cytochrome b5 occurrence in giant and other viruses belonging to the phylum Nucleocytoviricota" *NPJ Viruses*
56. Hwang, Dempsey, Tan et al. (1996) "A modular domain of NifU, a nitrogen fixation cluster protein, is highly conserved in evolution" *J Mol Evol*
57. Moniruzzaman, Martinez-Gutierrez, Weinheimer et al. (2020) "Dynamic genome evolution and complex virocell metabolism of globally-distributed giant viruses" *Nat Commun*
58. Kim, Makino, Amemura et al. (1993) "Molecular analysis of the phoH gene, belonging to the phosphate regulon in Escherichia coli" *J Bacteriol*
59. Okazaki, Nguyen, Nishihara et al. (2023) "A fast and easy method to co-extract DNA and RNA from an environmental microbial sample" *Microb Environ*
60. Comeau, Li, Tremblay et al. (2011) "Arctic Ocean microbial community structure before and after the 2007 record sea ice minimum" *PLoS One*
61. Bolyen, Rideout, Dillon et al. (2019) "Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2" *Nat Biotechnol*
62. Callahan, Mcmurdie, Rosen et al. (2016) "DADA2: high-resolution sample inference from Illumina amplicon data" *Nat Methods*
63. Bokulich, Kaehler, Rideout et al. (2018) "Optimizing taxonomic classification of marker-gene amplicon sequences with QIIME 2's q2feature-classifier plugin" *Microbiome*
64. (2024) "R: a language and environment for statistical computing"
65. Levins (1968) "Evolution in changing environments"
66. Krebs (1989) "Ecological methodology. HarperCollins College Division"
67. Chen, Zhou, Chen et al. (2018) "fastp: an ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
68. Li, Liu, Luo et al. (2015) "MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph" *Bioinformatics*
69. Langmead, Wilks, Antonescu et al. (2019) "Scaling read aligners to hundreds of threads on general-purpose processors" *Bioinformatics*
70. Kang, Li, Kirton et al. (2019) "MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies" *PeerJ*
71. Hyatt, Chen, Locascio et al. (2010) "Prodigal: prokaryotic gene recognition and translation initiation site identification" *BMC Bioinformatics*
72. Camargo, Roux, Schulz et al. (2024) "Identification of mobile genetic elements with geNomad" *Nat Biotechnol*
73. Yutin, Wolf, Raoult et al. (2009) "Eukaryotic large nucleocytoplasmic DNA viruses: clusters of orthologous genes and reconstruc tion of viral genome evolution" *Virol J*
74. Aylward, Moniruzzaman (2021) "ViralRecall-a flexible command-line tool for the detection of giant virus signatures in 'omic data" *Viruses*
75. Guo, Bolduc, Zayed et al. (2021) "VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses" *Microbiome*
76. Von Meijenfeldt, Arkhipova, Cambuy et al. (2019) "Robust taxonomic classification of uncharted microbial sequences and bins with CAT and BAT" *Genome Biol*
77. Eren, Esen, Quince et al. (2015) "Anvi'o: an advanced analysis and visualization platform for 'omics data" *PeerJ*
78. Mistry, Finn, Eddy et al. (2013) "Challenges in homology search: HMMER3 and convergent evolution of coiled-coil regions" *Nucleic Acids Res*
79. Nayfach, Camargo, Schulz et al. (2021) "CheckV assesses the quality and completeness of metagenomeassembled viral genomes" *Nat Biotechnol*
80. Olm, Brown, Brooks et al. (2017) "dRep: a tool for fast and accurate genomic comparisons that enables improved genome recovery from metagenomes through de-replication" *ISME J*
81. Mortazavi, Williams, Mccue et al. (2008) "Mapping and quantifying mammalian transcriptomes by RNA-Seq" *Nat Methods*
82. Aroney, Newell, Nissen et al. (2025) "CoverM: read alignment statistics for metagenomics" *Bioinformatics*
83. Kopylova, Noé, Touzet (2012) "SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data" *Bioinformatics*
84. Patro, Duggal, Love et al. (2017) "Salmon provides fast and bias-aware quantification of transcript expression" *Nat Methods*
85. Sievers, Higgins (2014) "Clustal Omega, accurate alignment of very large numbers of sequences" *Methods Mol Biol*
86. Capella-Gutiérrez, Silla-Martínez, Gabaldón (2009) "trimAl: a tool for automated alignment trimming in large-scale phylogenetic analyses" *Bioinformatics*
87. Price, Dehal, Arkin (2010) "FastTree 2--approximately maximum-likelihood trees for large alignments" *PLoS One*
88. Aylward, Moniruzzaman, Ha et al. (2021) "A phylogenomic framework for charting the diversity and evolution of giant viruses" *PLoS Biol*
89. Ha, Aylward (2024) "Automated classification of giant virus genomes using a random forest model built on trademark protein families" *NPJ Viruses*
90. Benjamini, Hochberg (1995) "Controlling the false discovery rate: a practical and powerful approach to multiple testing" *J Royal Stat Soc Ser B*
91. Aramaki, Blanc-Mathieu, Endo et al. (2020) "KofamKOALA: KEGG Ortholog assignment based on profile HMM and adaptive score threshold" *Bioinformatics*
92. Mistry, Chuguransky, Williams et al. (2021) "Pfam: the protein families database in 2021" *Nucleic Acids Res*
93. Buchfink, Reuter, Drost (2021) "Sensitive protein alignments at tree-of-life scale using DIAMOND" *Nat Methods*
94. Groussman, Blaskowski, Coesel et al. (2023) "MarFERReT, an open-source, version-controlled reference library of marine microbial eukaryote functional genes" *Sci Data*
95. Kazlauskas, Krupovic, Guglielmini et al. (2020) "Diversity and evolution of B-family DNA polymerases" *Nucleic Acids Res*
96. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol Biol Evol*
97. Hoang, Chernomor, Haeseler et al. (2018) "UFBoot2: improving the ultrafast bootstrap approximation" *Mol Biol Evol*
98. Letunic, Bork (2024) "Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724191&blobtype=pdf | # Bovine coronavirus enters HRT-18 cells via membrane fusion and clathrin-mediated endocytosis in a low pH-, dynamin-, cholesterol-, microtubule-, Rab7-, and Rab11-dependent manner
Chen Chen, Long Zhao, Nannan Su, Xingyu Peng, Boli Song, Liang Zhang, Kangkang Guo
## Abstract
Bovine coronavirus (BCoV) infection poses a significant threat to the global cattle industry due to its dual tropism for the respiratory and intestinal systems, causing substantial economic losses. Elucidating the molecular mechanisms of viral entry is critical for developing targeted interventions against BCoV. This study systematically investigates the entry mechanisms of BCoV in HRT-18 cells through different methods.Our data reveal that BCoV entry into HRT-18 cells is dependent on membrane fusion and clathrin-mediated endocytosis (CME). This process is dependent on dynamin, choles terol, microtubules, cathepsins, and low pH. In contrast, caveolin-mediated endocytosis, micropinocytosis, and TMPRSS2 do not contribute to BCoV entry. Furthermore, we identified Rab7 and Rab11 as key regulators of BCoV endocytosis. Silencing Rab7 and Rab11 significantly inhibited BCoV entry, while silencing Rab5 had no discernible effect. Confocal microscopy confirmed the co-localization of BCoV particles with Rab7 and Rab11, further supporting their role in the viral entry process. Collectively, our findings provide the first evidence that BCoV enters HRT-18 cells via membrane fusion and CME in a low pH-, dynamin-, cholesterol-, microtubule-, cathepsin-, Rab7-, and Rab11-depend ent manner. These findings advance our understanding of BCoV pathogenesis and may facilitate the development of novel antiviral strategies against this pathogen. IMPORTANCE Emerging and re-emerging coronaviruses are causing severe epidemics in both humans and animals worldwide. Bovine coronavirus (BCoV) is a major pathogen causing severe diarrhea and respiratory disease in cattle, leading to substantial economic losses in the livestock industry. However, the molecular mechanism of BCoV entry into cells remains poorly understood. Here, we reveal that BCoV enters HRT-18 cells via membrane fusion and clathrin-mediated endocytosis, and acidic environment, dynamin, cholesterol, microtubules, cathepsins, Rab7, and Rab11 are also required. This study represents the first report on the mechanism of BCoV cell entry, which advances the understanding of BCoV infection pathogenesis and provides potential targets for the development of novel antiviral drugs. KEYWORDS bovine coronavirus, membrane fusion, clathrin-mediated endocytosis, Rab7, Rab11 B ovine coronavirus (BCoV) is a globally prevalent pathogen that causes severe diarrhea and respiratory disease in cattle, resulting in substantial economic losses to the livestock industry (1). As a member of the family Coronaviridae, genus β-coronavirus, BCoV is an enveloped, single-stranded positive-sense RNA virus with the longest known RNA genome (~31 kb) (2, 3). This genome contains 13 open reading frames (ORFs) flanked by 5′ and 3′ untranslated regions. The ORF1 encodes a polyprotein, pp1a, which is transferred to pp1ab by ribosomes and then hydrolyzed by proteases into multiple
non-structural proteins, which are critical for viral replication and immune evasion (4). BCoV encodes five structural proteins, including spike (S), envelope (E), transmem brane glycoprotein (M), and nucleoprotein (N), and some coronavirus-specific hemagglu tinin-esterase (HE) proteins (5). The S protein mediates viral attachment, fusion, and entry (6). The N protein is the only one that binds to the RNA genome and is also involved in viral assembly and budding (7). The M protein plays an important role in BCoV assembly and constitutes the core of the virus together with the N protein (6). The E protein is located inside the M protein and acts together with the M protein in the assembly of viral particles (6,8). In addition, the E protein also plays a role in the release and pathogenicity of viral particles (8). The HE protein, containing lectin and esterase domains, enhances entry and tropism (9).
Viral infection is initiated by the specific interaction between viral envelope proteins and cellular receptors, followed by viral entry into host cells and the release of viral genomes for replication (9). Thus, viral entry into host cells is a critical determinant of successful infection. Current research indicates that coronaviruses utilize two primary pathways: direct membrane fusion at the plasma membrane or receptor-mediated endocytosis (10). The latter includes clathrin-mediated endocytosis (CME), caveolin-medi ated endocytosis (CavME), lipid raft-mediated endocytosis, micropinocytosis, as well as clathrin/caveolin-independent endocytosis . Studies have shown that transmissible gastroenteritis virus enters swine testis (ST) cells via both CME and CavME (11), while porcine epidemic diarrhea virus (PEDV) uses CME, CavME, and lipid raft-dependent endocytosis in African green monkey kidney cells (Vero) and intestinal porcine epithelial cell line-J2 (IPEC-J2) cells (12). Porcine deltacoronavirus (PDCoV) utilizes both macropino cytosis and CME for entry into porcine ileal epithelial cells (IPI-2I), and CavME for ST cells and porcine kidney (PK-15) cells (13,14). Severe acute respiratory syndrome-related coronavirus (SARS-CoV) employs CME and lipid raft-dependent pathways for host cell entry (15). Collectively, these studies demonstrate that coronaviruses exploit diverse endocytic pathways to invade host cells. However, the precise entry mechanism of BCoV into permissive cells remains poorly characterized.
The endosomal system consists of a series of interconnected membrane chambers responsible for cargo recognition, sorting, trafficking, and degradation, including early endosomes (EEs), late endosomes (LEs), and recycling endosomes (16). Accumulating evidence highlights the critical role of the endosomal system in life cycles, particu larly during the entry phase (16,17). Upon endocytosis, viruses are internalized into endosomes and undergo sequential sorting for intracellular transport. For SARS-CoV-2, the virus binds to the ACE2 receptors at the plasma membrane, triggering endocyto sis into Rab5-positive EEs and subsequent maturation into Rab7-positive LEs (18,19). Similarly, Japanese encephalitis virus exploits Rab5-and Rab11-dependent trafficking from EEs to recycling endosomes, creating a permissive environment for genome release (20). Collectively, these studies highlight the critical role of endosomal trafficking machinery in viral entry.
In this study, chemical inhibitors and RNA interference were employed to investigate which way and molecules were involved in BCoV entry into HRT-18 cells, a permissive cell line for BCoV infection. Our results revealed that BCoV enters HRT-18 cells via membrane fusion and the CME pathway, while CavME and micropinocytosis were not involved. This process was also dependent on an acidic environment, dynamin, cholesterol, microtu bules, and cathepsins. Additionally, Rab7 and Rab11 were identified as critical regulators of post-endocytic viral trafficking. These findings provide novel insights into the specific mechanisms of BCoV entry into HRT-18 cells, which may facilitate the development of targeted antiviral strategies.
## RESULTS
## BCoV entry into HRT-18 cells is low pH-dependent
To investigate the role of an acidic environment in BCoV entry into HRT-18 cells, cells were pretreated with non-toxic concentrations of NH 4 Cl (40 mM) and chloroquine (CQ, 80 µM), both endosomal acidification inhibitors, for 24 h prior to infection with BCoV (MOI = 1) (Fig. 1A). After 12, 24, and 48 h of infection, cells and supernatants were collected for viral proliferation detection. The results showed that treatment with NH 4 Cl or CQ significantly reduced viral RNA levels, N protein expression, and infectious virus titers compared to untreated cells (P < 0.05; Fig. 1B through E), indicating that an acidic environment is essential for BCoV infection. To determine whether an acidic environment is required for BCoV entry, cells were pretreated with NH 4 Cl or CQ and infected with BCoV at an MOI of 5 for 1, 2, or 3 h. The RT-qPCR analysis revealed a significant reduction in viral RNA copy numbers in inhibitor-treated cells compared to controls at all time points (P < 0.001; Fig. 1F), suggesting that NH 4 Cl and CQ inhibited BCoV entry into HRT-18 cells. To assess the effect of an acidic environment on BCoV attachment, HRT-18 cells were pretreated with CQ and NH 4 Cl for 24 h, followed by incubation with the virus at 4°C for 1 h to allow attachment without internalization. RT-qPCR analysis showed that neither CQ nor NH 4 Cl treatment affected BCoV attachment to the cell surface (P > 0.05; Fig. 1G). These results confirmed that an acidic environment is essential for BCoV entry into HRT-18 cells but not for viral attachment.
## BCoV entry into HRT-18 cells depends on dynamin
Dynamin, a GTPase critical for endocytosis (including clathrin-mediated and caveolinmediated pathways), has been implicated in viral entry (21,22). To investigate the potential role of dynamin in BCoV infection and entry, dynamin inhibitor dynasore was used. Cytotoxicity assays revealed that 20 µM dynasore was non-toxic to HRT-18 cells (Fig. 2A). Subsequently, cells were pretreated with 20 µM dynasore for 24 h and infected with BCoV (MOI = 1). Viral replication was assessed at 12, 24, and 48 h postinfection via viral RNA quantification, N protein analysis, and infectious virus titration. We observed that BCoV infection was markedly inhibited in dynasore-treated cells (P < 0.05; Fig. 2B through E). To determine whether dynamin affects BCoV entry, cells were pretreated with dynasore and infected with BCoV at an MOI of 5 for 1, 2, or 3 h. RT-qPCR analysis revealed a significant reduction in viral RNA copy numbers in dynasore-treated cells at all time points (P < 0.05; Fig. 2F), indicating that dynamin is required for the viral entry. Dynasore treatment had no effect on BCoV attachment to the cell surface, indicating that dynamin is not involved in the attachment process (P > 0.05; Fig. 2G). To further evaluate the role of dynamin in BCoV endocytosis, we used siRNA to silence dynamin expression. The silencing efficiency was assessed by RT-qPCR, and the siRNA with the highest silencing efficiency was selected for subsequent experiments (Fig. 2H). HRT-18 cells were transfected with dynamin-targeting siRNA for 48 h, followed by BCoV infection for 12, 24, and 48 h. The results showed that viral RNA copy numbers, N protein expression, and infectious virus titers were significantly reduced (P < 0.05; Fig. 2I through L), indicating that silencing dynamin markedly inhibited BCoV infection. The effects of dynamin silencing on BCoV attachment and entry were subsequently evaluated. RT-qPCR analysis showed a significant reduction in viral RNA copy numbers during the entry phase (P < 0.01; Fig. 2M), whereas no significant changes were observed during the attachment phase (P > 0.05; Fig. 2N). These findings indicate that dynamin is required for BCoV entry but does not affect the initial attachment to HRT-18 cells.
## BCoV entry into HRT-18 cells via membrane fusion
Membrane fusion is a crucial step in the entry of enveloped viruses into host cells, particularly for coronaviruses (10). To evaluate the role of membrane fusion in BCoV entry, we tested the novel fusion inhibitor SSAA09E3. The CCK-8 assays confirmed that 40 µM SSAA09E3 exhibited no cytotoxicity in HRT-18 cells (Fig. 3A). Cells were pretreated with 40 µM SSAA09E3 for 24 h prior to infection with BCoV (MOI = 1). Viral propagation was analyzed at 12, 24, and 48 h post-infection by quantifying viral RNA copy numbers, N protein expression, infection rates, and progeny virus titers. Figure 3B through E demonstrated that SSAA09E3 treatment significantly reduced BCoV gene copy numbers, N protein expression levels, viral infection rate, and progeny virus titer compared to controls (P < 0.01), indicating effective inhibition of BCoV propagation. To assess whether SSAA09E3 targets viral entry, cells were pretreated with the inhibitor and infected with BCoV (MOI = 5) for 1, 2, or 3 h. RT-qPCR analysis revealed a significant reduction in viral RNA copy numbers in SSAA09E3-treated cells at all time points (P < 0.01; Fig. 3F), confirming that SSAA09E3 inhibits BCoV entry. We also showed that inhibition of membrane fusion did not affect BCoV attachment to the cell surface (P > 0.05; Fig. 3G). Subsequently, flow cytometry analysis demonstrated that SSAA09E3 significantly inhibited BCoV entry by suppressing syncytium formation (Fig. 3H). Further microscopic visualization revealed that BCoV infection induces syncytium formation in HRT-18 cells and that SSAA09E3 treatment significantly inhibits this process (Fig. 3I). Collectively, these findings establish membrane fusion as an essential mechanism for BCoV entry into HRT-18 cells, while having no effect on viral attachment.
## BCoV entry depends on clathrin-mediated endocytosis
Clathrin-mediated endocytosis (CME), the classical endocytic pathway, is exploited by numerous viruses for cellular entry (23). To determine whether BCoV entry relies on CME, we used chlorpromazine (CPZ, 20 µM) (Fig. 4A), a specific inhibitor of CME, to pretreated HRT-18 cells for 24 h. Subsequently, cells were infected with BCoV (MOI = 1) for 12, 24, and 48 h. Viral proliferation was assessed by quantifying viral RNA copy numbers and N protein expression in cell lysates, and infectious virus titers in supernatants. The results revealed that CPZ significantly inhibited viral infection (P < 0.01) (Fig. 4B through E). To evaluate CPZ's effect on viral entry, pretreated cells were infected with BCoV (MOI of 5) for 1, 2, and 3 h. The results revealed a significant reduction in viral RNA copy numbers in CPZ-treated cells at all time points (P < 0.001; Fig. 4F), confirming inhibition of BCoV entry. Confocal microscopy further demonstrated co-localization of BCoV with clathrin during the entry phase (Fig. 4G). Subsequently, the effect of CPZ on BCoV attachment was evaluated. As shown in Fig. 4H, no significant difference in viral RNA copy numbers was observed between the control and CPZ-treated groups (P > 0.05), indicating that CPZ had no effect on BCoV attachment. To further confirm the role of CME in BCoV entry steps, clathrin heavy chain (CLTC) expression was silenced by siRNA. Silencing efficiency was evaluated by RT-qPCR, and the siRNA with the highest efficiency was selected for subsequent experiments (Fig. 4I). HRT-18 cells were transfected with siCLTC for 48 h, followed by BCoV infection. Samples were collected at 12, 24, and 48 h post-infection. Viral RNA copy numbers were quantified by RT-qPCR, N protein expression was analyzed by Western blotting, and viral infection rate and titers were determined by IFA andtissue culture infective dose 50 ( TCID 50 ) assay (P < 0.01; Fig. 4J through M). The results showed that silencing of CLTC significantly inhibited BCoV infection. Subsequently, the impact of CLTC silencing on BCoV attachment and entry was evaluated. RT-qPCR analysis revealed a significant reduction in viral RNA copy numbers during the entry phase (P < 0.001; Fig. 4N), but not in the attachment phase (P > 0.05; Fig. 4O). Collectively, these results indicate that BCoV entry into HRT-18 cells requires CME.
## BCoV entry into HRT-18 cells is caveolin-mediated endocytosis and macropi nocytosis independent
Caveolin-mediated endocytosis (CavME) is another important endocytic pathway, different from CME (24), that was investigated for its potential role in BCoV infection. To determine whether BCoV utilizes CavME, we employed the CavME inhibitor nystatin. A CCK-8 assay identified 80 µM nystatin as the maximum non-cytotoxic concentration (Fig. 5A). Next, HRT-18 cells were pretreated with nystatin (80 µM) for 24 h and infected with BCoV (MOI = 1) to assess the role of CavME in viral infection. Cell samples and supernatants were collected at 12, 24, and 48 h post-infection. Analyses were conducted using RT-qPCR, Western blotting, IFA, and TCID 50 . The results suggest that nystatin had no significant impact on BCoV proliferation (P > 0.05; Fig. 5B through E), suggesting that BCoV entry into HRT-18 cells does not rely on CavME.
Macropinocytosis, a non-specific endocytic pathway, has been implicated in viral entry for certain pathogens. To investigate whether micropinocytosis contributes to BCoV entry, the macropinocytosis inhibitor blebbistatin was utilized (25). Cells were pretreated with maximum safe concentration of blebbistatin (20 µM) for 24 h and subsequently infected with BCoV (MOI = 1). The BCoV gene copy numbers, protein expression levels, viral infection numbers, and progeny viral titers were measured by RT-qPCR, Western blot, IFA, and TCID 50 after infection for 12, 24, and 48 h. The results demonstrated no significant impact of blebbistatin treatment on BCoV propagation (P > 0.05; Fig. 6A through E). These findings collectively confirm that blebbistatin did not affect BCoV infection and indicate that BCoV entry into HRT-18 cells is independent of micropinocytosis.
## The effect of cholesterol on BCoV entry into HRT-18 cells
Cholesterol plays an essential role in the viral life cycle, and many coronaviruses require cholesterol in the viral envelope or host cell membrane for cellular entry (26). Here, methyl-β-cyclodextrin (MβCD), a cholesterol depletion agent, was used to analyze the role of cholesterol in BCoV infection and entry. The CCK-8 assays confirmed that 2.5 mg/mL MβCD exhibited no cytotoxicity in HRT-18 cells (Fig. 7A). Cells were pretreated with MβCD (2.5 mg/mL) for 24 h and infected with BCoV (MOI = 1). Viral infection was evaluated at 12, 24, and 48 h post-infection by quantifying viral RNA copy numbers, N protein expression, infection rates, and progeny virus titers (P < 0.05; Fig. 7B through E), indicating that cholesterol is required for BCoV infection. To assess the role of choles terol in BCoV entry and attachment, cells were pretreated with MβCD (2.5 mg/mL) for 24 h and then infected with BCoV. RT-qPCR analysis revealed that cholesterol depletion significantly reduced BCoV entry (P < 0.05; Fig. 7F) but had no effect on viral attachment (P > 0.05; Fig. 7G), indicating that BCoV entry into HRT-18 cells is cholesterol-dependent.
## Microtubule is required for BCoV entry into HRT-18 cells
Upon internalization, viruses traffic through endosomes along microtubule tracks. To investigate whether BCoV hijacks microtubules for intracellular transport, we disrup ted microtubule polymerization using colchicine, a potent microtubule depolymerizing agent. Firstly, the maximum safe concentration of colchicine was determined to be 600 nM by CCK-8 assay (Fig. 8A). Confocal microscopy revealed that 600 nM colchicine induced microtubule depolymerization, disrupting the compact microtubule network into fragmented structures (Fig. 8B). HRT-18 cells were pretreated with 600 nM colchicine for 24 h prior to infection with BCoV (MOI = 1). Viral replication was evaluated at 12, 24, and 48 h post-infection via RT-qPCR (viral RNA copy numbers), Western blotting (N protein expression), IFA (infection rates), and TCID 50 (progeny titers). Colchicine treatment significantly reduced BCoV propagation (P < 0.01; Fig. 8C through F). RT-qPCR analysis revealed that colchicine treatment significantly reduced BCoV entry (P < 0.05; Fig. 8G) but had no effect on viral attachment (P > 0.05; Fig. 8H), indicating that microtubule integrity is essential for BCoV entry into HRT-18 cells but dispensable for attachment.
## Cathepsins are essential for BCoV entry into HRT-18 cells
To investigate the role of host cell cathepsins in BCoV entry, HRT-18 cells were treated with E64d, a broad-spectrum cathepsin inhibitor. E64d inhibits host cell cathepsins and thereby prevents the cleavage of the coronavirus S protein, which interferes with membrane fusion between the virus and the endosomal membrane, ultimately blocking viral entry. Firstly, CCK-8 assays determined that the maximum non-cytotoxic concentra tion of E64d in HRT-18 cells was 80 µM (Fig. 9A). Based on this result, HRT-18 cells were treated with 80 µM E64d for 24 h and subsequently infected with BCoV (MOI = 1) for 12, 24, and 48 h. Compared to untreated controls, E64d-treated cells showed significantly decreased viral RNA copy numbers, N protein expression, infection rates, and progeny virus titers as determined by RT-qPCR, Western blotting, IFA, and TCID 50 (P < 0.05; Fig. 9B through E), suggesting that E64d effectively suppressed BCoV infection in HRT-18 cells. Then, cells were pretreated with E64d (80 µM) for 24 h and infected with BCoV (MOI = 5). Cell samples were collected during the attachment phase and at 1, 2, and 3 h post-entry. RT-qPCR analysis revealed that viral RNA copy numbers in the E64d-treated group were significantly reduced at all post-entry time points compared with the untreated group (P < 0.01; Fig. 9F), while no significant difference was observed during the attachment phase (P > 0.05; Fig. 9G). Together, the results demonstrated that cathepsin activity was essential for BCoV entry but dispensable for viral attachment.
## BCoV entry into HRT-18 cells is independent of TMPRSS2 activity
To further investigate whether BCoV could utilize serine protease-mediated activation at the plasma membrane to enter HRT-18 cells, we focused on TMPRSS2, a host protease known to cleave the coronavirus S protein at the cell surface. Firstly, CCK-8 assays confirmed that camostat, a TMPRSS2 inhibitor, was non-cytotoxic to HRT-18 cells at concentrations up to 100 µM (Fig. 10A). HRT-18 cells were then pretreated with 100 µM Compared to the untreated group, camostat treatment did not result in significant differences in viral RNA copy numbers, N protein expression levels, or infectious titers at any of the examined time points (P > 0.05; Fig. 10B through E), indicating that TMPRSS2 inhibition does not affect BCoV infection. To further assess the role of TMPRSS2, we examined the effects of camostat on BCoV attachment and entry and found that camostat had no impact on either process (P > 0.05; Fig. 10F andG). These results suggest that BCoV entry into HRT-18 cells does not depend on TMPRSS2 activity.
## Live-cell imaging reveals that CPZ and SSAA09E3 block BCoV entry
Based on our previous findings that BCoV entry is dependent on CME and membrane fusion, we sought to further validate these pathways by directly visualizing viral entry. To this end, three experimental groups were established: a CPZ-treated group to inhibit CME, an SSAA09E3-treated group to block membrane fusion, and an untreated control group. HRT-18 cells were pretreated with CPZ or SSAA09E3 for 24 h, followed by labeling of the cell membrane with 3,3´-dioctadecyloxacarbocyanine perchlorate (DiO). The virus was labeled with the lipophilic dye 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindodicarbocya nine (DiD). DiD-labeled BCoV was subsequently added and incubated with the cells at 4°C for 1 h. After incubation, the viral entry process was monitored for 5 min using a high-speed super-resolution confocal laser scanning microscope. In the untreated control group, most DiD-labeled BCoV particles gradually traversed the cell membrane and entered the cytoplasm (Fig. 11). In contrast, in both the CPZ-and SSAA09E3-treated groups, the majority of viral particles remained associated with the cell surface and failed to enter the cytoplasm (Fig. 11). These findings provide strong evidence that CPZ and SSAA09E3 effectively inhibited BCoV entry, thereby confirming the involvement of CME and membrane fusion in the entry process.
## The role of Rab proteins in BCoV infection of HRT-18 cells
Small GTPase Rab proteins play critical roles in viral entry by regulating endosomal trafficking. Small GTPases, such as Rab5, Rab7, and Rab11, are key regulators of early endosomes, late endosomes, and recycling endosomes, respectively (16). Here, interference with the expression of Rab5, Rab7, and Rab11 was used to verify the role of Rab-mediated endosomal transport in BCoV entry. Firstly, the silencing efficiency was determined by Western blotting (Fig. 12A through C). Cells were transfected with siRab5, siRab7, siRab11, or siNC for 48 h, followed by BCoV infection (MOI = 1). Viral proliferation was comprehensively evaluated at 12, 24, and 48 h post-infection using RT-qPCR, Western blotting, IFA, and TCID 50 . The results showed that the BCoV gene copy numbers, protein expression levels, viral infection numbers, and progeny viral titers were significantly decreased in Rab7-(P < 0.001; Fig. 12D through G) and Rab11-silenced cells (P < 0.05; Fig. 12H through K), but not in Rab5-silenced cells (P > 0.05; Fig. 12L through O). To further analyze BCoV entry, the RT-qPCR analysis revealed that Rab7 and Rab11 silencing significantly reduced viral RNA copy numbers (P < 0.05; Fig. 12Q andR),
## DISCUSSION
As obligate intracellular parasites, viral infectivity hinges on breaching host defense barriers and establishing efficient entry pathways (27). Elucidating the mechanisms of viral entry into cells is not only crucial for understanding the pathogenic mechanisms of viral infections but also contributes to the development of antiviral drugs. However, the way of BCoV entry into cells has not been reported. In this study, we provide the first evidence that BCoV enters HRT-18 cells via the membrane fusion and CME, requiring an acidic environment, dynamin, cholesterol, microtubules, cathepsins, and Rab7-and Rab11-mediated endosomal trafficking. Currently, viral entry mechanisms are primarily categorized into two classical pathways: membrane fusion and receptor-mediated endocytosis (10). Coronaviruses, including SARS-CoV-2 (12), PEDV (15), and mouse hepatitis virus type 2 (28) enter cells via CME and membrane fusion. Based on this precedent, we hypothesized that BCoV enters HRT-18 cells via endocytosis and membrane fusion. Following endocytosis, an (30). CME is a classical endocytic pathway and a major route for viral entry into cells. In this process, upon binding to host cell receptors, the virus triggers the recruitment of clathrin complexes to the inner side of the plasma membrane. This recruitment leads to the formation of invaginated structures that encapsulate viral particles, facilitating their internalization (31). To direct visualization of the effect of CPZ and SSAA09E3 on BCoV entry, live-cell imaging was performed and showed that both treatments inhibited the entry of DiD-labeled BCoV particles, which remained on the cell surface, further confirming that both CME and membrane fusion are essential for BCoV entry. Dynamin is a GTPase that assembles into a ring at the neck of the invaginated pit, and through its GTPase activity, it pinches off the endocytic vesicle from the cell membrane to form a clathrin-and caveolae-coated vesicle (32). Studies have shown that dynamin is required for the entry of multiple viruses, such as African swine fever virus (33), herpes simplex virus type 1 (HSV-1) (34), and rabies virus (35). Given the role of CME in BCoV entry, the role of dynamin in this process was further investigated. In this study, the use of a dynamin inhibitor and dynamin-targeting siRNA revealed that dynamin plays an important role in BCoV invasion of HRT-18 cells. Emerging evidence has revealed that multiple viruses, including SARS-CoV-2 (15, 36), human immunodeficiency virus (HIV) (37), Seneca Valley virus (SVV) (38), influenza virus (39), and Newcastle disease virus (40), enter cells via more than one endocytic pathway. It has also been found that such viruses invade different susceptible cells in different ways. For example, PDCoV enters PK-15 cells through the CavME (13), while in IPI-2I cells, it exploits both macropinocytosis and CME (14). Classical swine fever virus invades the porcine alveolar macrophage cell line 3D4/21 via CavME (41), but invades PK-15 cells via CME (42). To investigate whether CavME and micropinocytosis contribute to BCoV entry into HRT-18 cells, we treated cells with nystatin (a CavME inhibitor) and blebbistatin (a macropinocytosis inhibitor) and showed these two inhibitors did not block virus entry, suggesting that BCoV entry into HRT-18 cells was dependent on CME rather than CavME and micropinocytosis.
Cholesterol plays a crucial role in the entry of enveloped and non-enveloped viruses through diverse mechanisms (43). For influenza virus (44) and dengue virus (45), cholesterol is involved in the formation of lipid rafts in the cell membrane, which facilitates precise docking of the virus with cell surface receptors. For hepatitis C virus (46) and lymphocytic choriomeningitis virus (47), cholesterol influences membrane fluidity and rigidity, promoting membrane invagination and the formation of endocytic vesicles that encapsulate the virus during entry. Additionally, cholesterol regulates the fluidity and permeability of both the cell membrane and the viral envelope, enhancing membrane fusion. This mechanism is observed in viruses such as HIV (37) and SARS-CoV-2 (36) during their entry into host cells. Given the critical role of cholesterol in the entry of many viruses, it is reasonable to investigate whether cholesterol similarly contributes to BCoV entry. Here, the cholesterol-depleting agent MβCD was used to assess its potential impact on BCoV entry into HRT-18 cells. The results revealed that depletion of cell membrane cholesterol significantly blocked BCoV entry.
The microtubule plays indispensable roles in viral entry through two distinct mechanisms. First, viruses utilize microtubules to facilitate binding to host cell surface receptors (48). Multiple viruses, including vaccinia virus (49), HSV-1 (50), West Nile BCoV for 1 h, 2 h, and 3 h, and BCoV gene copy number was measured by RT-qPCR. (Q, R) Rab7 and Rab11 silencing inhibits BCoV entry. After silencing of Rab7 and Rab11, cells were infected with BCoV for 1 h, 2 h, and 3 h, and BCoV gene copy number was measured by RT-qPCR. (S, T, U) RT-qPCR was used to evaluate the effects of Rab7-, Rab11-, and Rab5-targeting siRNAs on BCoV attachment. Data are presented as the mean ± SD of three independent experiments (not significant, P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001). virus (WNV) (51), and respiratory syncytial virus (52), hijack microtubules for directional trafficking during entry. Our investigation demonstrated that pretreatment with the microtubule-disrupting agent colchicine significantly inhibited BCoV entry into HRT-18 cells, suggesting that microtubules are essential for BCoV entry.
Cathepsins are lysosomal cysteine proteases that facilitate the entry of various enveloped viruses by cleaving viral glycoproteins within endosomes. This proteolytic activation is essential for triggering membrane fusion between the viral envelope and the host endosomal membrane, enabling the release of the viral genome into the cytoplasm (53). Several viruses, including Ebola virus (54), SARS-CoV (55), and MERS-CoV (56), rely on cathepsin-mediated cleavage of their surface glycoproteins to mediate endosomal fusion. In our study, treatment of HRT-18 cells with E64d, a broad-spectrum cathepsins inhibitor, significantly inhibited BCoV entry. These findings indicate that cathepsin activity is essential for BCoV entry, likely by enabling spike protein activation and membrane fusion within endosomes.
TMPRSS2 facilitates viral entry by cleaving and activating S proteins at the plasma membrane, thereby promoting direct fusion between the viral envelope and the host cell membrane. However, in this study, inhibition of TMPRSS2 activity with camostat did not suppress viral entry or infection, indicating that TMPRSS2 was not required for BCoV-mediated membrane fusion. This entry mechanism is consistent with reports on other coronaviruses that do not efficiently exploit TMPRSS2 and instead rely on endosomal proteases for S protein activation (57,58).
Endosomal trafficking represents a critical pathway for viral internalization following entry, encompassing sequential progression through early endosomes, late endosomes, and recycling endosomes (16). These endosomes are mediated by Rab GTPases, with Rab5, Rab7, and Rab11 specifically orchestrating trafficking in early endosomes, late endosomes, and recycling endosomes (16). Coronaviruses, including porcine enteric alphacoronavirus (59) and infectious bronchitis virus (60), exemplify this paradigm by entering cells via clathrin-mediated endocytosis and undergoing sequential maturation through early and late endosomes. Here, silencing of Rab5, Rab7, and Rab11 by siRNA was performed to investigate the effect of endosomal trafficking on BCoV entry. The results showed that silencing of Rab7 and Rab11 potently inhibited BCoV entry into HRT-18 cells, while siRab5 had no significant effect on BCoV entry. Laser confocal microscopy further revealed the colocalization of BCoV with Rab7 and Rab11 during viral entry steps, confirming that BCoV traffics through late and recycling endosomes. Previous studies reported that porcine sapelovirus entry requires Rab7-dependent late endosomes and Rab11-dependent recycling endosomes (61), which is consistent with our findings. Notably, emerging studies revealed that SVV (62) and bluetongue virus (63) exclusively hijack late endosomes (Rab7) during entry, bypassing early endosomes. In summary, this study revealed that BCoV enters HRT-18 cells via membrane fusion and CME; after endocytosis, the virion is transported from late and recycling endosomes but not early endosomes. Dynamin, cholesterol, microtubules, cathepsins, and an acidic environment are also required for BCoV entry.
## MATERIALS AND METHODS
## Cells and virus
HRT-18 cells were obtained from the American Type Culture Collection and cultured in Dulbecco's Modified Eagle's Medium (DMEM) (Gibco, Grand Island, NY, USA) supplemen ted with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA) and 1% penicillinstreptomycin (Sigma-Aldrich, St. Louis, MO, USA). The cells were maintained in a 37°C incubator with 5% CO 2 . Bovine coronavirus (GenBank: OP866728.1) was isolated and preserved in our laboratory.
## Inhibitors
The inhibitors used in this study included SSAA09E3 (Cat HY-138102, MedChemExpress), a novel inhibitor that blocks the fusion of the viral membrane with the host cell membrane; CPZ (Cat C0982, Sigma), a clathrin-mediated endocytosis inhibitor; nystatin (Cat 475914, Sigma), a caveolae inhibitor that acts as a sterol-binding agent disrupting caveolae; blebbistatin (Cat 203391, Sigma), an inhibitor of micropinocytosis; dynasore (Cat T1848, TargetMol), a dynamin inhibitor; MβCD (Cat T4072, TargetMol), a cholesterol depletion inhibitor; chloroquine (Cat S6999, Selleck) and NH 4 Cl (Cat A9434, Sigma), a potent inhibitor of V-ATPase and a specific inhibitor of acidification of endosomal vesicles; colchicine (Cat HY-16569, MedChemExpress), which inhibits the polymerization of tubulin; E64d (Cat S7393, Selleck), a cathepsin inhibitor; and camostat (Cat HY-13512, MedChemExpress), a TMPRSS2 inhibitor.
## Cell viability assay
Cell viability was assessed using the CCK-8 assay. HRT-18 cells were seeded into 96-well plates and inoculated with SSAA09E3 (5, 10, 20, 40 µM), CPZ (5, 10, 20, 40 µM), nystatin (10, 20, 40, 80 µM), blebbistatin (5, 10, 20, 40 µM), dynasore (5, 10, 20, 40 µM), NH 4 Cl (10, 20, 40, 80 mM), chloroquine (20,40,80, 100 µM), MβCD (0.63, 1.25, 2.5, 5 mg/mL), colchicine (200, 400, 600, 800 nM), E64d (40, 60, 80, 100 µM), or camostat (40,80, 100, 120 µM) for 48 h at 37°C. Then, 10 µL of CCK-8 reagent was added to each well for 1 h at 37°C. The absorbance values were recorded at 450 nm using an Infinite M200 Pro system (Tecan, Männedorf, Switzerland). A viability of over 90% without morphological changes was considered the safe concentration.
## Western blotting
Proteins were extracted from HRT-18 cells using RIPA lysis buffer (Cat P0013B, Beyotime) and quantified using a BCA Protein Quantification Kit (Cat P0012, Beyotime). Protein samples were separated by SDS-PAGE and transferred onto a 0.22 µm PVDF membrane (Cat ISEQ00010, Merck Millipore). The membrane was blocked at room temperature for 2 h in TBST containing 5% non-fat milk. The primary antibodies used were GAPDH (Cat YM3029, Immunoway), Rab5 (Cat 66339-1-Ig, Proteintech), Rab7 (Cat R8779-25UL, Sigma-Aldrich), and Rab11 (Cat 67902-1-Ig, Proteintech). The anti-N protein antibody of BCoV was prepared and stored in our laboratory. The primary antibody was incuba ted overnight at 4°C. The HRP-conjugated goat anti-mouse/rabbit IgG (H+L) secondary antibody (Cat RS0001/RS0002, Immunoway) was incubated at room temperature for 1 h at a dilution of 1:12,000. After washing the membrane with TBST, ECL detection was performed. Protein bands were detected and analyzed using the Amersham Image Quant 800 Western blot imaging system (Cytiva, Sweden), and grayscale analysis was performed using ImageJ.
## TCID 50
The cell supernatants were serially diluted and used to infect HRT-18 cells in 96-well plates. The cells were incubated at 37°C in a 5% CO 2 incubator for 48 h. The cells were fixed with 4% paraformaldehyde for 20 min, washed with PBS, and then permeabilized with 0.3% Triton X-100 for 15 min. Afterward, the cells were blocked with 5% skim milk for 3 h. The BCoV N protein antibody (1:100) was incubated overnight at 4°C, and the goat anti-rabbit IgG (H+L) (Alexa Fluor 488) (Cat RS3211, Immunoway) antibody (1:200) was incubated for 1 h at room temperature in the dark. Nuclei were stained with Hoechst 33342 for 10 min. The fluorescence-positive wells were observed under an inverted fluorescence microscope (Axio Observer, ZEISS, Germany). Finally, the TCID 50 was calculated by the Reed-Muench method.
## Confocal microscopy
HRT-18 cells were seeded onto cell culture coverslips in 24-well plates and incubated for 12 h at 37°C in a 5% CO 2 incubator. (i) Cells were infected with BCoV for 1, 2, and 3 h, and the co-localization of CLTC, Rab7, and Rab11 with BCoV was detected. (ii) Cells were infected with BCoV for 24 and 48 h, and the co-localization of Rab7 and Rab11 with BCoV was detected. (iii) HRT-18 cells were treated with colchicine for 24 h to verify its inhibition of microtubule polymerization.
Cells were then washed three times with PBS and fixed with 4% paraformaldehyde at room temperature. Cells were then permeabilized with 0.3% Triton X-100 and blocked with 5% skim milk. The cells were incubated with BCoV antibody and mouse anti-Rab7 (Cat R8779-25UL, Sigma-Aldrich), mouse anti-Rab11 (Cat 67902-1-Ig, Proteintech), mouse anti-tubulin (Cat 66240-1-Ig, Proteintech), or mouse anti-CLTC (Cat 66339-1-Ig, Protein tech) overnight at 4°C. After three washes with PBS, the secondary antibodies Alexa Fluor 488 Anti-Mouse and Alexa Fluor 594 Anti-Rabbit were incubated for 1 h at room temperature in the dark. Subsequently, cells were incubated with DAPI (Cat BL105A, Biosharp) at 37°C for 10 min and washed three times with PBS. Finally, images were captured using a Leica TCS SP8 laser scanning confocal microscope (LSM510 META, ZEISS, Germany).
## Real-time quantitative PCR
Total RNA was extracted using the TRIzol method, and the RNA concentration was measured using a NanoDrop One (Thermo Fisher Scientific, Waltham, MA, USA). The RNA was reverse transcribed into cDNA using the Evo M-MLV Reverse Transcription Kit (Cat AG11728, Accurate Biology) according to the manufacturer's instructions. GAPDH was used as the internal reference gene, and the expression changes of BCoV, Rab5, Rab7, Rab11, dynamin, and CLTC genes were analyzed. Primer sequences are shown in Table 1. Real-time quantitative PCR was performed using the SYBR Green Pro Taq HS Premix qPCR Kit (Cat AG11701, Accurate Biology). The reaction mixture had a final volume of 20 µL, with the following amplification program: 95°C for 30 seconds, followed by 40 cycles of 95°C for 5 seconds and 60°C for 30 seconds. The melt curve analysis was conducted using the CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). Data were analyzed using the 2 -ΔΔCT method.
## siRNA and transfection
The siRNA was synthesized by Qingke Bio (Beijing, China), with three siRNA sequences designed for each target gene (dynamin, CLTC, Rab5, Rab7, and Rab11). The sequences are shown in Table 2. The siNC sequence was supplied and synthesized by the com pany (Qingke Bio, Beijing, China). HRT-18 cells were seeded into six-well plates, and transfection was performed when the cell confluence reached 60%. A total of 1.6 mL of DMEM medium containing 10% FBS was added to each well. Two sterile, clean centrifuge tubes were prepared. Then, 125 µL of Opti-MEM medium was added to each tube, and 100 pmol of siRNA was added to one tube, which was mixed gently using a pipette. To the second tube, 5 µL of Lipo6000 (Cat C0526FT, Beyotime) transfection reagent was added and mixed gently. After incubation at room temperature for 5 min,
## Primer
Sequence (5´-3´) the siRNA-containing solution was gently added to the Lipo6000 reagent-containing solution and mixed gently. The mixture was then incubated at room temperature for 15 min. The cells were incubated at 37°C for 48 h, and the interference efficiency was verified by Western blot. The siRNA with the highest interference efficiency was used for subsequent experiments.
$$GAPDH-F AAGGCTGTGGGCAAGG GAPDH-R TGGAGGAGTGGGTGTCG BCoV-F CGTTCTGGTAATGGCATCCTTA BCoV-R GTTTGCTTGGGTTGAGCTCTTCTA Rab5-F CCGACCTAGCAAATAAAAGAGC Rab5-R AGCGGATGTCTCCATGAATAAT Rab7-F GCATCCTAGCTTTTGATGTCAC Rab7-R CATTCTTGGCACTGACTTCAAA Rab11-F TTGTGGGCAATAAGAGTGATCT Rab11-R TGGAACATGAATAGGAACCACA CLTC-F CAATGGACCAAATAATGC CLTC-R GTGAACCAGGGTAGATGC Dynamin-F CAAGGATGAGGAGGAGAAAG Dynamin-R GTGTTGAAGATGGCGAAGA$$
## Virus invasion assay
For the virus attachment assay, cells were pretreated for 24 h with the maxi mum non-cytotoxic concentrations of chemical inhibitors (SSAA09E3, chlorpromazine, nystatin, blebbistatin, dynasore, NH 4 Cl, MβCD, chloroquine, colchicine, E64d, and camostat). Alternatively, cells were transfected with siRNAs targeting clathrin, dynamin, Rab5, Rab7, or Rab11. Following pretreatment, the cells were incubated with BCoV at an MOI of 5 at 4°C for 1 h to permit viral attachment without internalization. Unbound virus was removed by washing the cells three times with ice-cold PBS. Total RNA was then extracted, and viral genomic RNA copy numbers were quantified by RT-qPCR to assess the level of virus attachment.
For the viral entry assay, cells were pretreated for 24 h with the maximum non-cyto toxic concentrations of the respective compounds or transfected with siRNAs. Following pretreatment, the cells were infected with BCoV at a MOI of 5 and incubated at 4°C for 1 h to allow viral attachment without internalization. Unbound virus was removed by washing the cells three times with ice-cold PBS, after which the cells were incubated at 37°C for 1, 2, or 3 h to permit viral internalization. Subsequently, the cells were washed again three times with ice-cold PBS, and total RNA was extracted. Viral genomic RNA copy numbers were subsequently quantified by RT-qPCR.
## Fluorescent labeling and imaging of BCoV entry into HRT-18 cells
The virus was labeled with the lipophilic dye DiD (Thermo Fisher, V22887, USA), which incorporates into viral envelopes. DiD was added to the viral suspension and incuba ted at room temperature for 120 min. Unincorporated dye was subsequently removed using NAP-10 desalting columns (Cytiva, GE Healthcare). The DiD-labeled BCoV was then filtered through a 0.22 µm membrane filter (Millipore) and stored at -80°C until use in viral tracking assays.
To enable direct visualization of BCoV entry, HRT-18 cell membranes were labeled with the lipophilic dye DiO (Thermo Fisher, V22886, USA). After removal of excess dye, DiD-labeled BCoV was added and incubated with the cells at 4°C for 1 h. The entry process was subsequently imaged using a high-speed, super-resolution confocal laser scanning microscope (LSM980 with Airyscan2, Carl Zeiss Microscopy GmbH, Germany) equipped with a 63× oil immersion objective (NA 1.4).
## Flow cytometry and fluorescence microscopy for quantitative and visual assessment of BCoV-induced syncytium formation
Cell membranes were labeled with the lipophilic dye DiD, which integrates into the lipid bilayer. Upon membrane fusion between adjacent cells, the labeled membranes merge, resulting in larger multinucleated cells (syncytia) that exhibit enhanced and aggregated DiD fluorescence signals. Flow cytometry detects these changes as an increase in mean fluorescence intensity (MFI), reflecting the extent of membrane fusion. This method enables high-throughput, quantitative analysis of fusion events at the single-cell level. The experiment included six groups: (i) uninfected and untreated, (ii) uninfected and treated with SSAA09E3, (iii) BCoV-infected for 3 h without SSAA09E3, (iv) BCoV-infec ted for 3 h with SSAA09E3, (v) BCoV-infected for 24 h without SSAA09E3, and (vi) BCoV-infected for 24 h with SSAA09E3. Following treatment, HRT-18 cells were labeled with DiD to stain the plasma membrane, washed to remove excess dye, digested with trypsin (without EDTA), and analyzed by flow cytometry (BD high-speed cell sorter, APC channel). Data were processed using FlowJo v.10.
To directly visualize syncytium formation, fluorescence microscopy was performed. HRT-18 cells were allocated into three experimental groups as follows: (i) uninfected control cells (mock group), (ii) BCoV-infected cells, and (iii) BCoV-infected cells with SSAA09E3 treatment. Syncytium formation was assessed by observing overall morphol ogy with bright-field microscopy and by examining the BCoV N protein and nuclei with fluorescence microscopy.
## References
1. Saif (2010) "Bovine respiratory coronavirus" *Vet Clin North Am Food Anim Pract*
2. Lim, Choi, Lim et al. (2013) "Complete genome analysis of canine respiratory coronavirus" *Genome Announc*
3. Bidokhti, Tråvén, Ohlson et al. (2012) "Tracing the transmission of bovine coronavirus infections in cattle herds based on S gene diversity" *Vet J*
4. Chouljenko, Lin, Storz et al. (2001) "Comparison of genomic and predicted amino acid sequences of respiratory and enteric bovine coronaviruses isolated from the same animal with fatal shipping pneumonia" *J Gen Virol*
5. Clark (1993) "Bovine coronavirus" *Br Vet J*
7. Gupta, Singh, Singh et al. (2023) "Recent developments and future perspectives of vaccines and therapeutic agents against SARS-CoV2 using the BCOV_S1_CTD of the S protein" *Viruses*
8. Malik (2020) "Properties of coronavirus and SARS-CoV-2" *Malays J Pathol*
9. Hurst, Ye, Goebel et al. (2010) "An interaction between the nucleocapsid protein and a component of the replicasetranscriptase complex is crucial for the infectivity of coronavirus genomic RNA" *J Virol*
10. De Groot (2006) "Structure, function and evolution of the hemaggluti nin-esterase proteins of corona-and toroviruses" *Glycoconj J*
11. Wędrowska, Wandtke, Senderek et al. (2020) "Coronaviruses fusion with the membrane and entry to the host cell" *Ann Agric Environ Med*
12. (2025) *Full-Length Text Journal of Virology*
13. Wang, Li, Wang et al. (2020) "Dynamics of transmissible gastroenteri tis virus internalization unraveled by single-virus tracking in live cells" *FASEB J*
14. Wei, She, Wu et al. (2020) "PEDV enters cells through clathrin-, caveolae-, and lipid raft-mediated endocytosis and traffics via the endo-/lysosome pathway" *Vet Res*
15. Li, Xiao, Zhao et al. (2022) "Porcine deltacoronavirus (PDCoV) entry into PK-15 cells by caveolae-mediated endocytosis" *Viruses*
16. Fang, Zhang, Zhang et al. (2021) "Porcine deltacoronavirus enters porcine IPI-2I intestinal epithelial cells via macropinocytosis and clathrin-mediated endocytosis dependent on pH and dynamin" *J Virol*
17. Inoue, Tanaka, Tanaka et al. (2007) "Clathrin-dependent entry of severe acute respiratory syndrome coronavirus into target cells expressing ACE2 with the cytoplasmic tail deleted" *J Virol*
18. Mercer, Schelhaas, Helenius (2010) "Virus entry by endocytosis" *Annu Rev Biochem*
19. Dimitrov (2004) "Virus entry: molecular mechanisms and biomedical applications" *Nat Rev Microbiol*
20. Li, Moore, Vasilieva et al. (2003) "Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus" *Nature*
21. Benarroch, Juttukonda, Sabharwal et al. (2019) "Differential expression of Rab5 and Rab7 small GTPase proteins in placental tissues from pregnancies affected by maternal coronavirus disease" *Clin Ther*
22. Liu, Zhang, Li et al. (2017) "Rab5 and Rab11 are required for clathrin-dependent endocytosis of Japanese encephalitis virus in BHK-21 cells" *J Virol*
23. Lajoie, Nabi (2010) "Lipid rafts, caveolae, and their endocytosis" *Int Rev Cell Mol Biol*
24. Moreno-Layseca, Jäntti, Godbole et al. (2022) "Publisher Correction: Cargo-specific recruitment in clathrin-and dynaminindependent endocytosis" *Nat Cell Biol*
25. Mettlen, Chen, Srinivasan et al. (2018) "Regulation of clathrin-mediated endocytosis" *Annu Rev Biochem*
26. Sieczkarski, Whittaker (2002) "Dissecting virus entry via endocyto sis" *J Gen Virol*
27. Caì, Postnikova, Bernbaum et al. (2015) "Simian hemorrhagic fever virus cell entry is dependent on CD163 and uses a clathrin-mediated endocytosis-like pathway" *J Virol*
28. Christian, Haynes, Phillips et al. (1997) "Use of cyclodextrins for manipulating cellular cholesterol content" *J Lipid Res*
29. Sieczkarski, Whittaker (2005) "Viral entry" *Curr Top Microbiol Immunol*
30. Pu, Zhang (2008) "Mouse hepatitis virus type 2 enters cells through a clathrin-mediated endocytic pathway independent of Eps15" *J Virol*
31. Doyle, Busey, Iobst et al. (2024) "Endosomal fusion of pH-dependent enveloped viruses requires ion channel TRPM7" *Nat Commun*
32. Mas, Melero (2024) "Entry of enveloped viruses into host cells: membrane fusion" *Subcell Biochem*
33. Mcmahon, Boucrot (2011) "Molecular mechanism and physiological functions of clathrin-mediated endocytosis" *Nat Rev Mol Cell Biol*
34. Henley, Krueger, Oswald et al. (1998) "Dynaminmediated internalization of Caveolae" *J Cell Biol*
35. Hernaez, Alonso (2010) "Dynamin-and clathrin-dependent endocytosis in African swine fever virus entry" *J Virol*
36. Praena, Bello-Morales, Guerrero (2020) "Hsv-1 endocytic entry into a human oligodendrocytic cell line is mediated by clathrin and dynamin but not caveolin" *Viruses*
37. Gao, Wang, Zhao et al. (2019) "Entry of Challenge Virus Standard (CVS) -11 into N2a cells via a clathrinmediated, cholesterol-, dynamin-, pH-dependent endocytic pathway" *Virol J*
38. Palacios-Rápalo, Jesús-González, Cordero-Rivera et al. (2021) "Cholesterol-rich lipid rafts as platforms for SARS-CoV-2 entry" *Front Immunol*
39. Melikyan (2014) "HIV entry: a game of hide-and-fuse?" *Curr Opin Virol*
40. Liu, Shang, Ren et al. (2023) "Seneca Valley virus enters cells through multiple pathways and traffics intracellularly via the endolysosomal pathway" *J Gen Virol*
41. Sun, Whittaker (2013) "Entry of influenza virus" *Adv Exp Med Biol*
42. Zhao, Shi, Han et al. (2021) "Newcastle disease virus entry into chicken macrophages via a pHdependent, dynamin and caveola-mediated endocytic pathway that requires Rab5" *J Virol*
43. Ning, Gao, Zhou et al. (2016) "Caveolin-1-mediated endocytic pathway is involved in classical swine fever virus Shimen infection of porcine alveolar macrophages" *Vet Microbiol*
44. Shi, Liu, Zhou et al. (2016) "Entry of classical swine fever virus into PK-15 cells via a pH-, dynamin-, and cholesterol-dependent, clathrin-mediated endocytic pathway that requires Rab5 and Rab7" *J Virol*
45. Li, Yang, Mao et al. (2020) "Cholesterol-rich lipid rafts both in cellular and viral membrane are critical for caprine parainfluenza virus type3 entry and infection in host cells" *Vet Microbiol*
46. Sun, Whittaker (2003) "Role for influenza virus envelope cholesterol in virus entry and infection" *J Virol*
47. Carro, Damonte (2013) "Requirement of cholesterol in the viral envelope for dengue virus infection" *Virus Res*
48. Chamoun-Emanuelli, Pecheur, Simeon et al. (2013) "Phenothiazines inhibit hepatitis C virus entry, likely by increasing the fluidity of cholesterol-rich membranes" *Antimicrob Agents Chemother*
49. Shah, Peng, Carbonetto (2006) "Role of non-raft cholesterol in lymphocytic choriomeningitis virus infection via alpha-dystroglycan" *J Gen Virol*
50. Maccioni, Cambiazo (1995) "Role of microtubule-associated proteins in the control of microtubule assembly" *Physiol Rev*
51. Ploubidou, Moreau, Ashman et al. (2000) "Vaccinia virus infection disrupts microtubule organization and centrosome function" *EMBO J*
52. (2025) *Full-Length Text Journal of Virology*
53. Kim, Cianfrocco, Verhey et al. (2024) "The HSV-1 pUL37 protein promotes cell invasion by regulating the kinesin-1 motor" *Proc Natl Acad Sci*
54. Makino, Suzuki, Hasebe et al. (2014) "Establishment of tracking system for West Nile virus entry and evidence of microtubule involvement in particle transport" *J Virol Methods*
55. Cosentino, Marougka, Desquesnes et al. (2022) "Respiratory syncytial virus ribonucleoproteins hijack microtubule Rab11 dependent transport for intracellular trafficking" *PLoS Pathog*
56. Bosch, Bartelink, Rottier (2008) "Cathepsin L functionally cleaves the severe acute respiratory syndrome coronavirus class I fusion protein upstream of rather than adjacent to the fusion peptide" *J Virol*
57. Chandran, Sullivan, Felbor et al. (2005) "Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection" *Science*
58. Simmons, Gosalia, Rennekamp et al. (2005) "Inhibitors of cathepsin L prevent severe acute respiratory syndrome coronavirus entry" *Proc Natl Acad Sci*
59. Zhou, Pan, Zhang et al. (2016) "Glycopeptide antibiotics potently inhibit cathepsin L in the late endosome/lysosome and block the entry of Ebola virus, Middle East respiratory syndrome coronavirus (MERS-CoV), and severe acute respiratory syndrome coronavirus (SARS-CoV)" *J Biol Chem*
60. Burkard, Verheije, Wicht et al. (2014) "Coronavirus cell entry occurs through the endo-/lysosomal pathway in a proteolysis-dependent manner" *PLoS Pathog*
61. Millet, Whittaker (2014) "Host cell entry of Middle East respiratory syndrome coronavirus after two-step, furin-mediated activation of the spike protein" *Proc Natl Acad Sci*
62. Chen, Liang, Weng et al. (2023) "Porcine enteric alphacoronavirus entry through multiple pathways (caveolae, clathrin, and macropinocy tosis) requires Rab GTPases for endosomal transport" *J Virol*
63. Wang, Yuan, Sun et al. (2019) "Infectious bronchitis virus entry mainly depends on clathrin mediated endocytosis and requires classical endosomal/lysosomal system" *Virology (Auckl)*
64. Zhao, Cui, Yu et al. (2019) "Porcine sapelovirus enters PK-15 cells via caveolae-dependent endocytosis and requires Rab7 and Rab11" *Virology (Auckl)*
65. Liu, Shang, Ren et al. (2023) "Seneca Valley virus enters cells through multiple pathways and traffics intracellularly via the endolysosomal pathway" *J Gen Virol*
66. Patel, Mohl, Roy (2016) "Entry of bluetongue virus capsid requires the late endosome-specific lipid lysobisphosphatidic acid" *J Biol Chem*
67. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12598670&blobtype=pdf | # Reduced TLR3 and TLR9 Expression in Epidermodysplasia Verruciformis: Evidence From a Comparative Skin Study
Luis Alberto, Ribeiro Fróes, Cibele Conceição, Apóstolos Pereira, Lana Luiza Da Cruz, | Naiura, Vieira Pereira, | Walmar, Roncalli Pereira De Oliveira, Mirian Sotto, Alberto Luis, Fróes Ribeiro, Júnior
## Abstract
Epidermodysplasia verruciformis (EV) is a rare genodermatosis characterized by lifelong β-human papillomavirus (β-HPV) persistence, extensive flat warts, and increased risk of cutaneous squamous cell carcinoma. While TMC6, TMC8, and CIB1 mutations are recognized as genetic drivers, innate immune mechanisms contributing to HPV persistence remain incompletely defined. This study quantified the expression of Toll-like receptors (TLRs) 3, 4, 5, and 9 in normal skin and flat warts from patients with EV and immunocompetent individuals without EV (NEV). We performed immunohistochemical analysis on 135 formalin-fixed, paraffin-embedded specimens using standardized digital morphometry of epidermal keratinocytes. EV patients exhibited significantly reduced TLR3 and TLR9 expression in both normal skin and flat warts relative to controls, whereas TLR4 and TLR5 levels were comparable. Notably, flat warts from NEV individuals showed marked TLR3 upregulation relative to matched normal skin, whereas this response was absent in EV patients. These findings are consistent with an EV-associated epithelial innate-sensing phenotype. Our data suggest innate immune deficiencies may interact with previously described keratinocyte abnormalities, amplifying local immune dysfunction. These findings provide a framework for investigating TLRbased therapeutic approaches in EV.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.
## 1 | Introduction
Toll-like receptors (TLRs) are pattern-recognition receptors that detect conserved microbial motifs and signal through MyD88or TRIF-dependent pathways; in particular, TLR3 recognizes mainly double-stranded RNA, TLR4 recognizes lipopolysaccharide, TLR5 recognizes flagellin, and TLR9 recognizes unmethylated CpG DNA [1][2][3]. Human keratinocytes-key epithelial sentinels in skin-express TLR3, TLR4, TLR5, and TLR9 and mount ligand-induced inflammatory programs in vitro, underscoring that TLR expression in the epidermis is biologically meaningful even outside professional immune cells [4].
Epidermodysplasia verruciformis (EV) is a rare genodermatosis with lifelong susceptibility to cutaneous β-HPV infection and elevated squamous-cell carcinoma risk on sun-exposed skin [5]. EV results from biallelic loss of TMC6/TMC8 or CIB1, with CIB1-EVER1-EVER2 forming a keratinocyte-intrinsic complex for β-HPV restriction [6,7]. Genetic studies distinguish "typical" skin-restricted EV from "atypical" EV with broader immunodeficiency [8]. RNA-seq shows β-HPV transcripts enriched in lesional versus matched normal skin, reinforcing EV as a cutaneous β-HPV model [9].
Given the viral etiology of EV lesions and the role of TLRs in antiviral immunity, understanding TLR expression in these lesions may provide insights into disease pathogenesis. Studies outside EV show altered TLR expression in cutaneous viral lesions: common warts and molluscum contagiosum display increased epidermal TLR3/TLR9 versus normal skin [10]; verruca vulgaris shows elevated dermal TLR9/IRAK1 (interleukin-1 receptor-associated kinase 1) with increased plasmacytoid dendritic cells [11]. Moreover, EV shows altered epidermal differentiation compared with flat warts in non-EV individuals, highlighting disease-specific cutaneous programs [12].
While most TLR-HPV literature focuses on mucosal sites [13][14][15], TLR expression patterns in EV lesions remain uncharacterized, limiting understanding of innate immune dysfunction and therapeutic targets. We hypothesized that EV lesions display distinct TLR expression patterns reflecting the unique environment permitting persistent β-HPV infection. We therefore compared epidermal TLR3, TLR4, TLR5, and TLR9 expression in normal skin and flat warts from individuals with and without EV.
## 2 | Materials and Methods
## 2.1 | Study Design and Sample Selection
This retrospective comparative study was conducted at the Dermatopathology Laboratory, Hospital das Clínicas, Universidade de São Paulo. Formalin-fixed, paraffin-embedded (FFPE) skin samples were retrieved from institutional archives (2010-2024) and reviewed by two dermatopathologists. The study was approved by the Institutional Ethics Committee of Hospital das Clínicas, University of São Paulo, which waived informed consent due to exclusive use of archived, anonymized samples (approval number #50871415.2.0000.0068), in accordance with the Declaration of Helsinki.
Patients were classified as having epidermodysplasia verruciformis (EV) using strict clinicopathological criteria, including the characteristic clinical phenotype, histopathological features showing EV-type cytopathic changes [16]. compatible with β-HPV infection, and longitudinal follow-up confirming a chronic course (≥ 12 months). A subset of EV patients (n = 20) had undergone comprehensive HPV genotyping in a previous study [17], whereas the remaining EV cases fulfilled established diagnostic criteria without molecular characterization. For each participant, samples were categorized as flat wart or as clinically normal skin from unaffected areas; normal skin was obtained from anatomically matched sites to the wart sample within the same individual.
The analysis included 135 samples: 42 from EV patients (17 flat warts, 25 normal skin) and 93 from NEV individuals (21 flat warts, 72 normal skin). Demographic and anatomical data were extracted from medical records and grouped as head/neck, trunk, and limbs (Table 1).
## 2.2 | Immunohistochemistry
Tissue Sections (3-4 μm) from selected FFPE blocks were mounted on silanized slides, dried overnight at 37°C, deparaffinized in xylene, and rehydrated in graded ethanol. Antigen retrieval used citrate buffer (pH 6.0) in a pressure cooker (95°C, 20 min). Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol (10 min).
Slides were incubated overnight at 4°C with primary antibodies: anti-TLR3 (1:100, Abcam), anti-TLR4 (1:150, Santa Cruz Biotechnology), anti-TLR5 (1:100, Novus Biologicals), and anti-TLR9 (1:200, Santa Cruz Biotechnology). Detection employed a polymer-based system (REVEAL Biotin-Free DAB Detection System) with 3,3′-diaminobenzidine as chromogen. Slides were counterstained with Harris hematoxylin, dehydrated, cleared, and mounted with synthetic resin.
Positive controls included normal skin (TLR3, TLR4), molluscum contagiosum (TLR3, TLR5), and cutaneous leishmaniasis (TLR9) [10,18,19]. Negative controls substituted the primary antibody with isotype-matched nonimmune serum. The observed cellular distribution and compartment localization (predominantly epidermal keratinocytes) were concordant with published patterns for these targets [20], supporting specificity in our material. Analysis was restricted to the epidermal keratinocyte compartment; the dermal inflammatory infiltrate was not phenotyped and was typically sparse.
## 2.3 | Digital Image Acquisition and Quantification
Slides were digitized at 40× magnification using the Aperio ScanScope. TLR expression was quantified in ImageJ by color deconvolution and automatic thresholding. Five epidermal fields per sample were analyzed, avoiding artifacts. Results were expressed as percent stained area. Quantification was performed by a blinded researcher and reviewed by a dermatopathologist.
## 2.4 | Statistical Analysis
Analyses used Python and StatsModels. Mann-Whitney U tests followed Shapiro-Wilk assessment. Kruskal-Wallis tests assessed associations with sex and site. Multivariate linear regressions adjusted for diagnosis, age, sex, and site. Significance was set at p < 0.05. TLR3, TLR4, TLR5, and TLR9 expression was detected in the epidermis of both normal skin and flat warts from EV and NEV individuals. In EV flat warts, immunostaining was reduced in koilocytotic keratinocytes compared with adjacent nonkoilocytotic keratinocytes-most notably for TLR3-whereas NEV flat warts exhibited stronger staining in koilocytotic areas (Figure 1). High-magnification views (Figure 2) show that staining predominated in epidermal keratinocytes; dermal inflammatory cells were scarce and were not phenotyped.
## 3 | Results
Kruskal-Wallis testing indicated significant variation among groups for TLR3 (p < 0.001) and TLR9 (p < 0.001), but not for TLR4 (p = 0.485) or TLR5 (p = 0.071). Mann-Whitney U tests revealed lower TLR3 and TLR9 expression in EV patients compared to NEV controls (Figure 3). For TLR3, median values were 4.0% in EV versus 19.0% in NEV (p = 0.008), with similar differences in normal skin (p < 0.001). TLR9 showed median values of 5.6% in EV versus 7.0% in NEV (p = 0.007), extending to normal skin (p = 0.008). TLR4 and TLR5 showed no significant group differences (p = 0.927 and p = 0.499, respectively).
Within-group comparisons showed that NEV patients had significantly higher TLR3 expression in flat warts than in normal skin (p = 0.020), suggesting an adaptive response to viral infection. This upregulation was absent in EV patients (p = 0.167). For TLR9, no significant differences were observed between flat warts and normal skin in either group.
No sex-based differences in TLR expression were found. Expression heterogeneity was notable, with the highest coefficient of variation for TLR3 in the EV group (99.6%), suggesting variable receptor regulation across samples.
Multivariate models including diagnostic group, age, sex, and anatomical site as covariates were constructed for each TLR. EV diagnosis remained significantly associated with lower TLR3 expression (p = 0.025), while no other variables reached significance. For TLR9, NEV status (p = 0.031) and age (p = 0.011) were significant predictors, with the model showing substantial explanatory power (adjusted R² = 0.493). TLR4 and TLR5 models showed limited significance and low explanatory power. Sex had no significant effect in any model.
## 4 | Discussion
This study demonstrates significantly reduced TLR3 and TLR9 expression in both normal skin and flat warts of Epidermodysplasia Verruciformis (EV) patients compared to immunocompetent individuals (NEV). Because we assessed expression rather than signaling or virological outcomes, implications for β-HPV control or carcinogenesis cannot be inferred from these data. In contrast, NEV individuals exhibited higher TLR3 in flat warts than in matched normal skin, supporting a wart-associated induction. EV patients did not show this induction and maintained low TLR3 across both tissue types, consistent with the absence of wart-associated TLR3 upregulation observed in NEV.
TLR4 and TLR5 remained stable between groups. This aligns with reports of TLR expression in human keratinocytes and with their detection in cutaneous viral lesions [4,10] and, within this data set, indicates no between-group differences for these receptors.
TLR9 showed the lowest detectability overall, with a more pronounced reduction in EV patients. Notably, even clinically normal skin from EV patients exhibited lower TLR3/TLR9 than NEV controls, indicating a baseline difference rather than a purely lesion-associated change.
Together, TLR3 and TLR9 delineate an EV-specific epithelial expression profile. These reduced expression levels suggest differences in epithelial innate-sensing capacity; however, whether these differences are functionally relevant to β-HPV sensing or contribute to viral persistence and carcinogenesis was not tested here and requires further investigation. EV also exhibits abnormalities in epidermal differentiation markers (e.g., cytokeratins, involucrin, filaggrin, E-cadherin) compared with non-EV warts [12], suggesting disease-specific epithelial programs; possible links between differentiation state and TLR abundance were not assessed here.
Mechanistically, TLR3 recognizes double-stranded RNA in model systems; dsRNA production by cutaneous HPV has not been established. We quantified epithelial TLR protein expression; functional activation was not assessed. Epithelial TLR pathways can be biologically relevant in DNA-virus settings [21]. TLR3 expression has been reported in verruca vulgaris and molluscum contagiosum [10] and associated with HPV-related cervical disease [14]. In our cohort, NEV warts showed higher TLR3 than matched normal skin, whereas EV lacked this wart-associated induction-an epithelial expression phenotype rather than evidence of anti-HPV activity. In keratinocytes, engagement of TLR pathways can elicit type I interferon and NF-κB programs [1,4]. Multiple explanations remain plausible-including epigenetic regulation or host background-and warrant testing in keratinocyte systems.
For TLR9, keratinocytes respond to CpG stimulation by inducing inflammatory mediators, indicating functional signaling capacity [4]. In cutaneous HPV lesions, contextdependent patterns have been described: increased TLR9 and IRAK1 with unchanged TLR7/IRF7 (interferon regulatory factor 7) in common warts, and higher TLR3/TLR9 in warts than in normal skin [10,11]. Our findings extend this cutaneous context by demonstrating marked TLR9 reductions in EV patients across both normal skin and flat warts, representing a distinct epithelial expression phenotype. Beyond skin, TLR9 promoter polymorphisms have been associated with risk and prognosis in gastric cancer cohorts [22], although these findings do not address epithelial protein expression.
In contrast to the altered expression of endosomal TLRs (TLR3 and TLR9), TLR4 and TLR5 are surface-expressed receptors [1,2]. Correspondingly, epidermal TLR4 and TLR5 levels were stable between EV and NEV within our data set, in line with reports of constitutive expression in human keratinocytes and detection in cutaneous viral lesions [4,10].
While our findings are specific to cutaneous HPV disease, studies in other epithelial tumors have reported associations between tumor-cell TLR3 and clinical outcomes-favorable prognosis in triple-negative breast cancer and apoptosis upon experimental activation in non-small-cell lung cancerhighlighting clinical interest in epithelial TLR patterns while underscoring differences in disease context [23,24]. For TLR9, studies in common warts indicate context-dependent behavior [10,11]; our data address epidermal expression only.
Several limitations should be acknowledged. Despite careful case selection, our modest sample size may limit generalizability. While immunohistochemical detection assesses protein abundance at the tissue level, it does not provide direct evidence of functional signaling activity; observed TLR expression should not be interpreted as evidence of responsiveness to viral ligands. This caveat particularly applies to TLR3 in NEV warts, where we did not evaluate downstream readouts (e.g., IRF3 phosphorylation, interferonstimulated genes). Thus, higher TLR3 expression should not be construed as evidence of pathway activation.
The absence of molecular profiling constitutes another limitation. Host factors, including genetic variants, regulatory microRNAs, and epigenetic modifications, were not evaluated, constraining conclusions about the mechanistic basis of reduced TLR expression in EV. Although TLR polymorphisms have been linked to HPV-associated disease [25,26], their specific role in EV-related TLR dysregulation remains undefined.
Future investigations should incorporate functional assays to test whether differential receptor expression corresponds to downstream signaling competence. Ex vivo stimulation with TLR agonists could probe signaling capacity in EV versus NEV keratinocytes. Clinical implications, if any, would require dedicated studies, ideally integrating HPV genotyping and host genetic factors (e.g., TMC6/TMC8/CIB1). This combinationlower TLR3/TLR9 and absence of wart-associated induction in EV-defines a precise, testable framework for subsequent functional work on pathway competence in keratinocytes.
## References
1. Takeda, Kaisho, Akira (2003) "Toll-Like Receptors" *Annual Review of Immunology*
2. Miller, Modlin (2007) "Toll-Like Receptors in the Skin" *Seminars in Immunopathology*
3. Lester, Li (2014) "Toll-Like Receptors in Antiviral Innate Immunity" *Journal of Molecular Biology*
4. Lebre, Van Der Aar, Van Baarsen (2007) "Human Keratinocytes Express Functional Toll-Like Receptor 3, 4, 5, and 9" *Journal of Investigative Dermatology*
5. Orth (2006) "Genetics of Epidermodysplasia Verruciformis: Insights Into Host Defense Against Papillomaviruses" *Seminars in Immunology*
6. Ramoz, Rueda, Bouadjar et al. (2002) "Mutations in Two Adjacent Novel Genes Are Associated With Epidermodysplasia Verruciformis" *Nature Genetics*
7. Jong, Créquer, Matos (2018) "The Human CIB1-EVER1-EVER2 Complex Governs Keratinocyte-Intrinsic Immunity to β-Papillomaviruses" *Journal of Experimental Medicine*
8. Youssefian, Vahidnezhad, Mahmoudi (2019) "Epidermodysplasia Verruciformis: Genetic Heterogeneity and EVER1 and EVER2 Mutations Revealed by Genome-Wide Analysis" *Journal of Investigative Dermatology*
9. Saeidian, Youssefian, Huang "Whole-Transcriptome Sequencing-Based Concomitant Detection of Viral and Human Genetic Determinants of Cutaneous Lesions" *JCI Insight*
10. Ku, Kwon, Kim (2008) "Expression of Toll-Like Receptors in Verruca and Molluscum Contagiosum" *Journal of Korean Medical Science*
11. Tang, Zhu, Han et al. (2020) "Expression of Langerhans Cell and Plasmacytoid Dendritic Cell Markers, and Toll-Like Receptor 7/9 Signaling Pathway Proteins in Verruca Vulgaris Lesions" *Medicine*
12. Barcelos, Sotto (2009) "Involucrin, Filaggrin and E-Cadherin in Plane Warts and Epidermodysplasia Verruciformis Plane Wart-Type Lesions" *Comparative Analysis of the Expression of Cytokeratins*
13. Stanley (2010) "HPV-Immune Response to Infection and Vaccination" *Infectious Agents and Cancer*
14. Hasimu, Ge, Li et al. (2011) "Expressions of Toll-Like Receptors 3, 4, 7, and 9 in Cervical Lesions and Their Correlation With HPV16 Infection in Uighur Women" *Chinese Journal of Cancer*
15. Halec, Scott, Farhat et al. (2018) "Toll-Like Receptors: Important Immune Checkpoints in the Regression of Cervical Intra-Epithelial Neoplasia 2" *International Journal of Cancer*
16. Nuovo, Ishag (2000) "The Histologic Spectrum of Epidermodysplasia Verruciformis" *American Journal of Surgical Pathology*
17. De Oliveira, He, Rady (2004) "HPV Typing in Brazilian Patients Witn Epidermodysplasia Verruciformis: High Prevalence of EV-HPV 25" *Journal of Cutaneous Medicine and Surgery*
18. Baker, Ovigne, Powles et al. (2003) "Normal Keratinocytes Express Toll-Like Receptors (TLRs) 1, 2 and 5: Modulation of TLR Expression in Chronic Plaque Psoriasis" *British Journal of Dermatology*
19. Tuon, Fernandes, Pagliari et al. (2010) "The Expression of TLR9 in Human Cutaneous Leishmaniasis Is Associated With Granuloma" *Parasite Immunology*
20. Gupta, Wasnik, Mondal et al. (2024) "Critical Role of Keratinocytes in Cutaneous Immune Responses" *Exploration of Immunology*
21. Harte, Haga, Maloney (2003) "The Poxvirus Protein A52R Targets Toll-Like Receptor Signaling Complexes to Suppress Host Defense" *Journal of Experimental Medicine*
22. Wang, Xue, Yang et al. (2013) "TLR9 Promoter Polymorphism Is Associated With Both an Increased Susceptibility to Gastric Carcinoma and Poor Prognosis" *PLoS One*
23. Fan, Sui, Jin et al. (2023) "High Expression of TLR3 in Triple-Negative Breast Cancer Predicts Better Prognosis" *BMC Cancer*
24. Bianchi, Alexiadis, Camisaschi (2020) "TLR3 Expression Induces Apoptosis in Human Non-Small-Cell Lung Cancer" *International Journal of Molecular Sciences*
25. Pandey, Mittal, Srivastava (2011) "Evaluation of Toll-Like Receptors 3 (c.1377C/T) and 9 (G2848A) Gene Polymorphisms in Cervical Cancer Susceptibility" *Molecular Biology Reports*
26. Roszak, Lianeri, Sowińska et al. (2012) "Involvement of Toll-Like Receptor 9 Polymorphism in Cervical Cancer Development" *Molecular Biology Reports* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12515889&blobtype=pdf | # FcgRIIIA -activating antibodies in dengue virus infection reveals a distinct transient cross-reactive profile
Claudio Soto-Garita, Tatiana Murillo, Hartmut Hengel, Eugenia Corrales-Aguilar, Fernando Esquivel-Guadarrama, Rinkoo Gupta, Hoa Thi, My Vo
## Abstract
Dengue viruses belong to the genus Flavivirus and consist of a serocomplex of four serotypes (DENV-1, DENV-2, DENV-3, and DENV-4). As arthropod-borne viruses (arboviruses), their transmission is mediated primarily by the vector Aedes aegypti. Antiviral immune response is one of the most crucial factors influencing the progression from uncomplicated to severe dengue virus (DENV) infection. Two types of antibody responses are elicited during a DENV infection: one specific to the infecting serotype (serotype-specific or homotypic response) and another that cross-reacts with other serotypes (cross-reactive or heterotypic response). Both responses play roles in the protection against and in the induction of immunopathogenesis of DENV disease. In the case of the humoral immune response, the balance between protective and pathogenic effects mediated by antibodies (antibody-dependent enhancement, ADE) is highly dynamic and influenced by multiple factors. Although many downstream effector mechanisms depend on antibody recognition by Fcgamma receptors (FcgRs) present on immune effector cells, this interaction is traditionally not considered when evaluating antibody properties. Specifically, FcgRIIIA has been implicated in both protection and immunopathogenesis of virus infection. To assess its role within the humoral immune response to DENV, we took advantage of FcgRIIIA-CD3z reporter cells and tested receptor activation by polyclonal sera from individuals with past and acute DENV infections. In addition, the neutralizing capacity and the potential enhancement of infection were analyzed. The FcgRIIIA activation assay revealed a humoral profile distinct from neutralization and immunopotentiation, primarily mediated by crossreactive antibodies. Notably, this profile increases during the post-acute period but disappears within two years after infection. Because these two types of antibodies are found during both the cross-protective and disease-enhancing (immunopotentiation) phases, their exact function in each situation is still not clearly understood. The results of this study provide a valuable measurement of the effector function of anti-DENV antibodies, contributing to the understanding of their role in both protective and disease enhancing courses of DENV infection.
## 1 Introduction
Dengue virus (DENV) is an arthropod-borne virus (arbovirus) transmitted by Aedes aegypti and Aedes albopictus, being the former the most important vector. DENV is assigned to the family Flaviviridae and the genus Orthoflavivirus and poses a major global health burden in tropical and subtropical regions. The World Health Organization (WHO) estimates that between 100 to 400 million infections occur yearly and that half the world population is at risk of infection (1). Costa Rica is considered a hyperendemic country for dengue, with co-circulation of all four DENV serotypes and recurring outbreaks that continue to pose major public health concerns (2). DENV infection is characterized by an incubation period of 4 to 10 days after the mosquito bites and produces a spectrum of clinical manifestations. Although many of the infections are asymptomatic, it can produce a self-limited but debilitating clinical presentation characterized by high fever, headache, retroorbital pain, myalgia, arthralgia, nausea, vomiting, lymphadenopathy and rash. The major risk of DENV infection is for those patients who develop dengue-hemorrhagic fever (DHF) which can be death threatening (3). DHF has three phases: febrile, critical and recovery. In the critical phase the increase in capillary permeability leads to plasma leakage and hypovolemic shock with multiorgan failure, metabolic acidosis, disseminated intravascular coagulation and hemorrhage (4). Some critical patients can develop hepatitis, encephalitis, myocarditis, and severe hemorrhage without plasma leakage. In these cases, intravenous rehydration treatment can reduce mortality from 20% to 1% (3).
Four DENV serotypes (1)(2)(3)(4) exist, sharing between 60%-70% of their coding sequence (5). DENV pathogenicity in the human host can be partially explained by differences in viral virulence due to genotype and serotype (6). For instance, the Asian genotype of DENV-2 produces a more severe disease than the American genotype (7,8). Host factors are also implicated in the severity of the disease, including the humoral immune response. The immune response against DENV differs between serotypes, a serotypicspecific or homotypic response is produced against the infecting serotype while a cross-reacting or heterotypic response is generated against other serotypes (9). A heterotypic immune response provides protection for an estimated period of six months to three years while a homotypic immune response should give a lifetime protection (10). However, once a cross-reacting immune response cannot protect the host anymore, it can contribute to the immunopathogenesis of the disease by exacerbating inflammation through a cytokine storm or immunopotentiation (ADE) (11). The overproduction of cytokines produces endothelial cell damage increasing vascular permeability and plasma leakage characteristic of DHF (12). Complement activation and the production of a temporal autoimmune response may also occur (13,14).
Both the cellular and humoral heterotypic immune response may induce immunopathogenesis. Cross-reacting cytotoxic T cells are ineffective at controlling the infection and increase the production of cytokines (15). Antibody-dependent enhancement (ADE) of infection occurs when IgG antibodies bind the viral particles but are uncapable to neutralize them and instead, form immune complexes that bind to the Fcg receptors (FcgR) on immune cells, favoring viral infection of these cells followed by uncontrolled immune cell activation (16). Antibody specificity determines the risk of developing ADE. Antibodies targeting the I-and II-domain of the envelope (E) viral glycoprotein are highly serotype cross-reactive and associated with ADE (11,17). The tridimensional disposition of the epitopes and antibody concentration also has an impact on the development of ADE (18). Linear epitopes and neutralizing antibodies at low concentrations can favor ADE (19,20). Thus, serotype-specific, and cross-reactive antibodies may produce ADE depending on their concentration (18).
Fcg receptors (FcgRs) belong to the immunoglobulin superfamily and are expressed on the surface of various immune cells, including monocytes, macrophages, neutrophils, and NK cells. The three main classes-FcgRI (CD64), FcgRII (CD32), and FcgRIII (CD16)-differ in structure, cellular distribution, affinity for IgG subclasses, and the signaling pathways they activate (21). In the context of DENV infection, FcgRs play a dual role: they can mediate protective immune clearance or contribute to ADE, depending on the antibody characteristics and the receptor involved. Notably, FcgRIIIa (CD16a), expressed primarily on NK cells and some myeloid populations, has been implicated in both beneficial effector functions such as antibody dependent cell-mediated cytoxicity (ADCC) and potentially in facilitating ADE under certain conditions (22,23). Despite its relevance, the dynamics of FcgRIIIa activation during acute dengue infection remain poorly understood. In this study, we aim to characterize the FcgRIIIa-activating antibody profile in individuals with acute dengue infection, evaluate how it relates to other antibody effector functions and compare it to the profile found in convalescent patients.
## 2 Materials and methods
## 2.1 Serum samples
Two sets of serum samples were analyzed. A first set consisted of seven anonymous convalescent serum samples (S) collected in Golfito and Puntarenas, which represented DENV hyperendemic regions in Costa Rica, for a previous sero-epidemiological study during 2005-2006 (24). The second set of samples were collected from seven acute dengue adult patients with follow-up serial sample collections (Table 1). All sera were collected from non-severe dengue cases. Previous exposure to DENV infection was assessed with IgG detection in the acute sample with a commercial ELISA. Individuals with IgG antibodies against DENV during acute infection were categorized as non-primary infection (NP) and those where antibodies were not detected were classified as primary infection (P). Ethical approval for the use of human samples was given to the project B7360 in the resolution VI-3178-2017 by the Scientific Ethical Committee from the Vice rectory of Research of the University of Costa Rica.
## 2.2 Anti-DENV IgG and IgM detection
To detect IgG and IgM antibodies against DENV, two highly sensitive commercial ELISA kits were used (26): the Human Dengue IgG ELISA Test Kit (Diagnostic Automation, Cortez Diagnostics Inc., CA, USA) with 94.7% sensitivity and 97.4% specificity, and the Human Dengue IgM ELISA Test Kit (Diagnostic Automation, Cortez Diagnostics Inc., CA, USA) with 97.8% sensitivity and 93.5% specificity. Both assays were performed following the manufacturer's protocol. Optical density (OD) values were measured after a 25-minute reading at 450 nm and 630 nm using the Epoch spectrophotometer (BioTek, Vermont, USA).
## 2.3 Molecular detection and serotyping of DENV
Viral RNA was extracted from 200 ml of serum or urine using the MagNA Pure LC RNA Isolation Kit I (Roche, Basel, Switzerland) according to the manufacturer's instructions, using the MagNA Pure LC 2.0 extraction system (Roche, Basel, Switzerland). Detection and confirmation of DENV, ZIKV, and CHIKV were conducted on RNA samples using real-time reverse transcription PCR (RT-PCR) with Modular Diagnostic Kits for Dengue, Zika, and Chikungunya viruses, along with Multiplex RNA Master Mix on the LightCycler II (Roche, Basel, Switzerland), following the manufacturer protocol. Dengue serotyping was carried out following the protocol described by Lanciotti et al., using specific serotype controls (25).
## 2.4 Viral strains and cell lines
Dengue virus prototype strains all grown in the C6/36 cell line (ATCC ® CRL-1660 ™ RRID: CVCLZ230), donated by the Pedro Kourı Ínstitute in Cuba, were used in the K562 (ATCC ® CCL-243 ™ RRID: CVCL0004) immune enhancement and FcgRIIIA-CD3z activation assays (26). The strains were DENV-1 Angola (12 passages), DENV-2 Jamaica (19 passages), DENV-3 Nicaragua (13 passages) and DENV-4 Dominica (16 passages).
For neutralization assays, chimeric viruses (ChimeriVax -DENV1, DENV2, DENV3 and DENV4) produced by Sanofi Pasteur and grown in Vero cells (ATCC ® CCL-81 ™ RRID: CVCL0059) were used (27). These viruses are based on the yellow fever 17D vaccine backbone and express only the prM and E genes of each DENV serotype, thereby assessing the neutralizing activity of antibodies directed against the major structural antigens involved in viral entry. Using this approach restricts the readout to neutralization-relevant epitopes, thereby minimizing contributions from other viral proteins. These viruses were donated by Sanofi Pasteur through the CDC Arbovirus Reference Collection under a material transfer agreement (MTA).
## 2.5 Reporter cell BW: FcgRIII-z assay
The assay used to evaluate individual antibody-dependent activation of FcgRIII (CD16) involved co-culturing antigenbearing cells with BW5147 reporter cells that stably express chimeric FcgRIII-z chain receptors. These receptors trigger mouse IL-2 production upon receptor crosslinking by immune-complexed IgG, provided the opsonizing IgG is recognized by specific FcgR (26). This assay was standardized before in Corrales-Aguilar et al. (26). Briefly, to assess antibody-dependent activation of BW: FcgRIII-z reporter transfectants, Vero cells where infected with 0.1 multiplicity of infection (MOI) of each DENV serotype for a 72hour period, then virus was inactivated by UV-light. After inactivation, mock-infected and virus-infected cells were incubated with serial two-fold dilutions of human sera in D-MEM (Sigma-Aldrich, MO, USA). containing 10% (v/v) FCS (Thermo Fisher Scientific, MA, EE.UU.) for 30 minutes at 37 °C in a 5% CO 2 atmosphere. Non-bound IgG was removed by washing the cells three times with D-MEM containing 10% (v/v) FCS before co-culturing them with 100-000 BW: FcgRIII-z reporter cells per well for 16 to 24 hours at 37 °C in a 5% CO 2 atmosphere in RPMI medium (Thermo Fisher Scientific, MA, EE.UU.) supplemented
## 2.6 Focus reduction neutralization test
For the FRNT assay, ChimeriVax strains (YFV-DENV1, 2, 3, and 4), validated for viral neutralization studies (29), were used. A focus-reduction microneutralization assay (FRNT) was performed in flat-bottom 96-well plates (30). Serial two-fold dilutions of sera, starting at 1:40, were incubated for 1 hour at 37°C with viral stocks, adjusted to yield 30-200 foci per well in at least four wells. The mixture was then inoculated (50 mL/well) into confluent Vero cell monolayers and incubated for an additional hour to allow viral adsorption. The adsorption medium was replaced by 100 mL of 1.5% carboxymethylcellulose overlay medium to restrict infection. DENV-1, DENV-2, and DENV-3 were incubated at 37°C for 48 hours, while DENV-4 was incubated for 24 hours. Post-incubation, the overlay medium was removed, wells were washed with PBS (Thermo Fisher Scientific, MA, EE.UU.) and fixed with 100 mL of cold methanol per well. Plates were stored at -20°C for at least 24 hours. For focus visualization, immunostaining was performed using an anti-flavivirus group monoclonal antibody 4G2 (GeneTex, CA, USA. RRID: AB3074294) (1:600 dilution) followed by a secondary anti-mouse IgG antibody conjugated with peroxidase (1:600 dilution). The signal was developed using 3amino-9-ethylcarbazole (AEC) substrate, incubated for 30 minutes at room temperature in darkness. Foci were imaged using a stereoscope and manually counted with ImageJ software (RRID: SCR_003070). FRNT50 was determined in Prism 10 (GraphPad, San Diego, CA, USA) by nonlinear regression, identifying the dilution that reduced foci by 50% (FRNT50). High FRNT50 values indicate stronger neutralizing capacity against the tested DENV serotype.
## 2.7 Antibody dependent enhancement test
This study used the semi-adherent K562 cell line, which constitutively expresses FcgRIIa (31), based on a monolayer methodology (32). Plates were coated with fibronectin and 30-000 cells per well were added. Serial dilutions of test sera were mixed with DENV serotypes at a MOI of 0.5 (DENV4) to 0.1 (other serotypes) and incubated at 37 °C for 24 (DENV-4) to 48 (other serotypes) hours. Post-incubation, cells were fixed, immunostained with the 4G2 antibody and secondary anti-mouse peroxidaseconjugated antibodies and stained with AEC to visualize infected cells as described before. Infected cells, identified by a precipitated brown color, were observed under light microscopy, and the number of infected cells per 40X field was quantified using ImageJ software. The percentage of infection for all serial dilutions was plotted, and the level of immunopotentiation was determined based on the width of the curve. Samples that exhibited broad curves against more than one DENV serotype were considered to have a high level of immunopotentiation as defined in other studies (18). The magnitude of enhancement was interpreted based on the breadth and height of the curve: narrow, low curves were considered low enhancement, whereas broad curves with high percentages of infection across multiple dilutions indicated strong enhancement potential.
## 2.8 Statistical analysis
All assays were performed in triplicate unless otherwise indicated. Data are shown as individual values or as mean ± standard deviation (SD). For the FcgRIIIA activation assay, the cutoff for a positive response was defined as the mean IL-2 production of mock-infected cells plus three standard deviations. Neutralization titers (FRNT50) were determined by nonlinear regression analysis using GraphPad Prism 10 (GraphPad Software, San Diego, CA, USA). No formal hypothesis testing was performed due to the small sample size; instead, results are presented descriptively to illustrate individual antibody profiles over time.
## 3 Results
## 3.1 Serological and functional characterization of samples
Serum samples were classified into two main groups based on clinical and serological criteria: past infections and acute infections. The past infection group (S) consisted of asymptomatic individuals with serological evidence of prior DENV exposure, while the acute infection group included laboratory-confirmed cases of active dengue virus infection. Acute-phase samples were further subdivided into primary (P) and non-primary infections (NP), based on the presence or absence of anti-DENV IgG within the first seven days following symptom onset. The detection of IgG at this early stage was used as a proxy to distinguish primary infections from those that were likely secondary or beyond. Due to limitations in discriminating between secondary and tertiary or quaternary responses, all early IgG-positive acute cases were conservatively grouped as non-primary (NP) infections.
The samples from past infections presented highly diverse profiles depending on IgG antibody concentration measured as OD values. Samples with anti-DENV IgG optical density (OD) values below 0.500 displayed a monotypic neutralization profile, showing serotype-specific activity restricted to either DENV-3 (Figure 1, Sample 1 (S1)) or DENV-2 (Figure 1, Sample 5 (S5)).These specimens exhibited minimal ADE activity, revealed by the short breath of the curves against the four serotypes, and failed to induce significant activation of the FcgRIIIA-CD3z receptor, suggesting limited effector function in this group.
In contrast, samples with intermediate anti-DENV ELISA OD values (0.5-1.0) exhibited broader serotype recognition, neutralizing two (Figure 1, Samples 3, 7 (S3, S7)) or three (Figure 1, Sample 4 (S4)) DENV serotypes. Moderate ADE activity was observed across these samples. Notably, FcgRIIIA-CD3z activation was detected exclusively in S4. Interestingly, despite having the highest neutralizing titer against DENV-3, the strongest receptor activation in S4 was induced by DENV-1, highlighting a potential uncoupling between neutralization capacity and Fc-mediated effector activation.
Only two samples exhibited high anti-DENV IgG OD values (>1.0). Both (Figure 1, Samples 2, 6 (S2, S6)) neutralized three serotypes and displayed the highest levels of ADE and FcgRIIIA-CD3z activation among all specimens analyzed from the past infection cohort. S2 showed peak FcgRIIIA activation in response to DENV-4, with neutralization strongest against DENV-1. In contrast, in S6 the strongest receptor activation occurred in response to DENV-1, while the highest neutralization titer targeted DENV-3. These findings underscore the complex relationships among antibody specificity, enhancement potential, and Fc-mediated effector functions following natural DENV exposure.
## 3.2 Longitudinal analysis of serum samples from acute DENV infections
In the longitudinal study, the values for all antibody characterization assays for each patient were plotted across all collected samples (T1-T4) (Figure 2). In primary DENV infections (Figure 2, P1 and P2), the immune response followed classical kinetics, marked by the induction of anti-DENV IgG and a progressive increase in functional activity. Neutralization peaked at T3 timepoint, with strong titers against the infecting serotype (DENV-3). ADE activity rose during the T2 timepoints, with moderate levels persisting in the T3 subacute phase. FcgRIIIA-CD3z activation was largely absent, except for a minimal, abovethreshold response to DENV-4 in P2 at T3.
In non-primary infections (NP), antibody dynamics were more heterogeneous. In all cases, the infecting serotype was DENV-1. Patient NP2 showed detectable IgG during the acute phase, but without measurable neutralizing activity. By T2, neutralization peaked against DENV-2, and ADE activity increased notably. A low but detectable FcgRIIIA-CD3z activation signal was recorded in T3 against DENV-4 (Figure 2, NP2). For NP3, all measured antibody activities-including neutralization, ADE, and FcgRIIIA-CD3z activation-peaked at T2 and declined by T3 timepoint. Neutralizing responses were strongest against DENV-2 across all timepoints, followed by DENV-1, suggesting that DENV-2 was likely the priming serotype. FcgRIIIA activation in this case was restricted to DENV-4, which peaked at T2 and decrease by T3 (Figure 2,NP3).
NP5 was the only case with a fourth sample collected nearly five years post-infection (Figure 2, NP5). The acute-phase sample (T1) showed the highest IgG OD value among all evaluated samples, along with the strongest neutralizing response against DENV-2, followed by DENV-1, supporting DENV-2 as the primary infecting serotype. FcgRIIIA-CD3z activation was significant for DENV-1, DENV-2, and DENV-3, with DENV-1 showing the highest signal from T1. An elevated enhancing activity is seen for all serotypes in all time points. It should be noted that in this case the acute sample had a broad neutralizing activity, recognizing all four serotypes. Although antibody function remained relatively high through T3, all profiles declined markedly by T4.
Patient NP4 showed persistently high IgG OD values and a broad neutralization activity across all timepoints, being the strongest against DENV-2, suggesting this serotype as the primary exposure. ADE activity increased over time, and FcgRIIIA-CD3z activation was pronounced against DENV-1 from the acute phase through T3. A secondary, though significant, activation signal was also observed for DENV-4 (Figure 2, NP4). In the case of NP1, the neutralization profile also pointed to DENV-2 as the primary infecting serotype. Enhancing activity was initially low but increased by T2. FcgRIIIA-CD3z activation was undetectable in the acute phase, but increased significantly in the subacute sample, particularly in response to DENV-4, DENV-3, and DENV-2.
Collectively, these findings highlight the dynamic and individualized nature of DENV-specific antibody responses following natural infection. Primary infections showed a more predictable trajectory of rising neutralization and ADE activity, with minimal detection of FcgRIIIA activation. In contrast, nonprimary infections were characterized by broader serotype recognition, variable neutralization targets, and a more prominent engagement of FcgRIIIA-mediated triggering. Notably, patients NP4 and NP5-who exhibited the broadest neutralization profiles -were also the only individuals with the detectable FcgRIIIA activation against the infecting serotype (DENV-1) with peaks at relatively late time-points (T3) after symptom onset.
## 4 Discussion
In this pilot study, we determined for the first time distinct antibody effector functions and profiles by ELISA, FRNT, ADE test and FcgRIIIA activation assay across different immunological contexts of DENV 1-4 infection. Notably, in most cases, FcgRIIIA-CD3z activation did not consistently correlate with neutralization profiles, one explanation may be that the epitopes driving neutralization differ from those responsible for Fc-mediated functions (28). Neutralization is typically mediated by antibodies targeting structurally critical regions on the virion, such as quaternary epitopes recognizing multiple envelope (E) protein subunits or serotype-specific sites on the E protein domain III (33). By contrast, robust FcgRIIIA activation often arises from highly cross-reactive IgG antibodies against conserved epitopes that confer little DENV neutralization. Notably, many human anti-DENV antibodies dominantly target the precursor membrane (prM) protein and the conserved fusion-loop of E domain II; these antibodies are broadly cross-reactive among serotypes yet poorly neutralizing, even at high concentrations, and can still efficiently opsonize infected cells and virions, triggering FcgRIIIA (17,33). Indeed, the FcgRIIIA activation assay of this study utilized DENV-infected Vero cells that display both E and uncleaved prM on their surface, providing abundant targets for Fc binding in comparison to neutralization assay (34). In summary, the antigenic determinants of neutralization versus FcgRIIIA-mediated effector function only partially overlap, leading to an uncoupling dissection of these profiles in many samples.
Past infection data likely reflect a range of diverse time points post DENV infection (Figure 1). Samples with broader serotype reactivity and enhanced Fc-mediated function are consistent with non-primary infections or specimens taken within two years after exposure, when cross-reactive antibodies remain elevated (35). Longitudinal analysis of acute cases provides a clearer view on the kinetics of the humoral response and its associated effector functions. Individuals with secondary or multiple infections exhibited notably stronger FcgRIIIA-CD3z activation compared to primary cases. This increased activity reflects not only higher antibody titers but also qualitative differences in the IgG response, possibly due to subclass distribution and Fcg N297 glycosylation pattern as demonstrated in COVID-19 patients (36)(37)(38). DENV infection predominantly induces IgG1 and IgG3, both capable of engaging FcgRIIIA. IgG3 is short-lived and more potently neutralizing, while IgG1 is longer-lasting and subject to glycan modification (22). It has been described that afucosylation of IgG1 is more prominent in dengue secondary infections, and that elevated levels of afucosylated anti-E IgG1 are present early on severe dengue (22). Afucosylation significantly enhances FcgRIIIA binding (22) which may explain the difference observed between P and NP individuals.
Analysis of past infection samples revealed a consistent association between the breadth of serotype recognition by neutralization and the magnitude of FcgRIIIA-mediated effector activity. In acute infections, broadly neutralizing sera, typically from those with non-primary infection, tended to activate FcgRIIIA across multiple serotypes more robustly than narrow, typespecific sera. A broader neutralization profile implies a more extensive distribution of IgG bound to diverse epitopes on the virion surface or the infected cells membrane, thereby increasing the valency, defined as the multivalent engagement of antibodies with multiple epitopes, and the density of immune complexes (22). This configuration enhances the odds of cross-linking of FcgRIIIA on effector cells, a prerequisite for efficient receptor signaling (39). This finding is consistent with the concept that a minimum concentration and opsonization density of IgG must be achieved to overcome the activation threshold of FcgRIIIA. Prior studies of dengue immunity have noted that intermediate antibody levels can exacerbate infection (via ADE), but sufficiently high antibody levels confer protection (40,41). Analogously, only the samples with high IgG binding levels were potent in FcgRIIIA triggering, whereas those with modest titers did not (23,42). Thus, a higher abundance and breadth of antibodies likely ensures that FcgRIIIA is engaged in antiviral effector functions rather than in enhancing pathways.
The longitudinal FcgRIIIA activation profiles observed in individuals NP5 and NP4 provide valuable insight into the dynamics of Fc-mediated antibody responses during acute dengue infection. In both cases, a marked FcgRIIIA/CD16 activation signal was detected in response to the infecting serotype (DENV-1) during the acute phase, indicating the presence of FcgRIIIA-activating IgG early in infection. Interestingly, both individuals exhibited a transitory decline in activation at the T2 timepoint, followed by a peak in T3. This transient reduction may reflect in vivo engagement of FcgRIIIA-expressing effector cells, such as natural killer (NK) cells or monocytes, by IgG-virus immune complexes, leading to ADCC or phagocytosis and temporary clearance of activating antibodies in immune complexes (22,23,42,43). The increased CD16 activation signal observed in T3 may be due to clonal expansion against the epitopes recognized in the acute infection (44). Notably, while both individuals shared similar FcgRIIIA activation kinetics against the infecting serotype, they differed in their ADE profile: NP5 displayed high ADE in acute sample, while NP4 did not. This immune assessment enabled the distinction between FcgRIIIAactivating antibody profiles with low enhancing potential and those with strong enhancing activity.
Our study focused exclusively on FcgRIIIA activation profile, which does not capture the full range of FcgR-mediated effector mechanisms. Furthermore, FcgR polymorphisms such as FcgRIIA-H131R and FcgRIIIA-V158F, which affect the affinity of Fcg receptors for IgG subclasses, have been associated with increased susceptibility and protection against severe dengue, respectively (45,46). Therefore, a broader approach incorporating additional FcgRs, and their key polymorphic variants, along with FcgR reporter cell assay settings selective for certain ligands including soluble multimeric immune complexes and C reactive Protein isoforms (47,48), should be undertaken to evaluate the full effector potential of dengue-specific antibodies and to identify thresholds that help define the spectrum of clinical outcomes from DENV infection. The hyperendemic setting in Costa Rica, where multiple flaviviruses co-circulate, highlights the need for a broader viral panel to better interpret antibody profiles. This would allow for the inclusion of both severe and non-severe patients (2). Increasing the number of patients, outcomes of DENVdisease, and timepoints during the early acute and convalescent phases would provide a more detailed understanding of how FcgR activation evolves. This would also help to clarify its complex role in the dual nature of the humoral response in dengue infection.
Recent studies highlight the dual impact of FcgRIIIA interactions: afucosylated IgG1 enhancing FcgRIIIA binding has been linked to severe dengue (23,42,49,50), dengue immune complexes can activate NK cells and suppress ADE (51), and stronger FcgRIIIA-driven effector functions, including NK activation, associate with protection from symptomatic infection (22). While NK cell-based assays are highly informative to evaluate the protective role of CD16-activating antibodies, our reporter system allows the measurement of the broader fraction of antibodies capable of engaging FcgRIIIA, including those that may also contribute to immunopathogenic outcomes, since FcgRIIIA expression is not restricted to NK cells but includes monocytes implicated in infection and inflammation (52). This distinction provides a complementary view, revealing potentially different functional profiles of dengue antibodies. Additionally, Kao et al. recently revealed that CD8 T cells, which typically do not express Fcg receptors, can specifically induce the activating FcgRIIIa receptor in response to viral infections like COVID-19 and dengue (53). While FcgRIIIa expression closely follows the immune response timeline, its activation alone does not trigger CD8 T cell function; however, it synergizes with T cell receptor (TCR) stimulation to enhance activation (53). These findings uncover a novel costimulatory role for FcgRIIIa, showing how virus-induced antibodies can modulate CD8 T cell responses. By providing a scalable and reproducible way to measure FcgRIIIA engagement beyond natural killer and CD8 T cell functions, our assay offers a novel framework to characterize the balance between protective and pathogenic antibody responses.
Taken together, our data shows that neutralization and FcgRIIIA-mediated antibody functions against Dengue viruses are often uncoupled which has already been observed with other viral infections before (28). Furthermore, the different epitopes involved in each process may lead to distinct antibody functional profiles. Cross-reactive antibodies (e.g., anti-prM, fusion-loop) may not neutralize dengue virus effectively but still trigger immune effector mechanisms via Fc receptors. To better understand how antibody effector mechanisms and Fc-mediated immunity influence dengue outcomes, different FcgRs and their polymorphisms, distinct immune complex forms, more patients and defined timepoints of sampling should be studied. Our foremost rationale for using these tests will be to evaluate the functional quality of antibodies, especially cross-reactive ones, during different phases of dengue infection (acute and post-acute). This may help elucidate their dual role in both protection and immunopathogenesis, improving our understanding of disease progression and immune responses, and potentially guiding vaccine development by distinguishing between protective and pathogenic antibody profiles.
## Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
## References
1. Bhatt, Gething, Brady et al. (2013) "The global distribution and burden of dengue" *Nature*
2. Soto-Garita, Murillo, Chavez-Peraza et al. (2024) "Epidemiological, virological and clinical characterization of a Dengue/ Zika outbreak in the Caribbean region of Costa Rica 2017-2018" *Front Cell Infect Microbiol*
3. Khan, Yang, Lin et al. (2023) "Dengue overview: An updated systemic review" *J Infect Public Health*
4. Paz-Bailey, Adams, Deen et al. (2024) *Dengue. Lancet*
5. Holmes, Twiddy (2003) "The origin, emergence and evolutionary genetics of dengue virus" *Infect Genet Evol*
6. Rodriguez-Roche, Sanchez, Burgher et al. (2011) "Virus role during intraepidemic increase in dengue disease severity" *Vector Borne Zoonotic Dis*
7. Leitmeyer, Vaughn, Watts et al. (1999) "Dengue virus structural differences that correlate with pathogenesis" *J Virol*
8. Watts, Porter, Putvatana et al. (1999) "Failure of secondary infection with American genotype dengue 2 to cause dengue haemorrhagic fever" *Lancet*
9. St John, Rathore (2019) "Adaptive immune responses to primary and secondary dengue virus infections" *Nat Rev Immunol*
10. Reich, Shrestha, King et al. (2013) "Interactions between serotypes of dengue highlight epidemiological impact of crossimmunity" *J R Soc Interface*
11. Rothman (2011) "Immunity to dengue virus: a tale of original antigenic sin and tropical cytokine storms" *Nat Rev Immunol*
12. Dong, Moran, Chau et al. (2007) "High proinflammatory cytokine secretion and loss of high avidity cross-reactive cytotoxic T-cells during the course of secondary dengue virus infection" *PloS One*
13. Imrie, Meeks, Gurary et al. (2007) "Differential functional avidity of dengue virus-specific T-cell clones for variant peptides representing heterologous and previously encountered serotypes" *J Virol*
14. Ubol, Halstead (2010) "How innate immune mechanisms contribute to antibodyenhanced viral infections" *Clin Vaccine Immunol*
15. Mongkolsapaya, Dejnirattisai, Xu et al. (2003) "Original antigenic sin and apoptosis in the pathogenesis of dengue hemorrhagic fever" *Nat Med*
16. Halstead, Rourke (1977) "Antibody-enhanced dengue virus infection in primate leukocytes" *Nature*
17. Dejnirattisai, Jumnainsong, Onsirisakul et al. (2010) "Cross-reacting antibodies enhance dengue virus infection in humans" *Science*
18. De Alwis, Williams, Schmid et al. (2014) "Dengue viruses are enhanced by distinct populations of serotype cross-reactive antibodies in human immune sera" *PloS Pathog*
19. Rouvinski, Guardado-Calvo, Barba-Spaeth et al. (2015) "Recognition determinants of broadly neutralizing human antibodies against dengue viruses" *Nature*
20. Tsai, Liu, Tsai et al. (2018) "Seroprevalence of dengue virus in two districts of Kaohsiung City after the largest dengue outbreak in Taiwan since World War II" *PloS Negl Trop Dis*
21. Gessner, Heiken, Tamm et al. (1998) "The IgG Fc receptor family" *Ann Hematol*
22. Dias, Atyeo, Loos et al. (2022) "Antibody Fc characteristics and effector functions correlate with protection from symptomatic dengue virus type 3 infection" *Sci Transl Med*
23. Bournazos, Vo, Duong et al. (2021) "Antibody fucosylation predicts disease severity in secondary dengue infection" *Science*
24. Lee-Lui, Alfaro-Alvarado, Quesada-Johnson et al. (2005) "Prevalencia de anticuerpos contra virus dengue en el cantoń de Golfito (2005) y en el Distrito Central de Puntarenas"
25. Lanciotti, Calisher, Gubler et al. (1992) "Rapid detection and typing of dengue viruses from clinical samples by using reverse transcriptasepolymerase chain reaction" *J Clin Microbiol*
26. Corrales-Aguilar, Trilling, Reinhard et al. (2013) "A novel assay for detecting virus-specific antibodies triggering activation of Fcg receptors" *J Immunol Methods*
27. Guy, Guirakhoo, Barban et al. (2010) "Preclinical and clinical development of YFV 17D-based chimeric vaccines against dengue, West Nile and Japanese encephalitis viruses" *Vaccine*
28. Corrales-Aguilar, Trilling, Reinhard et al. (2016) "Highly individual patterns of virus-immune IgG effector responses in humans" *Med Microbiol Immunol*
29. Johnson, Kosoy, Hunsperger et al. (2009) "Evaluation of chimeric Japanese encephalitis and dengue viruses for use in diagnostic plaque reduction neutralization tests" *Clin Vaccine Immunol*
30. Dejnirattisai, Supasa, Wongwiwat et al. (2016) "Dengue virus sero-cross-reactivity drives antibody-dependent enhancement of infection with zika virus" *Nat Immunol*
31. Chawla, Chan, Zhang et al. (2013) "Dengue virus neutralization in cells expressing fc gamma receptors" *PloS One*
32. Konishi, Tabuchi, Yamanaka (2010) "A simple assay system for infectionenhancing and -neutralizing antibodies to dengue type 2 virus using layers of semiadherent K562 cells" *J Virol Methods*
33. Sarker, Dhama, Gupta (2023) "Dengue virus neutralizing antibody: a review of targets, cross-reactivity, and antibody-dependent enhancement" *Front Immunol*
34. Waldran, Kurtz, Gebo et al. (2024) "Dengue virus structural proteins are expressed on the surface of DENV-infected cells and are a target for antibody-dependent cellular phagocytosis" *Open Forum Infect Dis*
35. Salje, Cummings, Rodriguez-Barraquer et al. (2018) "Reconstruction of antibody dynamics and infection histories to evaluate dengue risk" *Nature*
36. Radovani, Nimmerjahn (2024) "IgG glycosylation: biomarker, functional modulator, and structural component" *J Immunol*
37. Hoepel, Chen, Geyer et al. (2021) "High titers and low fucosylation of early human anti-SARS-CoV-2 IgG promote inflammation by alveolar macrophages" *Sci Trans Med*
38. Ankerhold, Giese, Kolb et al. (2022) "Circulating multimeric immune complexes contribute to immunopathology in COVID-19" *Nat Commun*
39. Bruhns, Iannascoli, England et al. (2009) "Specificity and affinity of human Fcgamma receptors and their polymorphic variants for human IgG subclasses"
40. Katzelnick, Gresh, Halloran et al. (2017) "Antibody-dependent enhancement of severe dengue disease in humans" *Science*
41. Odio, Daag, Crisostomo et al. (2025) "Dengue virus IgG and neutralizing antibody titers measured with standard and mature viruses are protective" *Nat Commun*
42. Wang, Sewatanon, Memoli et al. (2017) "IgG antibodies to dengue enhanced for FcgRIIIA binding determine disease severity" *Science*
43. Chan, Zhang, Tan et al. (2011) "Rapid and massive virus-specific plasmablast responses during acute dengue virus infection in humans" *Proc Natl Acad Sci*
44. Mohsin, Mahmood, Amar et al. (2015) "Association of fcgRIIa polymorphism with clinical outcome of dengue infection: first insight from Pakistan" *Am J Trop Med Hyg*
45. Yamin, Kao, Macdonald et al. (2023) "Human FcgRIIIa activation on splenic macrophages drives dengue pathogenesis in mice" *Nat Microbiol*
46. Chen, Maul-Pavicic, Holzer et al. (2022) "Detection and functional resolution of soluble immune complexes by an FcgR reporter cell panel" *EMBO Mol Med*
47. Henning, Seer, Zeller et al. (2025) "Human Fcgreceptors selectively respond to C-reactive protein isoforms" *Front Immunol*
48. Teo, Tan, Loy et al. (2023) "Understanding antibody-dependent enhancement in dengue: Are afucosylated IgG1s a concern?" *PloS Pathog*
49. Thulin, Brewer, Sherwood et al. (2020) "Maternal anti-dengue igG fucosylation predicts susceptibility to dengue disease in infants" *Cell Rep*
50. Sun, Williams, Nagabhushana et al. (2019) "NK cells activated through antibody-dependent cell cytotoxicity and armed with degranulation/ IFN-g Production suppress antibody-dependent enhancement of dengue viral infection" *Sci Rep*
51. Ziegler-Heitbrock (2007) "The CD14+ CD16+ blood monocytes: their role in infection and inflammation" *J Leukoc Biol*
52. Kao, Pihlstrom, Niejadlik et al. (2025) "FcgRIIIa is a noncanonical costimulatory molecule for CD8 T cells" *Proc Natl Acad Sci U S A* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645921&blobtype=pdf | # Replication-driven HBV cccDNA loss in chimeric mice with humanized livers
Bai-Hua Zhang, Yuanping Zhou, Stephen Horrigan, Fabien Zoulim, Jianming Hu, Yong-Yuan Zhang
## Abstract
Hepatitis B virus (HBV) infection depends on the establishment of covalently closed circular DNA (cccDNA) and can be either transient or persistent. Persistent infection requires a noncytopathic viral phenotype, primarily achieved by limiting replication in infected cells. This study aimed to understand how most HBV-infected cells can avoid cytopathic consequences despite robust replication. Using chimeric mice with humanized livers, we assessed HBV replication kinetics and observed a progressive accumulation of viral products from early stages up to peak infection in infected cells, but the accumulation stopped increasing upon reaching the persistence phase, possibly resulting from the inhibition of HBV replication. We then examined HBV products, including cccDNA, transcription, and viral protein levels, to identify the potential point of inhibition, finding no clear suppression of viral transcription or protein synthesis. Quantitative analysis of cccDNA levels in bulk cells, single nuclei, and individual HBsAg-positive cells revealed that cccDNA was undetectable in a portion of infected cells. Our findings raise the possibility that infected cells may spontaneously clear cccDNA. This would stop HBV replication at its root and avoid the potential cytopathic consequences in infected cells due to uncontrolled viral replication. These results suggest that efficient cccDNA elimination might be achievable without directly targeting existing cccDNA.
IMPORTANCEThe primary barrier to curing chronic HBV infection is the persistence of covalently closed circular DNA (cccDNA), which is traditionally considered stable within infected cells. However, clinical observations have revealed that cccDNA can undergo frequent clearance and replacement in patients with chronic HBV infection. Building on these observations, our study demonstrated that cccDNA was undetectable in a portion of HBV-infected cells at different timepoints after peak infection in chimeric mice with humanized livers, suggesting that spontaneous cccDNA clearance may occur. These findings align with clinical data and indicate that effective cccDNA elimination may be possible without the need to target cccDNA itself directly. KEYWORDS persistent HBV infection, HBV replication, cytopathic effects, cccDNA clearance and replenishment H epatitis B virus (HBV) can cause persistent infection (1, 2). Establishing and maintaining HBV infection in hepatocytes requires the formation and persistence of episomal covalently closed circular DNA (cccDNA) molecules, which function as templates for viral transcription in the nucleus (3, 4). Therefore, a single copy of cccDNA in an infected cell is minimally required. The cccDNA molecules are assumed to be long-lived (5) because a chronic HBV (CHB) infection usually lasts for years or decades (6). CHB is thought to result from the failure of the host's immune system to clear the established infection (7). Current HBV cure strategies aim to directly eliminate or permanently silence cccDNA (5) or kill infected cells in the liver (8).
Clinical evidence indicates that the cccDNA population in CHB undergoes dynamic evolution. For example, the wild-type (WT) viral population in serum or cccDNA in the liver can be cleared and replaced by mutant populations under both untreated and treated conditions (9)(10)(11)(12)(13). The HBe minus mutant replaced WT, then disappeared and reappeared again during an 18-month duration under untreated conditions (9). Or during a 36-month study involving lamivudine treatment, cessation, and retreatment, the dominant viral population shifted from WT to drug-resistant mutant, then to WT, and was subsequently replaced again by the drug-resistant mutant (14). These are examples that HBV cccDNA in CHB is subject to continuous clearance and replacement under different conditions.
Many viruses replicate efficiently within host cells that produce high intracellular viral load that must be processed, trafficked, and released (15). However, infected cells often cannot efficiently secrete virions and other viral products, further building up the intracellular viral load that causes stress in various cellular organelles, ballooning infected cells, and stretching and disrupting the cellular membrane, culminating in cell damage/destruction (16). This combination of robust replication and inefficient secretion may result in cytopathic or lytic infections, where the host cell is eventually destroyed (17,18). Thus, a persistent viral infection in host cells is contingent upon noncytopathic effects of a virus, which are achieved mainly through controlling replication. For example, herpes simplex virus and human immunodeficiency virus (HIV) can establish persistent or latent infections through suppressing viral transcription after lytic or acute phase of infection (19)(20)(21)(22).
Similarly, HBV replicates robustly, causing progressive retention of viral products in infected cells (23,24). A continuous accumulation of viral products may also cause cytopathic effects. For instance, a common histopathology is ballooning degeneration in a portion of infected cells on liver sections (25), which resembles rounding of infected cells, a typical morphological manifestation of cytopathic effects observed in infected cultures (15). The retention of L protein in hepatocytes of HBV-transgenic mice causes a spectrum of pathologies, including hepatocyte enlargement and necrosis, and persistently elevated ALT levels. The severity of pathology is related to the concentra tion of intracellular envelope proteins (26). The intracellular HBsAg accumulation within smooth endoplasmic reticulum (ER) causes ER hyperplasia and displaces other organelles to the cell periphery, giving an appearance of "ground-glass" in some hepatocytes in CHB (27,28). Cytopathic effects in infected primary hepatocytes and acute liver injury during the in vivo infection were also observed with an L protein mutant (G133E) of duck hepatitis B virus (DHBV), a member of Hepadnaviridae family known for congenitally establishing persistent but noncytopathic infections (1,29). The observed cytopathic effects mainly resulted from increased intracellular levels of cccDNA, RNA, capsids, and rcDNA, as this L protein mutant impaired the production of enveloped virus (30,31).
However, HBV infection is largely noncytopathic (32), suggesting that most infected cells can avoid cytopathic consequences. In theory, these infected cells must have the ability to suppress HBV replication following early robust replication, thereby preventing potentially cytopathic consequences. As observed in the DHBV model, DHBV replication is effectively brought under control at the late phase of infection at least through inhibiting recycling-mediated cccDNA replenishment (30,33). This inhibition is mainly triggered by accumulating L protein levels in infected cells, which was later confirmed in HBV systems (34,35). These studies demonstrate that (i) the cellular accumulation of viral products could trigger the inhibition of replication and (ii) the control of cccDNA levels is employed for controlling viral replication in infected cells.
In case of failure to control cccDNA levels, continuous cccDNA accumulation may cause cytopathic effects, leading to cell destruction (cccDNA loss) and cell regeneration for de novo infection (cccDNA replenishment), which could explain cccDNA loss in cells with cytopathic effects.
We are interested in understanding how HBV replication is restricted in most HBV-infected cells that avoid potentially cytopathic consequences using uPA/SCID chimeric mice with humanized livers. In this study, we analyzed cccDNA levels at both the single-nucleus and single HBsAg-positive cell level and the bulk cell level. Our findings indicate that cccDNA may be spontaneously cleared from infected cells in the humanized livers of chimeric mice, aligning with clinical observations of continuous cccDNA clearance and replacement (9)(10)(11)(12)(13)(14). By confirming this clinical observation in our HBV model, we suggest that cccDNA elimination may be achievable by blocking cccDNA replenishment rather than directly targeting existing cccDNA.
## RESULTS
In this study, we characterized in vivo kinetics of HBV replication in uPA/SCID chimeric mice with humanized livers (36) and analyzed cccDNA levels using both bulk cells and single nuclei and HBsAg-positive cells.
This chimeric mouse model is known for supporting HBV persistent infection (37) that needs to be noncytopathic in most infected cells, though cytopathic effects involving a subset of cells were reported in this model (38). Similarly, we observed mostly normal histology of livers collected at various timepoints post-inoculation; however, focal areas of inflammatory infiltrates were present in some liver sections (Fig. 1). Therefore, the established persistent HBV infection, as well as dominantly normal liver histology in this model, supports that HBV infection in most cells is noncytopathic.
## In vivo replication kinetics suggested that the inhibition of HBV replication was possibly mediated through the clearance of cccDNA
To understand the kinetics of in vivo HBV replication, HBV-infected, untreated chimeric mice were euthanized on days 18, 45, 50, 52, 82, 99, 141, and 212 post-inoculation (pi) for assaying kinetic serum HBsAg and HBV DNA levels and intrahepatic HBV markers. There were two infection phases (Fig. 2A andB). The first was the phase of spread of infection to all infectible cells, in which both serum HBsAg and HBV DNA levels kept increasing after inoculation and peaked around day 82 pi. The second was the persistent infection phase, where HBV infection was maintained at a steady level. The observed kinetics of serum HBsAg and HBV DNA in this model recaptures the typical acute HBV infection that becomes persistent in humans (39).
The kinetics of intracellular accumulation of viral products also comprised two phases (Fig. 2C).
## Phase of progressive increase in accumulation
The first was the phase of a continuous rise in the accumulation of viral products. For instance, intrahepatic HBsAg levels increased from 230 copies per cell on days 18 to 110,000 copies on day 82 pi, which reflects a robust HBV replication and that the secretion of virions and subviral particles lagged behind unrestricted HBV replication during the accumulation phase.
## Phase of accumulation arrest
Following the peak on day 82 pi, the second phase began, where the increase in viral product accumulation stopped. During this phase, intracellular HBsAg levels remained steady around 100,000 copies per cell. Similarly, average cccDNA levels, as determined by qPCR using the standards calibrated with Absolute Q Digital PCR (ABQ dPCR), despite fluctuating twofold, stopped increasing over the next 130 days.
The steady levels of serum HBsAg and HBV DNA (Fig. 2A andB) suggest that the arrest in accumulation was not due to increased secretion of viral particles, assuming the cellular degradation of viral products and the clearance rates of viral particles from the blood remained stable, nor to significant cell death, which would have resulted in a reduction of both serum HBsAg and HBV DNA levels. Instead, the arrest likely resulted from an inhibition of replication, as accumulation would be expected to continue unabated if replication were not inhibited or to decline if the progressive accumulation persisted in most cells, potentially leading to significant injury to them. However, neither scenario was observed (Fig. 1 and2), suggesting that replication may have been actively restrained to prevent cytopathic consequences in most infected cells. This two-phase pattern indicates a shift from active viral replication to a controlled state in late phase infection, essential for maintaining a persistent and noncytopathic infection in most cells (30,33).
Our results also suggest that in vivo cccDNA kinetics includes two phases (Fig. 2C), and cccDNA may have been lost in both phases:
## Amplification phase
The total cccDNA level in the liver was amplified mainly by expanding the infection in the liver. The cccDNA level was increased from 0.00001 copies/cell (day 18 pi) to 0.35 copies/cell on day 82 pi (peak infection). Peaking infection means that all infectible cells must have been infected, as evidenced by the detection of HBsAg (Fig. 2D) in almost all Full-Length Text hepatocytes upon the peak. The cccDNA level is expected to be ≥1 copy/cell because a minimum of one copy of cccDNA is required in each infected cell. However, the average 0.35 copies/cell at peak infection implies approximately 1 copy of cccDNA per 3 infected cells, suggesting that cccDNA after the initial establishment may have been lost in a portion of infected cells.
## Maintenance phase
cccDNA was maintained at a steady level (0.35-0.6 copies/cell) to sustain HBV infection at a steady level upon reaching the peak. With an average cccDNA level of <1 copy/cell, the Poisson distribution predicted that some cells may contain >1 copy/cell and other cells may contain no cccDNA. This suggests that cccDNA was possibly spontaneously cleared from some cells, which may involve noncytopathic clearance mechanisms in addition to possible cytopathic clearance in a subset of cells during the maintenance phase. Therefore, a steady HBV infection level in the persistent infection phase is reached, possibly by establishing an equilibrium between the number of infected cells with cccDNA that maintain HBV replication and those that have lost cccDNA and ceased viral replication.
## HBV RNA transcription was found to be highly efficient in infected livers
We explored whether HBV replication was subjected to inhibition. Three key stages of the HBV replication can be targeted to inhibit replication after peak infection: cccDNA, viral transcription, and viral protein synthesis. We first analyzed HBV RNA levels in infected livers.
The rcDNA extraction procedures retain cytoplasmic RNAs (40,41) and were used for quantification of total HBV RNA levels by RT-qPCR.
Total HBV RNA levels were assayed in 80 rcDNA samples from four untreated livers (n = 20 per liver) to compare the relative RNA transcription efficiency between two phases of infection. Mice 842 and 836 were sacrificed on day 50 pi and 52 pi, respectively, representing the amplification phase. Mice 831 and 987 were sacrificed on day 141 pi and 218 pi, representing the maintenance phase.
Cytoplasmic HBV RNA levels exceeded 1,000 copies per cell in 74 of 80 samples (92.5%), ranging from 1,044 copies in sample 842.5 to 8,489 copies in sample 987.3. The remaining six samples had levels below 1,000 copies per cell, ranging from 351 copies in sample 831.19 to 981 copies in sample 842.13 (Fig. 3A1 through A4). The parallel kinetics observed in serum HBsAg and HBV DNA levels (Fig. 2A andB) during the two phases of HBV infection in this model suggest that HBV RNAs are predominantly transcribed from cccDNA. HBV RNA metabolism plays an important role in determining the steady-state levels of HBV RNA within infected cells. For example, pgRNA encapsi dated within nucleocapsids is degraded during reverse transcription by the RNase H activity of HBV polymerase. In addition, HBV RNAs are also subject to degradation by host RNA decay pathways, and a fraction of pgRNA and HBx RNA may be packed into virions and secreted from cells (42). The detected HBV RNA reflects the remaining pool after metabolism. Despite these ongoing degradation and export processes, we observed that the intracellular HBV RNA levels remained relatively high given the low cccDNA levels, indicating that HBV RNAs are efficiently and continuously transcribed from cccDNA.
We used the ratio of RNA copies/cell to cccDNA copies/cell to measure the relative efficiency of RNA transcription from cccDNA. The high ratios of average RNA copies per cell to average cccDNA copies per cell in the same samples were notable (Fig. 3B1, B2, B3 and B4). Average ratios ranged from 6,480 to 12,232 in mice 836 and 842 representing the amplification phase and ranging from 7,687 in mouse 831 to 11,049 in mouse 987 representing the maintenance phase (Fig. 3B). These findings suggest that a single copy of cccDNA may undergo transcription up to 10,000 times, indicating efficient RNA transcription through repeated utilization of a single or few copies of cccDNA in HBV-infected cells. Notably, no discernible differences in relative transcription efficiency were observed between the two phases of infection (Fig. 3B). The observed steady HBV RNA levels may reflect ongoing de novo infection that replenishes HBV RNA in some cells that lost cccDNA, subsequently HBV RNA. In addition, there was a positive correlation between average cccDNA and HBV RNA levels (R 2 = 0.93, Fig. 3C), suggesting that total HBV RNA level is largely determined by cccDNA level in the infected cells of this model. Unlike the latent phase of HIV infection, at which viral RNA transcription in reservoir cells is inhibited, the efficient RNA transcrip tion suggests that the suppression of RNA transcription is an unlikely mechanism to stop HBV replication, at least in this chimeric mouse system.
## No significant reduction in cellular envelope protein levels after peak infection
We compared HBsAg levels at both the peak stage (day 82 pi) and the post-peak stage (after day 82 pi). No marked difference in HBsAg levels was observed between these phases (Fig. 2A, C, andD), suggesting that the inhibition of viral protein synthesis is unlikely to be utilized as a mechanism to suppress HBV replication in this model. This is consistent with maintaining an efficient RNA transcription during the same phases described above.
## Average cccDNA levels after peak infection were less than 1 copy/cell
We extended the analysis of cccDNA levels in untreated mice to obtain the range of cccDNA levels in livers. To avoid non-representative findings by a single or a few samplings, we routinely sampled each liver 20 times, resulting in 220 cccDNA samples from 11 livers collected between days 82 and 253 pi. The highest average cccDNA level was 2.5 copies/cell, while the lowest was 0.003 copies/cell among the 220 cccDNA samples (Fig. 4A).
The average cccDNA levels in 28 (12.7%) of the 220 cccDNA samples were >1 copy/ cell, whereas there were <1 copy/cell in the remaining 192 (87.3%) cccDNA samples, indicating that some cells may not contain cccDNA at different time points. The average cccDNA level in 220 cccDNA samples was 0.5 copies/cell.
The detected cccDNA level per liver varied considerably from 1.2 to 0.16 copies/cell among 11 livers. An average cccDNA level >1 copies/cell (1.2 copies/cell) was only detected in 1 of the 11 livers (Fig. 4B).
The ABQ dPCR method enables direct quantification of target molecules and serves as a reference standard for quantification. We compared cccDNA copies detected by qPCR and ABQ dPCR in 40 cccDNA samples from mice 802 and 805. As shown in Fig. 4C, the differences between the two assays are minimal, typically within a twofold range or less. Such variations are expected between two different assay platforms. This consis tency indicates that the qPCR standards are well aligned with ABQ dPCR quantification, and the low cccDNA copies per cell detected are unlikely to result from underestimation by our qPCR assay.
We also evaluated the robustness of the cccDNA qPCR assay by retesting 20 cccDNA samples from mice 802 and 805 four and seven times, respectively. The results showed only fractional variations in mouse 802 samples, while up to twofold variations were exhibited in mouse 805 samples (Fig. 4D1 andD2), demonstrating the qPCR's reliability across repeated measurements.
## cccDNA loss was detected at the single-nucleus level
One of the criteria used by the vendor PhoenixBio for selecting uPA/SCID chimeric mice with humanized livers is a liver replacement index (RI) of >70% (36). Among the 57 mice received, only four had RI between 76% and 78% and ranged between 80% and 93% in the remaining 53 mice, approximately 76% and 93% of liver cells being human liver cells. As bulk cells are routinely used for cccDNA quantification in our assay, we used an average of 30% non-human liver cells to normalize the calculated cccDNA copies/cell. The actual number of non-human liver cells varied in each sample, which may have impacted the calculated copies/cell. We sought to quantitatively detect cccDNA copies at the single-nucleus level (41) to corroborate the absence of cccDNA in some infected cells.
HBV rcDNA is 100-fold to 1,000-fold more abundant than cccDNA and is also expected to be delivered to the nucleus for cccDNA conversion (43). We simultaneously detected cccDNA and rcDNA in each nucleus using Absolute Q duplexing digital PCR (ABQ duplexing dPCR), and the detected rcDNA was used as an HBV infection marker. The strategy, principle, and specificity of the simultaneous detection of cccDNA and rcDNA are described in detail in Materials and Methods and Fig. 5. However, this rcDNA detection may underestimate rcDNA copies if the rcDNA molecules in the nuclei were already undergoing a repair process that had filled the gap region of the plus strand, as they could be linearized by NcoI and excluded from the detection by duplex dPCR.
Nuclei from three livers, harvested on day 141, 218, or 253 pi from untreated mice 831, 987, and 907, respectively (Fig. 6F), were deposited at one nucleus per well in a 96-well plate using a BD FACSAria II. The total number of analyses performed is listed in Table 1.
Detected cccDNA and rcDNA at a single nucleus cccDNA was detected as cccDNA only or coexisting with rcDNA (Fig. 6A andB), and rcDNA was detected coexisting with cccDNA or rcDNA only (Fig. 6B, E1B, E2B, E3B and E4B).
## cccDNA copies per nucleus
The cccDNA was detected as a single copy in most of the cccDNA-positive nuclei. Twenty (66.7%) of the 30 cccDNA-positive nuclei in mouse 831 contained only a single copy, while the remaining 10 nuclei had >1 copy, ranging from two to eight (Fig. 6C).
In mouse 987, 41 (75%) of the 55 cccDNA-positive nuclei contained a single copy of cccDNA, while 14 (25%) nuclei had >1 copy. In mouse 907, the cccDNA was detected as a single copy in 34 (77%) of the 44 cccDNA-positive nuclei, while the remaining 10 nuclei contained >1 copy, ranging between 2 and 6 copies/nucleus. Thus, ≥2/3 of the detected cccDNA-positive nuclei contained only a single cccDNA copy.
## rcDNA copies per nucleus
A single copy of rcDNA was detected in 24 (57%) of the 42 rcDNA-positive nuclei in mouse 831, and the remaining 18 (43%) nuclei had 2-11 copies/nucleus (Fig. 6D). In mouse 987, a single copy of rcDNA was detected in 41 (60%) of the 66 rcDNA-positive nuclei, while the remaining 25 rcDNA-positive nuclei contained 2-8 copies/cell. In mouse 907, rcDNA was detected as a single copy in 13 (52%) of the 25 rcDNA-positive nuclei, and the remaining 12 (48%) nuclei had >1 copy of rcDNA, ranging from 2 to 19 copies/ nucleus.
## cccDNA-/rcDNA+ nuclei
Despite the potential to underestimate rcDNA copies in the nuclei by our assay, a portion of infected cells from 27%, 47%, to 55% of the nuclei among the three livers (Table 2) had no detectable cccDNA, whereas rcDNA was detectable in the same nuclei (Fig. 6E). Ours (Fig. 2A andB) as well as published infection kinetics data (44) in this model show that peak infections are usually reached on days 82-90 pi, implying that all infectible human liver cells are likely already infected prior to days 90 pi. HBsAg and HBcAg staining showed that most cells were positive (Fig. 2D and6F) in mice 831, 987, and 907 sections. Thus, cccDNA-/rcDNA+ cells likely represent either a loss of cccDNA from infected cells in which rcDNA was delivered through recycling or cccDNA is yet to be formed with the rcDNA delivered from de novo infection in recently generated uninfected cells after cccDNA loss. The detected cccDNA-/rcDNA+ nuclei corroborated our bulk cell-based finding that cccDNA may have been spontaneously lost from a portion of the infected cells. The cccDNA loss observed during the maintenance phases of HBV infection, along with the presence of cccDNA-/rcDNA+ nuclei in three livers harvested after peak infection, suggests ongoing dynamic cccDNA turnover.
## Heterogeneous HBV-infected population demonstrated by FACS and ABQ duplexing dPCR-based analysis
To further investigate cccDNA status in infected cells, we isolated HBsAg-positive cells through perfusion of the livers of two untreated HBV-infected chimeric mice (Mice ID 516 and 518) on day 228 pi. HBV infection reached the peak on day 84 pi and became persistent and maintained steady levels thereafter in these two mice (Fig. 7A andB). Isolated cells were stained with a rabbit anti-HBs antibody (LSBio LS-C683282), followed by a FITC-conjugated goat anti-rabbit IgG (LSBio LS-C60878) to gate FITC-positive cells for sorting. We first analyzed the distribution pattern of isolated cell populations from two uninfected mice with humanized livers (Mice ID 4607 and 4610 provided by Phoenix Bio) and two infected mice (ID 504 and 514) using standard side scatter (SSC-A) and forward scatter (FSC-A). In contrast to uninfected cells that are mainly distributed in the diagonally left region, the distribution of infected cells was spread into the diagonally right region, indicating an increase in both cellular granularity and size in a portion of infected cells. Notably, much larger sizes with twofold to threefold higher FSC values among the infected cells in the diagonally right region were detected in both infected mice compared to uninfected mice (Fig. 7C andD). Infected cells with different levels of virions, subviral particles, and capsids, factors that impact the extent of cellular granularity and size, are expected to have broader SSC and FSC values. There was also a major dense population with lower SSC but stretched along FSC to some extent (arrow in Fig. 7C andD). Therefore, a broader distribution of infected cell populations likely reflects a heterogeneous HBV-infected population with varying levels of viral components. Such a wider distribution of infected cells was also detected in mice 516 and 518 (Fig. 7E showing mouse 518 FACS). Cells in Fig. 7E were further divided into three populations (P2-P4) based on increasing SSC values (Fig. 7F). HBsAg-positive cells within each of the P2-P4 populations were further gated based on FITC signals, resulting in P5-P7 (Fig. 7G1 through G3). Cells with higher SSC values indeed exhibited higher mean fluorescence intensity (MFI) (Table 3), indicating increased FITC intensity, that is, relatively higher HBsAg levels with the increased SSC values.
To assess cccDNA status in HBsAg-positive cells from the P5 and P6 populations, gated HBsAg-positive cells were individually sorted into wells of 96-well plates. Due to the limited availability of cell suspension, P7 cells could not be collected. The DNA released from each well was then analyzed using ABQ duplexing dPCR to simultaneously detect both cccDNA and rcDNA within the same cells.
The cccDNA status in HBsAg-positive cells differed significantly between the P5 and P6 populations (Table 4).
In P5, most HBsAg-positive cells (67%) had detectable rcDNA but lacked detecta ble cccDNA (HBsAg+/rcDNA+/cccDNA-; Table 4; Fig. 7H). This pattern resembles the above-described nuclei that were rcDNA+/cccDNA-, supporting our finding that cccDNA was absent in a subset of infected cells. It also indicates that cccDNA could be the first viral DNA species cleared from these cells.
In P6, most HBsAg-positive cells (78%) contained detectable cccDNA but no detectable rcDNA (Table 4; Fig. 7I). This pattern may resemble the nuclei containing only cccDNA. These cccDNA-positive cells contained higher levels of HBsAg without detectable rcDNA, which could have been efficiently secreted as virions.
The difference in cccDNA-only and rcDNA-only cell populations between P5 and P6 was statistically significant (P = 9.14E-14; Table 4). The distinct patterns of HBsAg-positive cells observed in P5 and P6 provide additional support for our assay.
Combined results from FACS and duplexing dPCR revealed diverse states of viral components in infected cells isolated from the livers with persistent infection. They also provide new evidence that a subset of HBV-infected cells, which were positive for HBsAg and rcDNA, did not have detectable cccDNA. Given that cccDNA was absent in these freshly isolated HBsAg-positive cells, distinct from dead cells or debris per FACS gating, this finding suggests that the detected cccDNA loss in these cells was unlikely due to liver injury.
## Infrequent human Ki67 RNA expression indicated a low rate of human liver cell proliferation in humanized livers
Human Ki67 was chosen as a marker to examine the proliferation of human liver cells in our model and its potential association with observed cccDNA loss in the humanized livers of chimeric mice. This selection was made based on the previous staining of human Ki67 protein in the nuclei of human hepatocytes for proliferation studies within the same model (45). Ki67 RNA levels were measured using RT-qPCR. HepG2 cells and their derivative HepG2.2.15 cells are considered moderately proliferative, typically doubling in number every 2 days in culture. We used their cellular RNA as a control for Ki67 RNA expression, which averaged 32 copies per cell (n = 3). We then analyzed Ki67 RNA levels in 380 cytoplasmic RNA samples derived from 19 humanized livers across four experiments: four livers collected on day 257 post-infection (pi) from experiment 1, three livers on day 218 pi from experiment 2, eight livers on days 18, 45, 50, 52, 82, 99, 141, and 212 pi from experiment 3, and four livers on day 230 pi from experiment 4.
Ki67 RNA levels were below one copy per cell in 355 of 380 samples (93%), and above one copy per cell in the remaining 25 samples (7%). Of these, 20 samples were from mouse 834, 4 from mouse 806, and 1 from mouse 905 (Fig. 8A).
At the liver level, the average Ki67 RNA was <1 copy per cell in 18 of the 19 liv ers, representing a 44 to 1,748 folds lower than HepG2/HepG2.2.15 cells (Fig. 8B). The remaining mouse, 834, exhibited an average of 4.1 copies per cell, still about eight-fold lower than HepG2/HepG2.2.15 cells.
These results indicate that Ki67 RNA expression was infrequent both across different time points during infection and among the four experimental cohorts.
If cell proliferation were a primary factor driving cccDNA loss, one would anticipate an inverse correlation between Ki67 RNA and cccDNA levels. However, no such correlations a Chi-square value (X 2 ) is 55.54 and P-value is 9.14 × 10 -14 in comparison of the differences in the detected cccDNA only and rcDNA only containing cells between P5 and P6 populations.
were detected between Ki67 RNA and cccDNA levels among 160 samples from 8 livers collected on days 18, 45, 50, 52, 82, 99, 14, and 212 pi, respectively (Fig. 8C).
## Normal serum alanine transaminase levels indicated the absence of signifi cant liver injury during the infection course
Alanine transaminase (ALT) activity was assessed in serial serum samples from three untreated mice. ALT activity exhibited fluctuations over a 6-month duration, spanning from day 21 to day 207 pi, and remained below 40 U/L in all samples (Fig. 8D). This suggests that the observed reduction in cccDNA level occurred in the absence of significant liver injury. We also monitored the potential loss of human liver cells in humanized livers by tracking kinetic serum human albumin levels, which remained steady (Fig. 8E). These findings are consistent with the observed low levels of human Ki67 RNA in the liver and normal serum ALT levels.
## DISCUSSION
We acknowledge that HBV cccDNA is conventionally viewed as stable in infected cells (5). Thus, cccDNA persistence in HBV infection can be simply explained by the longevity of cccDNA molecules. However, we also appreciate the extensive reports of clinical observations demonstrating the replacement of wild type with mutant population under treated or untreated conditions (9)(10)(11)(12)(13). Such turnover of viral population is also documented in the DHBV persistent infection model (46,47). Furthermore, spontaneous cccDNA loss in DHBV-infected cultures was independently observed by two groups (48,49). Thus, a dynamic cccDNA state is shared among members of the Hepadnaviri dae family. Consistent with the clinically observed frequent cccDNA turnover (9-13), NA-treated human patients have demonstrated a 1-2.9 log reduction in cccDNA levels (50)(51)(52)(53)(54)(55). Similar cccDNA reduction responses to NA therapy were also described in WHV-infected woodchucks (56), DHBV-infected ducks (41,57), and HBV-infected uPA/ SCID chimeric mice (45). Thus, the observed cccDNA reduction occurs across different species, constituting a broad biological base for this phenomenon. Since NAs do not directly target cccDNA molecules, these observations suggest that cccDNA may be spontaneously cleared from infected cells that did not experience cytopathic effects, in addition to a subset of cells that may have undergone cytopathic effects or spontaneous turnover that lost cccDNA through cell destruction or division.
Therefore, the notion of cccDNA longevity could not explain the dynamic state of cccDNA observed across several hepadnavirus species.
In this study, we detected spontaneous cccDNA loss through analysis of both bulk cells and single nuclei and single HBsAg-positive cells isolated from humanized livers of chimeric mice. Our findings confirm the observed cccDNA loss/reduction in human HBV infection (9)(10)(11)(12)(13) as well as in WHV (56) and DHBV infection models (41,57) and in uPA/SCID chimeric mice model ( 45) under treated and untreated conditions.
Our findings do not dispute an overall stable cccDNA level established after peak infection (Fig. 2C); rather, they provide an alternative explanation for how the persis tence of cccDNA is maintained. Dynamic changes take place underneath the cccDNA persistence, that is, the early cccDNA pool is cleared either noncytopathically or through cell death/division, then the lost cccDNA pool is replenished. Therefore, the steady cccDNA level or cccDNA persistence is maintained by ongoing cccDNA replenishment after its loss (Fig. 9). We have also established preclinical proof that blocking cccDNA replenishment leads to progressive cccDNA elimination by >100-fold from infected livers (manuscript in preparation). Such therapeutic/functional evidence supports the notion that cccDNA persistence is mainly maintained by its ongoing replenishment, and the property of spontaneous cccDNA loss can be leveraged into a progressive cccDNA elimination.
Our findings appear to be able to explain both cccDNA persistence in CHB and clinically observed dynamic cccDNA clearance/replacement. They may also offer insights into why most infected hepatocytes avoid cytopathic consequences despite supporting early robust viral replication.
Liver injury certainly causes cccDNA loss, especially in untreated patients with CHB. However, the main scenario is different among the treated patients. For instance, the kinetics progressively replacing WT viral population were in parallel with progressively normalizing ALT kinetics, and the spread of the drug-resistant mutant occurred during a normal ALT period in four patients treated with lamivudine (14), suggesting that the observed replacements are probably independent of cell injury. This also seems to be held true for the observed 1-2.9 log cccDNA reduction (50)(51)(52)(53)(54)(55) during NAs treatment that normalizes ALT levels in >80% treated patients, though spontaneous cell turnover or injury that led to cccDNA loss may still have occurred in a subset of cells of their livers.
Our findings, including largely normal histology, consistently normal ALT levels, infrequent Ki67 RNA expression, stable human albumin levels over infection course, and the presence of HBsAg-positive/rcDNA-positive cells lacking detectable cccDNA, suggest that spontaneous cccDNA loss occurs primarily through non-cytolytic mechanisms (58) in most cells that avoided cytopathic consequences, with a minor contribution from cytopathic clearance in a subset of cells with cytopathic effects during persistent HBV infection established in this chimeric mouse model. We observed that the intracellular accumulation of viral products was likely halted at the late stage of infection (Fig. 2C). Our explanation is that HBV replication was probably inhibited. The analysis of levels of cccDNA, viral RNAs, and viral proteins indicates that the suggested inhibition may result from spontaneous cccDNA loss. This is why we assume the observed cccDNA loss is mainly driven by ongoing replication that causes progressive accumulation of viral products, which may successfully trigger the inhibition of replication at the cccDNA level, as reported in DHBV model (30,33). In the case of unsuccessful inhibition of viral replication, cccDNA loss may also occur through cell injury in a subset of cells.
The limitation of this study is the lack of an estimate for the proportion of cccDNA loss attributable to cell injury and spontaneous cell turnover. Future studies are needed to further investigate the contribution of non-replication-related factors, such as natural hepatocyte turnover and proliferation to cccDNA decline in this model. A definitive measurement of cccDNA turnover in vivo, ideally at the single-cell level, is an important next step to generate additional evidence for cccDNA loss.
This study employed the Absolute Q digital PCR (ABQ dPCR) instrument to quantify cccDNA copies at single-nucleus and HBsAg-positive cell level. The ABQ dPCR is capable of detecting a single copy of the target gene. The dPCR assay is particularly suitable for detecting a low abundance of target molecules like cccDNA, as they can be individually distributed in a total of 20,480 microchambers. Each microchamber either contains zero or a single copy of cccDNA, reduces competition among templates, dilutes out inhibitory factors, and minimizes variability.
While we are confident in the single-copy detection sensitivity of ABQ dPCR, we cannot entirely rule out the possibility that some undetectable cccDNA in certain cells may reflect the limits of detection.
Early in our study, we carefully investigated all potential technical sources of variability, including reagent quality, extraction protocols, sample handling and storage conditions, and qPCR assay performance and standard calibration, and found no evidence of procedural inconsistencies that could account for the observed differences.
Importantly, our flow cytometry-based analysis of HBsAg-positive cells revealed two distinct subpopulations, suggesting that the absence of detectable cccDNA is not solely due to technical limitations.
Average cccDNA levels varied considerably among 11 livers. We believe the variability is more likely due to biological and virological factors, particularly heterogeneity in the transcriptional activity of individual infected hepatocytes. Specifically, the relative usage of HBV promoters (core vs. surface) may differ among cells, leading to divergent transcriptional profiles. As observed in human liver biopsies, some cells are core-positive only (indicating preferential pgRNA transcription, consequently higher rcDNA levels, and potentially more active rcDNA-to-cccDNA recycling or higher cccDNA levels), while other cells exhibit only HBsAg staining (likely reflecting dominant S RNA transcription and limited pgRNA expression, resulting in lower rcDNA and cccDNA levels) (59). The proportions of these transcriptionally distinct hepatocyte subpopulations may vary among animals, contributing to the liver-to-liver variability in average cccDNA levels.
Our FACS results revealed that a portion of infected liver cells was widely distrib uted in the diagonally right regions compared to an uninfected liver cells popula tion, suggesting a heterogeneous infected cell population with varying levels of viral products. In addition, we observed that only about 7% of HBsAg-positive cells con tained both cccDNA and rcDNA. This pattern is reproducible when sorting from the same regions with low SSC values from two additional infected humanized livers. We acknowledge, however, that cells from other regions were not gated and analyzed in this study, and the proportions may differ. Interestingly, this finding resembles immunohisto chemical observations in liver sections from both HBeAg-positive and -negative patients, where fewer than 10% of infected hepatocytes showed dual positivity for HBsAg and HBc proteins, and only detected in HBeAg-positive patients (59). This supports the possibility that most infected cells are mutually exclusive in terms of expressing HBsAg and HBc, potentially reflecting mutually exclusive transcription of pgRNA versus S RNA. Such exclusiveness may partially explain why only a small proportion of HBsAg-positive cells harbor both cccDNA and rcDNA. These observations may also suggest that turnover of cccDNA and rcDNA in infected hepatocytes could be rapid, implying that persistent HBV infection could be interrupted and terminated if new rounds of infection are durably blocked, which worked efficiently when this new strategy was tested in our chimeric mouse model (manuscript in preparation). However, our analyses were preliminary and limited to establishing proof for the absence of detectable cccDNA in some HBsAg+/ rcDNA+ cells. More detailed analysis is warranted in future studies.
Our findings prompt us to rethink the cccDNA elimination strategy. Directly targeting cccDNA or killing infected cells is currently thought to be the only pathway for cccDNA elimination and complete cure of chronic HBV infection. Our results provide an additional strategy that does not require direct targeting of cccDNA molecules and may achieve cccDNA elimination through durable blocking of cccDNA replenishment.
## MATERIALS AND METHODS
## Animals and HBV infection
uPA/SCID chimeric mice were supplied by PhoenixBio USA (New York, NY, USA). All mice were kept in housing cages (TP107, One Corporation, Osaka, Japan) in a BSL-2 room with controlled temperature at 23°C and 12 hour-light/dark cycle. All animals were fed with γ-radiated CRF1 food and autoclaved water ad libitum. An HBV inoculum, prepared from mouse serum (project no H01-108 animal 4) by diluting viremia of 5E9 HBV DNA copies/mL to 2E7 HBV DNA copies with PBS in 100 µL volume, was administered intravenously (tail vein) to each chimeric mouse.
## Monitoring HBV infection and human albumin level in blood
Blood was collected tri-weekly for quantification of serum HBV DNA (qPCR, see below), HBsAg (GS HBsAg EIA 32591, Bio-Rad), and human albumin (Human albumin ELISA kit E-80AL, Immunology Consultants Laboratory) levels by ELISA per instructions.
## ALT activity in serum
Serum ALT activity was assessed using the alanine transaminase colorimetric assay kit (Cayman Chemical item no 700260) according to the detection manual. Absorbance values were measured at 340 nm once per minute for 10 min, and the resulting 10 absorbance values were plotted against time. Due to limited serum volumes, a modifica tion was made: the 20 µL serum sample was adjusted to 10 µL and compensated with 10 µL of H2O. Consequently, in the calculation formula, the 0.02 mL of serum sample was adjusted to 0.01 mL accordingly.
## Analysis of intrahepatic HBV DNA
Each liver was randomly sampled 20-40 times by cutting 20-40 mg liver tissue (weighed and recorded) and placed in a disposable micro-homogenizer (BioMasher, Takara cat no: 9790B) in 500 µL of an isotonic buffer (154 mM Tris-HCl, pH 7.5, 1 mM EDTA, and 0.05% Triton X-100) with 10 strokes. The homogenized tissue suspension was spun for 2 min at 14,000 rpm. The aqueous phase (about 400 µL) was transferred to a new microtube for isolation of replicative intermediates (RI) while the nucleic pellet remained in the tube for cccDNA isolation.
## Extraction of rcDNA from the 400 µL supernatant by following procedures
Mix lysates with 110 µL of proteinase K (final concentration: 0.5 mg/mL) and 1% SDS, and incubate at 50°C for 1 h. Add 500 µL phenol, vortex, chill on ice for 3 min, and centrifuge at 14,000 rpm for 2 min. Transfer the supernatant to a new tube, add 1,000 µL of 100% ethanol for precipitation, and centrifuge at 14,000 rpm for 15 min. Wash the pellet with 1,000 µL of 100% ethanol, followed by centrifugation at 14,000 rpm for 10 min. Remove residual ethanol and air-dry the pellet for 5 min. Dissolve the pellet in 200 µL of 10:1 TE buffer (pH 7.4). The resulting rcDNA is ready for qPCR analysis.
## Extraction of cccDNA from nucleic pellets by following procedures
The nuclear pellet was suspended in 200 µL of 10:1 TE buffer containing 0.05% Triton X-100 (pH 7.4), followed by the addition of 200 µL of 6% SDS-0.1 M NaOH solution and incubation at 37°C for 15 min. Next, 100 µL of 3 M potassium acetate (pH 5.07) was added, mixed thoroughly, chilled on ice for 5 min, and centrifuged at 14,000 rpm for 2 min to remove KSDS-protein-ssDNA complexes. The supernatant was transferred to a new tube, extracted with 500 µL of phenol, and centrifuged again at 14,000 rpm for 2 min. The recovered supernatant was supplemented with 5 µL of glycogen (4 µg/µL, total 20 µg), followed by the addition of 1,000 µL ethanol and centrifugation at 14,000 rpm for 15 min. The pellet was washed with 1,000 µl ethanol, centrifuged at 14,000 rpm for 10 min, and then dissolved in 100 µL EcoRI buffer at 37°C for 15 min before heat inactivation at 80°C for 20 min. The resulting cccDNA samples were used for qPCR.
## RT-qPCR detection of total HBV RNA in cytoplasmic samples
Total HBV RNA levels were determined in each of 80 rcDNA samples prepared from untreated mice 842 and 836 (representing the amplification phase, N = 40), 831 and 987 (representing the maintenance phase, N = 40). Specifically, 20 cccDNA and 20 rcDNA samples from each liver were tested. Average RNA concentrations were approximately 0.5 µg/µL in rcDNA samples. A260/280 ratios varied narrowly between 1.98 and 2.08. All RNA samples were 10-fold diluted and then 2 µL were used for RT-qPCR with TaqMan Fast Virus 1-Step Master Mix (Thermo Fisher 4444432) and primers/probe located in the S gene (rcDNA for qPCR in Table 3). rcDNA was also detected in the same rcDNA samples without RT in the same plates with the same primers/probe for RNA detection. The detected HBV RNA levels were approximately 10 times higher than those of rcDNA in the same samples. The net RNA copies are plotted after subtracting rcDNA copies in the same samples. The ratios of RNA copies/cell to cccDNA copies/cell were calculated using the total net RNA copies (nuclear RNA copies + cytoplasmic RNA copies).
## RT-qPCR detection of human Ki67 RNA levels in 380 cytoplasmic samples
One set of pre-stocked human Ki67 RNA primers/probe system (FAM-MGB, Hs01032435_g1) was purchased from ThermoFisher Scientific. This detection system generates a 179 bp-long amplicon that was isolated for the preparation of qPCR standards. Two microliter from each cytoplasmic rcDNA sample containing cellular RNAs was used for RT-qPCR detection of human Ki67 RNA with the same TaqMan Fast Virus 1-Step Master Mix (Thermo Fisher 4444432). Correlation analysis between Ki67 RNA and cccDNA in the same samples was conducted using scatter plots, generating correlation trendlines and R 2 values.
The rcDNA samples were consistently stored at -20°C, and all handling procedures were conducted in a biosafety cabinet with an air blower on. All tips, plates, tubes, and solutions used were nuclease-free. The Ki67 RNA levels in 20 rcDNA samples
## qPCR of serum HBV DNA and intrahepatic rcDNA and cccDNA
Primers and probe sequences for the detection of serum HBV DNA and intracellular rcDNA by qPCR are listed in Table 5, while primer sequences for the detection of cccDNA flank the gap region, and the probe is placed immediately after the DR1 sequence (Table 6). The specificity of the listed cccDNA primers and probe can discriminate against rcDNA amplification by 300-fold to 6,000-fold. qPCR was performed using TaqMan fast advanced master mix (ThermoFisher cat no:4444558) in a QuantStudio 3 instrument (ThermoFisher cat no: A28136) that accommodates 0.1 mL 96-well hard-shell plate. All standards used for qPCR were calibrated with the Absolute Q digital PCR.
## Absolute Q (ABQ) digital PCR of cccDNA
Absolute Q (ABQ) digital PCR of cccDNA was performed to validate cccDNA copies/cell initially computed by qPCR, using the ThermoFisher Absolute Q Digital PCR system (cat. no. A52864). Briefly, a 9.1 µL reaction mix was prepared consisting of 1.8 µL of 5 × DNA dPCR mix (ThermoFisher cat. no. A52490), 0.5 µL of 20 × primers/probe mix (final concentrations: 900 nM each primer and 250 nM probe), 1 µL of cccDNA sample, and 5.8 µL of DNase-and RNase-free water. A 9 µL aliquot of the reaction mix was then loaded into a single well of a microfluidic array plate (MAP; ThermoFisher cat. no. A53301). Digital PCR was carried out with the following cycling conditions: 10 min preheating at 96°C, followed by 40 cycles of 5 s at 96°C and 15 s at 60°C. Data reports were generated using QuantStudio Absolute Q Digital PCR software. The sensitivity of dPCR is a single copy per microchamber, and the result is deemed valid if the Rox fluorescent signal was read in >19,000 of 20,480 microchambers for each sample.
## Procedures and principles of simultaneous detection of cccDNA and rcDNA by ABQ duplexing digital PCR in the same nuclei
For detection of both cccDNA and rcDNA in the same nuclei, 20-30 mg of liver tissue was homogenized in 500 µL of homogenization buffer (10 mM Tris-HCl (pH 7.5), 3 mM MgCl₂, 0.25 M sucrose, and 0.05% Triton X-100), and nuclei were pelleted by centrifuga tion and resuspended in homogenization buffer containing 2 µg/mL ethidium bromide. Individual nuclei were then singly sorted and deposited into wells of 96-well plates, digested with proteinase K (0.5 mg/mL, 60 min), and heat-inactivated at 80°C for 15 min. The released HBV DNA was subsequently linearized by NcoI digestion and subjected to ABQ duplexing dPCR for simultaneous detection of cccDNA and rcDNA.
## Principles of simultaneous detection of cccDNA and rcDNA by ABQ duplexing digital PCR (ABQ dPCR) in the same nuclei
The ABQ dPCR instrument can simultaneously detect four fluorescent signals of FAM, VIC, ABY, and JUN/Cy5, which allows detecting four different targets in the same reaction (Multiplexing). The emission wavelengths of FAM and Cy5 are 517 and 670 nm, respectively, and there is no overlap in their wavelength spectrum. Thus, FAM was selected for labeling the cccDNA probe and Cy5 for labeling the rcDNA probe to detect both molecules in the same reaction, called duplexing dPCR. The specificity of cccDNA detection is provided through cccDNAspecific primers that flank the gap region in the HBV genome and the cccDNAspecific probe that is placed immediately after the DR I sequence (Table 6). Figure 5E shows an example of its specificity for cccDNA and rcDNA detection.
Linearized cccDNA template cannot generate a fluorescent signal with the rcDNA primers/probe detection system HBV DNA released from each of the deposited nuclei was subjected to NcoI digestion to exclude cccDNA from detection with rcDNA primers/probe. Plus strand in rcDNA is only partially synthesized, containing a single-stranded gap of 600-2,100 nucleotides at 3′ end (60). Since NcoI is located between nt1372 and 1376, close to the 3′ end of the plus strand. Therefore, it is most likely present in a single-strand sequence in rcDNA molecules (60). Thus, NcoI will linearize cccDNA but cannot cut rcDNA (Fig. 5A andB).
The NcoI linearized cccDNA sequence starts with C at nt1373 (5′) and ends with C at nt1372 (3') (Fig. 5C). The rcDNA forward primer will bind the 3′ end of the linearized cccDNA, but the rcDNA probe binds its 5′ end. The Taq DNA polymerase that binds the forward primer at the 3′ end cannot reach the probe that is located at the 5′ end, thus cannot cut off the 1st base C with Cy5 dye through its 5′-3′ exonuclease activity, and the Cy5 fluorescent signal cannot be generated (Fig. 5D). Cy5 fluorescent signal will be generated if both rcDNA forward primer and probe bind a continuous template comprising nt1345 to nt1454 sequentially, that is, F primer binds upstream of the probe binding position (Fig. 5C), which only occurs in rcDNA after NcoI cut. Thus, rcDNA, but not cccDNA, will be specifically detected with rcDNA primers/probe following NcoI cut.
We used an HBV DNA plasmid (an ADW subtype monomer cloned into the Psp65 vector) that was linearized by NcoI as a surrogate cccDNA molecule to validate that cccDNA was not detected by the rcDNA probe/primers. When only the cccDNA sample (approximately 8-10 copies/μL) was used for the test, 8 copies of cccDNA molecules were only detected by the FAM-labeled cccDNA probe; however, no positive signal was detected by the Cy5-labeled rcDNA probe in duplexing dPCR (Fig. 5E1A andB). When only rcDNA (approximately 30 copies of rcDNA/μL, extracted from an infected liver) was included for test, 27 copies of rcDNA were detected only by Cy5-labeled rcDNA probe but not by FAM-labeled cccDNA probe in duplexing dPCR (Fig. 5E2A andB). When the above cccDNA and rcDNA samples were mixed for test, 8 copies of cccDNA and 30 copies of rcDNA were detected by FAM-labeled cccDNA probe and Cy5-labeled rcDNA probe, respectively, in duplexing dPCR (Fig. 5E3A andB).
## Detected rcDNA molecules in the nuclei were not non-specifically bound to nuclei
To evaluate the possibility that the detected rcDNA in sorted individual nuclei was nonspecifically bound to the nuclear membrane during the preparation of nuclei suspension through homogenization that released virions and capsids into the lysate, we prepared nuclei suspensions from two livers of two uninfected chimeric mice (animal ID HKB-043-020 or B20 and HKB-043-046 or B46 purchased from PheonixBio). Each nuclei suspension was divided into two vials; one was directly used for sorting, and the other was mixed with lysate (containing 0.05% Triton X-100) from mouse 987, who was an untreated control with average 870 copies of rcDNA/cell for 20 minutes, then removed the lysate and dissolved in isotonic buffer (154:1 TE with 0.05% Triton X-100) for sorting. The sorted nuclei from four vials were subjected to duplexing ABQ dPCR. Table 4 shows no significant differences in detecting nuclei with Cy5 intensity ≥500 between two nuclei suspensions mixed with mouse 987 lysate and the two nuclei suspensions without mixing, suggesting that detected rcDNA molecules in sorted nuclei likely derived from released virions and capsids that nonspecifically bound nuclei, which is consistent with the concept that HBV capsid mainly utilizes cellular transport machineries, but not diffusion or passive trapping to reach nuclear membrane where interaction between nuclear localization signal on capsid and nuclear import receptors occurs (61, 62) (Table 5).
## Isolating individual HBsAg-positive cells via FACS
Individual hepatocytes were isolated from each liver of mice 516 and 518 on day 218 pi using a two-step collagenase perfusion method (63). Each liver was perfused at 37°C for 10 min at 1.5 mL/min with Ca 2+ -free and Mg 2+ -free Hanks' balanced salt solution (CMF-HBSS) containing 200 mg/mL ethylene glycol tetraacetic acid (EGTA), 1 mg/mL glucose, 10 mM N-2-hydroxyethylpiperazine-N′-2-ethane sulfonic acid (HEPES). The perfusion solution was then changed to CMF-HBSS containing 0.05% collagenase (Gibco 17101-015), 0.6 mg/mL CaCl2, 10 mM HEPES, and continued for 17-23 min at 1.5 mL/min. The liver was dissected and transferred to a dish; liver cells will be gently disaggregated in the dish with CMF-HBSS containing 10% bovine Alb, 10 mM HEPES. The disassociated cells were centrifuged three times (50 × g, 5 min). The pellet was suspended in PBS with 0.05% Triton X-100.
Cell concentration in the suspension was adjusted to 5 × 10 6 /mL, was first stained with 1/100 diluted rabbit anti-HBs antibody (LSBio LS-C683282), then with 1/500 diluted goat anti-rabbit IgG conjugated with FITC (LSBio LS-C60878). After washing, the stained cells were suspended in PBS with 0.05% Triton X-100 for sorting.
Before sorting, a FACS analysis using SSC-A and FSC-A was performed with the Aurora CS Sorter located at the Flow Cytometry Shared Service, University of Maryland, School of Medicine in Baltimore, MD. The gating strategy included gating cells first based on SSC-A values, then using FITC signals to gate HBsAg-positive cells for sorting. Individual HBsAg-positive cells were singly sorted into wells of 96-well plates. The sorted cells were subjected to the same treatment as described for the sorted nuclei before duplexing dPCR detection.
## Immunohistochemical staining of HBsAg and HBcAg on sections
Briefly, formalinfixed paraffinembedded liver sections were cut at a thickness of 5 µM and used for HBsAg staining after deparaffinization, proteinase K digestion, and inactivation of endogenous peroxidase with 3% hydrogen peroxide. Rabbit anti-HBs antibody (LS-C683282, LSBio) or rabbit anti-HBcAg antibody (LS-C170914) and goat anti-rabbit IgG conjugated with HRP (LS-C316062, LSBio) were used as primary and secondary antibodies, respectively. DAB chromogen kit (ACH500-IFU, CP Lab Chemicals) was used for color development.
## Calculation of HBsAg copies/cell
To calculate copy numbers of HBsAg per cell, cellular HBsAg in liver lysates was determined with qHBsAg ELISA, resulting in IU per 500 µL lysate, which was converted to ng/500 µL, then to copies/cell.
Major parameters for the calculation include the following:
i. 1 IU of HBsAg ADR subtype is approximately equal to 1 ng (HBsAg ELISA manual, Bio-rad). ii. 1 ng of HBsAg equals 2.74e8 copies using 2.16 MDa molecular weight of a HBsAg particle (64) iii. 1 mg of liver tissue contains approximately 1.39e5 cells (65)
## Statistical analysis
HBsAg (IU/mL), HBV DNA (copies/mL), and antibody levels (µg/mL) are expressed as mean ± standard deviation (SD). Average intracellular rcDNA and cccDNA levels are expressed as copies/cell. The number of cells per sampling was calculated using sample weight (mg) multiplied by 1.39E5 cells per mg liver tissue (65), then normalized by a factor of 0.7 by considering 70% of liver cells are human hepatocytes based on the Replacement Index (76%-93%) among the chimeric mice provided by PheonixBio. The formula to calculate copies/cell is listed below:
Copies/cell = Total rcDNA or cccDNA copies in a liver sample Sample weight mg × 1.39e5 × 0.7
## References
1. Summers (1981) "Three recently described animal virus models for human hepatitis B virus" *Hepatology*
2. Seeger, Mason, Lai (2020) "Molecular biology of hepatitis viruses"
3. Summers, Mason (1982) "Replication of the genome of a hepatitis B--like virus by reverse transcription of an RNA intermediate" *Cell*
4. Tuttleman, Pourcel, Summers (1986) "Formation of the pool of covalently closed circular viral DNA in hepadnavirus-infected cells" *Cell*
5. Alter, Block, Brown et al. (2018) "A research agenda for curing chronic hepatitis B virus infection" *Hepatology*
6. Seto, Lo, Pawlotsky et al. (2018) "Chronic hepatitis B virus infection" *Lancet*
7. Guidotti, Chisari (2006) "Immunobiology and pathogenesis of viral hepatitis" *Annu Rev Pathol*
8. Fanning, Zoulim, Hou et al. (2019) "Therapeutic strategies for hepatitis B virus infection: towards a cure" *Nat Rev Drug Discov*
9. Brunetto, Giarin, Oliveri et al. (1991) "Wild-type and e antigen-minus Full-Length Text Journal of Virology November"
10. "hepatitis B viruses and course of chronic hepatitis" *Proc Natl Acad Sci*
11. Carman, Jacyna, Hadziyannis et al. (1989) "Mutation preventing formation of hepatitis B e antigen in patients with chronic hepatitis B infection" *Lancet*
12. Jiang, Wu, Kuhnhenn et al. (2019) "Formation of semi-enveloped particles as a unique feature of a hepatitis B virus PreS1 deletion mutant" *Aliment Pharmacol Ther*
13. Chen, Jia, Wang et al. (2022) "Analysis of entire hepatitis B virus genomes reveals reversion of mutations to wild type in natural infection, a 15 year follow-up study" *Infect Genet Evol*
14. Huang, Zhou, Cai et al. (2021) "Rapid turnover of hepatitis B virus covalently closed circular DNA indicated by monitoring emergence and reversion of signature-mutation in treated chronic hepatitis B patients" *Hepatology*
15. Lau, Khokhar, Doo et al. (2000) "Long-term therapy of chronic hepatitis B with lamivudine" *Hepatology*
16. Albrecht, Fons, Boldogh et al. (1996) "Effects on cells"
17. (2007) "Cytopathic effects of viruses protocols"
18. Koyuncu, Hogue, Enquist (2013) "Virus infections in the nervous system" *Cell Host Microbe*
19. Louten (2022) "Virus replication"
20. Stevens, Cook (1971) "Latent herpes simplex virus in spinal ganglia of mice" *Science*
21. Rock, Fraser (1983) "Detection of HSV-1 genome in central nervous system of latently infected mice" *Nature*
22. Jordan, Bisgrove, Verdin (2003) "HIV reproducibly establishes a latent infection after acute infection of T cells in vitro" *EMBO J*
23. Karn, Stoltzfus (2012) "Transcriptional and posttranscriptional regulation of HIV-1 gene expression" *Cold Spring Harb Perspect Med*
24. Naoumov, Portmann, Tedder et al. (1990) "Detection of hepatitis B virus antigens in liver tissue. A relation to viral replication and histology in chronic hepatitis B infection" *Gastroenterology*
25. Chu, Liaw (1995) "Membrane staining for hepatitis B surface antigen on hepatocytes: a sensitive and specific marker of active viral replication in hepatitis B" *J Clin Pathol*
26. Macsween, Anthony, Scheuer et al. (1994) *Pathology of the liver*
27. Chisari, Filippi, Buras et al. (1987) "Structural and pathological effects of synthesis of hepatitis B virus large envelope polypeptide in transgenic mice" *Proc Natl Acad Sci*
29. Hadziyannis, Gerber, Vissoulis et al. (1973) "Cytoplasmic hepatitis B antigen in "ground-glass" hepatocytes of carriers" *Arch Pathol*
30. Gerber, Hadziyannis, Vissoulis et al. (1974) "Electron microscopy and immunoelectronmicroscopy of cytoplasmic hepatitis B antigen in hepatocytes" *Am J Pathol*
31. Mason, Seal, Summers (1980) "Virus of Pekin ducks with structural and biological relatedness to human hepatitis B virus" *J Virol*
32. Lenhoff, Summers (1994) "Construction of avian hepadnavirus variants with enhanced replication and cytopathicity in primary hepatocytes" *J Virol*
33. Lenhoff, Luscombe, Summers (1999) "Acute liver injury following infection with a cytopathic strain of duck hepatitis B virus" *Hepatology*
34. Ganem, Prince (2004) "Hepatitis B virus infection--natural history and clinical consequences" *N Engl J Med*
35. Summers, Smith, Huang et al. (1991) "Morphogenetic and regulatory effects of mutations in the envelope proteins of an avian hepadnavirus" *J Virol*
36. Lentz, Loeb (2011) "Roles of the envelope proteins in the amplification of covalently closed circular DNA and completion of synthesis of the plus-strand DNA in hepatitis B virus" *J Virol*
37. Gao, Hu (2007) "Formation of hepatitis B virus covalently closed circular DNA: removal of genome-linked protein" *J Virol*
38. Tateno, Kawase, Tobita et al. (2015) "Generation of novel chimeric mice with humanized livers by using hemizygous cDNA-uPA/SCID mice" *PLoS One*
39. Ishida, Chung, Imamura et al. (2014) "HBV infection in humanized chimeric mice has multipha-sic viral kinetics from inocula tion to steady state and an HBV half-life of 1 hr: 1715" *Hepatology*
40. Sugiyama, Tanaka, Kurbanov et al. (2009) "Direct cytopathic effects of particular hepatitis B virus genotypes in severe combined immunodefi ciency transgenic with urokinase-type plasminogen activator mouse with human hepatocytes" *Gastroenterology*
41. Keating, Heitman, Wu et al. (2014) "Cytokine and chemokine responses in the acute phase of hepatitis B virus replication in naive and previously vaccinated blood and plasma donors" *J Infect Dis*
42. Zhang, Summers (2004) "Rapid production of neutralizing antibody leads to transient hepadnavirus infection" *J Virol*
43. Zhang, Zhang, Theele et al. (2003) "Singlecell analysis of covalently closed circular DNA copy numbers in a hepadnavirus-infected liver" *Proc Natl Acad Sci*
44. Stadelmayer, Diederichs, Chapus et al. (2020) "Full-length 5'RACE identifies all major HBV transcripts in HBV-infected hepatocytes and patient serum" *J Hepatol*
45. Tuttleman, Pugh, Summers (1986) "In vitro experimental infection of primary duck hepatocyte cultures with duck hepatitis B virus" *J Virol*
46. Ishida, Chung, Imamura et al. (2018) "Acute hepatitis B virus infection in humanized chimeric mice has multiphasic viral kinetics" *Hepatology*
47. Allweiss, Volz, Giersch et al. (2018) "Proliferation of primary human hepatocytes and prevention of hepatitis B virus reinfection efficiently deplete nuclear cccDNA in vivo" *Gut*
48. Zhang, Summers (1999) "Enrichment of a precore-minus mutant of duck hepatitis B virus in experimental mixed infections" *J Virol*
49. Pult, Abbott, Zhang et al. (2001) "Frequency of spontaneous mutations in an avian hepadnavirus infection" *J Virol*
50. Civitico, Locarnini (1994) "The half-life of duck hepatitis B virus supercoiled DNA in congenitally infected primary hepatocyte cultures" *Virology (Auckl)*
51. (2025) *Full-Length Text Journal of Virology*
52. Guo, Wang, Barrasa et al. (2003) "Conditional replication of duck hepatitis B virus in hepatoma cells" *J Virol*
53. Werle-Lapostolle, Bowden, Locarnini et al. (2004) "Persistence of cccDNA during the natural history of chronic hepatitis B and decline during adefovir dipivoxil therapy" *Gastroenterology*
54. Wong, Yuen, Ngai et al. (2006) "One-year entecavir or lamivudine therapy results in reduction of hepatitis B virus intrahepatic covalently closed circular DNA levels" *Antivir Ther (Lond)*
55. Wursthorn, Lutgehetmann, Dandri et al. (2006) "Peginterferon alpha-2b plus adefovir induce strong cccDNA decline and HBsAg reduction in patients with chronic hepatitis B" *Hepatology*
56. Lutgehetmann, Volz, Quaas et al. (2008) "Sequential combination therapy leads to biochemical and histological improvement despite low ongoing intrahepatic hepatitis B virus replication" *Antivir Ther (Lond)*
57. Boyd, Lacombe, Lavocat et al. (2016) "Decay of ccc-DNA marks persistence of intrahepatic viral DNA synthesis under tenofovir in HIV-HBV co-infected patients" *J Hepatol*
58. Lai, Wong, Ip et al. (2017) "Reduction of covalently closed circular DNA with long-term nucleos(t)ide analogue treatment in chronic hepatitis B" *J Hepatol*
59. Zhu, Yamamoto, Cullen et al. (2001) "Kinetics of hepadnavirus loss from the liver during inhibition of viral DNA synthesis" *J Virol*
60. Addison, Walters, Wong et al. (2002) "Half-life of the duck hepatitis B virus covalently closed circular DNA pool in vivo following inhibition of viral replication" *J Virol*
61. Guidotti, Rochford, Chung et al. (1999) "Viral clearance without destruction of infected cells during acute HBV infection" *Science*
62. Aggarwal, Odorizzi, Brodbeck et al. (2023) "Intrahepatic quantification of HBV antigens in chronic hepatitis B reveals heterogene ity and treatment-mediated reductions in HBV core-positive cells" *JHEP Rep*
63. Summers, Connell, Millman (1975) "Genome of hepatitis B virus: restriction enzyme cleavage and structure of DNA extracted from Dane particles" *Proc Natl Acad Sci*
64. Blondot, Bruss, Kann (2016) "Intracellular transport and egress of hepatitis B virus" *J Hepatol*
65. Gallucci, Kann (2017) "Nuclear import of hepatitis B virus capsids and genome" *Viruses*
66. Yamasaki, Kataoka, Kato et al. (2010) "In vitro evaluation of cytochrome P450 and glucuronidation activities in hepatocytes isolated from liver-humanized mice" *Drug Metab Pharmacokinet*
67. Gilbert, Beales, Blond et al. (2005) "Hepatitis B small surface antigen particles are octahedral" *Proc Natl Acad Sci*
68. Sohlenius-Sternbeck (2006) "Determination of the hepatocellularity number for human, dog, rabbit, rat and mouse livers from protein concentration measurements" *Toxicol In Vitro* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12203388&blobtype=pdf | # From Immunity to Oncology: Itaconic Acid as a Driver in HBV-Induced HCC
Shahab Mahmoudvand, | Sheida, Behzadi Sheikhrobat, Somayeh Shokri, Hossein Baghi
Dear Editor, Itaconic acid (ITA) is an immunomodulatory mammalian metabolite secreted from primary macrophages that dramatically increases upon activation. The metabolite plays a significant role in epigenetic regulation, influencing immune responses and disease progression. In recent years, ITA has gained attention due to its anti-microbial and immunomodulatory activities [1]. However, the 'yin and yang' role of itaconate should not be overlooked because it can promote tumour growth [2]. This letter offers an overview of ITA's epigenetic role, which may provide new strategies for treating hepatocellular carcinoma (HCC), particularly in patients infected with the hepatitis B virus (HBV). The interplay between HBV and host epigenetic mechanisms is crucial in developing HCC, suggesting that interventions targeting these pathways could be beneficial.
ITA has been identified as a modulator of histone modifications, particularly through a process known as lysine itaconylation. This modification plays a significant role in various biological processes, including immune responses and cancer progression [3]. Itaconate promotes the expression of immune checkpoint proteins like programmed cell death protein 1 (PD-1) and T cell immunoglobulin and mucin domain 3 (TIM-3) by enhancing histone modifications at the Eomesodermin (EOMES) promoter, contributing to CD8 + T-cell exhaustion in HCC [4].
Research indicates that itaconate promotes CD8 + T-cell exhaustion via epigenetic induction, which may exacerbate HCC development in HBV-infected individuals. Thimme et al. [5] recently explained that CD8 + T cells are crucial in controlling HBV infection but are functionally impaired during chronic HBV infection.
CD8+ T cells play a crucial role in the context of HBV infection. Specifically, during acute-resolving HBV infection, these cells serve as the primary effector cells that facilitate viral clearance and contribute to the pathogenesis of the disease. CD8+ T cell exhaustion refers to impaired function and diminished numbers of T cells, significantly contributing to advancing HBV infection [6]. Recent progress in exploring exhausted T cells during chronic HBV infection has provided novel insight into the possibility of immunotherapy for this disease [7]. Gu et al. demonstrated an epigenetic connection between itaconate and HCC, indicating that focusing on immune-responsive gene 1 (IRG1), which is responsible for the synthesis of itaconate, or on itaconate itself could represent a promising approach for the treatment of HCC. Furthermore, combining T-cell and anti-PD1 therapy may offer potential curative effects [4]. Conversely, while lysine itaconylation presents a promising area of research, the complexity of viral interactions with host modifications indicates that further studies are necessary to elucidate its specific role
## References
1. Aso, Kono, Kanda (2023) "Itaconate Ameliorates Autoimmunity by Modulating T Cell Imbalance via Metabolic and Epigenetic Reprogramming" *Nature Communications*
2. Chen, Dowerg, Cordes (2023) "The Yin and Yang of Itaconate Metabolism and Its Impact on the Tumor Microenvironment" *Current Opinion in Biotechnology*
3. Lang, Siddique (2024) "Control of Immune Cell Signaling by the Immuno-Metabolite Itaconate" *Frontiers in Immunology*
4. Gu, Wei, Suo (2023) "Itaconate Promotes Hepatocellular Carcinoma Progression by Epigenetic Induction of CD8(+) T-Cell Exhaustion" *Nature Communications*
5. Thimme, Wieland, Steiger (2003) "CD8(+) T Cells Mediate Viral Clearance and Disease Pathogenesis During Acute Hepatitis B Virus Infection" *Journal of Virology*
6. Bosch, Kallin, Donakonda (2024) "A Liver Immune Rheostat Regulates CD8 T Cell Immunity in Chronic HBV Infection" *Nature*
7. Urbanek-Quaing, Chou, Gupta (2024) "Enhancing HBV-Specific T Cell Responses Through a Combination of Epigenetic Modulation and Immune Checkpoint Inhibition" *Hepatology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12866110&blobtype=pdf | # Anti-IL-15 treatment reduces acute lentivirus inflammation and signaling in the brain
Daniel Ram, Mohan Gopalakrishnan, Malika Aid, Brandon Keele, R Reeves, C Tan, Mohan Raja, Gopalakrishnan, Kyle Kroll, Jasmine Miftahof, Omar Aristizabal, Eunice Kayitare Gikundiro, Caitlin Davis, Marco Hefti, Kimberly Fiock, Brook Tilahun, Yvette Umutoniwase, Kate Loidolt, Christine Fennessey, Noe Mercado, Victoria Harper-Alexander, Rhianna Jones, Griffin Woolley, Valerie Varner, Michelle Lifton, Steven Bosinger, Dan Barouch
## Abstract
Highlights• Peripheral pre-administration of αIL-15 alters CNS immune responses to acute SIV infection• T cell quantities in the brain are reduced following peripheral αIL-15
## In brief
Ram and Gopalakrishnan et al. demonstrate that peripheral cytokine modulation alters the brain's inflammatory environment and immune response to retroviral infection. These results demonstrate that a favorable central nervous system environment can be induced by intravenous administrations of antibodies. This work will inform and guide therapeutic strategies for HIV-associated neurocognitive disorder.
## INTRODUCTION
Despite advancements in viral suppression with antiretroviral therapies, HIV-associated neurocognitive disorder (HAND) remains an ongoing chronic health problem in approximately half of people living with HIV (PLWH) worldwide. 1,2 HIV infects the brain within days during the acute phase of infection and causes neurological symptoms. 3 Viral entry into the central nervous system (CNS) elicits an early immune response, driving neuroinflammation and eventually leading to the development of HAND. 4 Understanding the mechanisms of neuroinflammation and identifying the contributing factors during acute infection is needed to prevent HAND. Studying simian immunodeficiency virus (SIV) infection in a rhesus macaque (RM) model, we have previously demonstrated that an elevated inflammatory immune response occurs during acute SIV infection in the brain, mirroring findings in human cerebrospinal fluid (CSF). 5 In addition, we found that inflammatory responses even precede reliable virus detection in the brain, suggesting an impact of the peripheral immune response on the CNS. Thus, we sought to elucidate the connection between the peripheral immune response and CNS neuroinflammation during acute lentiviral infection.
Interleukin-15 (IL-15) is a versatile cytokine produced by myeloid lineage cells that mediates both inflammatory and protective immune reactions against viral, bacterial, and parasitic pathogens by serving as a mediator between the innate and adaptive immune systems. IL-15 plays an important role, specifically in HIV and SIV infections, by contributing to the proliferation and activation of natural killer (NK) cells and antigen-specific cytotoxic T cells [6][7][8][9] to control viremia. 10 IL-15 is also important in diseases of the CNS, where increases in serum and CSF IL-15 have been observed in patients with active multiple sclerosis. 11 IL-15's role in neuroinflammation is particularly relevant to understanding the pathogenesis of HAND and other CNS disorders associated with viral infections. Peripheral IL-15 administration in experimental autoimmune encephalomyelitis (EAE) mice resulted in prolonged increases in the number of pro-inflammatory infiltrating T cells in the brain, exacerbating the disease. 12 While these studies provide some evidence of increased inflammatory immune cells infiltrating the CNS following peripheral IL-15 administration, it remains unclear whether CNS inflammation could conversely be reduced by therapeutically depleting IL-15.
Recent work showed that treatment with a rhesusized monoclonal antibody against IL-15 (αIL-15) prior to infection with SIV mac239X resulted in IL-15 depletion in vivo, total NK cell ablation, minimal effects on T cell quantities, and subsequently led to increases in plasma viremia and a modest increase in the inflammatory response. 13 However, the impact of this αIL-15 pretreatment on CNS immune and inflammatory responses to acute SIV infection remains unclear. To address this gap, we aimed to investigate several key aspects of the effects of peripheral IL-15 depletion on CNS immune and inflammatory responses to acute SIV infection.
Our study sought to understand how peripheral IL-15 depletion prior to infection influences viral replication and immune responses in the CNS during acute SIV infection (Figure S1). We were further interested in determining whether peripheral modulation of NK cells could alter CNS immune and inflammatory responses to acute SIV infection. Additionally, we aimed to elucidate the specific effects of αIL-15 pretreatment on CNS viral pathogenesis and associated immune and inflammatory responses during this critical early stage of infection.
To address these research objectives, we treated RM with αIL-15 prior to infection with SIV mac 239X to determine the effects of IL-15 neutralization on CNS viral pathogenesis and the associated immune and inflammatory response to acute SIV infection. This approach also allowed us to investigate the broader question of whether peripheral modulation of the immune system can influence CNS immune and inflammatory responses.
## RESULTS
## Histology analysis
Brain tissues from acutely infected RM did not show any significant histological abnormalities at either day 7 or day 14 postinfection. Hematoxylin and eosin staining of tissue sections from different brain regions revealed no remarkable differences between the treatment groups (Figure S2).
## Viral load in the brain is independent of plasma viral load
All RMs infected with SIV in this analysis cohort had detectable viral loads in the plasma. αIL-15-treated RM euthanized at 7 days post-infection (dpi) (n = 2) had a mean viral load of 4.57 ± 0.41E10 SIV Gag (Group-Specific Antigen) RNA copies/ mL, while αIL-15-treated RM euthanized at 14 dpi had an increased mean viral load of 8 ± 0.25 E10 SIV Gag RNA copies/mL (range = 7.68-8.39E10; Table 1). Quantitative PCR performed to detect SIV in the brain revealed that SIV Gag RNA was below the limit of detection (LOD) at 7 dpi for both αIL-15-treated and untreated animals. However, at 14 dpi, SIV Gag RNA was consistently detected in the frontal cortex (subcortical white matter) of both αIL-15-treated and untreated animals, as well as in the basal ganglia and thalamus, collected from three RMs (Rh11, Rh23, and Rh24), which also showed detectable virus (Table 1). Importantly, there was no significant difference in viral loads between αIL-15-treated and untreated control animals, suggesting that IL-15 depletion did not directly affect viral replication in the brain or plasma at these early time points.
## Viral seeding in the brain exhibits compartmentalized clonal expansion
To analyze virus presence in different brain regions, we performed RNA in situ hybridization on the frontal cortex, thalamus, and basal ganglia of RM brains (Figure 1). Quantification of SIV-positive cells showed no remarkable differences in SIV quantities between brain regions of αIL-15-treated animals at the two time points (7 and 14 dpi) or in untreated controls at the respective time points. Interestingly, sequence analysis of the viral clones in each brain region showed that the distribution of clonal populations of SIV appears to be compartmentalized in a region-specific manner (Figure S3), with different subsets of viral clones populating distinct regions of the brain. This compartmentalization suggests that viral seeding in the brain may occur through distinct events, leading to the establishment of region-specific viral populations.
αIL-15 treatment alters microglia and astrocyte responses to SIV infection IL-15 expression in brain tissue was not significantly different between αIL-15-treated groups and untreated groups (Figure 2). While animals with SIV infection had higher levels of IL-15 expression in brain tissue, especially by day 14, the differences did not reach statistical significance. Astrocytes were quantified by GFAP (Glial Fibrillary Acidic Protein) staining in brain tissue and did not show significant changes with αIL-15 treatment or with SIV infection (Figure 2). Microglia were quantified by Iba-1 (Ionized calcium-binding adapter molecule 1) staining and also did not show significant changes with αIL-15 treatment or with SIV infection. SIV-infected groups had higher numbers of astrocytes expressing IL-15 than uninfected groups, although αIL-15treated groups were not significantly different from untreated groups. Microglial expression of IL-15 was minimal in the brain and did not change with αIL-15 treatment or with infection. However, microglia (stained with Iba-1) revealed changed morphology following αIL-15 treatment (Figure 3). Specifically, we observed that treatment with αIL-15 resulted in increased numbers of ramified microglia in SIV-infected animals, defined as an increased average number of dendritic extensions per cell body in the analyzed areas. While αIL-15 treatment alone did not differ from the untreated group, the αIL-15-treated group with 14 days of SIV infection had more ramified microglia than the untreated group with 14 days of infection. In addition, we also examined neurons in the cortex and did not find significant differences with SIV infection or αIL-15 treatment (Figure S4).
## αIL-15 treatment altered T cell quantities in the brain
Peripheral αIL-15 administration significantly abrogated SIV-infection-induced changes in T cell quantities in the brain. In the αIL-15 untreated group of RM, CD3 + T cells (both CD3 + CD4 + and CD3 + CD4 -) were lower in quantity at 7 dpi compared with the uninfected and untreated group. T cell numbers were higher at 14 dpi, reaching quantities almost equal to those of uninfected RM. However, in αIL-15-treated RM, there was no such trend observed in CD3 + T cells in general or in the CD3 + CD4 -T cell subset. αIL-15-treated uninfected RM had significantly fewer CD3 + CD4 - T cells in brain tissue compared with untreated and uninfected RM. This difference was not observed between αIL-15-treated and untreated RM infected with SIV at either day 7 or day 14. CD3 + CD4 + T cells exhibited a similar response in the brain parenchymal region, while there were slightly lower numbers of CD3 + CD4 + T cells in the perivascular region at 7 dpi compared with uninfected αIL-15-treated controls; however, this difference was not statistically significant (Figure S5). Interestingly, CD3 + CD4 + T cells in the perivascular space of the untreated group at day 14 of infection appeared to be similar in quantity to those of the untreated and uninfected group.
## αIL-15 treatment altered immune cell activation transcriptomics in the brain
Bulk RNA-seq of frontal cortex tissue from RM treated with αIL-15 demonstrated a complex modulation of immune and inflammatory response gene pathways in response to SIV infection (Figure 4A). Several specific gene pathways were differentially regulated, revealing a nuanced effect of αIL-15 treatment on the CNS immune response. These included IL-6 signaling, as well as the IFN and M1 gene pathways (Figure 4B), which were downregulated following αIL-15 treatment, as were genes associated with oxidative phosphorylation and mitochondrial function (Figure 4C). These RNA transcript changes suggest a shift in the inflammatory profile of the CNS in αIL-15-treated animals.
## αIL-15 treatment altered inflammatory cytokine expression in the brain
To further delineate which cells produced the observed inflammatory and anti-inflammatory cytokines, we performed double immunohistochemistry (IHC). While perivascular and parenchymal CD163 + /CD68 + macrophage quantities were not altered by αIL-15 treatment, their expression of TGF-β was significantly increased in macrophages in the perivascular regions (Figure 5). Proinflammatory cytokine expression in microglia was also modulated by αIL-15 treatment. In contrast, the number of IL-6-expressing microglia in macaques treated with αIL-15 was not significantly different from that of those without treatment.
Those infected with SIV after αIL-15 treatment showed significantly fewer microglia expressing IL-6 compared with those infected with SIV without treatment at both day 7 and day 14.
(Figure 5). These findings indicate that αIL-15 pretreatment has diverse effects on different cell populations in the CNS, potentially altering the balance between pro-and anti-inflammatory responses to acute SIV infection.
## αIL-15 treatment altered inflammatory cytokine expressions in plasma and correlated with myeloid and T cell responses in the frontal cortex and thalamus
We had previously quantified cytokines in the plasma of this cohort of macaques. 13 Correlating these soluble analyte concentrations with the myeloid cell gene transcriptomic signatures from the frontal cortex, we observed that while the myeloid gene pathway signatures correlated positively with the proinflammatory cytokines detected in the plasma of untreated and SIV-infected RM (Figure S6A), those treated with αIL-15 and infected with SIV showed decreased and inverse correlations (Figure S6B). During acute SIV infection, there are increased numbers of perivascular macrophages in the frontal lobe contributing to the myeloid cell gene signature. To examine this further, we quantified IL-18 expression on microglia and macrophages in the thalamus by IHC (Figure S7). Neither acute SIV infection nor αIL-15 treatment resulted in significantly altered IL-18 expression by microglia or macrophages. Next, we examined G-CSF expression by microglia and macrophages in the thalamus. We found a near-significant (p = 0.05) increase in microglial G-CSF expression on day 14 after infection in the αIL-15-treated group (Figure S8). Correlating brain T cell gene signature pathways with soluble analytes in plasma from day 14 samples, we also observed a significant positive correlation with inflammatory cytokines in frontal cortex samples from the αIL-15-untreated groups with SIV infection (Figure S9A) and decreased and inverse correlations in the αIL-15-treated groups with SIV infection (Figure S9B). We further examined CD137 and Lag 3 expression on CD3 + T cells in the thalamus and did not find any significant differences with SIV infection or α IL-15 treatment (Figure S10). These data indicate that the presence of IL-15 in the periphery is associated with correlative CNS myeloid and T cell migration and proinflammatory responses in acute lentiviral infection.
## αIL-15 treatment altered the BBB
We assessed blood-brain barrier (BBB) endothelial integrity by measuring both brain vessel endothelial cell (CD31) expression intensity and by quantifying RM IgG in the brain parenchyma, as previously described. 14 Higher CD31 intensity is indicative of more tightly packed endothelial cells lining the vessels. In contrast, higher quantities of IgG detected in the brain parenchyma indicate loss of BBB integrity, as IgG has leaked across a compromised barrier. While the control group showed a negative correlation between CD31 intensity and IgG quantities, the correlation was no longer significant in the αIL-15-treated group. Additionally, the αIL-15-treated uninfected group showed increased CD31 intensity, whereas intensity was lower in the αIL-15-treated SIV-infected group (Figure S11).
To further assess BBB integrity, we quantified BBB junctional proteins claudin, occludin, and zonula occludens-1 (ZO-1) (Figure 6). αIL-15 treatment did not change the expression of these junctional proteins in uninfected RM. RM treated with αIL-15 treatment had significantly less occludin and ZO-1 expression on day 7 after SIV infection compared with the untreated SIV-infected group. These differences were not present between the treated and the untreated groups at day 14.
## Activated RNA gene pathways correlated to mechanisms of brain inflammation
We conducted single-sample gene set enrichment analysis (ssGSEA) to correlate the gene pathways detected in the bulk brain tissue RNA-seq with PCR and histopathological findings. We observed different correlations in the group treated with αIL-15 (Figures S11 andS12). Of significance (p < 0.05), in the control group, the gene pathways of NK cells and M2 macrophages were positively correlated with plasma and brain viral loads, quantities of T cells, TGF-β, and IL-2G receptors in the brain. In contrast, in the αIL-15-treated group, the correlations of these two immune cell gene pathways were reduced: NK cell pathways correlated with T cells and TGF-β, whereas the macrophage M2 gene pathway correlated with plasma viral load and TGF-β. The infection-induced activated cytokine pathways, including IL-6, TGF-β, IFN, and TNF-α, correlated positively with plasma viral load (IL-6), IL-2B and IL-2G (TGF-β), macrophages (IFNA), brain viral load, macrophages and T cells (IFNG), and plasma viral load (IFN II and IFN III). The activated cytokine gene pathways in the αIL-15 group also showed different correlations with viral loads and histopathological findings. The IL-6 gene pathways correlated with IHC-quantified IL-6 in brain tissues. The TGF-β gene pathway correlated negatively with IL-2B, IL-2G, and brain viral loads. The antiviral IFN correlated positively with plasma viral loads and TGF-β. Overall, treatment with αIL-15 greatly reduced NK cell gene pathway correlations due to the depletion of these cells, decreased macrophage gene pathway correlations, and altered TGF-β gene pathway correlations, potentially due to the activation of this pathway in a different group of cells.
## DISCUSSION
We hypothesized that neuroinflammation during acute SIV infection could stem from the presence of virus in the brain, proinflammatory cytokines breaching the BBB, or immune cells interacting directly with CNS-resident immune cells or indirectly via cytokine release. This study examined the effect of αIL-15 pretreatment on the CNS response to acute SIV infection at two early time points (7 and 14 dpi). Our findings suggest that the quantities of virus in the CNS may not be the sole driver of neuroinflammation at these early time points, as higher viral loads in brain tissues did not correlate with elevated inflammatory markers compared to tissues with lower viral loads. RM pretreated with peripheral IL-15 exhibited similar CNS SIV viral quantities as those in no-pretreatment infected controls but displayed distinct immune and inflammatory responses. This suggests that αIL-15-mediated immune modulations in the periphery may influence acute inflammatory changes in the brain, as evidenced by analyses of cytokine and gene expression. It should be noted that, while αIL-15 pretreatment did not seem to impact viral quantities in the CNS at these early time points, longer monitoring periods might be necessary to detect such alterations.
Although the role of NK cells in the brain following SIV infection has yet to be fully characterized, they are found to play a significant regulatory role in the murine EAE model and CNS plaques of patients with multiple sclerosis. 15 Prior work in pig-tailed macaques with the neurovirulent swarm virus (SIV/ 17E-Fr and SIV/DeltaB670) linked a strong blood NK cell functional response to improved CNS outcomes. 16 Given the limitations of RM-specific reagents, we are unable to identify NK cells in brain tissues conclusively. However, our research indicates that αIL-15 treatment had implications extending beyond the regulation of viral replication in acute SIV infection. Notably, as αIL-15 treatment could also entail a degree of peripheral CD8 + T cell loss, we cannot fully dismiss the potential involvement of these cells in the observed effects. However, the measured peripheral total T cell quantities did not change with treatment in our cohort. 13 Our bulk RNA-seq data from frontal cortex tissues showed that peripheral αIL-15 treatment generally decreased the antiviral and anti-inflammatory gene pathways, yet altered macrophage activation toward M1 states. This is further supported by increased gene signatures for IL-6, IL-1β, and TNFα (Figure 4) in the αIL-15 treatment group. It is unclear whether the changes are due to IL-15 modulation monocytes in the periphery prior to their migration into the brain tissue or if this is due to a lack of IL-15 stimulation of monocytes/ macrophages in either the blood or brain tissue. This of monocytes/macrophages toward the M1 state was also associated with increased microglial changes and a surprising reduction in IL-6-positive monocytes/macrophages, as determined by IHC. Though the microglia appeared to exhibit greater ramifications following αIL-15 treatment, these different findings may be due to the single time point examination of the macrophage/monocyte activation process, which represents a continuum between M1 and M2 states.
ssGSEA, in which we correlated the RNA-seq data of activated gene pathways to SIV viral loads in plasma and brain tissues, as well as our histological analysis of brain tissues, further confirmed our findings that treatment with αIL-15 abolished the positive correlations of immune cells (NK cells, T cells, and macrophages) and antiviral activated gene pathways to brain tissue findings. However, the RNA-seq was performed on bulk frontal cortex tissues. Future investigations using spatial transcriptomics and single-cell RNA-seq from multiple brain regions are needed to better define the dynamic immune and inflammatory changes in the brain. Recently, the FDA approved the use of an IL-15 agonist to stimulate an immune response as part of the treatment for a subtype of bladder cancer. 17 In brain tissue, we did not detect a direct decrease of IL-15 from peripheral treatment with αIL-15. The consequent immune and inflammatory changes in the brain are most likely due to indirect effects of decreased peripheral IL-15. Our data indicate that the conof modulating IL-15 pathways in the blood should be examined in end organs, including the brain. For example, RM treated with αIL-15 and infected with SIV had lower quantities of occludin and ZO-1 7 days after infection. While these differences were not detected between groups infected for 14 days, decreases in occludin in CSF have been associated with HAND in PLWH. 18 The modulation of T cells in various brain tissues and perivascular or parenchymal spaces was also interesting. Although it appears that the general trend was the detection of fewer CD3 + cells per high-power field following αIL-15 treatment, CD3 + CD4 + T cells (as well as CD3 + CD4 -T cells) were at similar levels to control samples from 7 to 14 dpi. Since αIL-15 treatment is expected to remain effective for several weeks to months following administration, this may indicate an expansion of an IL-15-independent T cell population responding to increased viral presence in the brain at 14 dpi. The αIL-15-treated group had fewer CD3 + CD4 -T cells in the brain, even though total peripheral T cell quantities did not decrease. 13 Loss of these cells in the brain may also contribute to the observed decrease in neuroinflammatory response during acute infection but may hinder viral clearance over the long term. Interestingly, αIL-15 treatment resulted in a substantial upregulation of genes associated with oxidative phosphorylation compared to non-depleted animals (Figure 4C). These pathways may alter the inflammatory environment through mTOR signaling, 19 although further studies are required to establish a causal link.
Sequencing analysis of the barcoded virus in this study indicated that, during the acute phase of infection, once a virus seeds a particular region of the brain, it is retained in that location and spreads locally to cells (macrophages, microglia, and others) within the same space but generally does not spread outside the original site within the limited time frame of our study. Our experimental design did not allow us to assess whether specific viral clones could eventually spread to other regions of the brain over time. Longer infection duration studies are needed to determine viral spread within or between the compartmentalized regions of the brain. Nevertheless, because the virus used for infection was barcoded, we were able to determine the distribution of some viral clones in different brain regions. In regions where we were able to identify the virus, we observed a preference for single clones per region, suggesting that viral seeding is limited and individual lineages can define the predominant population within a small anatomic region. These distinct populations may evolve independently due to local immune pressures or other microenvironmental factors, potentially influencing the course of infection and associated neuropathology in different brain regions. There was no obvious difference in virus seeding between individual sites of the brain in the no-treatment control and the αIL-15-treated samples, although increasing the sample size will be necessary for a more conclusive determination.
Our data showed that peripheral modulation of the immune system prior to acute SIV infection can alter anti-viral immune and inflammatory responses in multiple brain regions. However, the effect of peripheral immune modulation on mitigating HAND, a chronic condition, requires further assessment. Despite this, these findings provide evidence that therapeutics that modulate inflammation in the blood have the potential to decrease inflammation in the brain, thereby aiding in the management of HAND and other neuroinflammatory diseases.
## Limitations of the study
While infection with SIV closely models human infection and immune response, a limitation of non-human primate studies is the small number of animals in each group. We had originally designed the study for at least three animals in each group. However, one of the day 7 αIL-15-treated animals did not become infected, thus reducing the number in that group to two. This study primarily focused on the frontal cortex because it constitutes the largest part of the brain, providing sufficient tissue material for both RNA-seq and IHC/ RNAscope analyses. Analysis of the basal ganglia and thalamus will expand understanding of neuroinflammation. Furthermore, this study did not include neurobehavioral or cognitive assessments, though such changes may not be detectable in RM during acute infection. Lastly, administration of αIL-15 was carried out prior to SIV infection; thus, we are limited in our assessment of the therapeutic use of αIL-15 after infection and how this might influence CNS immune and inflammatory responses.
## RESOURCE AVAILABILITY
## Lead contact
Requests for further information, resources, and reagents should be directed to the lead contact, C. Sabrina Tan (Sabrina-tan@uiowa.edu).
## Materials availability
Regents generated in this study will be made available upon request. Payments and/or a materials transfer agreement may be required.
## Data and code availability
• RNA sequencing data have been deposited in GEO (accession number GSE310364) and are publicly accessible at https://www.ncbi.nlm.nih. gov/geo/query/acc.cgi?acc=GSE310364.
• This paper does not report original code.
• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
## EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS
This study was based on analysis of SIV mac239x infected adult male RM (Macaca mulatta) of Indian origin, ages 4-10 years. The animals were housed and cared for at Biomere Inc. (Worcester, MA) under the rules and regulations of the Committee on the Care and Use of Laboratory Animal Resources and, in agreement with the American Association for Accreditation of Laboratory Animal Care standards and Guide for the Care and Use of Laboratory Animals. 20 Animals were fed a standard diet.
## METHOD DETAILS Animals and virus infection
Animal care and viral infection procedures were conducted as previously described by our group. 13 Peripheral administration of αIL-15 antibody (Lot: SM17-25; kindly provided by the NIH NHP Reagent Resource) was given intravenously at day -21 (20mg/kg) and day -7 (10mg/kg) relative to SIV mac239X inoculation prior to infection, based on prior study protocols. 13,21 The animals were infected intrarectally with a single dose of 300,000-500,000 IU of SIV mac 239X (Table S1). This virus stock contained nine isogenic and sequence-discriminable molecularly tagged variants (variants A to I) plus wild-type (WT) SIV mac 239X (10 in total) within a single inoculum stock, with equal proportions of all genotypes. 22 RM were necropsied either at 7 or 14 days post-infection (dpi) (Figure S1). The right-sided heart was perfused with cold PBS during necropsy to decrease blood cell contamination in the brain tissues.
## Tissue collection and processing
Brains were harvested, and tissue samples (roughly 1 cm × 1 cm × 1 cm) were collected from the frontal cortex of one of the hemispheres and cryopreserved in RNAlater for assessing viral load and for RNA-Seq analysis. The other hemisphere was stored in 10% neutral buffered formalin for 14 days, after which the hemisphere was sliced and tissue samples were collected from different brain regions (frontal cortex, thalamus, basal ganglia), embedded in paraffin, and sectioned at 10 μm thickness for in situ hybridization (ISH) and immunohistochemistry (IHC) studies. The sectioned tissues on glass slides were stored under vacuum conditions until used. A tissue section from each brain region was also stained with H&E for routine histological analysis.
## Viral load quantification
Peripheral viral loads were quantified as previously described. 23 In brief, the known amount of RNA extracted from cell-free plasma was reverse-transcribed to cDNA using gag-specific primers, followed by RNase treatment at 37 • C for 20 min. Then, cDNA amplification was performed using a 7300 ABI Real-Time PCR system (Applied Biosystems) according to the manufacturer's protocol (TaqMan Fast Advanced Master Mix User Guide; publication 4444605). All reactions were run in triplicate, and a SIVgag RNA standard was included.
Brain viral load was determined by quantifying both DNA and RNA through qPCR as described earlier. 24 In short, total RNA (using the RNeasy 96 QIAcube HT kit) and genomic DNA (using the QIAamp 96 QIAcube HT kit) were isolated from frozen brains according to the manufacturer's protocol (Qiagen). Following RNA extraction, the total RNA was converted to cDNA using Superscript III VILO (Invitrogen) using the manufacturer's specifications. Viral Gag copies were determined using the QuantStudio 6 Flex system (Applied Biosystems) following previously optimized thermocycler settings, with the following protocol repeated for 45 cycles: 95 • C for 20 s for initial denaturation, followed by 95 • C for 1 s and 60 • C for 20 s. SIV Gag-specific primers and probes that were used for the assays are as follows: sGag21 (forward) GTCTGCGTCATCTGGTGCATTC, sGag22 (reverse) CACTAGGTGTCTCTGCACTATCTGTTTTG, and sGag23 (probe) 5 ′ FAM-CTTCCTCAGTGTGTTTCACTTTCTCTTCTGCG-BHQ-3 ′ . RT-PCR assays were run in duplicate, whereas the viral DNA assays were run in triplicate. To calculate SIV Gag DNA and RNA copies, standards were used. For preparing RNA standards, the AmpliCapMax T7 High Yield Message Maker kit (Cell Script) was used, followed by RNA purification using an RNA Clean and Concentrator kit (Zymo Research). RNA standards were prepared in log dilutions, and standards were used in each RT-PCR assay. For the SIV DNA PCR assay, an RPP30 control was also included. Viral RNA load was calculated as RNA copies per microgram of total input RNA. LOD for the SIV RNA assay was one copy per μg of total RNA input. Viral DNA load was calculated as SIV Gag DNA copies per million cells, after normalizing DNA copies to the total number of input cells. Each cell has two copies of the RPP30 gene. LOD for the SIV DNA assay was eight copies per million cells.
## Viral clonal analysis
Viral RNA was extracted from the tissues using a Qiagen Viral RNA kit. Next-generation sequencing analysis was performed as previously described. 13,25 Detection and quantification of SIV RNA-positive cells using RNAscope Cells positive for SIV RNA (in basal ganglia, thalamus, and frontal cortex) were detected and quantified using RNAscope Multiplex Fluorescent V2 and RNAscope 2.5 HD Assay-Brown Kits (Advanced Cell Diagnostics, Inc.), respectively. SIV gag probes against the gag region of SIV purchased from Advanced Cell Diagnostics, Inc., were used for quantifying SIV RNA-positive cells by RNAscope 2.5 HD Assay-Brown Kit following the manufacturer's instructions.
Single and sequential staining through IHC Manual IHC with different chromogenic substrates was used for single-antigen detection or for detecting two antigens simultaneously using a sequential double-staining approach. Immune cell markers, including CD3 (clone F7.2.38), CD4 (clone 4B12), and GFAP (pAb number: Z0334) from Agilent-Dako; CD163 (clone EDHu1) from BioRad; CD68 (clone 298807) from R&D Systems; Iba-1 (pAb Catalog: 019-19741) from FUJIFILM Wako, were used for single antigen detection. Cells positive for proinflammatory IL-6 (ab219758) and anti-inflammatory TGF-β (EPR21143), both from Abcam, Claudin-5 (EPR7583), occludin (OC-3F10), zonula occludens -1 (ZO1-1A12), IL-15 (E-4), IL-18 (PIPA5110679), G-CSF (BVD13-3A5), Lag 3 (EPR20261), CD137 (EPR25096-57), NeuN (ab104225), CXCL12 (EPR1216), were identified by staining for either of the cytokines followed by staining for one of the immune cell markers mentioned above. Standard IHC protocol was followed. Tissue sections stained for Iba-1 were also used to assess the change in microglia morphology. Morphology analysis was carried out by counting the number of spines attached to the cell body per 40x high-power frame and averaging the number of spines by the number of microglial cell bodies per frame. For each slide, an average of 15 randomly selected frames was analyzed. The spine counting was performed manually by different individuals who were blinded to the sample name/group to minimize person-to-person bias in the counting, and the distribution of data was found to be unbiased.
## Microscopy and image analysis
RNAscope detection and imaging were performed using a Zeiss LSM 880 laser scanning confocal microscope, with at least 20 high-power fields (HPFs) per slide imaged at 20X magnification. Slides from single and double IHC were imaged in a Zeiss Axio Imager M1 microscope. At least 20 HPFs at 20X or 40X were captured for each section. For each section, 15 HPF images were randomly chosen for quantification. While counting the total number of positively stained cells in single or double IHC, the positively stained cells were also sub-classified based on their location: as perivascular (V) if they lined the blood vessel, or parenchymal (P) if they were present in the brain parenchyma. Cell counts are expressed as an average calculated across the number of images quantified (15 fields of view). Expression of GFAP and Iba-1 was quantified using the Keyence microscope analysis program by quantifying the intensity of staining, performing pixel analysis of colocalizations, and normalizing to DAPI staining in each field of view.
Library preparation and bulk RNA-Seq analysis RNA was isolated from bulk brain frontal cortex tissues, processed, and analyzed as previously described. 5 Single-sample Gene Set Enrichment Analysis (ssGSEA) was performed following the protocol described by Pranali (2021), https://rpubs.com/pranali018/ SSGSEA. ssGSEA scores, which represent up-or down-regulated gene pathway scores per sample, were correlated with clinical measures using the corrplot package 26 and visualized using both corrplot and ComplexHeatmap. 27 Luminex data analysis Luminex data were obtained from a previously published manuscript utilizing this animal cohort. 13 Luminex values were loaded into R v4.4.2. Single-sample Gene Set Enrichment Analysis (ssGSEA) scores were calculated for each bulk RNA-seq sample. Correlations between Luminex analyte values and ssGSEA scores were calculated using the cor() function from the base stats package. P-values were obtained with the cor.mtest() function from the corrplot 26 package. Correlations of interest were visualized with the corrplot functions.
## QUANTIFICATION AND STATISTICAL ANALYSIS
Unless otherwise mentioned, GraphPad Prism 9.0 was used for all statistical analyses. A nonparametric Kruskal-Wallis test was used for analysis, and multiple comparisons were assessed using Dunn's post hoc analysis. Unless otherwise stated, data are presented as mean ± standard deviation (SD). p values (multiplicity-adjusted p values for multiple comparisons) were considered significant when *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001, while ns indicates p > 0.05, which means no significance.
## References
1. Wei, Hou, Su et al. (2020) "The Prevalence of Frascati-Criteria-Based HIV-Associated Neurocognitive Disorder (HAND) in HIV-Infected Adults: A Systematic Review and Meta"
2. Zenebe, Necho, Yimam et al. (2022) "Worldwide Occurrence of HIV-Associated Neurocognitive Disorders and Its Associated Factors: A Systematic Review and Meta" *Analysis. Front. Psychiatry*
3. Hellmuth, Fletcher, Valcour et al. (2016) "Neurologic signs and symptoms frequently manifest in acute HIV infection" *Neurology*
4. Ulfhammer, Ede ´ N, Antinori et al. (2022) "Cerebrospinal Fluid Viral Load Across the Spectrum of Untreated Human Immunodeficiency Virus Type 1 (HIV-1) Infection: A Cross-Sectional Multicenter Study" *Clin. Infect. Dis*
5. Gopalakrishnan, Aid, Mercado et al. (2021) "Increased IL-6 expression precedes reliable viral detection in the rhesus macaque brain during acute SIV infection" *JCI Insight*
6. Ma, Caligiuri, Yu (2022) "Harnessing IL-15 signaling to potentiate NK cell-mediated cancer immunotherapy" *Trends Immunol*
7. Harwood, Connor (2021) "Therapeutic Potential of IL-15 and N-803 in HIV/SIV Infection" *Viruses*
8. Younes, Freeman, Mudd et al. (2016) "IL-15 promotes activation and expansion of CD8+ T cells in HIV-1 infection" *J. Clin. Investig*
9. Barrenas, Hansen, Law et al. (2021) "Interleukin-15 response signature predicts RhCMV/SIV vaccine efficacy" *PLoS Pathog*
10. Bernard, Alsulami, Pavey et al. (2022) "NK Cells in Protection from HIV Infection" *Viruses*
11. Rentzos, Cambouri, Rombos et al. (2006) "IL-15 is elevated in serum and cerebrospinal fluid of patients with multiple sclerosis" *J. Neurol. Sci*
12. Laurent, Deblois, Clenet et al. (2021) "Interleukin-15 enhances proinflammatory T-cell responses in patients with MS and EAE" *Neurol Neuroimmunol Neuroinflamm*
13. Woolley, Mosher, Kroll et al. (2023) "Natural Killer Cells Regulate Acute SIV Replication, Dissemination, and Inflammation, but Do Not Impact Independent Transmission Events" *J. Virol*
14. Byrnes, Busman-Sahay, Angelovich et al. (2023) "Chronic immune activation and gut barrier dysfunction is associated with neuroinflammation in ART-suppressed SIV+ rhesus macaques" *PLoS Pathog*
15. Belien, Goris, Matthys (2022) *Natural Killer Cells in Multiple Sclerosis: Entering the Stage. Front. Immunol*
16. Shieh, Carter, Blosser et al. (2001) "Functional analyses of natural killer cells in macaques infected with neurovirulent simian immunodeficiency virus" *J. Neurovirol*
17. Mullard (2024) "First-in-class IL-15 receptor agonist nabs FDA approval for bladder cancer" *Nat. Rev. Drug Discov*
18. Bai, Bono, Borghi et al. (2024) "Association between tight junction proteins and cognitive performance in untreated persons with HIV" *AIDS*
19. Mao, Van Hoef, Zhang et al. (2016) "IL-15 activates mTOR and primes stress-activated gene expression leading to prolonged antitumor capacity of NK cells" *Blood*
20. "Guide for the Care and Use of Laboratory Animals. The National Academies Collection: Reports funded by National Institutes of Health"
21. Okoye, Degottardi, Fukazawa et al. (2019) "Role of IL-15 Signaling in the Pathogenesis of Simian Immunodeficiency Virus Infection in Rhesus Macaques" *J. Immunol*
22. Del Prete, Park, Fennessey et al. (2014) "Molecularly tagged simian immunodeficiency virus SIVmac239 synthetic swarm for tracking independent infection events" *J. Virol*
23. Whitney, Hill, Sanisetty et al. (2014) "Rapid seeding of the viral reservoir prior to SIV viraemia in rhesus monkeys" *Nature*
24. Cadena, Ventura, Abbink et al. (2021) "Persistence of viral RNA in lymph nodes in ARTsuppressed SIV/SHIV-infected Rhesus Macaques" *Nat. Commun*
25. Hensley-Mcbain, Berard, Manuzak et al. (2018) "Intestinal damage precedes mucosal immune dysfunction in SIV infection" *Mucosal Immunol*
26. Wei, Simko "R package 'corrplot': Visualization of a Correlation Matrix (Version 0.92) 2021"
27. Gu, Eils, Schlesner (2016) "Complex heatmaps reveal patterns and correlations in multidimensional genomic data" *Bioinformatics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817144&blobtype=pdf | # A comprehensive database for biological data derived from sewage in five European cities
Ágnes Becsei, Patrick Munk, Alessandro Fuschi, Saria Otani, József Stéger, Dávid Visontai, Krisztián Papp, Christian Brinch, Ravi Kant, Ilya Weinstein, Olli Vapalahti, Miranda De Graaf, Claudia Schapendonk, Jeroen Roelfsema, Maaike Van Den Beld, Roan Pijnacker, Eelco Franz, Patricia Alba, Antonio Battisti, Alessandra De Cesare, Valentina Indio, Fulvia Troja, Tarja Sironen, Chiara Oliveri, Frédérique Pasquali, Ivan Liachko, Benjamin Auch, Colman O'cathail, Krisztián Bányai, Magdolna Makó, Péter Pollner, Marion Koopmans, Istvan Csabai, Daniel Remondini, Frank Aarestrup
## Abstract
Sewage metagenomics is a powerful tool for proactive pathogen surveillance and understanding microbial community dynamics. To support such efforts, we present a highly curated and accessible longitudinal dataset of 239 sewage samples collected from five European cities. The dataset, processed through metagenomic sequencing, includes rich analytical outputs such as taxonomic profiles, identified antimicrobial resistance genes, assembled contigs with annotated origins, metagenome-assembled genomes with functional gene annotations, and metadata. Given the computational intensity and time required to reproduce such analyses, we share this dataset to promote reuse and advance research. In addition to the metagenomic data, qPCR was used to identify specific pathogens, and Hi-C sequencing was performed on a subset of the samples to strengthen genomic linkage analysis. Central to this resource is a publicly available PostgreSQL database, designed to facilitate efficient exploration and reuse of the data. This comprehensive database allows users to perform targeted queries, subset data, and streamline access to this extensive resource.
## Introduction
Sewage surveillance has been used for monitoring public health for more than four decades, initially focusing on detecting infectious disease threats such as polio [ 1 ], and more recently, it was successfully used for surveillance of SARS-CoV-2 [ 2 , 3 ]. However, these efforts have primarily relied on targeted PCR-based or targeted sequencing methods. As the cost of next-generation sequencing (NGS) has decreased, this technology has become more accessible for routine use [ 4 ]. NGSbased metagenomics enables sequencing of entire microbial communities [ 5 ] in both clinical [ 6 ] and environmental samples [ 7 ]. When combined with strategically selected sampling points (e.g. sewage treatment plants) [ 8 ], NGS-based surveillance has the potential to provide simultaneous detection of a broad range of pathogens and antimicrobial resistance genes (ARGs), offering a powerful, data-rich complement to traditional diagnostic and surveillance methods.
Antimicrobial resistance (AMR) poses a significant threat to global health, making it essential to understand its epidemiological patterns to predict the emergence and spread of ARGs.
In 2016, we initiated the global sewage project that involved collection and metagenomic sequencing of sewage from various regions around the world. A key focus of our work has been identifying and quantifying ARGs. Our findings reveal major diversity in the quantity of ARGs [ 9 ], and statistical analyses suggest that AMR trends are influenced by indicators related to the national health system and level of sanitation [ 10 , 11 ]. A longitudinal study on AMR in Copenhagen reveals that AMR levels are relatively stable over time [ 12 ]. In addition, we have explored the global composition of the bacteriome [ 8 , 13 ]. We recovered high-and medium-quality bacterial genomes referred to as metagenome-assembled genomes (MAGs) from sewage and developed a workflow for quantification and correlation to investigate their dynamics over time and geography [ 8 ]. Sewage metagenomics combined with urban virome surveillance provides a baseline for a catch-all early warning system for emerging pathogens [ 14 ]. Finally, sewage metagenomics also proved effective for studying the distribution of human mtDNA haplogroups in sewage, providing insights into human population genetics [ 15 ].
Although the raw sequencing data, as well as available contig, bin, and MAG sequences from these projects, are publicly accessible via platforms such as ENA [ 16 ] or NCBI SRA [ 17 ], utilizing the datasets for other applications would still require expensive bioinformatic computation and duplicate efforts. This highlights the need for more efficient data sharing practices in scientific research. While platforms such as MGnify [ 18 ] offer an automated pipeline for the analysis and archiving of microbiome data, they do not aim to organize different types of analytical output in an interconnected, structured way and do not support sharing of one's own analytical outputs.
Beyond these technical limitations, the continued use of statements of "data availability upon reasonable request" in scientific papers is still in place although deemed inefficient and unacceptable [ 19 ]. To address this issue and effectively avoid duplications, it is crucial to adhere to the FAIR principles [ 20 ] of findability, accessibility, interoperability, and reusability. This requires not only a commitment to data sharing but also the development of structured, user-friendly, open-source databases that are reliable, robust, and performant.
In this paper, we present a showcase dataset organized into a PostgreSQL database, which includes detailed metagenomic profiles, ARGs, and other microbiome-related data from sewage samples collected across five European cities: Copenhagen, Rotterdam, Budapest, Rome, and Bologna over 9-19 months with variable time frames between the cities (2019)(2020)(2021). This includes the raw data from shotgun metagenomic sequencing, assembled contigs, reconstructed MAGs, and qPCR (for pathogens) done on 230 samples, as well as Hi-C sequencing and analysis results for 24 samples. Although the dataset contains 239 sewage samples, certain samples from Copenhagen were sequenced two or three times, leading to a total of 278 metagenomic sequencing samples.
Shotgun sequencing is a technique where DNA is randomly fragmented into small pieces, and each fragment is sequenced independently. The main advantage of this method in bacterial metagenomics compared to targeted sequencing methods (e.g. 16S rRNA sequencing) is that shotgun sequencing uses a non-targeted approach to capture representative data from the entire DNA content of a sample [ 21 ].
Once the sequencing data is generated, we assemble the reads into contigs-longer, contiguous sequences formed by overlapping fragments [ 22 ]. During metagenomic binning, contigs likely originated from the same organism or genome were grouped together. Bins representing complete or near complete genomes are referred to as MAGs [ 23 ].
Proximity-ligation techniques like Hi-C are employed to identify interactions between DNA molecules within the same bacterial cell. This process involves crosslinking DNA fragments that are spatially close in three dimensions, followed by digestion and the formation of chimeric junctions. The resulting DNA fragments are then sequenced using shotgun sequencing. Reads from these Hi-C fragments provide information on the connections between non-contiguous DNA sequences from the same cell. These connections are then used in clustering algorithms to determine which DNA fragments originate from each individual cell. This means Hi-C can potentially link plasmids containing ARGs to their host by physically connecting them to the host genome [ 24 ].
## Methods
## Data generation
DNA and sequencing data Untreated sewage samples were collected routinely and during the COVID-19 outbreak between 2019 and 2021 from sewage treatment plants in Rome, Bologna, Rotterdam, Budapest, and Copenhagen. Samples were collected every 2 weeks to obtain time-series data from these European cities. DNA was extracted from these samples, and sequencing data were generated. The sequencing data were derived from the study by Becsei and Fuschi et al. [ 8 ]. qPCR qPCR for the detection of selected bacterial pathogens and parasitic protozoans was conducted in six multiplex reactions using a total reaction volume of 20 μl with SensiFast (Bioline, GC Biotech, Waddinxveen, Netherlands) mastermix, primer and probe concentrations of 0.5 and 0.25 μM, respectively, except for G. lamblia (primers: 0.25 μM, probe: 0.05 μM).
The list of primers used is available on Figshare [ 25 ]. PCRs were run on a Lightcycler 480 II instrument (Roche Diagnostics, Basel, Switzerland) using 45 cycles of 5 seconds of denaturation at 95 • C and 30 s of annealing at 60 • C using Phocid herpesvirus as internal control.
## Reference-based metagenomic data
Trimmed reads of all libraries were mapped using kma [ 26 ] (v1.2.8) with paired-end and singleton files as input against the ResFinder [ 27 ] database (commit = 3eedbde) and against a custom genomic database, which was constructed as previously described in the work of Osakunor et al. [ 28 ].
Settings of KMA allowed mapping only one query sequence per template and with default penalty values. Resulting mapstat files summarizing abundances in each sample were loaded into the database and are available on Figshare (see the 'Data records' section).
## Assembly-based metagenomic data
In our previous work, reads were assembled to contigs, and MAGs were reconstructed for each individual sample and sampling site via a binning of the contigs followed by the selection of high-and medium-quality MAGs and species-level dereplication. Taxonomic classification, quality information, and quantification results for each MAG in 239 samples were obtained from this work. For detailed description of these workflows refer to Becsei and Fuschi et al. [ 8 ].
We used the PPR-Meta [ 29 ] (v. 1.1) tool to analyze all contigs obtained from assembly workflows. PPR-Meta, a deep learning-based computational tool classifies metagenomic fragments into phages, plasmids, or chromosomal origins.
In our previous study, assembly of reads from individual samples and pooled reads from each sampling site, followed by binning, resulted in 21 708 166 contigs binned into 34 725 bins. Through dereplication of medium-and high-quality genomes, we identified 2332 distinct prokaryotic species among the MAGs. To label and identify relevant genomic features of these MAGs, they were annotated using Prokka [ 30 ] (v1.14.6) with default settings.
Hi-C sequencing 250 mg from each sewage pellet (12 from Copenhagen and 12 from Rotterdam) was prepared for Hi-C sequencing protocol using the ProxiMeta™ Hi-C Kit Protocol v4.0 (Phase Genomics, Seattle, USA). Briefly, sewage pellets were resuspended in formaldehyde solution to reach a 1% volume/volume concentration for crosslinking. These samples were left at room temperature for 20 min with periodic gentle stirring. To stop the crosslinking, glycine (from the ProxiMeta™ Hi-C Kit) was added to a final concentration of 1% volume/volume, and the samples were again left at room temperature for 20 min with occasional stirring.
A Hi-C library was constructed using ProxiMeta Hi-C Microbiome v4.0 (Phase Genomics, Seattle, USA) following the manufacturer's instructions, which are based on Hi-C protocol [ 31 ]. Cross-linked DNA from each sample was cut using Sau3AI and MlucI enzymes (from the ProxiMeta™ Hi-C Kit) and then proximity-ligated with biotin-tagged nucleotides to form chimeric molecules from different genomic regions that were close together in the original cells. These chimeras were isolated with streptavidin beads and further processed using the ProxiMeta library preparation reagents. The result-ing Hi-C metagenomes were sequenced. The ProxiMeta analysis pipeline (Phase Genomics, Seattle, USA) was employed for analysis of the data [ 24 ]. We refer to the resulting groups of contigs as genome clusters. These clusters represent potential genomic sequences from the same organism.
## Results
## Summary of the dataset
The metadata provided comprises GPS coordinates, the names of the sewage treatment plants, and for some cities, temperature and pH measurements. The dataset incorporates outputs of bioinformatic analyses, including abundances of ARGs and different taxa obtained through reference-based classification of sequencing reads. We identified 961 distinct ARGs. Most of these gene hits confer resistance to betalactam, aminoglycoside, and tetracycline antibiotics, which are also the most well-represented categories in the ResFinder database [ 27 ].
Aligning fragments to a genomic database produced hits across all three superkingdoms (Archaea, Bacteria, and Eukaryota) and identified 3431 distinct genera. Metagenomic assemblies are presented in the form of contigs, contig bins, and MAGs, accompanied by relevant quality assessments and annotations. Gene annotations of the dereplicated mediumand high-quality MAGs reveal a maximum of 9233 and a minimum of 346 coding sequences (CDS) per MAG. Among the 2332 MAGs, 56 encode the 23S, 16S, and 5S rRNA genes, along with tRNAs for at least 18 of the 20 possible amino acids, and have completeness > 90% and contamination < 5%, thus fulfilling the MIMAG criteria [ 32 ] for highquality MAGs.
The taxonomic classification results for the selected MAGs revealed 879 distinct bacterial genera. Analyzing the possible sources of the contigs identified 14 989154 contigs of plasmid origin, 8346 618 of phage origin, and 44 515 839 of bacterial chromosomal origin. The qPCR tests cover results for 10 bacterial species and 6 protozoan species. ProxiMeta analysis yielded 3105 genome clusters.
An overview of the data and how it is organized is presented in Fig. 1 .
## Validation of acquired metadata for the samples
All metadata was provided by the involved partners. Metadata had three obligatory entries: sample type, sampling dates and geographical location, which all partners must provide. There were additional entries that partners provided, when possible, e.g. time, temperature and pH. Metadata information was validated with the following: Geographical origin of sample identifiable via openstreetmap.org and sampling date with a specific format: yyyy-mm-dd.
## Validation for MAGs to be accepted in the final dataset
Validation of MAGs is outlined in our previous work [ 8 ]. Our goal was to construct a comprehensive, non-redundant, and environmentally representative reference genome dataset covering all sewage samples. A wide array of 34 725 MAGs, originating from two distinct sources: 23 082 genomes from binned contigs of each single sample analysis, and 11 643 genomes from binned contigs of co-assembled samples by site. The primary goal was the selective retention of medium- Step 3: FastQ files derived from direct shotgun sequencing underwent processing through different metagenomic workflows. These involved mapping trimmed reads against both a genomic sequence database and the Resfinder ARG database, the assembly of reads into contigs, and subsequent binning of these contigs to extract MAGs. Taxonomic identification of MAGs was accomplished using the Genome Taxonomy DataBase (GTDB). Step 4: The outputs from various approaches, along with additional analysis results, were integrated into a PostgreSQL database. Quality-filtered fastQ files and fasta files of contigs, MAGs, and Hi-C clusters were uploaded to the ENA. A subset of tables from the database is available on the Figshare repository (see the 'Data records' section). quality genomes, defined by a contamination level ≤10% and a completeness ≥50%. This selection process yielded a refined collection of 12 687 MAGs. Then we identified and removed duplicate MAGs, culminating in the final collection of 2332 genomes.
## Analytical outputs are stored in a SQL database
Results in the form of summary Tables of the analyses were organized into a PostgreSQL database referred to as sewage database. During the schema design of the sewage database we balanced between the dogma of canonical data representation and their usability. The simplified diagram of the database is depicted in Fig. 2 . The meta information of the samples is stored in two separate Tables to minimize data repetition. These are the 'meta' and the 'location' Tables. The Table 'meta' contains additional information about the samples including collection date, sample type, DNA purification method, IDs for the sample record in the European Nucleotide Archive (ENA) and a reference to the collection site. The 'replica' column indicates which sequencing runs were technical replicates of the same sewage sample. The closely related Table, 'location', details the sites where samples have been taken, including their country, city, the GPS location information and the name of the plant.
Abundance Tables hold the results of the three different analysis pipelines. Abundance Tables for the ARG classification and genomic reference-based classification approaches ('resfinder_gene_abundance', 'resfinder_class_abundance', 'genomic_abundance') contain the number of reads aligned to each resistance gene or genome. Abundance Table for the high-and medium-quality dereplicated MAG collection called 'mag_abundance'. This Table contains the number of bases aligned to each MAG per sample.
The results of Hi-C sequencing analysis, including details on the genome clusters, are located in the Table 'proximeta', and qPCR results can be found in Table 'qpcr'.
## Validation of the sewage database
Validation of the database comprises two main components. Technical validation of the database includes ensuring uni- The 'meta' table and the 'location' table encompass metainformation. The tables 'resfinder_gene_abundance' and 'resfinder_class_abundance' joined with detailed gene information ('node_resfinder_gene', 'resfinder_gene_phenotype') summarize the results of the AMR gene identification. The taxonomic classification of fragments using a genomic database is presented in the 'genomic_abundance' table. Results derived from the assembly-based metagenomic workflow include the contig table containing basic contig information and the potential origin of each contig. This 'contig' table is linked to the 'mag' table providing the association of contigs with their respective MAG, while taxonomy and quality information for MAGs can be found in the 'gtdb' and 'checkm2' tables. Results from Hi-C sequencing analysis are in table 'proximeta'. The table 'qpcr' stores the results of identification of pathogens through qPCR.
formity and consistency in data types, lengths, and formats across corresponding fields in different Tables. Verifying the integrity of primary and foreign keys used for establishing relations between Tables, ensuring referential integrity, and preventing orphaned records. We applied optimization strategies to avoid data redundancy and inconsistencies, particularly when dealing with repeated information across Tables and within a Table . An example includes the introduction of the 'gtdb_warning' Table to store each warning message uniquely. Additionally, uniform meta information such as 'sequencing platform' shared across all samples was not loaded into the 'meta' Table, while making this information available within the database description. We also took advantage of the use of unique constraints the database engine offers to ensure cleanliness of the data. For instance, in Table 'mag', it is impossible to store any ad-hoc combination of properties 'meta_id', 'loca-tion_id', and 'bin_number', their three-tuple must be unique, which conforms to the definition of the mag. In the database, a version of taxonomy is also represented. There are instances where different taxa share the same name. To uniquely identify a taxon, both the taxon name and its rank (e.g. genus, species, etc.) must be considered together, ensuring that each taxon is distinct.
Under content validation, we conducted a series of diverse queries to ensure that the structural design allows an effective data retrieval. We evaluate the accuracy and validity of the retrieved data from the database in comparison to the original input and expected outcomes. We replicated some of the analysis presented in Becsei and Fuschi et al. [ 8 ] study using data retrieved from the database.
## Discussion
This study provides a robust and unprecedented dataset on sewage metagenomics from five European cities, addressing the critical need for comprehensive and accessible data to monitor microbial communities, AMR, and pathogens. By utilizing advanced sequencing techniques such as shotgun metagenomics and Hi-C sequencing, alongside targeted pathogen detection through qPCR, the dataset offers both breadth and depth for microbial surveillance.
The inclusion of 239 samples, processed across diverse workflows, underscores its value as a resource for studying microbial ecology, microbial adaptation, population dynamics, and potential pathogens in urban environments.
A central feature of this study is the development of a Post-greSQL database designed to accommodate and organize the vast array of data generated. This database serves as an accessible and flexible platform that integrates metadata, analytical outputs, and derived genomic features, enabling efficient querying and data exploration. Structured to minimize redundancy and ensure data integrity, it provides a robust foundation for researchers to engage with the data. Tables within the database store essential information, such as sample metadata, ARG abundance, taxonomic profiles, and MAG-related data, which are linked via carefully constructed schema. Furthermore, the inclusion of Hi-C sequencing results and qPCR data in dedicated Tables makes it possible to explore genomic linkages and validate findings across different analytical methods.
Our long-term goal is to support the standardization of organizing and storing metagenomic data. We aim to develop a workflow that facilitates the integration of analytical outputs from various metagenomic pipelines and other connecting data into a unified database structure that can be accessed and re-used by others.
The database's design also supports future updates, allowing for the inclusion of additional samples or analytical outputs, ensuring its relevance for sewage studies, and expanding applications over time.
Future efforts should also focus on integrating these types of datasets with clinical or epidemiological data [ 33 ] to provide a more comprehensive understanding of healthrelated trends. Additionally, developing user-friendly pipelines or visualization tools tailored to this dataset could enhance accessibility and usability for a broader scientific audience.
In conclusion, this study represents a significant step toward standardized and accessible metagenomic surveillance. The integration of a thoughtfully designed database not only ensures reusability but also sets a benchmark for data sharing and analysis in sewage metagenomics. This concept supports the transition from fragmented spreadsheet based data management to a standardized, queryable platform.
## References
1. Hovi, Shulman, Avoort (2012) "Role of environmental poliovirus surveillance in global polio eradication and beyond" *Epidemiol Infect*
2. Izquierdo-Lara, Elsinga, Heijnen (2021) "Monitoring SARS-CoV-2 circulation and diversity through community wastewater sequencing, the Netherlands and Belgium" *Emerg Infect Dis*
3. Izquierdo-Lara, Heijnen, Munnink (2023) "Rise and fall of SARS-CoV-2 variants in Rotterdam: comparison of wastewater and clinical surveillance" *Sci Total Environ*
4. Aarestrup, Brown, Detter (2012) "Integrating genomebased informatics to modernize global disease monitoring, information sharing, and response" *Emerg Infect Dis*
5. Qin, Li, Raes (2010) "A human gut microbial gene catalogue established by metagenomic sequencing" *Nature*
6. Hasman, Saputra, Sicheritz-Ponten (2014) "Rapid wholegenome sequencing for detection and characterization of microorganisms directly from clinical samples" *J Clin Microbiol*
7. Tringe, Rubin (2005) "Metagenomics: DNA sequencing of environmental samples" *Nat Rev Genet*
8. Becsei, Fuschi, Otani (2024) "Time-series sewage metagenomics distinguishes seasonal, human-derived and environmental microbial communities potentially allowing source-attributed surveillance" *Nat Commun*
9. Munk, Brinch, Møller (2022) "Genomic analysis of sewage from 101 countries reveals global landscape of antimicrobial resistance" *Nat Commun*
10. Hendriksen, Munk, Njage (2019) "Global monitoring of antimicrobial resistance based on metagenomics analyses of urban sewage" *Nat Commun*
11. Njage, Van Bunnik, Munk (2023) "Association of health, nutrition, and socioeconomic variables with global antimicrobial resistance: a modelling study" *Lancet Planet Health*
12. Brinch, Leekitcharoenphon, Duarte (2020) "Long-term temporal stability of the resistome in sewage from Copenhagen" *mSystems*
13. Jespersen, Munk, Johansen (2023) "Global within-species phylogenetics of sewage microbes suggest that local adaptation shapes geographical bacterial clustering" *Commun Biol*
14. Nieuwenhuijse, Munnink, Phan (2020) "Setting a baseline for global urban virome surveillance in sewage" *Sci Rep*
15. Pipek, Medgyes-Horváth, Dobos (2019) "Worldwide human mitochondrial haplogroup distribution from urban sewage" *Sci Rep*
16. Leinonen, Akhtar, Birney (2011) "The European nucleotide archive" *Nucleic Acids Res*
17. Leinonen, Sugawara, Shumway (2011) "The sequence read archive" *Nucleic Acids Res*
18. Richardson, Baldi (2023) "MGnify: the microbiome sequence data analysis resource in 2023" *Nucleic Acids Res*
19. Tedersoo, Küngas, Oras (2021) "Data sharing practices and data availability upon request differ across scientific disciplines" *Sci Data*
20. Wilkinson, Dumontier, Ijj (2016) "The FAIR guiding principles for scientific data management and stewardship" *Sci Data*
21. Sharpton (2014) "An introduction to the analysis of shotgun metagenomic data" *Front Plant Sci*
22. Setubal (2021) "Metagenome-assembled genomes: concepts, analogies, and challenges" *Biophys Rev*
23. Kang, Li, Kirton (2019) "MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies" *PeerJ*
24. Stalder, Mo (2019) "Linking the resistome and plasmidome to the microbiome" *ISME J*
25. Ágnes, Patrick (2024) "An extensive metagenomic dataset of sewage samples across five European cities: a resource for analyzing microbial communities, antimicrobial resistance and pathogens" *Figshare*
26. Clausen, Aarestrup, Lund (2018) "Rapid and precise alignment of raw reads against redundant databases with KMA" *BMC Bioinf*
27. Florensa, Kaas, Clausen (2022) "ResFinder-an open online resource for identification of antimicrobial resistance genes in next-generation sequencing data and prediction of phenotypes from genotypes" *Microb Genomics*
28. Osakunor, Munk, Mduluza (2020) "The gut microbiome but not the resistome is associated with urogenital schistosomiasis in preschool-aged children" *Commun Biol*
29. Fang, Tan, Wu (2019) "PPR-Meta: a tool for identifying phages and plasmids from metagenomic fragments using deep learning" *GigaScience*
30. Seemann (2014) "Prokka: rapid prokaryotic genome annotation" *Bioinformatics*
31. Lieberman-Aiden, Van Berkum, Williams (2009) "Comprehensive mapping of long range interactions reveals folding principles of the human genome" *Science*
32. Bowers, Kyrpides, Stepanauskas (2017) "Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea" *Nat Biotechnol*
33. Karkman, Berglund, Flach (2020) "Predicting clinical resistance prevalence using sewage metagenomic data" *Commun Biol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12764368&blobtype=pdf | # Incidence of Cytomegalovirus (CMV) Infection in After Kidney Transplant Patients: A Systematic Review and Meta-Analysis
Gleice Jesus, | Róger Costa, Gabriel Franco, | Cleyde, C Marconi, Maria Arriaga, Eduardo Netto
## Abstract
Kidney transplantation is recognised by the World Health Organisation as the most effective therapy for end-stage renal disease, offering substantial improvements in survival and quality of life. However, the immunosuppression required to prevent graft rejection predisposes recipients to opportunistic infections, among which cytomegalovirus (CMV) remains a leading cause of morbidity and mortality. To determine the incidence of CMV infection in kidney transplant recipients and identify clinical and laboratory predictors and associated risk factors. A systematic review and meta-analysis, registered in PROSPERO (CRD42024524165), was conducted in accordance with PRISMA guidelines. PubMed, Web of Science, and LILACS databases were searched up to 6 March 2024 for cohort studies reporting CMV incidence post-kidney transplantation. Methodological quality was assessed using the Newcastle-Ottawa Scale (NOS). Publication bias was evaluated using funnel plots, Egger's test, Kendall's tau, and the fail-safe N. Fifteen studies met the inclusion criteria, comprising 3888 patients with a mean age of 45.1 years and a mean follow-up of 7.14 months. The pooled incidence of CMV infection was 42% (95% CI: 30%-54%), with high heterogeneity (I 2 = 98.75%; p < 0.001). Sensitivity analysis confirmed the robustness of the results, and no significant publication bias was detected. The main risk factors were: advanced age, paediatric age (< 5 years), donor-recipient serodiscordance (D+/R-), posttransplant lymphopenia, and reduced cell-specific immunity defect. CMV infection remains a common and clinically significant complication following kidney transplantation. With a high incidence and strong association with specific laboratory and clinical predictors. Individualised prevention strategies and early virological and immunological monitoring, especially in highrisk groups, are essential to reduce morbidity and mortality and preserve graft function.
## 1 | Introduction
Kidney transplantation is recognised by the World Health Organisation (WHO) as the most effective therapeutic option for patients with end-stage renal disease, significantly improving quality of life and increasing survival rates to 83% compared with dialysis [1]. Despite these benefits, the immunosuppression required to prevent graft rejection markedly increases susceptibility to opportunistic infections, affecting up to 57.1% of recipients within 6 months post-transplant. Among these, cytomegalovirus (CMV) infection is a leading cause of morbidity and mortality, with an overall incidence of 8.8% between the third and sixth months after transplantation [2] and a ninefold higher risk of graft rejection (odds ratio [OR] = 8.9; 95% CI: 2.8-28.1; p = 0.001) [3]. Incidence rates vary between populations: 52% in children [4], 60.7% in older adults [5], and up to 75% in donor-recipient serodiscordance (D+/R-) cases [6], the latter representing one of the main risk factors for primary infection and symptomatic disease. Over 3 decades, the incorporation of sensitive laboratory methods has changed the way CMV is detected after kidney transplantation. The diagnosis of infection can be carried out by serological methods (IgG, IgM), pp65 antigenemia assays, viral culture, or molecular techniques [7]. Quantitative polymerase chain reaction (qPCR) is particularly valuable for early detection of viral replication [4], enabling timely therapeutic interventions and reducing progression to symptomatic disease [8]. It offers superior sensitivity and specificity compared with pp65 antigenemia [1,2,5], particularly in leukopenic patients [4,7].
Cytomegalovirus (CMV) infection is one of the most significant infectious complications in the post-kidney transplant period, associated with high morbidity and mortality rates and an increased risk of graft rejection. Previous studies indicate incidences ranging from 8.8% to over 70%, with significant variations according to sociological, clinical, and laboratory factors. This variability makes it difficult to define accurate and universal estimates, in addition to limiting the implementation of prevention and monitoring protocols that are appropriate for all clinical contexts. Unlike syntheses focused on immunosuppressive regimens, this review anchors the incidence of CMV in laboratory markers, emphasising comparable early detection and risk stratification across centres. This, the present systematic review and meta-analysis aims to estimate the incidence of CMV infection in kidney transplant recipients, synthesise evidence based on laboratory markers that allow detection, in addition to showing the evolution of diagnostic methods and the usefulness of infection predictors to anticipate events, and finally, identify associated risk factors. An integrated understanding of these metrics can guide more effective preventive and preemptive surveillance strategies, with the potential to reduce clinical complications, preserve graft function, and optimise the prognosis of transplant patients.
## 2 | Methodology
## 2.1 | Protocol Registration
## 2.3 | Eligibility Criteria
After screening the articles found through the search strategy, those that adequately met the inclusion and exclusion criteria were selected. For inclusion, cohort and case-control studies published up to March 2024 that evaluated CMV infection in living patients after kidney transplantation, presenting data on the incidence of infection, were selected. Systematic reviews, meta-analyses, letters to the editor, monographs, studies on infection by other viruses, transplantation of other organs, comparative incidence of medications, and studies without full text available were excluded.
## 2.4 | Study Selection
Articles were imported into Rayyan [11] for screening. After duplicate removal, three reviewers (G.J., R.C., C.M.) independently assessed titles and abstracts. Discrepancies were resolved through consensus meetings. In cases where an agreement could not be reached, a fourth reviewer acted as a tie-breaking mediator. Studies that met the criteria were read in full to confirm eligibility.
## 2.5 | Data Extraction
Four authors (G.J., R.C., C.M., G.F.) extracted data into an Excel spreadsheet, including: author, publication year, title, country, study design, sample size, mean age, population, positive cases, follow-up duration, incidence, time to infection onset, risk factors, diagnostic methods, laboratory variables and conclusions. The data were then included in a Excel worksheet.
## 2.6 | Risk of Bias Assessment
The methodological quality of the included observational studies was assessed using the Newcastle-Ottawa Scale (NOS) [12] tool, as recommended for cohort and case-control studies. The instrument evaluates three parameters: selection, comparability and outcome/exposure, as can be seen in Table 1. The evaluation was made by one author (G.J.) and verified by two others (G.F. e M.A.). Risk of bias was then categorised as high, moderate, or low. The researchers resolved any discrepancies by jointly re-evaluating a paper.
## 2.7 | Statistical Analysis
For the data analysis of this meta-analysis, the Jamovi Project programme (Jamovi version 2.4.8) was used. The incidence of CMV infection and their respective 95% confidence intervals (CI) were used as the ratio between the number of positive patients and the total number of patients included in the study. The incidence was expressed as a percentage. Heterogeneity between studies was assessed using the I 2 statistic. Publication bias was assessed using the funnel plot, Egger's test, and Fail-Safe N.
## 3 | Results
## 3.1 | Search Strategy
A total of 755 potentially eligible articles were identified for inclusion in this review. After removing duplicates (n = 44), 711 articles were screened by title and abstract using Rayyan [10]. Of these, 679 articles were excluded, and the remaining 32 were selected for full-text reading. Subsequently, 17 studies were excluded: 7 for not providing sufficient incidence data, 5 for presenting comparative medication incidence, 2 for addressing kidney graft rejection incidence, 1 for reporting incidence in deceased patients, 1 for associating incidence with cancer, and 1 for being a meta-analysis. After completing the eligibility assessment, 15 articles were included in the systematic review and meta-analysis, as shown in Figure 1.
Table 2 of the general characteristics of the studies shows the sample size (N) of each study, the total of them was 3888 (median of 159), the mean age was 45.1 and the mean follow-up time was 7.14 months.
## 3.2 | Risk Bias
To assess the methodological quality and risk of bias of the observational studies included in this systematic review, the Newcastle-Ottawa Scale (NOS) [12] was used, and all 15 articles included were cohort articles and were endorsed by the NOS cohort model. The scale assigned up to 9 points distributed among the parameters.
In the Table 1, the studies evaluated obtained scores ranging from 6 to 9 points. The presence of publication bias was also assessed using the funnel plot, as shown in Figure 2, in addition to the Fail-Safe N = 11.326 (p < 0.001), Kendalls Tau = 0.067 (p = 0.770) and Egger's Regression = 0.633 (p = 0.527) tests. These results confirm that there is no evidence of publication bias, demonstrating the stability and reliability of this meta-analysis.
The funnel plot analysis was complemented by Kendall's Tau (value 0.124), Egger's regression (value 1.394), and Fail-Safe N (363.000) tests. And despite the slight asymmetry in the funnel plot, none of the complementary tests showed significant evidence of publication bias, and the high Fail-Safe N suggests that the results of the meta-analysis are robust.
## 3.3 | Statistical Analysis
The present meta-analysis, consisting of 15 studies, identified a pooled incidence of CMV infection of 0.42 (95CI: 0.30%-0.54%), considering follow-up periods ranging from 1 to 24 months. These results demonstrate a robust and clinically relevant finding. The I 2 value was 98.75% (p < 0.001), showing high heterogeneity among the studies. Therefore, the meta-analysis was carried out under a random-effects model. Figure 3 presents the individual results of the studies included in the meta-analysis, as well as the estimated pooled effect. It is observed that, although some studies present wider confidence The estimate evidenced reflects the aggregate prevalence of infection among the studies, considering methodological and sample variations. The studies with the greatest weight in the analysis, such as Sousa et al. [1], Pouteil-Noble et al. [14], Hemmersbach-Miller et al. [5] corroborate this finding, reinforcing the consistency of the data and the reliability of the results.
Despite the heterogeneity between the studies, possibly due to differences in diagnostic methods, populations evaluated and follow-up periods, as demonstrated by Salazar et al. [4], Sousa et al. [1,2] and Feng et al. [20], the findings support the clinical relevance of CMV infection in the context of kidney transplantation. The body of evidence supports the importance of virological monitoring and the implementation of appropriate preventive strategies, especially in higher-risk groups, such as patients with D+/R serology-, children, and the elderly.
Also in Figure 3, it is observed that two works by the same author, Sousa et al. [1] and Sousa et al. [2], carried out in different years, presented very different incidences of CMV. The 2010 study recorded an incidence of 0.13 (95 CI: 11%-15%), while the 2021 study had a significantly higher incidence of 0.70 (95 CI: 66%-74%). A possible explanation for this discrepancy is the age profile of the populations analysed: in 2021, the authors included older patients (mean of 52 years), while in 2010 they evaluated younger patients (mean of 41 years). Considering that advanced age is a recognized risk factor for CMV infection, this demographic difference may have had a direct influence on the higher incidence observed in Sousa's study in 2021 [2].
## 3.4 | Laboratory Predictors and Diagnostic
## Performance
In classic and contemporary series, qPCR detected CMV before serology, including as the only positive method in some cases, supporting its use as an early detection tool [16]. The pp65 antigenemia appeared, on average, on day 45. Cellular specific anti-CMV immunity measured early predicted subsequent events, with a higher incidence of infection in patients with no or weak response [3]. ALC-D28 (absolute lymphocyte count 28th day) < 1100/μL showed useful operational PPV (83% NPV) and a 3.3-fold higher risk of CMV in the 1st year [21].
Positive IgM and viruria were associated with moderate/severe forms (ratio of 3.3 for severity with positive IgM) [14]. Low levels of MASP-2 in the pre-transplant period were associated with the development of CMV disease in the first 12 weeks (p = 0.028) [6].
In parallel, AECA (Anti-endothelial antibodies) rose 1-4 weeks Reviews in Medical Virology, 2026 after detection of CMV DNA and remained high for months, suggesting endothelial injury and possible impact on rejection [15]. Some studies presented risk factors or reduction factors for infection among the patients studied, as shown in Table 1. Alakulppi et al. [17] observed a lower incidence when the donor has the IL-10 genotype (-1082 AA), with a p-value = 0.036. Feng et al. [20], showed more variations in risk factors.
The synthesis of clinical and laboratory predictors (Table 1) highlights important characteristics of early detection. qPCR identified subclinical infection before antigenemia, including as the only positive method in some cases [16]. Shiina et al. [21] It was observed that the absolute lymphocyte count on the 28th post-transplant day (ALC-D28 < 1100/μL) was associated with a higher incidence of CMV (HR = 3.32; VPN = 83%). In the study by Fernandéz-Ruiz et al. [3] it was seen that reduced anti-CMV specific cellular responses, assessed in the pre-and postimmediate, predicted subsequent CMV events. And components of the complement lectin pathway, especially MASP-2, have been associated with early CMV disease [6]. Together, these markers allow for reproducible risk stratification and guide surveillance and preemptive intervention (Table 3).
## 4 | Discussion
Our findings reinforce that surveillance strategies anchored in laboratory predictors allow us to estimate the incidence and anticipate the risk of CMV infection in a comparable manner. Older studies, based on serology or antigenemia, probably underestimated the incidence and detected the events later [13, 14]. On the other hand, the progressive incorporation of molecular methods such as qPCR, since the end of the 90s [16] and monitoring of specific cellular immunity has shifted the identification of cases to earlier stages of the post-transplant period and revealed subclinical infection. This methodological evolution explains part of the heterogeneity observed and, at the same time, justifies the inclusion of historical series to understand the trajectory of the disease over almost 3 decades [3,21] (Table 4).
Based on the studies analysed, CMV infection continues to be a frequent event after kidney transplantation with a wide variation in incidence, according to diagnostic method, population profile, and follow-up window. The study by Sousa et al. [2], reported an incidence of 8.8% between the third-and sixthmonths post-transplant, while Feng et al. [20], found rates of 47% among recipients with unfavourable immunological and serological risk factors. Despite its high frequency, studies suggest that when identified and treated early and treated appropriately, CMV infection does not compromise the survival of the actual graft. When the infection is not controlled, it can cause acute rejection, viral nephropathy, and graft failure [1]. Shiina et al. [21] showed that patients with low cellular immunity had CMV infection of more worrisome clinical condition (HR = 3.32; 95% CI = 1.08-10.2).
Stratification by subgroups reinforced the clinical applicability of the findings. Children and the elderly had a higher incidence and severity of events [4,5]. D+/R-serodiscordance remained the main risk factor for primary infection and disease [14]. In recipients of donors with expanded criteria, CMV was the most common complication, independently associated with infection by multidrug-resistant microorganisms [1]. Such patterns not only reduce heterogeneity through more homogeneous comparisons, but also enrich the clinical relevance of the review by guiding who to monitor more closely, when, and with which tools.
The absence of direct analysis by immunosuppressive regimens can be understood as a limitation, but it was a conscious methodological option to avoid confusion due to the great variation of regimens in studies. Instead, we emphasise laboratory markers that are comparable over time and useful for preemptive treatments: serial qPCR as the screening axis [16], pp65 [13] and ALC-D28 [21] to qualify initial risk, specific cellular immunity to modulate intensity [3] or duration of surveillance, and IgM/viruria as severity alarms [20].
In summary, by centring incidence on laboratory predictors and arranging markers in post-transplant timelines, this review explains historical heterogeneity, standardises comparison between studies, and delivers actionable parameters for pre-emptive surveillance and risk stratification in post-kidney transplant care.
## 5 | Conclusion
This review confirms that CMV infection remains one of the most frequent and clinically relevant infectious complications in kidney transplant recipients, presenting mainly in the first few months after the procedure. The estimated pooled incidence was 42% (95% CI: 30%-54%), with high heterogeneity (I 2 = 98.75%), reflecting substantial differences between studies in terms of population profiles, immunosuppressive regimens employed, diagnostic methods, and prophylaxis strategies adopted.
By refocusing the analysis on laboratory markers, this work offers a practical and standardisable path for surveillance: serial qPCR as the basis of early detection, pp65 antigenemia as the typical operational window. ALC-D28 < 1100/μL as a simple screening point to intensify follow-up, MIC to modulate the frequency and duration of monitoring, and, as complementary signs, low MASP-2 pre-transplant and elevated AECA after viraemia, suggesting greater susceptibility and possible vascular impact. Reduced IMR: Higher incidence of events at 12 months Monitor CMI in the pre-and post-immediate; modular qPCR frequency/prophylaxis Hemmersbach-Miller [5] Elderly (≥ 65 years old) Elevated risk of infections in year 1
Enhanced surveillance protocols; integration with qPCR and clinical assessment This approach is independent of the variability of immunosuppressive regimens between centres and eras and speaks directly to the care routine: it allows for earlier action, personalisation of risk, and protection of the graft. Sensitivity analysis in modern scenarios where qPCR is the gold standard reinforces the robustness of these conclusions.
## References
1. De Sousa, Da Fonseca, Taminato (2021) "Infectious Events in Kidney Transplant Recipients From Deceased Expanded Criteria Donors: A Prospective Cohort" *Revista da Escola de Enfermagem*
2. Sousa, Galante, Barbosa et al. (2010) "Incidence and Risk Factors for Infectious Complications in the First Year After Transplantation" *Jornal Brasileiro de Nefrologia*
3. Fernández-Ruiz, Giménez, Vinuesa (2019) "Regular Monitoring of Cytomegalovirus-Specific Cell-Mediated Immunity in Intermediate-Risk Kidney Transplant Recipients: Predictive Value of the Immediate Post-Transplant Assessment" *Clinical Microbiology and Infection*
4. Salazar, Alba, Deluchi (2008) "Cytomegalovirus Infection and Disease in Children Undergoing Solid Organ Transplantation"
5. Hemmersbach-Miller, Alexander, Sudan et al. (2019) "Infections After Kidney Transplantation: Does Age Matter?" *Clinical Transplantation*
6. Sagedal, Thiel, Hansen et al. (2008) "Impact of the Complement Lectin Pathway on Cytomegalovirus Disease Early After Kidney Transplantation" *Nephrology Dialysis Transplantation*
7. Wu, Qian, Yang (2014) "Simultaneous Monitoring of CMV and BKV by Quantitative PCR in Renal Transplant Recipients" *Journal of Virological Methods*
8. Watzinger, Ebner, Lion (2006) "Detection and Monitoring of Virus Infections by Real-Time PCR" *Molecular Aspects of Medicine*
9. Schiavo (2019) "PROSPERO: An International Register of Systematic Review Protocols" *Medical Reference Services Quarterly*
10. Page, Mckenzie, Bossuyt (2021) "The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews" *BMJ*
11. Ouzzani, Hammady, Fedorowicz et al. (2016) "Rayyan-a Web and Mobile App for Systematic Reviews" *Systematic Reviews*
12. Wells, Shea, O'connell et al. (2000) "The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses"
13. Boland, De Gast, Hené (1990) "Early Detection of Active Cytomegalovirus (CMV) Infection After Heart and Kidney Transplantation by Testing for Immediate Early Antigenemia and Influence of Cellular Immunity on the Occurrence of CMV Infection" *Journal of Clinical Microbiology*
14. Pouteil-Noble, Ecochard, Bosshard (1992) "Cytomegalovirus ( CMV) Excretion as a Factor in the Severity of CMV Disease in Kidney and Simultaneous Kidney and Pancreas Transplantation" *Transplant International*
15. Toyoda, Galfayan, Galera et al. (1997) "Cytomegalovirus Infection Induces Anti-Endothelial Cell Antibodies in Cardiac and Renal Allograft Recipients" *Transplant Immunology*
16. Costa, Miranda, Alves et al. (1999) "Detection of Cytomegalovirus Infections by PCR in Renal Transplant Patients" *Brazilian Journal of Medical and Biological Research*
17. Alakulppi, Kyllönen, Salo et al. (2006) "The Impact of Donor Cytokine Gene Polymorphisms on the Incidence of Cytomegalovirus Infection After Kidney Transplantation" *Transplant Immunology*
18. Cervera, Lozano, Saval (2007) "The Influence of Innate Immunity Gene Receptors Polymorphisms in Renal Transplant Infections" *Transplantation*
19. Watcharananan, Louhapanswat, Chantratita et al. (2012) "Cytomegalovirus Viremia After Kidney Transplantation in Thailand: Predictors of Symptomatic Infection and Outcome" *Transplantation Proceedings*
20. Feng, Yang, Wang (2016) "Incidence and Risk Factors for Cytomegalovirus Infection in Patients With Kidney Transplantation: A Single-Center Experience" *Transplantation Proceedings*
21. Shiina, Kawabe, Suehiro (2023) "Peripheral Blood Absolute Lymphocyte Count as a Predictor of Cytomegalovirus Infection in Kidney Transplant Recipients" *Transplantation Proceedings* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817652&blobtype=pdf | # Longitudinal Monitoring of Systemic Cytokines After Mild Zika Virus Infection Revealed an Association Between Th17 Polarization and Clinical and Serological Outcomes
Solène Marquine, Marie Mura, Franck De Laval, Gilda Grard, Cyril Badaut, Sébastien Briolant
## Abstract
Zika virus (ZIKV) is a neurotropic virus that can cause a variety of neurological manifestations, ranging from mild forms to severe disorders like Guillain-Barré syndrome and congenital Zika syndrome. The pathophysiology of these complications is not fully understood, but they have been linked to host immune responses, particularly a proinflammatory Th1/Th17 profile. In this study, the kinetics of 14 cytokines were characterized in ZIKV-infected patients recruited in French Guiana in 2016-2017. Cytokine concentrations were quantified using a multiplexed bead-based immunoassay in serum samples collected sequentially from 36 patients during the first month after symptom onset. This longitudinal follow-up provides chronological information on the immune response to mild-to-moderate ZIKV infection, with an early antiviral response dominated by IFN-γ, TNF-α and regulated by IL-10, followed by a peak of Th1 and then Th17-associated cytokines that persists for up to 1 month. The early presence of IL-17A, IL-21, and IL-23 was positively correlated with the maximum amplitude of the serological response (total anti-ZIKV IgG and seroneutralization titers), but also with the duration of neurological symptoms (paresthesia and muscle strength decrease), highlighting the bivalent role of Th17 immune response in ZIKV pathogenesis.
men more frequently [7]. When maternal infection occurs during pregnancy, this neurotropic and cardiotropic virus can severely impair fetal brain development and is also associated with cardiac septal defects [8], heart failure [9], and myocarditis [10], leading to congenital Zika syndrome and microcephaly, which were reported at high frequency during the Latin American outbreak [11]. The severe neurological disorders have been associated with specific immune responses [12] such as elevated levels of CXCL-10 (also termed IP-10) or Th17 cytokines [13] but further characterization is required. ZIKV transmission occurs mainly through bites of Aedes mosquitoes, but also via sexual contact, blood transfusion, and vertical transmission [14]. The infection triggers local inflammation at the inoculation site, where innate immune cells secrete type I/II interferons and proinflammatory cytokines, thereby restricting viral replication and initiating the adaptive response [15,16]. Strong activation of T lymphocytes has been reported [17], with early IFN-γ production by circulating CD8 + [18] and CD4 + T cells known to orchestrate the immune response with different T helper (Th) profiles, as well as a contribution of natural killer (NK) cells that eliminate infected targets in the early stages of the disease [19] in cooperation with cytotoxic T lymphocytes. Despite these observations, detailed analyses of immune cell populations during ZIKV infection remain limited. Most studies have assessed immune responses by measuring the levels of cytokines, chemokines or growth factors in the serum [20] or cerebrospinal fluids [21] of ZIKV-infected patients. Nonsevere infections have been associated with Th1, Th2, Th17, and Th9 polarization [22,23], but the reported responses vary depending on the immune mediators analyzed, the timing of sample collection, and cohort characteristics such as age, sex distribution (with a majority of women), and geographic origin (South of America [24] and Asia [23,25]). Coinfections with other arboviruses, frequent in endemic areas, further complicate interpretation [26,27]. Consequently, additional studies are required to better define the kinetics of ZIKV-induced immune responses and their contribution to disease severity.
The clinical observations and serological results of a year-long longitudinal follow-up of a cohort of ZIKV-infected patients recruited in French Guiana (ZIFAG cohort), between February 2016 and November 2017, have already been described [1,28]. The most common symptoms were itchy skin rashes, asthenia and/or headaches, but paresthesia, reduced muscle strength, and areflexia were also transiently reported [1]. In the present work, the aim was to study the kinetics of 14 circulating cytokines in the serum of 36 patients with mild to moderate symptoms following a ZIKV infection, up to 1 month after symptom onset, in order to explore the polarization of the immune response and its association with clinical and serological outcomes.
## 2 | Material and Methods
## 2.1 | Study Design
Patients in the ZIFAG cohort attended 12 medical consultations spread over 1 year after the onset of symptoms; a precise description of the intensity and duration of all their symptoms was carried out in the first month, along with a standardized clinical examination, and venous blood samples were also taken [1]. The present cytokine study focused solely on the first 2-6 serum samples taken from 36 of these patients (all of them have been vaccinated against yellow fever, sex-ratio: 2, median age: 39 years, 95% CI [35][36][37][38][39][40][41][42], min: 26, max: 63), that is, between Days 0 and 28 after symptom onset. The presence of ZIKV was confirmed in these samples by RT-PCR [1]. Serum viral load has already been measured and described in Matheus et al. [29]. The control group comprised 67 ZIKV-uninfected males (all of them have been vaccinated against yellow fever, median age: 31 years, 95% CI [27][28][29][30][31][32][33], min: 19, max: 48), confirmed by negative molecular biology and serological tests in another study [30]. The presence of a previous infection with an Orthoflavivirus in patients as well as the amplitude of anti-ZIKV immunoglobulins (optical density ratio = OD (target)/OD (blank)) and seroneutralization titers were assessed as previously described [28]. Their kinetic characteristics, maximum amplitude, and the day when this amplitude was reached, were determined by extrapolation from experimental data (ODr and tiers), with a curve plotted using the Wood equation [31].
## 2.2 | Cytokine Assay
A bead-based multiplexed immunoassay was performed to measure the concentrations (pg/mL) of 14 cytokines (TNF-α, IFN-γ, IL-10, IL-12, IL-17A, IL-2, IL-21, IL-23, IL-1β, IL-5, IL-4, IL-6, GM-CSF, IL-13) using the MILLIPLEX map Kit HSTCMAG-28SK (Millipore, Billerica, MA, USA), in accordance with the manufacturer's instructions. Each cytokine present in serum samples was captured by magnetic beads (BPLX MAG COOH, Luminex Inc.) coated with specific antibodies, then complexed with a biotinylated detection antibody and fluorescent streptavidin. Fluorescent signals were detected using the MAGPIX instrument and xPONENT software (Luminex Corp., Austin, TX, USA). Fluorescence intensity was converted to cytokine concentration using standard curves plotted with a 5-parameter logistic model. For statistical analysis, samples whose fluorescence values could not be interpolated from the standard curve were replaced by half the minimum detection limit.
## 2.3 | Statistical Analysis
The Kruskal-Wallis test with Dunn's post hoc test and the Benjamini-Hochberg procedure were used to compare the different time points and the control group. Correlation matrices (R Studio, v4.2.2) were constructed to relate these cytokine assay results to different variables previously described, such as IgG directed against ZIKV (Pearson), ZIKV microneutralization (Pearson), or duration of the symptoms (Spearman). Regarding the time-dependent cytokine heatmap, the cytokine measurements were centered (mean subtracted) and scaled (by standard deviation). Each cytokine concentration was then summarized by its median value at each chosen time period. These transformed values were represented using a color code implemented with the heatmap R library. Cytokine hierarchical clustering relied on an agglomeration method of Euclidean distances between rows. Principal component analysis (PCA) was performed with the FactoMineR R package (scale unit = TRUE) and using the following variables: the age of the ZIKV-infected patients, all cytokine measurements, the time points, and the presence or absence of a past Orthoflavivirus infection. The samples were visualized in PCA space using the factoextra package with PC1 and PC2 as coordinates.
## 3 | Results
## 3.1 | Serum Cytokine Kinetics During the First Month After Symptom Onset
The early cytokine response (from Days 0 to 3 after symptom onset) was dominated by the antiviral profile (Figures 1 and2A, and Supporting Information S1: Figure 1). As the median incubation period for ZIKV infection was estimated at 6.8 days (95% CI [5.8-7.7 days]) [32], this early time point corresponds to 6-11 days after infection. Hierarchical clustering of the response kinetics associated the early elevation of IFN-γ with TNF-α and IL-10 (Figures 1 and2A,B). During the second week postonset of symptoms, the Th1-Th17 response was dominant, particularly IL-12 and IL-17A (Figure 1 and Supporting Information S1: Figure 1). Three weeks after the onset of symptoms (Days 15-21), a peak in IL-2 and IL-21 was observed (Figure 1 and Supporting Information S1: Figure 1). Four weeks after symptom onset, IL-23 still showed a statistically significant difference compared to the control group (p < 0.01, Kruskal-Wallis test with Dunn's post hoc test and Benjamini-Hochberg correction), while all the other cytokines had returned to homeostasis (Figure 2C and Supporting Information S1: Figure 1). IL-23 did not peak, but increased steadily up to 1 month in ZIKV-infected individuals in the cohort (Figure 2C and Table 1). The presence of a previous infection with an Orthoflavivirus in patients has no impact on the cytokine profile, as illustrated by a PCA (Supporting Information S2: Figure 2). A Wilcoxon test was used to assess gender differences for each cytokine at each time point. No statistically significant differences were observed between male and female patients (Supporting Information S3: Supplementary File).
## 3.2 | Correlation Between Circulating Cytokine Concentrations and Serological Outcomes
Correlations between cytokine concentrations and previously published serological data from the same patients infected with ZIKV [28] were investigated. The maximum amplitude of the anti-ZIKV IgG response (median value of eight (ODr), Level max values in Table 1 in Marquine et al. [28]) was positively correlated with cytokine concentrations, mainly IFN-γ, IL-10, IL-12, IL-17A, and IL-5 (Figure 3A). At an early stage (from Days 0 to 3 after symptom onset), levels of IFN-γ (R 2 = 0.49, p = 0.01, Pearson), IL-12 (R 2 = 0.46, p = 0.02, Pearson), IL-1β (R 2 = 0.47, p = 0.02, Pearson), IL-2 (R 2 = 0.47, p = 0.02, Pearson), and IL-17A (R 2 = 0.52, p = 0.006, Pearson) were statistically significantly correlated with the magnitude of the anti-ZIKV IgG response (Figure 3A). With regard to maximum seroneutralization titers (median value of 151, Level max values in Table 1 in Marquine et al. [28]), a statistically significant positive correlation was observed with IL-21 in the early stage (Days 0-3) (R 2 = 0.66, p < 10 -3 , Pearson) and late stage (> 21 days) (R 2 = 0.84, p < 10 -3 , Pearson) (Figure 3B).
## 3.3 | Correlation Between Circulating Cytokine Concentrations and Clinical Outcomes
In the current cohort, the median age was 40 years [IQR 34 -45] and the majority of patients were male (71%). ZIKV infections were mild to moderate, with the main symptoms being rash (92%), asthenia (73%), and headache (69%). No patient developed severe neurological symptoms, but 51% experienced areflexia during the first month after symptom onset, and 20% complained of decreased muscle strength and paresthesia [1] for 2-17 days (min-max) and 1-7 days, respectively (Figure 4). These durations were statistically significantly and positively correlated with early concentrations of IL-1β (ρ = 0.56, p = 0.003, Spearman), IL-2 (ρ = 0.49, p = 0.007, Spearman), and IL-21 (ρ = 0.44, p = 0.03, Spearman) (Figure 5), and more generally with the Th1-Th17 response, whereas an early elevation of Th2 cytokines (IL-5, IL-13) was statistically significantly and positively correlated with headache duration (ρ = 0.57, p = 0.003 and ρ = 0.65, p < 0.001, respectively, Spearman) (Figure 5). Age was positively correlated with the duration of arthralgia (ρ = 0.51, p = 0.004) and initial viral load with myalgia (ρ = 0.53, p = 0.003), but no link was established with neurological disorders or cytokine concentration.
## 4 | Discussion
In the present study, serum cytokine kinetics during the first month after symptom onset were performed in 36 ZIKVinfected patients with mild to moderate symptoms. The early cytokine response was dominated by an antiviral profile characterized by elevated levels of IFN-γ and TNF-α, which act synergistically to promote antiviral functions [33], as well as IL-10, a key regulator of excessive inflammation. The latter mainly displays anti-inflammatory and resolutive properties protecting the host against tissue damage during the acute phase of infection [34]. Previous cohorts have also reported an increase in serum IL-10 levels during the early phase of ZIKV infection [20,25]. However, some deleterious effects of dysregulated IL-10 have been described, such as the promotion of autoimmunity [35] and elevated serum levels have been observed in patients with severe Guillain-Barré Syndrome [36]. During the second week after symptom onset, the Th1-Th17 response was dominant, particularly IL-12 and IL-17A. IL-12 may have different roles, ranging from induction of Th1-T cell responses to enhancement of NK and CD8 + T cells cytotoxicity. IL-17A induces Th17 polarized responses, whose antiviral functions can be extended: Th1response enhancement, activation and survival of cytotoxic CD8 + T cells, as observed during West Nile virus infection [37], and protective inflammatory response with low killing activity of cytotoxic CD8 + T cells [38]. But IL-17 can also promote viral infections and tissue damage [39]. Its excessive activation could therefore contribute to the pathophysiology of Guillain-Barré syndrome [40], a fairly common neurological complication associated with ZIKV infection [3,6]. Furthermore, this severe peripheral neuropathy appears much earlier following the onset of ZIKA disease symptoms than after infection with other arboviruses such as Chikungunya [5]. In a case-control study of a ZIKV-associated Guillain-Barré syndrome outbreak in French Polynesia [6], the neurological complication appeared 6 days [IQR 4-7] after symptom onset. This corresponds to the increase in the Th17 immune response observed in our follow-up cohort.
However, other immune mediators seem to be linked to the development of severe neurological complications [21,41], which could differ since only nonsevere clinical manifestations were observed during the 1-year follow-up.
Three weeks after the onset of symptoms, a peak in IL-2 and IL-21 was observed. IL-2 is a pleiotropic cytokine involved in pro-and anti-inflammatory T cell differentiation and homeostasis [42], while IL-21 is produced by NK T cells, CD4 + Th17 T cells, and, more potently, follicular helper T (Tfh) cells, which promote germinal center formation, B cell differentiation, and immunoglobulin production [43]. A higher level of serum IL-21 was also observed in another cohort of ZIKVinfected patients after the acute phase of the disease [25], as well as during primary or secondary dengue infections [44]. Four weeks after the onset of symptoms, IL-23, a key cytokine for the maintenance and expansion of Th17 cell, was consistently increased for up to 1 month in ZIKV-infected individuals in the cohort compared to the control group.
Associations between circulating cytokine concentrations and serological outcomes at the early stage of infection were evaluated. IL-17A was positively correlated with the maximum amplitude of anti-ZIKV IgG response, and IL-21 was positively correlated with the maximum seroneutralization titers, both in the early and late stage of infection. IL-17A and IL-21 are both markers of Th17 response and their early presence may be critical for the B cell response, especially IL-21 at an early stage, which may sustain germinal center and B cell maturation [43]. It could be hypothesized that individuals with higher seroneutralization titers may have prolonged germinal center reaction, with persistence of IL-21 at a later stage following infection.
Relationships between circulating cytokine concentrations and clinical outcomes of patients infected with ZIKV were explored. The duration of decreased muscle strength and paresthesia was positively correlated with the Th1-Th17 responses (IL12, IL-1β, IL-2, IL-23, and IL-21), while the duration of headaches was correlated with an early elevation of Th2 cytokines (IL-5, IL-13). None of them correlated with viral load, and may rely on an excessive immune response. Other studies [25,27] have also reported a positive correlation between IL-5 levels and headaches during ZIKV infection. In contrast, we did not observe any association between high IL-10 levels and arthritis [27] or myalgia [25] in the present study. Nevertheless, the duration of these symptoms correlated with age and viral load in our study, suggesting that they were driven by viral activity. The positive correlation of Th17 cytokines with the magnitude of the humoral response, as well as the duration of neurological symptoms suggests a bivalent role of Th17 response in ZIKV pathophysiology, as it may induce an excessive immune response. Antigen mimicry, bystander activation, viral neurotropism, and cytotoxicity have been hypothesized as mechanisms underlying the development of autoimmune neurological conditions associated with ZIKV infection [45]. In a recent mouse model of inflammatory neuropathy, a high accumulation of CD4 + T cells that secrete IL-21 was observed in peripheral nerves. A combination of techniques (single-cell RNA sequencing, histology, and cytometry with intracellular cytokine staining) was used to characterize these cells as peripheral Tfh-like cells, some of which also secreted IFN-γ and IL-10. Knocking down the IL21 receptor protected the animals developing. This was associated with a decrease in the infiltration of pathological CD4 + T cells, as well as a reduction in the number of innate cells, which are also involved in this process [46]. In humans, the number of circulating Tfh cells increased in a subgroup of patients with Guillain-Barré syndrome [47]. Though, IL21 is critical for the development of Tfh cells [48] and for the maturation and differentiation of B cells. A positive correlation has also been observed between elevated serum levels of IL-21 and antibody levels in patients infected with the dengue virus [44]. However, the regulation of the fate of B cells is complex since IL-21 could also induce the apoptosis of naïve B cells [49]. An important limit of our study is the absence of cellular data to refine the origin of the circulating cytokines and the functional status of cell subtypes, as not only time but location and origin matter. In a pediatric cohort of ZIKV-infected children, PBMC phenotyping using CyTOF identified CD14 + monocytes expressing CD169 as key factors of the immune response, as well as CXCL10 upregulation. In this study, there was no impact of prior Orthoflavivirus infection on the innate immune response to ZIKV, in accordance with our data obtained in an adult population [50]. Another CyTOF characterization of PBMC, isolated from viremic adults in endemic areas (blood donors followed over 3 months), revealed a coordinated immune cellular signature associated with higher titers of ZIKV-neutralizing antibodies. During the acute phase of ZIKV infection, the number of specific innate and adaptive cell types temporarily increased, including intermediate CD14 + CD16 + monocytes, CD69 + NK, HLA-DR + CD38 + nonnaïve CD8 + T cells, Th1 CD4 + T cells, and Tbet + plasma cells [51]. Unfortunately, this study neither assessed circulating Tfh cell subpopulations very precisely nor included cytokine measurements (in serum or intracellularly), that would have helped to interpret the origin of circulating IL-21 and other Th17 cytokines.
Overall, this longitudinal cytokine monitoring provides important chronological information on the immune responses to mild-to-moderate ZIKV infection: (i) an early antiviral response dominated by IFN-γ and TNF-α, controlled by an early increase in the immune modulator IL-10; (ii) a Th-1 response that peaks between 1 and 2 weeks after symptom onset, followed by a Th17 response that peaks 1 week later; (iii) resolution of the Th1 response by the third week, but persistence of the Th17 response for up to a month. This response pattern did not differ with regard to dengue serological status. Finally, the early increase in Th17 cytokines was positively correlated with maximal IgG response and seroneutralization titers, which may have a beneficial effect on infection, but was also correlated with the duration of neurological symptoms and may therefore have deleterious effects for the patient. Given that Guillain-Barré syndrome has been associated with ZIKV infection in many countries, and that the maternal Th1/Th17 profile after ZIKV infection has been implicated in congenital Zika syndrome in Brazil [13], the balance and early increase of the Th17 response to ZIKV should be considered of the greatest importance. Further research is still needed to better understand its precise role in the pathogenesis of long-term symptoms following ZIKV infection and to identify specific immune signatures that could predict the clinical outcome of the disease.
## References
1. "Note: All cytokines concentrations are in pg/mL"
2. "a D, days after symptom onset, median and"
3. De Laval, Aubigny, Mathéus (2018) "Evolution of Symptoms and Quality of Life During Zika Virus Infection: A 1-Year Prospective Cohort Study" *Journal of Clinical Virology*
4. Leonhard, Lant, Jacobs (2018) "Zika Virus Infection in the Returning Traveller: What Every Neurologist Should Know" *Practical Neurology*
5. Muñoz, Parra, Pardo (2017) "Neurological Implications of Zika Virus Infection in Adults" *Journal of Infectious Diseases*
6. Scatularo, Ballesteros, Saldarriaga (2022) "Zika Heart: A Systematic Review" *Trends in Cardiovascular Medicine*
7. Brito Ferreira, Militão De Albuquerque, De Brito (2020) "Neurological Disease in Adults With Zika and Chikungunya Virus Infection in Northeast Brazil: A Prospective Observational Study" *Lancet Neurology*
8. Cao-Lormeau, Blake, Mons (2016) "Guillain-Barré Syndrome Outbreak Associated With Zika Virus Infection in French Polynesia: A Case-Control Study" *Lancet*
9. Santos, Rodriguez, Almiron (2016) "Zika Virus and the Guillain-Barré Syndrome Case Series From Seven Countries" *New England Journal of Medicine*
10. Orofino, Passos, De Oliveira (2018) "Cardiac Findings in Infants With In Utero Exposure to Zika Virus-A Cross Sectional Study" *PLoS Neglected Tropical Diseases*
11. Li, Armstrong, Zhao (2022) "Zika Virus Infection Downregulates Connexin 43, Disrupts the Cardiomyocyte Gap Junctions and Induces Heart Diseases in A129 Mice" *Journal of Virology*
12. Li, Hou, Armstrong (2025) "Zika Virus Infection in Neonatal Mice Disrupts Connexin 43 and Induces Cardiac Inflammation, Implicating Viral Myocarditis in Neonatal Pathogenesis" *Journal of Virology*
13. Musso, Ko, Baud (2019) "Zika Virus Infection After the Pandemic" *New England Journal of Medicine*
14. Foo, Chen, Chan (2018) "Biomarkers and Immunoprofiles Associated With Fetal Abnormalities of ZIKV-Positive Pregnancies" *JCI Insight*
15. Fialho, Veras, Jesus (2023) "Maternal Th17 Profile After Zika Virus Infection Is Involved in Congenital Zika Syndrome Development in Children" *Viruses*
16. Gregory, Oduyebo, Brault (2017) "Modes of Transmission of Zika Virus" *Journal of Infectious Diseases*
17. Peng, Yang, Simons (1996) "Immunologic Mechanisms in Mosquito Allergy: Correlation of Skin Reactions With Specific IgE and IgG Antibodies and Lymphocyte Proliferation Response to Mosquito Antigens" *Annals of Allergy, Asthma & Immunology*
18. Hamel, Dejarnac, Wichit (2015) "Biology of Zika Virus Infection in Human Skin Cells" *Journal of Virology*
19. Pardy, Richer (2019) "Protective to a T: The Role of T Cells During Zika Virus Infection" *Cells*
20. Samri, Bandeira, Gois (2024) "Comprehensive Analysis of Early T Cell Responses to Acute Zika Virus Infection During the First Epidemic in Bahia, Brazil" *PLoS One*
21. Maucourant, Nonato Queiroz, Corneau (2021) "NK Cell Responses in Zika Virus Infection Are Biased Towards Cytokine-Mediated Effector Functions" *Journal of Immunology*
22. Kam, Leite, Lum (2017) "Specific Biomarkers Associated With Neurological Complications and Congenital Central Nervous System Abnormalities From Zika Virus-Infected Patients in Brazil" *Journal of Infectious Diseases*
23. Almeida, Ferreira, Sonon (2021) "Cytokines and Soluble HLA-G Levels in the Acute and Recovery Phases of Arbovirus-Infected Brazilian Patients Exhibiting Neurological Complications" *Frontiers in Immunology*
24. Tappe, Pérez-Girón, Zammarchi (2016) "Cytokine Kinetics of Zika Virus-Infected Patients From Acute to Reconvalescent Phase" *Medical Microbiology and Immunology*
25. Lum, Lye, Tan (2018) "Longitudinal Study of Cellular and Systemic Cytokine Signatures to Define the Dynamics of a Balanced Immune Environment During Disease Manifestation in Zika Virus-Infected Patients" *Journal of Infectious Diseases*
26. Fares-Gusmao, Rocha, Sippert et al. (2019) "Differential Pattern of Soluble Immune Markers in Asymptomatic Dengue, West Nile and Zika Virus Infections" *Scientific Reports*
27. Petphong, Kosoltanapiwat, Limkittikul (2023) "Detection of Anti-ZIKV NS1 IgA, IgM, and Combined IgA/IgM and Identification of IL-4 and IL-10 as Potential Biomarkers for Early ZIKV and DENV Infections in Hyperendemic Regions"
28. Alves, Magalhães, Santos (2025) "Coinfection With Chikungunya and Zika Results in Mild Disease and Distinct Inflammatory Response" *Npj Viruses*
29. Sánchez-Arcila, Badolato-Correa, De Souza (2020) "Clinical, Virological, and Immunological Profiles of DENV, ZIKV, and/ or CHIKV-Infected Brazilian Patients" *Intervirology*
30. Marquine, Briolant, Claverie "Zika Virus-Specific and Orthoflavivirus-Cross-Reactive IgGs Correlate With Zika Virus Seroneutralization Depending on Prior Dengue Virus Infection" *PLoS Neglected Tropical Diseases*
31. Matheus, De Laval, Moua (2017) "Zika Virus Persistence and Higher Viral Loads in Cutaneous Capillaries Than in Venous Blood" *Emerging Infectious Diseases*
32. De Laval, Matheus, Maquart (2016) "Prospective Zika Virus Disease Cohort: Systematic Screening" *Lancet*
33. Denis, Garnier, Claverie (2023) "The Wood Equation Allows Consistent Fitting of Individual Antibody-Response Profiles of Zika Virus or SARS-CoV-2 Infected Patients" *Heliyon*
34. Fourié, Grard, Leparc-Goffart et al. (2018) "Variability of Zika Virus Incubation Period in Humans" *Open Forum Infectious Diseases*
35. Wong, Goeddel (1986) "Tumour Necrosis Factors α and β Inhibit Virus Replication and Synergize With Interferons" *Nature*
36. Ouyang, Rutz, Crellin et al. (2011) "Regulation and Functions of the IL-10 Family of Cytokines in Inflammation and Disease" *Annual Review of Immunology*
37. Smith, Allard, Wang et al. (2018) "IL-10 Paradoxically Promotes Autoimmune Neuropathy Through S1PR1-Dependent CD4 + T Cell Migration" *Journal of Immunology*
38. Hayat, Asad, Munni (2024) "Interleukin-10 Promoter Polymorphisms and Haplotypes in Patients With Guillain-Barré Syndrome" *Annals of Clinical and Translational Neurology*
39. Acharya, Wang, Paul
40. (2017) "Cell Cytotoxicity To Facilitate West Nile Virus Clearance" *Journal of Virology*
41. Hamada, Garcia-Hernandez, Reome (2009) "Tc17, a Unique Subset of CD8 T Cells That Can Protect Against Lethal Influenza Challenge" *Journal of Immunology*
42. Ma, Yao, Peng et al. (2019) "The Protective and Pathogenic Roles of IL-17 in Viral Infections: Friend or Foe?" *Open Biology*
43. Debnath, Nagappa, Murari et al. (2018) "IL-23/IL-17 Immune Axis in Guillain Barré Syndrome: Exploring Newer Vistas for Understanding Pathobiology and Therapeutic Implications" *Cytokine*
44. Li, Yang, Wang (2023) "Extensive Cytokine Biomarker Analysis in Serum of Guillain-Barré Syndrome Patients" *Scientific Reports*
45. Ross, Cantrell (2018) "Signaling and Function of Interleukin-2 in T Lymphocytes" *Annual Review of Immunology*
46. Ettinger, Sims, Fairhurst (2005) "IL-21 Induces Differentiation of Human Naive and Memory B Cells Into Antibody-Secreting Plasma Cells" *Journal of Immunology*
47. Vivanco-Cid, Maldonado-Rentería, Sánchez-Vargas et al. (2014) "Dynamics of Interleukin-21 Production During the Clinical Course of Primary and Secondary Dengue Virus Infections" *Immunology Letters*
48. Acosta-Ampudia, Monsalve, Castillo-Medina (2018) "Autoimmune Neurological Conditions Associated With Zika Virus Infection" *Frontiers in Molecular Neuroscience*
49. Seyedsadr, Bang, Mccarthy (2024) "A Pathologically Expanded, Clonal Lineage of IL-21-Producing CD4+ T Cells Drives Inflammatory Neuropathy" *Journal of Clinical Investigation*
50. Che, Qiu, Jin et al. (2016) "Circulating Memory T Follicular Helper Subsets, Tfh2 and Tfh17, Participate in the Pathogenesis of Guillain-Barré Syndrome" *Scientific Reports*
51. Vogelzang, Mcguire, Yu et al. (2008) "A Fundamental Role for Interleukin-21 in the Generation of T Follicular Helper Cells" *Immunity*
52. Kim, Manara, Grassmann (2025) "IL-21 Shapes the B Cell Response in a Context-Dependent Manner" *Cell Reports*
53. Michlmayr, Kim, Rahman (2020) "Comprehensive Immunoprofiling of Pediatric Zika Reveals Key Role for Monocytes in the Acute Phase and No Effect of Prior Dengue Virus Infection" *Cell Reports*
54. Mccarthy, Odorizzi, Lutz (2022) "A Cytotoxic-Skewed Immune Set Point Predicts Low Neutralizing Antibody Levels After Zika Virus Infection" *Cell Reports* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12203700&blobtype=pdf | # Diagnostic utility of real-time RT-PCR for chikungunya virus detection in the acute phase of infection: a retrospective study
Anupa Sajith, Varsha Iyengar, Prasad Varamballi, Chiranjay Mukhopadhyay, Sudheesh Nittur
## Abstract
Background: Chikungunya fever is a viral disease spread by Aedes mosquitoes, reported in over 110 countries. In India, it was first detected in Kolkata in 1963 and is now widespread. Diagnosis often relies on detecting anti-chikungunya IgM antibodies, but these may not be present during the acute phase as viremia can last up to 8 days, leading to underreporting. The present study aims to assess the diagnostic use of real-time RT-PCR for detecting chikungunya-specific nucleic acid in serum samples during the early stages of infection. Methodology: A retrospective cross-sectional study was conducted using archived samples collected as a part of 'the hospital-based Acute Febrile Illness (AFI) surveillance project' . AFI cases having fever ≤8 days and arthralgia without any aetiology between 2016 and 2018 from Karnataka, Kerala, and Tamil Nadu were included in the study. Samples were subjected to nucleic acid extraction followed by chikungunya real-time RT-PCR using a standardized in-house protocol. The samples that tested positive for Chikungunya by real-time RT-PCR were further tested for detecting anti-Chikungunya IgM antibodies using enzyme-linked immunosorbent assay. Demographic characterization of the cases was performed using SPSS version 20 and GraphPad Prism version 10. Results: Out of a total of 646 samples tested, 31 samples (4.79%) were positive by real-time RT-PCR for chikungunya virus, 20 of which had a Ct value of <30, indicating a relatively high viral load. Among the 31 serum samples tested for anti-Chikungunya IgM antibodies, only one showed a positive result. Demographic analysis showed that 67% of cases were male and 32% were female, respectively. Clinical data analysis showed that most of the cases presented with cough (87%), headache (80.6%), myalgia (77.4%), coryza (70.9%), vomiting (58%), and abdominal pain (38.7%).
Conclusion:The current study findings highlight the importance of screening patients with fever for up to 8 days and arthralgia for not only detecting IgM antibodies against chikungunya using ELISA but also for chikungunya virus-specific nucleic acid through real-time PCR/nucleic acid amplification techniques, or other methods. Not performing laboratory tests to screen for chikungunya-specific nucleic acid or antigen may result in underreporting of chikungunya cases, thereby impacting effective control measures and management of cases.
## 1. Introduction
Chikungunya fever is an arthropod-borne viral illness that has gained significant international attention since the early 2000s. This arboviral illness, which is spread by the Aedes mosquito, is characterized by symptoms like fever and joint pain. The name 'chikungunya' comes from the word 'kungunyala' , which in the Kimakonde language means 'to dry up or become contorted' . The disease's characteristic arched posture is caused by rheumatologic symptoms known as 'chikungunya' [1]. The United Republic of Tanzania was the first nation to detect the virus in 1952, with more countries in Asia and Africa following. The first outbreak in urban areas was documented in India in the 1970s and in Thailand in 1967. There have been progressively constant and extensive chikungunya virus (CHIKV) outbreaks since 2004. This is mainly due to changes caused to the virus that increase the likelihood of infection transmission by Aedes albopictus mosquitoes. Currently, more than 110 nations in Asia, Africa, Europe, and the Americas have been found to harbour the chikungunya virus. India published the first chikungunya fever (CHIKF) reports in 1963. Following a 32-year hiatus, the chikungunya virus reemerged in India in 2005, leading to the most significant epidemic ever reported by 2006. The disease is currently prevalent throughout the nation, with outbreaks causing massive declines in production and the economy [2].
Over 160,000 cases of chikungunya disease and over 50 deaths have been documented globally as of March 31, 2024. The following nations have the most cases reported: Brazil (161794), Paraguay (5105), Bolivia (182), Argentina (272). In Asia, the most chikungunya cases were reported in India (154), Thailand (182), Pakistan (36), and Malaysia (13). Senegal was the only African nation that reported cases [3] of chikungunya infection in 2024 [4].
In India, according to the National Center for Vector Borne Diseases Control (NCVBDC), in 2018 Karnataka had the greatest number of laboratory-confirmed chikungunya cases (2546) followed by Gujarat (1290), Madhya Pradesh (1609) and Maharashtra (1009). In 2019, Karnataka had 3664 chikungunya confirmed cases, Maharashtra had 1646, and Telangana had 1358 confirmed cases. From 2020 to 2023, Karnataka had a higher number of confirmed cases (7736) than Gujarat (6664) and Maharashtra (6097), compared to other regions. Laboratory confirmation of chikungunya cases was done using chikungunya IgM MAC ELISA through the National Center for Vector Borne Diseases Control laboratory network [5,6].
The beginning of chikungunya virus disease in symptomatic patients usually occurs 4-8 days (with a range of 2-12 days) following an infected mosquito bite. It is characterized by a sudden onset of fever that is often accompanied by intense joint pain. Usually lasting a few days, the joint pain can be severe and continue for weeks, months, or even years at a time. The symptoms of chikungunya can vary from fever to complicated neurological conditions like encephalitis in rare cases [1]. Chikungunya disease is mostly found in regions where dengue fever is endemic and is transmitted by the same vector, the Aedes mosquito. Both occur in similar tropical and subtropical areas and present with similar symptoms. This overlap can sometimes result in misdiagnosis of chikungunya, as it may be mistaken for dengue fever. Considering the similar clinical characteristics of dengue and chikungunya, diagnostic testing is crucial, particularly in dengueendemic areas [7]. Given the absence of specific antiviral treatments and the restricted availability of approved vaccines, especially for older adults due to safety concerns, it is necessary to emphasize the importance of accurate diagnostics to differentiate chikungunya from other febrile illnesses like dengue. Accurate diagnostics are pivotal for appropriate patient management and to prevent the misuse of treatments [8].
Several diagnostic tests are available for chikungunya, which include virus isolation, molecular assays and serological tests [9]. Among these methods, anti-Chikungunya IgM ELISA is the most commonly used diagnostic test [3]. Anti-chikungunya IgM antibodies typically begin to appear near the end of the first week of illness. Therefore, a convalescent-phase sample should be collected if the acute-phase sample tests negative to rule out the diagnosis. However, the viremia of chikungunya cases lasts up to 8 days, and it is difficult to detect anti-chikungunya-specific IgM antibody during the initial stages of infection. Numerous RT-PCR assays have been developed for detecting chikungunya virus, which are highly sensitive and specific. Therefore, detecting chikungunya-specific nucleic acid during the early phase of infection is imperative to estimate the extent of the current outbreak/epidemic and implement appropriate control measures [10,11]. In this study, we aimed to evaluate whether molecular diagnostics could be an useful tool during the acute phase of chikungunya virus infection.
## 2. Materials and methods
## 2.1. Selection of samples
A retrospective cross-sectional study was conducted using archived serum samples from Karnataka, Kerala, and Tamil Nadu collected as part of 'the hospital-based Acute Febrile Illness (AFI) surveillance study between 2016 and 2018. Participants provided written consent, or for minors, their guardians provided consent through an assent form, which was collected during the hospitalbased acute febrile illness study conducted between 2016 and 2018. The study was carried out following the principles outlined in the 2013 Helsinki Declaration [12] and after obtaining clearance from the Institutional Ethical Committee (IEC: 185/2024). All experiments were performed following relevant guidelines and regulations. A total of 689 serum samples were selected retrospectively and archived using the purposive sampling method, using the following inclusion criteria. This includes patients having fever ≤8 days and arthralgia without any etiology from South Indian states such as Karnataka, Kerala and Tamil Nadu from the period 2016 to 2018 (Figure 1) at the time of blood collection. Out of 689 serum samples selected, 43 were excluded due to insufficient volume. The remaining 646 samples collected during the acute stage of infection were retrieved from the biobank and used for the study.
## 2.2. Pooling of serum samples and nucleic acid extraction
All 646 serum samples were pooled by taking 5 samples in each pool, constituting 128 pools. The pooling strategy of the serum samples was performed following guidelines by ICMR [13]. Serum samples were pooled by mixing 50 microliters from each of 5 individual serum samples to create a pool with a total volume of 200 µL. From each pool, 140 µL was taken and subjected to nucleic acid extraction using a commercially available nucleic acid extraction kit (QIAamp Viral RNA nucleic acid extraction kit, catalogue number: 52906, Qiagen™, Germany) as per the manufacturer's instructions.
## 2.3. Detection of chikungunya-specific nucleic acid using in-house chikungunya real-time RT-PCR targeting E1 gene (CHIK E1)
Extracted nucleic acid was subjected to chikungunya real-time RT-PCR, previously described by Edwards CJ et al. targeting the E1 gene [14]. The following in-house standardized protocol was used: a final reaction mixture with a total volume of 20 μL, which includes 0.2 μL of nuclease-free water, 11.5 μL of buffer mix (2X), 1.0 μL of each forward and reverse primers (10 µM), 0.5 μL of probe (10 µM), 0.8 μL enzyme mix (25X) and 5 µL of extracted nucleic acid. The AgPath-ID™ One-Step RT-PCR reagents (Applied Biosystems, USA) were used to perform the real-time RT-PCR. The cycling conditions used were 50 °C for 30 min, 95 °C for 15 min, 40 cycles of 95 °C for 15s and 58 °C for 30s. The reaction was performed in Applied Biosystems QuantStudio™ 6 Flex real-time PCR System using 6-FAM as reporter, NFQ-MGB as quencher and ROX as passive reference dye. The positive pools were selected, and individual samples from positive pools were further subjected to nucleic acid extrac tion and real-time RT-PCR using the standardized protocol (Figure 2).
## 2.4. Detection of anti-chikungunya IgM antibody using ELISA
The real-time RT-PCR positive serum samples were subjected to anti-chikungunya IgM ELISA using CHIKjj Detect TM IgM ELISA (catalogue no. DA6110, InBios International, Inc.,307 Westlake Ave N, Suite 300, Seattle, WA 98109 USA).
## 2.5. Demographic and clinical data analysis
Demographic data were collected and summarized using appropriate descriptive statistics: mean ± standard deviation (SD) for normally distributed continuous variables, median with interquartile range (IQR) for non-normally distributed continuous variables, and frequencies with percentages for categorical variables. Data analysis was performed using SPSS version 20. The chi-square test was employed to compare the distribution of clinical symptoms between chikungunyapositive and negative cases. Statistical analyses were conducted using GraphPad Prism version 10, and a p-value <0.05 was considered statistically significant.
## 3. Results
## 3.1. Detection of chikungunya-specific-nucleic acid using real-time RT-PCR
Out of 646 serum samples tested, 31 were positive by chikungunya real-time RT-PCR. All the positive samples had a Ct value ranging from 13 to 35. However, all 31 cases had a fever duration of less than five days (Table 1). Chikungunya infection was detected in 4.79% of patients by chikungunya real-time RT-PCR in South India (Karnataka, Kerala and Tamil Nadu).
## 3.2. Detection of anti-chikungunya IgM antibody among real-time RT-PCR positive cases using ELISA
Only one out of 31 real-time RT-PCR positive serum samples tested positive for anti-chikungunya IgM antibody using ELISA. According to the kit interpretation, Immune status ratio (ISR) value ≥1 was considered as reactive (positive). Only one sample had an ISR value of 1.03, which had Ct value of 33.2 in real-time RT-PCR and had a history of fever for 5 days.
## 3.3. Clinical data analysis of real-time RT-PCR positive and negative cases
Among the 31 positive samples, all patients exhibited fever and arthralgia (100%), which was our inclusion criteria. Around 25 (80.6%) patients have shown headache, and 24 (77.4%) showed myalgia. Some patients were presented with abdominal pain (12/31, 38.7%) and vomiting (18/31, 58%). Few cases presented with neck stiffness (6/31,19.3%) and rashes in 2 (6.45%) cases. Apart from this, most of the cases presented with atypical clinical manifestations, like respiratory symptoms such as cough in 27 (87%) and coryza in 22 (70.9%) cases. Table 2 distribution of clinical symptoms among chikungunyapositive (n = 31) and chikungunya-negative (n = 615) individuals (Table 2). Fever and joint pain were reported in 100% of cases in both groups; however, the difference in fever occurrence reached statistical significance (p = 0.0368), likely influenced by the large sample size of the negative group.
## presents the
Vomiting was observed significantly more frequently among chikungunya-positive cases (58.1%) compared to negative cases (30.2%), with a p-value of 0.0023. Additionally, the presence of maculopapular rash was significantly higher in the positive group (6.5%) versus the negative group (0.8%), with a p-value of 0.0403.
Other symptoms, including cough, headache, myalgia, coryza, abdominal pain, and neck stiffness, did not differ significantly between the groups (p > 0.05).
These findings suggest that vomiting and the presence of maculopapular rash may serve as distinguishing clinical features in chikungunya-positive patients within the studied cohort.
## 3.4. Demographic data analysis of positive samples (n = 31)
The median age was 16 with an inter quartile range (IQR) of 10-30. Among these, 21 (67%) were male and remaining 10 (32%) were female. The cases were categorized into different age groups as per [15], aged from 5 to 9 years comprised of 16.12% (5/31), 10-14 were 22.5% (7/31), 15-19 were 22.5% (7/31), 20-24 were 6.45% (2/31), 25-29 were 9.6% (3/31), 30-34 were 6.45% (2/31) and remaining 5 cases (16.1%) were aged between 35 and 64s. All the reported cases were from the period 2017 and most of the cases were from Tamil Nadu. From total of 31 positives, 24 cases were from Denkanikottai (67%) and Anchetty (9.6%) regions of Krishnagiri district of Tamil Nadu state. Mananthavady region of Wayanad district, Kerala had 4 cases (12.9%) and Shivamogga district of Karnataka. which includes Soraba with 2 cases (6.4%) and Sagara with 1 (3.2%) case respectively.
## 4. Discussion
Chikungunya is an arboviral disease, which can cause severe joint pain and other symptoms. This can slowly lead to chronic joint pain and other neurological diseases like encephalitis in rare cases. Based on different studies conducted on chikungunya prevalence, studies have shown that the prevalence of chikungunya is high in Southern India [16]. The highest number of laboratory-confirmed cases was recorded in 2016, and this was followed by 2017 and 2019. The highest number of confirmed cases was reported from Karnataka, Maharashtra and Delhi. In 2019, a total of 12205 cases (14.9%) with laboratory confirmation of chikungunya was reported in 21 Indian states and 3 Union territories among 81914 patients who were clinically suspected of having chikungunya. Karnataka reported the highest number of cases of CHIKV (3664), with Maharashtra (1646), Telangana (1358), and Uttarakhand [1] reporting the lowest number of cases [17].
In our present study, we found that Southern India (Kerala, Karnataka and Tamil Nadu) has shown 4.79% of chikungunya infection among fever with arthralgia cases, and earlier studies conducted in 2012 also showed a RT-PCR and/or IgM-ELISA positivity with a rate of 49.3% [18]. There were reports of chikungunya outbreak in 2017 in Tamil Nadu and 2019-2020 in Kerala which showed PCR positivity rate of 90% which is followed by Karnataka, Telangana and Maharashtra with a PCR positivity rate of 15% [19]. In our study, we retrospectively analyzed the serum samples from a study conducted between 2016 and 2018, which included samples from Tamil Nadu, where an outbreak of chikungunya was reported in 2017, which correlates with the previous findings [11]. These findings indicate that Southern India continues to be endemic to chikungunya and has a high potential for future outbreaks, thus highlighting the importance of public health vigilance.
In our study, we noticed that most of the cases showed fever and arthralgia, which is the typical representation of chikungunya. However, our study also showed more respiratory symptoms of around 70-80%, which is a rare symptom of chikungunya. All these cases were negative for other respiratory viruses such as influenza virus, respiratory syncytial virus, rhinovirus, adenovirus, parainfluenza virus, coronavirus, human metapneumoviruses, as well as common respiratory bacterial pathogens. A case study reported in 2017 by Abhijeet has shown an unusual presentation of chikungunya having respiratory distress syndrome in a patient who was diagnosed with chikungunya fever [17]. Studies have shown that chikungunya can cause respiratory, renal, cardiovascular, neurological, ocular, neonatal infection with vertical transmission and skin manifestation [20]. While most people do not consider chikungunya to be life-threatening, atypical clinical manifestations can result in substantial morbidity and have been documented, particularly during epidemics [21]. Therefore, our findings indicate the importance of screening patients with respiratory symptoms with arthralgia for chikungunya viruses apart from common respiratory viruses. Here in our study, all the chikungunya real-time RT-PCR positive samples were negative for dengue NS1 antigen and anti-dengue IgM antibody by ELISA, as well as dengue-specific nucleic acid by real-time RT-PCR as most of the chikungunya affected areas overlap with dengue endemic regions, and this can cause the mosquito vector to carry both the agents. Moreover, they have indistinguishable symptoms and there may be chances of both the diseases getting misdiagnosed [7].
We also performed anti-chikungunya IgM antibody detection ELISA for real-time RT-PCR positive samples (n = 31), however, only one serum sample turned out to be positive with an index value just above the cut-off value. This sample had a Ct value of 33.2 in real-time RT-PCR with a history of fever for 5 days. These findings indicate the importance of screening for chikungunya-specific nucleic acid using real-time RT-PCR or other nucleic acid amplification techniques during the initial stage of infection, especially during the viremic phase. One limitation of our study is that we could not perform whole-genome sequencing of the positive cases due to limited funding. Moreover, the use of retrospective samples from 2016 to 2018 restricts our ability to conclude the current epidemiological status of chikungunya in India.
## 5. Conclusion
In conclusion, this study findings highlight the importance of screening patients with fever lasting up to 8 days and arthralgia for not only detecting IgM antibodies against chikungunya using ELISA but also for chikungunya virus-specific nucleic acid through realtime PCR/nucleic acid amplification techniques, or other methods capable of detecting chikungunya virus antigens. Such comprehensive screening is essential for accurately assessing the prevalence of the disease within a population, particularly in endemic regions, along with serological testing. Not performing molecular diagnostic/antigen detection techniques to screen for chikungunya-specific nucleic acid may result in the underreporting of chikungunya cases. Therefore, a com bined serological and molecular diagnostic approach can enhance case detection and provide a more accurate estimate of the disease burden.
Comprehending the prevalence of chikungunya enables effective utilization of resources, public health management and the development of targeted efforts by authorities, such as immunization campaigns or vector control strategies [22].
## References
1. Varikkodan, Kunnathodi, Azmi (2023) "An overview of Indian biomedical research on the Chikungunya virus with particular reference to its vaccine, an unmet medical need" *Vaccines*
2. "National%20Guidelines %20for%20Clinical%20Management%20of%20Chikungunya %20Fever%202023.pdf"
3. Kaur, Singh, Sidhu (2018) "Coinfection of chikungunya and dengue viruses: a serological study from North Western region of Punjab, India" *J Lab Physicians*
4. Johnson, Russell, Goodman (2016) "Laboratory diagnosis of chikungunya virus infections and commercial sources for diagnostic assays" *J Infect Dis*
5. Skalinski, Santos, Paixão (2023) "Chikungunya seroprevalence in population-based studies: a systematic review and meta-analysis"
6. Kumar, Kamaraj, Khan (2021) "Seroprevalence of chikungunya virus infection in India, 2017: a crosssectional population-based serosurvey" *Lancet Microbe*
7. (2013) "World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects" *World Medical Association*
8. Barathidasan, Sharmila, Raj (2022) "Pooled sample testing for COVID-19 diagnosis: evaluation of bi-directional matrix pooling strategies" *J Virol Methods*
9. Edwards, Welch, Chamberlain (2007) "Molecular diagnosis and analysis of Chikungunya virus" *J Clin Virol*
10. Abhold, Wozniak, Mulcahy (2023) "Demographic, social, and behavioral correlates of SARS-CoV-2 seropositivity in a representative, population-based study of Minnesota residents" *PLOS One*
11. (2021) "The Translational Research Consortia (TRC) for Chikungunya Virus in India. Current status of chikungunya in India" *Front Microbiol*
12. Ray, Ratagiri, Kabra (2012) "Chikungunya infection in India: results of a prospective hospital based multi-centric study" *PLOS One*
13. Rs, Suresh, Saini (2019) "Clinical and virological characteristics of patients during the outbreak of chikungunya in Thiruvananthapuram" *J Clin Virol Plus*
14. Singh (2017) "Acute respiratory distress syndrome: an unusual presentation of chikungunya fever viral infection" *J Glob Infect Dis*
15. Rajapakse, Rodrigo, Rajapakse (2010) "Atypical manifestations of chikungunya infection" *Trans R Soc Trop Med Hyg*
16. Schwartz, Albert (2010) "Biology and pathogenesis of chikungunya virus" *Nat Rev Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12674863&blobtype=pdf | # Global Transmission and Evolution of Chikungunya Virus: Origins, Adaptive Mutations, and Intercontinental Spread of the Three Genotypes
Yujia Hao, Qingmiao Fan, Fan Yu, Fei Xu, Huiling Qin, Yuge Yuan, Wenzhou Ma, Duo Zhang, Chengcheng Peng, Nan Li, Pengpeng Xiao
## Abstract
Chikungunya virus (CHIKV) is an arthropod-borne virus that has caused several major outbreaks around the world and is becoming increasingly harmful. Although significant progress has been made in understanding the global epidemiology and transmission of CHIKV, a systematic description of the transmission history of its three genotypes is still lacking. To address this gap, this study integrates multiple bioinformatics approaches to explore their origin, evolution, and transmission dynamics. We analyzed publicly available CHIKV genomes from NCBI to elucidate the genetic evolution and transmission potential of these genotypes. Phylogeographic and molecular evolutionary analyses showed that the West African (WA) genotype originated in Nigeria and spread exclusively within Africa; the Eastern/Central/South African (ECSA) genotype originated in Tanzania and spread globally; and the Asian genotype originated in Thailand, spread throughout Asia, Oceania, and the Americas, exhibiting the highest evolutionary rate among the three genotypes. We also identified 15 positively selected sites and 10 nonconservative mutation sites with altered hydrophobicity across CHIKV proteins, all of which need further investigation into their effects on viral protein function. The data from this study are important for understanding the transmission history of the three genotypes of CHIKV, providing new targets for CHIKV antiviral therapy and ideas for developing effective prevention and control measures in the future.
## 1. Introduction
Chikungunya virus (CHIKV) is a positive-stranded RNA virus of Alphavirus transmitted by Aedes with a global distribution. The genomic structure of CHIKV contains two open reading frames (ORFs) encoding nonstructural proteins (NSP1, NSP2, NSP3, and NSP4) and structural proteins (capsid protein [C], E3, E2, 6K, and E1) [1]. Although CHIKV infection may be asymptomatic, symptomatic patients may develop chikungunya fever (CHIKF), which is characterized by sudden onset of high fever (39-40°C), maculopapular rash, and persistent joint pain with hyperviremia and antigenemia. The course of the disease usually lasts 1-3 weeks, and some patients may develop chronic arthritis that lasts for several months and up to several years [2,3]. Since its emergence in Africa, CHIKV has spread rapidly, causing more than 10 million cases in more than 125 countries or territories over the past two decades and posing a global public health threat [4].
CHIKV originated in Africa and was first isolated in Tanzania in 1953 [5,6]. Phylogenetic analysis identified three major lineages of CHIKV: West African (WA), Eastern/Central/South African (ECSA), and Asian genotypes [7]. Since its discovery, CHIKV has been spreading and causing sporadic outbreaks in sub-Saharan Africa [8]. After emerging in Kenya in 2004, the ECSA strain of CHIKV spread to the Indian Ocean islands, triggering an unprecedented outbreak, notably on Réunion Island. It subsequently reached Asia and India, and eventually caused indigenous transmission in European countries, such as Italy and France [9][10][11][12]. Another major outbreak occurred in December 2013, when the Asian genotype of CHIKV first emerged on the Caribbean island of Saint Martin and spread to 22 countries across the Caribbean, Central, and South America within just 9 months, resulting in hundreds of thousands of cases [13]. In 2014, the ECSA was reported in northeastern Brazil, where the strain is still spreading today as the most prevalent strain [14].
CHIKV is transmitted mainly through the bites of infected mosquitoes. Depending on the region, transmission occurs via the sylvatic cycle and the urban cycle. In Africa, CHIKV is mainly transmitted in the sylvatic cycle, involving wild nonhuman primates and forest-dwelling Aedes [15]. Occasionally, the virus jumps to populations living near forested areas. Similar zoonotic transmission has also been reported in Asia [16]. The urban cycle is the transmission of CHIKV between humans and mosquitoes in cities, with Aedes aegypti (Ae. aegypti) and Aedes albopictus (Ae. albopictus) being the main vectors in this cycle. The urban cycle of CHIKV has been associated with several large CHIKV epidemics on different continents, including Asia, Europe, and North America [17]. The distributions of Ae. aegypti and Ae. albopictus are projected to expand due to environmental adaptation and increasing human mobility, raising the likelihood of CHIKV spread to new regions and underscoring its growing threat to global health [18,19].
High vector competence and widespread human susceptibility have facilitated the continued expansion and adaptive evolution of CHIKV. Evidences from the 2014 outbreak in Latin America suggests that CHIKV may cause a greater burden than any other arthropod-borne virus [20,21]. Advances in DNA sequencing and bioinformatics have enabled extensive viral genome surveillance, improving our understanding of the genetic diversity and transmission dynamics of emerging viruses [22]. In this study, we explored the epidemiology and evolution of CHIKV during its emergence using global whole genome sequences to further elucidate the genetic evolution and transmission potential of CHIKV.
## 2. Methods
2.1. Whole Genome Sequence Searching and Typing. As of November 13, 2023, CHIKV whole-genome sequences longer than 11 kb were downloaded from NCBI (https://www.ncbi. nlm.nih.gov/). Collection country, year, and host information were recorded (Supporting Information 1: Table S1). Sequences were uploaded to the CHIKV typing website (https://www.genomedetective.com/app/typingtool/ chikungunya/) to obtain genotype information for all the sequences. Country distribution and time-series maps were visualized using the R package.
## 2.2. Host Composition and Vector Distribution.
As of April 5, 2024, the worldwide distribution of Ae. albopictus and Ae. aegypti from 1953 to 2024 was obtained using GBIF (https:// www.gbif.org/). Additional host information from NCBI sequences included Homo sapiens (n = 802), Ae. aegypti (n = 16), Ae. albopictus (n = 11), Aedes furcifer (n = 4), Macaca fascicularis (n = 4), Aedes luteocephalus (n = 2), Culex quinquefasciatus (n = 2), Aedes dalzieli (n = 1), Aedes africanus (n = 1), Aedes opok (n = 1), Anopheles (Ceilia) funestus (n = 1), sentinel mouse (n = 1), and Chiroptera (n = 1). Host composition was visualized using the R language.
2.3. Phylogenetic Analysis. A total of 929 whole genome sequences were aligned using MAFFT [23] and the aligned sequences were evaluated and edited in Aliview [24]. The evolutionary distances were computed using the optimal GTR + F+ R8 model, the phylogenetic tree was constructed with the maximum likelihood (ML) method. Bootstrap values are given for 1000 replicates. Phylogenetic trees were generated using IQTREE v2.3.5 [25] and visualized and annotated at iTOL [26] (https://itol.embl.de/). Outer rings were added to display genotype and host information, and branches were colored by country.
## 2.4. System Dynamics
Reconstruction. The E1 gene of CHIKV is the basis for the classification of CHIKVs into three genotypes: Asian, WA, and ECSA, and its sequences are relatively conserved [7]. In addition, a single mutation in the E1 gene has previously caused a large-scale outbreak of CHIKV [27], and E1 sequences have been used for the phylogenetic analyses of CHIKV [28]. We selected the whole-genome sequences of the ECSA and Asian genotypes with available country and time information (~30% of each genotype), and all available WA sequences with such information (ECSA = 138, Asian = 119, and WA = 12; Supporting Information 2: Table S2). After alignment with MAFFT, E1 sequences were obtained for subsequent analysis using Aliview. To avoid the effects of recombination before assessing temporal signals and inferring time-scale phylogenies, recombination analyses were performed on the E1 of CHIKV to investigate potential recombination events using the Phi test in RDP4 [29], using RDP, GENECONV, Bootscan, Maxchi, Chimera, Si Sscan, Phylpro, LARD, and 3Seq to identify recombination events. Temporal signal detection was performed on nonrecombinant sequences to test the relationship between genetic diversity and sampling date, using TempEst [30] to regress the root tip genetic distances on the ML tree against the exact sampling date and to remove outlier sequences based on the estimated root tip distances. To obtain more robust rate estimates, the Bayesian Monte Carlo Markov chain (MCMC) method implemented in the BEAST software package v1.10.4 [31] was used to first select the optimal nucleotide model for the sequences based on the Bayesian information criterion (BIC) using the ModelFinder plug-in in the PhyloSuite [32], and later to run the file using the Beauti configuration. The uncorrected lognormal relaxation molecular clock model and the Bayesian Skyline merger tree prior were selected. Multiple runs of MCMC were merged using LogCombiner v1.10.4 with 200 million generations per group and sampling every 20,000 generations. Convergence of the relevant parameters was assessed using Tracer v1.6 [33] (effective sample sizes [ESSs] for relevant model parameters >200, few ESSs > 100) and skyline plots of the three genotypes were obtained. The posterior distributions of the trees obtained from the BEAST analysis were analyzed using TreeAnnotator v1.10.4 to obtain maximum spectral confidence (MCC) trees after having a 10% burnin. The MCC trees were visualized using FigTree v1.4.4 (http://tree.bio.ed.ac. uk/software/figtree/). 2.5. Spatial and Temporal Dynamic Analysis. To track the spatial and temporal dynamics of the three major CHIKV genotypes globally, a phylogeographic analysis was performed using SPREAD3 v0.9.7.1 [34] by applying an asymmetric substitution model of positional transformations and estimating positional diffusivity using the Bayesian stochastic search for variable selection (BSSVS) model, with the Bayes factor (BF) test selected to identify significant links between geographic regions. Circos plots and maps were used to visualize the probable routes. Circos plots were created using the Cnsknowall website (https://www.cnsknowall.com) and the maps were drawn by R packages.
2.6. Amino Acid Variation Analysis. To detect selection on all CHIKV proteins, coding region sequences were analyzed for recombination, and a total of 815 sequences were involved in the positive selection site analysis after removing recombinant sequences and sequences containing parsimonious and ambiguous bases. A ML tree based on available sequences was reconstructed using Datamonkey [35] (http://www. data monkey.org/). Methods used to examine positive amino acid sites included single-likelihood-ratio ancestor counting (SLAC), fixed-effects likelihood ratio (FEL), evolutionary mixed-effects modeling (MEME), and fast unconstrained Bayesian inferential selection (FUBAR). The significance levels for SLAC, FEL, and MEME were set at a p-value threshold of 0.1, and the significance level for FUBAR was set at a p-value threshold of 0.9, we considered a site detected by more than two algorithms as positively selected, and the amino acid variability of the detected positively selected sites was visualized using SEQLOGO plots from the OmicShare (https://www.omicshare.com/tools). The Shannon entropy online analysis tool (http://www.hiv.lanl.gov/content/seque nce/ENTROPY/entropy_one.html) was used to search for mutational hotspots in CHIKV, selecting nonconservative mutation sites with changes in the nature of amino acids among those with a Shannon entropy of more than 0.7, showing that mutations occurring in ≥20% mutations in the sequence. Amino acid sites with changes in amino acid hydrophobicity were selected among this sites and bar graphs were drawn using the R package. Homology modeling of consensus sequences for each protein was performed using the I-TASSER website [36] (https://zhanggroup.org/I-TASSE R/), and protein models were visualized using UCSF ChimeraX [37].
## 3. Results
3.1. Spatial and Temporal Distribution of CHIKV. We obtained 929 whole-genome sequences (>11 kb) from NCBI, of which 892 had collection dates: 12 WA, 485 ECSA, and 395 Asian genotypes (Supporting Information 1: Table S1). Plotting according to the collection time, it can be seen from Supporting Information 3: Figure S1 that after about 2005, the number of CHIKV sequences began to grow gradually, and there was a big outbreak of Asian genotypes in 2014, and two peaks of outbreaks of ECSA genotypes. In addition, a total of sequences with national information were counted, including 14 WA, 500 ECSA, and 398 Asian genotypes. The wholegenome sequencing data of CHIKV now covers over countries and regions worldwide, specifically, including countries and regions in North America and the Caribbean, 18 countries in Africa, 15 countries in Asia, 10 countries in Oceania, 7 countries in South America, and 4 countries in Europe. Geographically, the sequences of Asian genotype are primarily found in the Americas, Asia, and Oceania, while the sequences of the WA genotype are mainly distributed in Africa. The sequences of the ECSA genotype are the most widespread and has become the dominant lineage circulating globally (Figure 1).
## 3.2. Host Composition and Transmission of CHIKV.
According to the statistics of the data downloaded from NCBI, it was found that about 94.7% of the hosts of CHIKV were humans (Figure 2A). Among the non-human hosts, mosquitoes accounted for a larger and more diverse proportion, with a total of nine species of mosquitoes, including seven species of Aedes, one Anopheles, and one Culex (Figure 2B). The host distribution of the three genotypes can be seen in the host exosphere of the evolutionary tree (Figure 1A), where the WA genotype has six hosts, including humans, mosquitoes, bats, and rats, which corresponds to the fact that the WA genotype is found mainly in the sylvatic cycle. The ECSA and the Asian genotypes are found mainly in the urban cycle, and the transmission is mainly sustained through the mosquito-human link. From the distribution of the two main vectors of CHIKV, Ae. albopictus, and Ae. aegypti, the two mosquito species are widely distributed in the world (Figures 1A and2C, D), with a lesser distribution in the African region, which also confirms that the transmission of CHIKV in Africa mainly relies on the forest cycle (Figure 2E). The wide distribution of the main vectors also implies that CHIKV has the potential to spread worldwide.
## 3.3. Evolutionary Dynamics of Three
Genotypes. E1 sequences intercepted from whole-genome sequences were analyzed phylogenetically on a time scale. The root-to-tip distance analysis revealed a positive correlation between genetic distance and sampling time in the selected datasets of the WA (R 2 = 0.7192, correlation coefficient = 0.8481), ECSA (R 2 = 0.6772, correlation coefficient = 0.8229), and Asian (R 2 = 0.9099, correlation coefficient = 0.9539) genotypes. These findings accord with the molecular clock theory, thereby validating the dataset's appropriateness for phylogenetic molecular clock analysis (Supporting Information 4: Figure S2).
From the MCC trees obtained, it can be seen that the WA genotype originating strain was isolated from Nigeria with the mean time of the most recent ancestor (TMRCA) dating back to 1948 (95% HPD: 1820-1964), the ECSA genotype originating strain was isolated from Tanzania with the TMCRA dating Transboundary and Emerging Diseases back to 1933 (95% HPD: 1917-1946), and the Asian genotype originating strain was isolated from Thailand with the TMCRA dating back to November 1954 (95% HPD: October 1948-November 1957) (Figures 3 and4A). Estimating the evolutionary rate and 95% HPD of the three major lineages with a relaxed molecular clock model, the evolutionary rate of the WA genotype was 2.23 × 10 -4 (95% HPD:
1.09 × 10 -7 -4.11 × 10 -4 ) substitutions/site/year, and the evolutionary rate of the ECSA genotype was 3.31 × 10 -4 (95% HPD: 2.67 × 10 -4 -4.01 × 10 -4 ) substitutions/site/year, and 4.09 × 10 -4 (95% HPD: 3.01 × 10 -4 -5.16 × 10 -4 ) substitutions/site/ year for the Asian genotype, which could be seen that the Asian genotype had a higher evolutionary rate (Figure 4B). Using the Skyline model to estimate the effective population size by time and 95% HPD intervals, it can be seen that the population size of the ECSA genotype peaked around 1960 and then declined slightly around 1997 until it reached a second peak around 2010, which corresponds to the history of the ECSA genotype outbreaks. The Asian genotype peaked around 2014, while the WA genotype showed minimal fluctuation with wide confidence intervals (Figure 4C). of Asian genotype in Asia and Oceania, but there is no transmission route to the Americas. It is speculated that the results may be inaccurate because the datasets involved in the analysis are not complete (Figure 5B,C). The combination of the MCC tree and the circos plot shows that the ECSA genotype spread around the world, from Tanzania to African countries, spread widely among African countries, and then spreads to Asia, Europe, the Americas, and Oceania, with localized circulations between continents as well. The most probable transmission routes based on BF >3.0 and PP >0.5
show that the Democratic Republic of the Congo is an important hub for the spread of the CHIKV ECSA genotype from Africa, and Italy is an important hub for the spread of the genotype from Europe (Figure 5D,E).
3.5. Amino Acid Variation. We performed selection pressure and mutation analysis of CHIKV. Detection of positively selected sites in individual proteins identified 15 codons subject to positive selection, with four sites subject to positive selection detected in the E1 protein, three in the NSP1 protein, two in the NSP3 protein, and one in each of the NSP2 protein, NSP4 protein, C protein, E3 protein, E2 protein, and 6K protein (Figure 6A and Table 1). A statistical analysis of the amino acid distribution of each positive selection site revealed that sites 253 in NSP1, 218 in NSP2, 457 in NSP3, 81 in C, 47 in 6K, and 211 in E1 showed variability across genotypes (Figure 6B and Supporting Information 5: Table S3). Shannon entropy analysis of individual proteins revealed variable sites in the sequences, and entropy values above the threshold of 0.7 indicated that these sites were highly variable (Supporting Information 6: Table S4). The amino acid hydrophobicity changed at the highly mutated sites of each protein were amino acids 128, 253, 488, and 507 of NSP1, 91 of NSP4, 33 of E3, 5, 205, and 318 of E2, and 211 of E1 (Figure 7). In the positive selection site analysis and Shannon entropy analysis of individual proteins, it was found that both the number of positive selection sites and the number of high mutation sites that undergo nonconservative mutations were higher in the NSP1 protein, the NSP3 protein had the highest number of high mutation sites that undergo nonconservative mutations, and the E1 protein had the highest number of positive selection sites. Among them, 253 in NSP1, 457 in NSP3, 33 in E3, 47 in 6K, and 211 in E1, which are all positive selection sites and high mutation sites that undergo nonconservative mutations.
## 4. Discussion
4.1. Transmission History of the Three Genotypes. Previous studies have considered the global evolution and spread of global CHIKV sequences across the globe [28]. In our study, the three genotypes of CHIKV were studied separately using phylogenetic, Bayesian, and phylogeographic analyses to better clarify the evolutionary and dispersal history of the three lineages. The evolutionary dynamics and spread of the ECSA genotype confirmed that the ECSA genotype is the dominant genotype, capable of spreading widely around the world, and that there is a tendency for the ECSA genotype to spread to non-endemic areas as human activities increase across geographical boundaries and as climate change occurs. However, the severe scarcity of genomic data for the WA genotype hinders a comprehensive elucidation of its transmission dynamics from the available sequences. Consequently, critical gaps persist, impeding both our understanding of and response to recurrent outbreaks. From the comparison of the evolutionary characteristics of the three major genotypes, it is evident that the Asian genotype has a high evolutionary rate and has shown an alarming range and speed of spread since it spread to the Americas, which confirms the pandemic potential of this genotype. The results of the study were also allowed to verify that the large-scale outbreak experienced in French Polynesia from October 2014 to January 2015 may have been introduced to French Polynesia from the Americas and not from other Pacific countries [38]. Our study identified hub countries for outward transmission of the ECSA genotype and the possibility that transmission of the Asian genotype to non-endemic areas of the world may be associated with travel cases and their indigenous transmission, which provides new ideas for putting in place controls for the re-emergence of CHIKV outbreaks.
## 4.2. Amino Acid Variation in Nonstructural Proteins.
Given that a single substitution causes significant differences in the specificity and pathogenicity of the alphavirus vectors, we performed a selection pressure versus mutation analysis of CHIKV. CHIKV encodes four nonstructural proteins (NSP1, NSP2, NSP3, and NSP4) that are involved in a range of processes, including genome replication and immune regulation. Among the four viral nonstructural proteins, NSP1 is the only membrane-associated protein responsible for the localization of the viral replication complex and anchoring to the replication site on the membrane [39]. an important effect on the binding capacity of NSP1 [40]. Amino acid 211 in the positively selected state is located in one of the membrane-attached (MA) loops of NSP1, and the NSP1 protein is clustered by two intertwined MA loops, where positively charged residues form a cluster around the membrane insertion site and establish electrostatic interactions with phosphatidylserine lipids enriched in the inner leaflet of the plasma membrane, which are essential for RNA replication (Blue sites represent positive selection sites, orange represents amino acids at the highly mutated site in the consensus sequence, red represents the site that is both highly mutated and positively selected, and green represents the mutated amino acids). [41]. Amino acids 128 and 171 in NSP1 have not yet been studied to confirm their functional impact, in addition, the entire Cterminal tail after amino acid 474 in NSP1 is disordered, and we found three highly mutated amino acid sites in this range, whose functional impact remains to be investigated. The NSP2 protein is a key protein for viral proliferation and possesses both RNA triphosphatase and helicase activities in its N-terminal domain [42]. The positive selection site 218 found in NSP2 is located in the N-terminal deconjugate enzyme structural domain of the NSP2 protein, where the amino acids of the ECSA genotype are different from those of the other two genotypes, and the function of this site may be related to deconjugate enzyme function [43].
The NSP3 protein has been recognized as essential for viral replication and adaptation to its host [44], and both the positive selection site and the hypermutation site found in NSP3 are located in the C-terminal hypervariable structural domain (HVD), which is disordered in NSP3 and serves as a platform for interactions with a wide range of host proteins. The linear motifs contained in the HVD of CHIKV NSP3 have been shown to recruit defined host protein families into the formation of functional viral replication complexes [45]. There is a direct link between CHIKV RNA synthesis and NSP3 phosphorylation, and elimination of phosphorylation sites in HVD inhibits CHIKV replicase activity [46]. Whereas the variants at the highly mutated loci identified in this study were all associated with threonine, the functional impact of these loci needs to be further verified.
The NSP4 is an RNA-dependent RNA polymerase (RdRp) that is a core subunit of the viral replication complex [47]. Amino acid 473 in NSP4 is a positive selection site found in NSP4, and positions 91, 107, and 506 are highly mutated. The N-terminal region of NSP4 (100 residues long) forms a partially unstructured structural domain, which is necessary for the proper functioning of NSP4, and can be targeted to block its polymerase activity, thus inhibiting viral replication in the host cell [48]. The C-terminal RdRp structural domain also plays a key role in catalyzing genomic RNA replication and transcription [49]. These sites identified in this study can be further investigated for these functions.
4.3. Amino Acid Variation in Structural Proteins. CHIKV contains five structural proteins, C, E3, E2, E1, and 6K proteins. The main function of the C protein is to form nuclear capsids capable of self-cleaving from structural polyproteins prior to genomic RNA binding [50]. Our study identified amino acid 81 as a positive selection site in the C protein, which is located in the N-terminal structural domain of C and is involved in genomic RNA encapsulation. Although the amino acid 81 is not a highly mutated site in the C protein, we found that the WA genotype differs from the other two genotypes at this site. It has been found that nucleolus localization sequences located between 60 and 99, which are involved in nuclear import, were identified in the N-terminal region of the C protein, mutations in the C protein NoLS have been shown to attenuate replication in mammalian and mosquito cells [51], but the functional impact of amino acid differences at this site remains to be verified.
The E1 and E2 glycoproteins are primarily responsible for membrane fusion and viral entry into host cells, where E2 interacts with and attaches to cellular receptors, and E1 is involved in fusion of viral and cellular cell membranes. The E1 protein contains three distinct β-structural domains, structural domains I, II and III (DI, DII, and DIII), and DII has a highly conserved hydrophobic fusion loop, which, when the pH of the endosome is lowered, it is exposed to the surrounding environment, triggering the E1 rearrangement and leading to the insertion of the hydrophobic fusion loop into the cell membrane. Our study found that amino acids 4, 211, 291, and 321 in E1 were under positive selection, with the 211 also being a high mutation site. The E1-K211E mutation has been shown to enhance the fitness of Ae. aegypti [52]. Amino acid 211 of E1 is the most variable among the ECSA strains, and in addition to the K211E mutation, it has been found that K211T plays a role in viral attachment to cells [53]. Amino acids 145 and 304 are also mutation sites with high entropy values, and no studies have yet demonstrated the functional impact of mutations at these sites. Among them, amino acid 145 is located in the DII domain, and it is worthwhile to investigate the effect of mutation at this position on the function of E1 for cell membrane fusion. The E2 protein also consists of three distinct structural domains, structural domains A, B, and C. The receptor binding site is located in structural domain A, whereas structural domain B is located in the outermost layer of each spicule and shields the fusion loop on structural domain II of E1, and structural domain C is located closest to the viral membrane [54]. In the E2 protein we found three highly mutated sites, sites 5, 205, and 318 and one site under positive selection, site 246. It can be seen from the 3D structure of the protein that 205 is located in structural domain B and 318 is located in structural domain C. The functional implications need to be further explored.
Compared with other proteins, the E3 and 6K proteins have been less studied. The E3 protein plays a central role in the regulation of the E2 protein folding and binding to the E2-E1 dimer [55]. In the E3 protein we found that amino acid 33 is both positively selected and highly mutated, and the amino acids at this site are different in all three genotypes. The 6K proteins are highly hydrophobic, cysteine-rich acylated proteins with a wide variety of functions, ranging from participation in envelope protein processing to membrane permeabilization, viral outgrowth, and viral assembly [56]. In the 6K protein we identified a positively selected site 47 that differed in the Asian genotype from the other two genotypes. These need to be further investigated to determine whether variation at this site affects the CHIKV phenotype. 4.4. Application Prospect of Amino Acid Variation Sites. The global spread of the three major genotypes of CHIKV confirms that the virus is expanding at a remarkable rate. Millions of people in tropical and subtropical regions are already affected or at high risk of infection, and countries with moderate climates may experience severe outbreaks due to the presence of mosquito vectors throughout the year. The global upsurge in CHIKV outbreaks and transmission is closely linked to the emergence of viral mutations that allow for increased viral adaptations to existing and novel vectors, the leading to increased transmission. Structural proteins encapsulate viral nucleic acids and coordinate viral particle assembly, and nonstructural proteins play key roles in viral replication, translation, and mediating host immune escape. Thus, both structural and nonstructural proteins are promising targets for the development of anti-CHIKV antiviral drugs. The first Chikungunya vaccine, Ixchiq, is now available, but it was developed based on the ECSA genotype and there are no specific antiviral drugs for CHIKV infection, which highlights the importance of monitoring the CHIKV genome [57]. We analyzed the selection pressure analysis and mutational hotspots of individual CHIKV proteins with the aim of providing new perspectives for the development of antiviral drugs for the treatment of CHIKV. The relevant sites obtained from the selection pressure analysis and mutation hotspot analysis of CHIKV proteins need to be further investigated by computational methods such as molecular dynamics and experimental methods, such as reverse genetics in order to clarify whether these sites can be targets for the control of CHIKV. 4.5. Limitations. Limitations of our study, where we based our statistical analyses only on CHIKV genome-wide data currently publicly available in NCBI, suggest that our analyses may have missed transmission events involving unsampled countries or countries that did not upload sequence information. For example, only 12 data containing country and temporal information were involved in the analysis for the WA genotype, and the lack of sampling resulted in the phylogenetic analysis of the WA genotype not clearly reflecting the phylogenetic results of the WA genotype, and also in the results obtained from phylogeographic analyses, the Nigerian and Côte d'Ivoire's were geographically closer to each other, but not predicting the paths of transmission between the two. In our phylogeographic analysis, we were unable to obtain specific sampling points for all sequences based on the information available, and then used the coordinates of each country's capital as the sampling location for the sequences. In addition, due to the limitation of computational resources, only part of the E1 gene sequences of CHIKV were sampled for the analysis of evolutionary dynamics and phylogeography in this study, and although we have tried our best to ensure that all the information of countries and time is included in the process of data selection, the final results are still deficient. Genomics studies can be combined with epidemiological data and vector distribution data to better track the spread and evolution of CHIKV, our study was limited by computational resources and limited publicly available information that prevented the results of our phylogeographic analyses from adequately demonstrating the spread of CHIKV, for example, our results lack the current prevalence of CHIKV in Brazil [58][59][60], which requires further data collection and analysis.
## 5. Conclusion
Our study, despite its limitations, was able to provide new insights into the origin, evolutionary dynamics, and global transmission pathways of CHIKV. The impact of CHIKF on human health has increased dramatically over the past two decades. Although its mortality rate is low, it has a high incidence of long-term disability, which constitutes a significant health risk [61]. As CHIKV continues to evolve and expand its geographic distribution, communication between continents has become more rapid and frequent. As CHIKV genomewide data continues to be updated, we are better able to track the spread and evolution of CHIKV, and the potential discovery of new targeted control measures to help reduce its public health burden, through genomics studies. As the CHIKV genome-wide data continue to be updated, genomic studies will allow us to better track the spread and evolution of CHIKV and identify new targeted control measures. Combining genomic studies with epidemiological data (e.g., incidence rates, outbreak timelines, and demographic information) will allow us to more clearly define how genetic variation affects viral transmission patterns and public health, and thus identify new targeted control measures to help reduce its public health burden. In addition, we can explore new directions in drug and vaccine development based on positive selection or mutations during evolution.
## References
1. Frolov, Frolova (2022) "Molecular Virology of Chikungunya Virus" *Current Topics in Microbiology and Immunology*
2. Calvo, Archila, López et al. (2021) "Rediscovering the Chikungunya Virus" *Biomédica*
3. Silva, Dermody (2017) "Chikungunya Virus: Epidemiology, Replication, Disease Mechanisms, and Prospective Intervention Strategies" *Journal of Clinical Investigation*
4. De Souza, Fumagalli, De Lima (2024) "Pathophysiology of Chikungunya Virus Infection Associated With Fatal Outcomes" *Cell Host & Microbe*
5. Robinson (1955) "An Epidemic of Virus Disease in Southern Province, Tanganyika Territory, in 1952-53. I. Clinical features" *Transactions of the Royal Society of Tropical Medicine and Hygiene*
6. Robinson (1956) "An Epidemic of a Dengue-Like Fever in the Southern Province of Tanganyika" *The Central African Journal of Medicine*
7. Powers, Brault, Tesh et al. (2000) "Re-Emergence of Chikungunya and O'nyong-Nyong Viruses: Evidence for Distinct Geographical Lineages and Distant Evolutionary Relationships" *Journal of General Virology*
8. Wahid, Ali, Rafique et al. (2017) "Global Expansion of Chikungunya Virus: Mapping the 64-Year History" *International Journal of Infectious Diseases*
9. Sergon, Njuguna, Kalani (2004) "Seroprevalence of Chikungunya Virus (CHIKV) Infection on Lamu Island"
10. Salje, Cauchemez, Alera (2016) "Reconstruction of 60 Years of Chikungunya Epidemiology in the Philippines Demonstrates Episodic and Focal Transmission" *Journal of Infectious Diseases*
11. Wimalasiri-Yapa, Stassen, Huang (2019) "Chikungunya Virus in Asia -Pacific: A Systematic Review" *Emerging Microbes & Infections*
12. Hakim, Annisa, Gazali et al. (2022) "The Origin and Continuing Adaptive Evolution of Chikungunya Virus" *Archives of Virology*
13. Morrison (2014) "Reemergence of Chikungunya Virus" *Journal of Virology*
14. De Souza, De Lima, Simões Mello (2023) "Spatiotemporal Dynamics and Recurrence of Chikungunya Virus in Brazil: An Epidemiological Study" *The Lancet Microbe*
15. Tsetsarkin, Chen, Weaver (2016) "Interspecies Transmission and Chikungunya Virus Emergence" *Current Opinion in Virology*
16. Althouse, Guerbois, Cummings (2018) "Role of Monkeys in the Sylvatic Cycle of Chikungunya Virus in Senegal" *Nature Communications*
17. Weaver, Lecuit (2015) "Chikungunya Virus and the Global Spread of a Mosquito-Borne Disease" *New England Journal of Medicine*
18. Girard, Nelson, Picot et al. (2020) "Arboviruses: A Global Public Health Threat" *Vaccine*
19. Bellone, Lechat, Mousson (2023) "Climate Change and Vector-Borne Diseases: A Multi-Omics Approach of Temperature-Induced Changes in the Mosquito" *Journal of Travel Medicine*
20. Cardona-Ospina, Diaz-Quijano, Rodríguez-Morales (2015) "Burden of Chikungunya in Latin American Countries: Estimates of Disability-Adjusted Life-Years (DALY) Lost in the 2014 Epidemic" *International Journal of Infectious Diseases*
21. Montalvo Zurbia-Flores, Reyes-Sandoval, Kim (2023) "Chikungunya Virus: Priority Pathogen or Passing Trend?" *Vaccines*
22. Robishaw, Alter, Solano (2021) "Genomic Surveillance to Combat COVID-19: Challenges and Opportunities" *The Lancet Microbe*
23. Katoh, Standley (2013) "MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability" *Molecular Biology and Evolution*
24. Larsson (2014) "AliView: A Fast and Lightweight Alignment Viewer and Editor for Large Datasets" *Bioinformatics*
25. Minh, Schmidt, Chernomor (2020) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Molecular Biology and Evolution*
26. Letunic, Bork (2024) "Interactive Tree of Life (iTOL) v6: Recent Updates to the Phylogenetic Tree Display and Annotation Tool" *Nucleic Acids Research*
27. Zeller, Van, Bortel et al. (2016) "Chikungunya: Its History in Africa and Asia and Its Spread to New Regions in 2013-2014" *Journal of Infectious Diseases*
28. Deeba, Haider, Ahmed (2020) "Global Transmission and Evolutionary Dynamics of the Chikungunya Virus" *Epidemiology and Infection*
29. Martin, Murrell, Golden et al. (2015) "RDP4: Detection and Analysis of Recombination Patterns in Virus Genomes" *Virus Evolution*
30. Rambaut, Lam, Carvalho et al. (2016) "Exploring the Temporal Structure of Heterochronous Sequences Using TempEst (Formerly Path-O-Gen)" *Virus Evolution*
31. Drummond, Rambaut (2007) "BEAST: Bayesian Evolutionary Analysis by Sampling Trees" *BMC Evolutionary Biology*
32. Zhang, Gao, Jakovlić (2020) "PhyloSuite: An Integrated and Scalable Desktop Platform for Streamlined Molecular Sequence Data Management and Evolutionary Phylogenetics Studies" *Molecular Ecology Resources*
33. Rambaut, Drummond, Xie et al. (2018) "Posterior Summarization in Bayesian Phylogenetics Using Tracer 1.7" *Systematic Biology*
34. Bielejec, Baele, Vrancken et al. (2016) "SpreaD3: Interactive Visualization of Spatiotemporal History and Trait Evolutionary Processes" *Molecular Biology and Evolution*
35. Weaver, Shank, Spielman et al. (2018) "Datamonkey 2.0: A Modern Web Application for Characterizing Selective and Other Evolutionary Processes" *Molecular Biology and Evolution*
36. Yang, Zhang (2015) "I-TASSER Server: New Development for Protein Structure and Function Predictions" *Nucleic Acids Research*
37. Pettersen, Goddard, Huang (2021) "UCSF ChimeraX: Structure Visualization for Researchers, Educators, and Developers"
38. Aubry, Cao-Lormeau (2019) "History of Arthropod-Borne Virus Infections in French Polynesia" *New Microbes and New Infections*
39. Mounce, Cesaro, Vlajnić (2017) "Chikungunya Virus Overcomes Polyamine Depletion by Mutation of nsP1 and the Opal Stop Codon To Confer Enhanced Replication and Fitness" *Journal of Virology*
40. Gottipati, Woodson, Choi (2020) "Membrane Binding and Rearrangement by Chikungunya Virus Capping Enzyme nsP1" *Virology*
41. Zhang, Law, Law et al. (2021) "Structural Insights Into Viral RNA Capping and Plasma Membrane Targeting by Chikungunya Virus Nonstructural Protein 1" *Cell Host & Microbe*
42. Mastalipour, Gering, Coronado et al. (2025) "Novel Peptide Inhibitor for the Chikungunya Virus nsP2 Protease: Identification and Characterization" *Current Research in Microbial Sciences*
43. Law, Utt, Tan (2019) "Structural Insights Into RNA Recognition by the Chikungunya Virus nsP2 Helicase" *Proceedings of the National Academy of Sciences*
44. Dominguez, Shiliaev, Lukash (2021) "NAP1L1 and NAP1L4 Binding to Hypervariable Domain of Chikungunya Virus nsP3 Protein Is Bivalent and Requires Phosphorylation" *Journal of Virology*
45. Lukash, Agback, Dominguez "Structural and Functional Characterization of Host FHL1 Protein Interaction With Hypervariable Domain of Chikungunya Virus nsP3"
46. (2020) *Journal of Virology*
47. Teppor, Žusinaite, Merits (2021) "Phosphorylation Sites in the Hypervariable Domain in Chikungunya Virus nsP3 Are Crucial for Viral Replication" *Journal of Virology*
48. Chen, Tan, Zheng (2017) "Chikungunya Virus nsP4 RNA-Dependent RNA Polymerase Core Domain Displays Detergent-Sensitive Primer Extension and Terminal Adenylyltransferase Activities" *Antiviral Research*
49. Singh, Kumar, Yadav et al. (2018) "Deciphering the Dark Proteome of Chikungunya Virus" *Scientific Reports*
50. Tan, Lello, Liu (2022) "Crystal Structures of Alphavirus Nonstructural Protein 4 (nsP4) Reveal an Intrinsically Dynamic RNA-Dependent RNA Polymerase Fold" *Nucleic Acids Research*
51. Jacobs, Taylor, Herrero et al. (2017) "Mutation of a Conserved Nuclear Export Sequence in Chikungunya Virus Capsid Protein Disrupts Host Cell Nuclear Import" *Viruses*
52. Thomas, Rai, John et al. (2013) "Chikungunya Virus Capsid Protein Contains Nuclear Import and Export Signals" *Virology Journal*
53. Agarwal, Sharma, Sukumaran et al. (2016) "Structural Proteins of Chikungunya Virus Enhance Fitness in Aedes aegypti"
54. Rangel, Mcallister, Dancel-Manning et al. (2022) "Emerging Chikungunya Virus Variants at the E1-E1 Interglycoprotein Spike Interface Impact Virus Attachment and Inflammation" *Journal of Virology*
55. Chen, Phuektes, Yeo (2024) "Attenuation of Neurovirulence of Chikungunya Virus by a Single Amino Acid Mutation in Viral E2 Envelope Protein" *Journal of Biomedical Science*
56. Sjöberg, Lindqvist, Garoff (2011) "Activation of the Alphavirus Spike Protein is Suppressed by Bound E3" *Journal of Virology*
57. Ramsey, Mukhopadhyay (2017) "Disentangling the Frames, the State of Research on the Alphavirus 6K and TF Proteins" *Viruses*
58. Wang, Wang, Leng et al. (2024) "Drugs Targeting Structural and Nonstructural Proteins of the Chikungunya Virus: A Review" *International Journal of Biological Macromolecules*
59. Rodrigues, Souza, Fonseca (2020) "Genomic Surveillance of the Chikungunya Virus (CHIKV) in Northeast Brazil After the First Outbreak in 2014" *Revista DA Sociedade Brasileira De Medicina Tropical*
60. Machado, De Morais-Sobral, Campos (2019) "Genome Sequencing Reveals Coinfection by Multiple Chikungunya Virus Genotypes in a Recent Outbreak in Brazil" *PLoS Neglected Tropical Diseases*
61. Xavier, Alcantara, Fonseca (2023) "Increased Interregional Virus Exchange and Nucleotide Diversity Outline the Expansion of Chikungunya Virus in Brazil" *Nature Communications*
62. Mourad, Makhani, Chen (2022) "Chikungunya: An Emerging Public Health Concern" *Current Infectious Disease Reports* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12647605&blobtype=pdf | # Risk factors associated with cytomegalovirus reactivation and disease in critically-ill COVID-19 and non-COVID-19 patients, concomitantly admitted to intensive care
Maya Korem, Efrat Orenbuch-Harroch, Matan Cohen, Esther Oiknine-Djian, Ayala Livneh, Orit Caplan, Sigal Sviri, Peter Vernon Van Heerden, Dana Wolf
## Abstract
Critically-ill patients are at increased risk for cytomegalovirus (CMV) reactivation, associated with adverse clinical outcomes. Given the surge in intensive care unit (ICU) admissions during the COVID-19 pandemic and the continued burden of critical illness associated with the ongoing circulation of SARS-CoV-2, we sought to resolve risk factors for CMV reactivation and disease within the broader ICU patient population including those with and without COVID-19, to identify common and potentially distinct contributors to CMV reactivation and disease in this vulnerable setting. This prospective study included 208 adult ICU (85 COVID-19, and 123 concomitant non-COVID-19) patients, monitored weekly for CMV DNAemia. CMV reactivation was categorized as any detectable DNAemia or as clinically-significant reactivation characterized by high-level DNAemia (≥ 1000 IU/mL) and/or CMV disease. Overall, 29.8% of ICU patients experienced CMV reactivation, with 10.6% exhibiting clinicallysignificant reactivation. COVID-19 ICU patients had significantly higher rates of any CMV reactivation (40% vs. 23%, p = 0.009), high-level DNAemia (18% vs. 2%, p = 0.001), and CMV disease (12% vs. 1%, p = 0.001) compared to concomitant non-COVID-19 patients. Risk factors associated with clinicallysignificant CMV reactivation in ICU patients included septic shock, lower absolute lymphocyte count, high-dose steroid use, multiple blood transfusions, and COVID-19. CMV reactivation correlated with prolonged ventilation, hospitalization, and ICU stay, and increased in-hospital mortality. The high rates of clinically-significant CMV reactivation in both COVID-19 and non-COVID-19 ICU patients and the identified risk factors, along with the worse clinical outcomes linked to CMV reactivation, highlight the need for vigilant monitoring of CMV reactivation and for consideration of early antiviral treatment in ICU patients at risk, and support future interventional trials.
dysregulated inflammation, studied mainly in patients with bacterial sepsis, could favor CMV reactivation in this setting 1,3 . Moreover, CMV reactivation in ICU patients has been associated with adverse clinical outcomes 1,[3][4][5] .
Sepsis and longer hospitalization have been recognized as key risk factors for CMV reactivation in criticallyill patients (who are immunocompetent at baseline), while additional specific risk groups for CMV reactivation have not been consistently identified, and have been mostly described in the 'pre-COVID-19' era [1][2][3] .
The COVID-19 pandemic dramatically increased the burden of critical illness. To date, the continued circulation of SARS-CoV-2 following the decline of the pandemic, has remained a significant cause of hospitalizations and ICU admissions 6 . Severe COVID-19 is often characterized by profound immunological perturbations 7 , which, together with the use of immunosuppressive therapies, could underlie the frequently reported CMV reactivation in critically-ill COVID-19 patients [8][9][10] .
The surge in ICU admissions during the COVID-19 pandemic presented us with an opportunity to resolve risk factors associated with CMV reactivation and disease in the expanded ICU patient population and to further directly compare the rates of CMV reactivation and disease in COVID-19 versus concomitant non-COVID-19 ICU patients.
## Methods
## Study design and patient population
This prospective observational study included critically-ill adults (≥ 18 years) who were admitted to the two ICU's at Hadassah-Hebrew University Medical Center in Jerusalem, Ein-Kerem Campus (a 900-bed tertiary hospital) from September 1, 2020, to December 31, 2021. Included were COVID-19-positive and COVID-19negative (both PCR confirmed; 11) ICU patients who were screened at least once a week for CMV DNAemia as part of their routine care.
The demographic, clinical, and laboratory data of patients admitted to ICU were collected from the medical files (see Table 1 for analyzed parameters and definitions).
All methods were performed in accordance with the relevant guidelines and regulations. The study was approved by the committee on research involving human subjects of the Hebrew University-Hadassah Medical School (# 0984 -20-HMO) and informed consent was waived off.
## CMV reactivation and disease
CMV DNAemia was detected and quantified as described 12 with some modifications. DNA was extracted from whole blood (MagnaPure, Roche). RT-PCR was performed using CMV glycoprotein B/IE1 primers/probes, and a housekeeping gene (ERV3) with TaqMan Fast Advanced Master Mix (Applied Biosystems) on the QuantStudio 5 Real-Time PCR instrument (Applied Biosystems).
CMV reactivation was categorized as any detectable DNAemia, or high-level DNAemia (≥ 1000 IU/mL). This threshold reflected our clinical practice (above which we consider antiviral treatment). CMV disease was diagnosed using standard criteria 13 .
## Statistics
Statistical analyses were performed using SPSS version 30. Group comparisons were conducted using chi-square test for categorical variables and t-test/Mann-Whitney test for quantitative continuous variables. Time-dependent Cox-regression model was employed to assess the impact of CMV reactivation on mortality. Receiver operating characteristic (ROC) analysis was used to evaluate total lymphocyte count (TLC) thresholds as predictors of CMV reactivation. All statistical tests were two-tailed; A p-value ≤ 0.05 was considered statistically significant.
## Results
During the study period, 85 and 123 concomitant ICU patients with and without COVID-19, respectively, were regularly screened for CMV DNAemia while admitted in ICU. The demographic, clinical, and CMV reactivation parameters, as compared between COVID-19 and non-COVID-19 ICU patients, are presented in Table 1. The groups were comparable in background comorbidity index and were equally screened for CMV during the ICU stay. Despite lower SOFA scores, COVID-19 patients experienced higher rates of septic shock, bacteremia, candidemia, pneumonia, and acute respiratory distress syndrome (ARDS), compared to the non-COVID-19 group, with higher requirements for invasive ventilation, inotropic support, and high-dose systemic steroid treatment (received over longer duration).
The outcome of COVID-19 ICU patients was worse in terms of in-hospital mortality, length of ICU stay, and duration of ventilation (Table 1).
Overall, CMV reactivation at any level, and clinically-significant reactivation (high-level reactivation and/ or disease), were detected in 29.8% (62/208) and 10.6% (22/208) of ICU patients, respectively. Importantly, the COVID-19 ICU group had significantly higher rates of any CMV reactivation, high-level CMV DNAemia, and CMV disease, compared to concomitant non-COVID-19 group (40% vs. 23%, p = 0.009; 18% vs. 2%, p > 0.001; 12% vs. 1%, p < 0.001, respectively), with longer time from ICU admission to CMV reactivation (21 ± 13 vs. 14 ± 11 days, p = 0.04, respectively) (Table 1).
Analysis of the combined cohort of ICU patients with and without COVID-19 (N = 208) identified variables associated with any-, and with clinically-significant CMV reactivation, including: septic shock, high-dose steroids administration, COVID-19, the number of blood transfusions, and lower ALC. A median ALC of 575/ µL discriminated between patients with and without clinically significant CMV reactivation (specificity 67.4%, sensitivity 93.3%). CMV reactivation was associated with prolonged duration of ventilation, hospitalization, and ICU stay, and in-hospital mortality was higher among patients with CMV reactivation at any level (Table 2). A time-dependent analysis assessing the impact of CMV reactivation on mortality (while accounting for the In the group of COVID-19 ICU patients (N = 85), the main variables associated with clinically-significant CMV reactivation were the number of blood transfusions, lower ALC, and the duration of ventilation, hospitalization, and ICU stay (Table 3).
A total of 8 patients with pneumonitis (6 patients with COVID-19 and 2 patients without COVID-19) were examined for pulmonary-tract CMV reactivation in bronchoalveolar lavage (BAL) fluid (Supplementary Table S1). Local pulmonary CMV reactivation was detected in 5 of the 6 COVID-19 patients. All five had concurrent systemic reactivation. Negative BAL CMV results were obtained in 3 patients who had no detectable DNAemia (Supplementary Table S1).
## Discussion
This study analyzed data from a prospective cohort of concomitant ICU patients with and without COVID-19, monitored for CMV DNAemia, to determine risk factors associated with CMV reactivation and disease.
The main factors associated with clinically-significant CMV reactivation in the overall ICU patient group (N = 208) were septic shock, lower ALC, high-dose steroid use, multiple blood transfusions, and COVID-19 (Table 2). Although the SOFA proxy-measure of severity did not correlate with CMV reactivation, the observed risk factors support the link between baseline illness severity and CMV reactivation risk 1 . Furthermore, in agreement with previous reports 4 , our findings suggest that CMV reactivation may be associated with worse clinical outcome measures [i.e., association with increased duration of ventilation, hospitalization, and ICU stay (Table 2), and time-dependent association with mortality]. While most studies on CMV reactivation in the ICU setting focus on systemic reactivation (i.e., CMV DNAemia), local reactivation in the pulmonary tract could potentially contribute to disease severity. In this regard, it is noteworthy that pulmonary reactivation was detected in five of five COVID-19 ICU patients with concurrent systemic reactivation, who were tested for CMV in BAL. This albeit anecdotal finding, suggests that local pulmonary CMV reactivation may be under-recognized in critically-ill patients.
The association of CMV reactivation with high-dose steroid treatment and lower ALC in ICU patients is biologically plausible, given the central role of T-cells in CMV immune control. It further highlights common immune suppression pathways leading to CMV reactivation in ICU and transplant patients, in whom lymphopenia is known to predict recurrent CMV disease 14 . We also showed that a median ALC value of 575/ µL could identify patients at risk for clinically-significant reactivation. While predictive ALC thresholds should be further defined in ICU patients, this readily available laboratory parameter could be used to monitor and identify critically-ill patients at increased risk for CMV reactivation. Our finding that the receipt of multiple blood transfusions was associated with higher rates of CMV reactivation expands a few previous reports in critically-ill patients 2,3,9 . Mechanistically, although a low risk for CMV transmission via blood products remains (despite leukodepletion) 13 , it is more likely that the immunosuppressive effect of blood transfusions per se 15 could enhance CMV reactivation. As SARS-CoV-2 variants continue to circulate, critically-ill COVID-19 patients are regularly managed in general ICUs. The reported CMV reactivation rates in these patients vary 8,9 . Here we show that COVID-19 ICU patients, despite lower baseline SOFA scores, exhibit higher rates of CMV reactivation and disease compared to concomitant non-COVID-19 patients (Table 1). We further identified lower ALC and multiple blood transfusions as main variables associated with clinically-significant CMV reactivation in COVID-19 ICU patients (Table 3), together suggesting an immunosuppression-mediated mechanism. Interestingly, we have shown that CMV reactivation is significantly delayed in COVID-19 compared to non-COVID-19 ICU patients (Table 1). While this observation could be related to earlier ICU admission (during the disease course) or increased early non-CMV related mortality in COVID-19 patients, it is tempting to speculate that the late-phase immunosuppression characterizing the course of severe COVID-19, could underlie this temporal difference.
Our study was limited by the single-center source of the data, and the inclusion of patients who were regularly tested for CMV. Nonetheless, the cohort, combining for the first time concomitant COVID-19 and non-COVID -19 patients, and the analysis of well-defined virologic and clinical outcomes of CMV reactivation, support the generalizability and clinical implications of our findings.
## Conclusions
In conclusion, we showed the high rates of clinically-significant CMV reactivation in critically-ill ICU patients, higher in COVID-19 compared to concomitant non-COVID-19 patients, and identified patients at increased risk for CMV reactivation and disease. Together with the worse clinical outcomes linked to CMV reactivation, the findings highlight the need for vigilant monitoring of CMV reactivation and consideration of early antiviral treatment in ICU patients at risk, and support future interventional trials.
## References
1. Imlay, Limaye (2020) "Current Understanding of cytomegalovirus reactivation in critical illness" *J. Infect. Dis*
2. Limaye (2008) "Cytomegalovirus reactivation in critically ill immunocompetent patients" *JAMA*
3. Osawa, Singh (2009) "Cytomegalovirus infection in critically ill patients: a systematic review" *Crit. Care Lond. Engl*
4. Imlay (2021) "Risk factors for cytomegalovirus reactivation and association with outcomes in critically ill adults with sepsis: a pooled analysis of prospective studies" *J. Infect. Dis*
5. Papazian (0134) "Cytomegalovirus reactivation in ICU patients" *Intensive Care Med*
6. Wiegand (2025) "Estimating COVID-19 associated hospitalizations, ICU admissions, and in-hospital deaths averted in the united States by 2023-2024 COVID-19 vaccination: a conditional probability, causal inference, and multiplier-based approach" *Vaccine*
7. Tay, Poh, Rénia et al. (2020) "The trinity of COVID-19: immunity, inflammation and intervention" *Nat. Rev. Immunol*
8. Schinas (2023) "Targeting CMV reactivation to optimize care for critically ill COVID-19 patients: A review on the therapeutic potential of antiviral treatment" *Viruses*
9. Imoto (2019) "Incidence and potential risk factors of human cytomegalovirus infection in patients with severe and critical coronavirus disease" *J. Infect. Chemother. Off J. Jpn Soc. Chemother*
10. Tassaneeyasin (2024) "Prevalence and risk factors of cytomegalovirus reactivation in critically ill patients with COVID-19 pneumonia" *PloS One*
11. Barak (2021) "Lessons from applied large-scale pooling of 133,816 SARS-CoV-2 RT-PCR tests" *Sci. Transl Med*
12. Vorontsov (2022) "Amniotic fluid biomarkers predict the severity of congenital cytomegalovirus infection" *J. Clin. Invest*
13. Ljungman (2017) "Guidelines for the management of cytomegalovirus infection in patients with haematological malignancies and after stem cell transplantation from the"
14. Gardiner (2018) "Absolute lymphocyte count: a predictor of recurrent cytomegalovirus disease in solid organ transplant recipients" *Clin. Infect. Dis. Off Publ Infect. Dis. Soc. Am*
15. Vamvakas (1016) "Is white blood cell reduction equivalent to antibody screening in preventing transmission of cytomegalovirus by transfusion? A review of the literature and meta-analysis" *Transfus. Med. Rev* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12870148&blobtype=pdf | # Author affiliations: CSIRO Australian Centre for Disease
Victoria Australia, ( Halpin, J Barr, S Edwards, M Tripp, B Blanch-Lázaro, G Habarugira, Universiti Malaya, Kuala Lumpur, L.-Y Chang
H enipaviruses, exemplified by Hendra virus (HeV) and Nipah virus (NiV), represent critical threats to global health security, given their broad host geographic range, repeated spillover into humans and domestic animals, high case-fatality rates (CFR; 57% CFR for HeV (1) and 80% CFR for NiV [2]), and limited medical countermeasures, creating substantial challenges for outbreak preparedness. Since its emergence in 1994, HeV has caused devastating outbreaks in Australia with infection primarily in horses and humans; there were >105 equine cases in which the animals died or were euthanized and 7 human cases, including 4 deaths (3). NiV was identified in 1998 after it caused a large outbreak in Malaysia and Singapore that affected pigs and humans, resulting in 110 human deaths (1). Since then, NiV outbreaks have been reported in Bangladesh and India, with near-annual human outbreaks in Bangladesh since 2001. An outbreak consistent with NiV infection also occurred in the Philippines, although full viral sequence and isolates were not obtained (1).
The scientific community has recently mobilized efforts to address those challenges through policy frameworks, research roadmaps, investments in vaccines and therapeutics, and collaborative scientific exchange. The World Health Organization (WHO) prioritized Henipavirus nipahense as a prototype pathogen within the Paramyxoviridae family (4). In parallel, 15 subject matter experts published an update to the 2019 WHO research and development (R&D) roadmap for accelerating the development of medical countermeasures to enable effective and timely responses to NiV outbreaks (5). The WHO South-East Asia Regional Office (SEARO) synthesized
## Integrating Prevention and Response at the Crossroads of Henipavirus Preparedness, Hendra@30 Conference, 2024
Kim Halpin, Raúl Gómez Román, Emmie de Wit, Alison J. Peel, Jonathan H. Epstein, Jennifer Barr, Sarah J. Edwards, Melanie N. Tripp
## ONLINE REPORT
Diseases caused by henipaviruses, exemplified by Hendra virus and Nipah virus, pose a serious risk to public health because of their epidemic potential and high casefatality rates and the paucity of medical countermeasures to mitigate them. In December 2024, a group of 150 scientists from 16 countries convened in Geelong, Victoria, Australia, to mark the 30th anniversary of the discovery of Hendra virus. The Hendra@30 conference built upon its predecessor conference held in 2019 in Singapore, Nipah@20, by expanding its program across broader disciplines and integrating sessions on human sociology and disease ecology into the main scientific discussions. We describe key highlights from Hendra@30 and reflect on 4 key elements that have advanced henipavirus research and medical countermeasures research and development. We propose that integrating bat ecology into henipavirus research blueprints will enable development of ecologic countermeasures that prevent spillover and will complement existing preparedness and response efforts with evidence-based prevention strategies.
those global research and policy initiatives into a strategy for the prevention and control of NiV in the Southeast Asia region, developed through expert consultation and published in 2023 (6). CSIRO and the Coalition for Epidemic Preparedness Innovations hosted the Hendra@30 Henipavirus International Conference in Geelong, Victoria, Australia, for 150 scientists from 16 countries (7); the conference marked the 30th anniversary of the first recognized outbreak of HeV disease (1,3). The Hendra@30 conference provided a forum to reflect on the history of henipavirus spillovers, to review 3 decades of scientific findings, and to understand the current challenges in developing diagnostics, therapeutics, and vaccines against henipaviruses in general (Table ). Whereas its predecessor conference, Nipah@20, held in 2019 in Singapore, focused mostly on advances in virology and medical countermeasures (8), Hendra@30 expanded its program to incorporate bat ecology and behavioral determinants of virus transmission, as well as landscape and climatic drivers of spillover. That interdisciplinary approach brought disease ecologists and social scientists into scientific discussions alongside virologists, veterinarians, and public health experts. In addition to the scientific presentations, Hendra@30 included a tour of the local laboratory, a field visit to the Geelong flying fox colony, and a dedicated session featuring survivors of HeV and NiV outbreaks who shared their personal experiences with the viral diseases and how the outbreaks impacted their families and livelihoods. One survivor described her ongoing struggles with neurologic sequelae, 16 years after HeV infection. The daughter of a patient who succumbed to NiV illness has pursued a career in science, focusing her research on the virus that profoundly affected her family. Last, the first veterinarian to encounter cases of HeV during the initial outbreak shared insights into the emotional challenges faced while addressing the loss of both horses and human lives. Those shared experiences added important dimensions, ensuring that the ecologic context and the human impact of henipaviruses remained central to the meeting's scientific discussions.
All countries that have had henipavirus or henipalike outbreaks (Australia, Bangladesh, India, Malaysia, the Philippines, and Singapore) were represented at the conference; their delegations accounted for 66% (99/150) of all attendees. Reflecting the conference's commitment to inclusive leadership, the organizing committee comprised three quarters women (17/23 [74%]) and more than one third early-career scientists (8/23 [35%]) and prioritized diverse representation among speakers. To encourage broad participation by the henipavirus research community, organizers provided travel grants to 14 emerging scholars from low-and middle-income countries.
The proceedings from the conference are now publicly available (7). Here, we highlight 4 key elements from work presented at the conference and from published work that have substantially advanced henipavirus R&D since 2020 and will likely have a profound impact on henipavirus prevention, preparedness, and response (PPR) efforts over the next 5 years.
## Key Takeaways
## Progress in the Development of Henipavirus Medical Countermeasures
In the 5 years between the Nipah@20 and Hendra@30 conferences, SARS-CoV-2 emerged, and the COVID-19 pandemic was declared. The urgent need for COVID-19 medical countermeasures resulted in the advancement of several diagnostics, vaccines, and therapeutics that will likely advance henipavirus R&D (9). For instance, now that several COVID-19 mRNA vaccines are approved for either emergency or routine use in various countries, regulatory familiarity with mRNA platforms could simplify review of mRNA henipavirus vaccines. Preclinical and phase I clinical data for mRNA-1215 were presented, describing early preclinical evidence on cross-protection against NiV and HeV infections. Cross-protection against heterologous henipaviruses had been previously observed with other vaccine platforms, including the HeV protein-subunit and adjuvanted vaccine licensed for use in horses in Australia (10). Presenters reviewed the history of that vaccine along with the continued clinical development for human use, including the prospect of delivery via microneedle patches. Although clinical data in humans were presented only for the mRNA vaccine platform, additional veterinary and preclinical data from other vaccine approaches were also reported, including data in foals from mares immunized with the Equivac HeV protein subunit vaccine; data from mice immunized with a polyphosphazene adjuvanted microneedle patch vaccine based on the HeV protein subunit platform; data in Syrian golden hamsters immunized with a Nipah viral replicon vaccine lacking the NiV fusion (F) protein; data in mice and ferrets immunized with self-amplifying, replicon RNA Nipah vaccines, delivered in lipid inorganic nanoparticles; and data in African green monkeys immunized with single-cycle, recombinant vesicular stomatitis virus-based vaccine expressing the NiV glycoprotein (G).
Several small molecules and therapeutics against henipaviruses under development, mostly in preclinical development with limited clinical data (9). Data were presented at the Conference for novel nanobodies (DS90), new combinations of small molecules (dexamethasone and remdesivir), new methods for the discovery and screening of compounds with in vitro antiviral activity, and immunoglobulins recovered from antibody-secreting cells isolated from humans vaccinated with mRNA-1215. A systematic review has identified well designed clinical efficacy trials and in vivo pharmacokinetic and pharmacology studies as 2 bottlenecks needed to move products down the clinical pipeline (11).
Overall, Hendra@30 hosted 16 scientists presenting data on medical countermeasures. However, much of the progress in the development of such countermeasures has also been the result of progress in other areas of henipavirus R&D, including understanding the viruses at the molecular level.
## Progress in Understanding the Molecular Mechanisms of the Viruses
The conference offered 18 presentations on studies of viral infection mechanisms across virology/immunology, pathogenesis, and rapid oral sessions, including several studying the nuclear trafficking of henipavirus proteins. The mechanisms to enter the nucleus and associated functions such as viral budding (in matrix protein) and immune modulation (in W protein) were discussed in several presentations. Those functions are often important for viral pathogenesis, and therefore, unravelling those pathways could provide novel drug targets (12).
The G and F proteins expressed on the surface of infected cells are a target for the immune system and, hence, medical countermeasures. Cross-reactivity between HeV and NiV has been identified for certain mAbs (including m102.4 and 5B3) against those proteins. Given the recent expansion of henipaviruses and parahenipaviruses, experts discussed the diversity of the G and F proteins across those genera, which holds implications for potential broad-spectrum vaccines and mAbs targeting those proteins.
A second HeV genotype was identified in 2021 (13,14); at the conference, 2 presentations focused on how that genotype might compare to HeV-g1. One study revealed key differences in disease outcomes from experimental challenge in African green monkeys; HeV-g2-infected animals showed reduced severity of respiratory and neurologic disease compared with the original genotype, leading to improved survival outcomes. The molecular mechanism behind those differences is unclear but may be linked to decreased replication efficiency of HeV-g2, which was described in both presentations, and reduced ability to inhibit the interferon induction response, described in the second presentation.
Studying henipaviruses in a standard animal model, African green monkeys, enabled researchers to examine pathogenesis, including the ways the virus reaches the central nervous system; histopathology suggests the blood-brain barrier or the olfactory bulb. Other in vitro and in vivo models in use explore neuropathogenesis using cerebral organoids and a hamster model of infection. Multiple conference presentations covered various aspects of viral infection mechanisms. That field still has gaps in research, including the extent to which the mechanisms of infection vary across henipavirus genotypes and species. Understanding mechanisms of infection is important, given the known diversity of the viruses in the henipavirus genus.
## Progress in Understanding Henipavirus Genetic Diversity
Understanding of henipavirus diversity has expanded dramatically through metagenomic and metatranscriptomic sequencing. Although published studies have been dominated by detections of henipa-like viruses in shrews and rodents (genus Parahenipavirus) (15,16), dozens of new bat henipaviruses were pre-sented at the conference. First, data were presented on the largest survey of bat henipaviruses from Australian flying foxes, incorporating samples collected during 2018-2021 from sites in southeast Queensland and northeast New South Wales, Australia. That work identified 24 new henipavirus species, all with complete or near-complete genomes, and revealed 3 distinct henipavirus clades. Clade 1 henipaviruses included 4 of the 5 existing known henipaviruses (Hendra, Nipah, Cedar, and Ghana viruses), whereas clade 2 included the remaining known henipaviruses (Angavokely virus, discovered in urine collected from fruit bats in Madagascar [17], and Salt Gully virus, the discovery of which was presented at the conference), plus 5 new species with complete genomes. Clade 3 comprised an entirely novel clade of henipaviruses with 11 complete genomes, 5 near-complete genomes, and 3 partial L genes. Only a few of those newly identified viruses have been isolated as of 2025.
Additional evidence of remarkable diversity emerged from studies on Rousettus spp. Egyptian rousette bats in South Africa; 18 putative henipavirus species were identified in those studies. Efforts to understand how genetic diversity translates into antigenic diversity include the creation of a library of soluble G and F proteins from various henipaviruses, including the Langya virus and Angavokely virus F proteins, yielding data on the multimeric diversity of these proteins. Overall, despite major advances in characterizing the extensive genetic diversity of henipavirus and related viruses, particularly those in Australia and South Africa, the implications for crossspecies transmission and potential for disease emergence of these new species remain poorly understood.
## Progress in Understanding Henipaviruses in the Context of Bat Ecology and Bat Health
Hendra@30 featured 20 oral and poster presentations examining the ecologic pathways of henipavirus transmission from bats to spillover hosts (7). Sessions covered disease ecology (viral load and diversity, community ecology, bridging animal hosts, and anthropogenic drivers of spillover), surveillance (in bats, humans, and horses), behavioral determinants of transmission (from bats to animals and humans), and within-host infection dynamics via experimental infection studies in Pteropus, Rousettus, and Artibeus spp. bats.
A unifying theme of the presentations was the connection between habitat preservation, bat health, and spillover risk, building upon published work demonstrating that HeV spillover is driven by rapid environmental changes (18). Research presented used climatic, ecologic, land cover, and flowering data to generate Bayesian network models to accurately predict clusters of HeV spillovers over a 25-year period in Australia. Those studies demonstrated that flying foxes are responding to habitat loss by persistently adopting behaviors previously observed only during climate-driven periods of acute nutritional stress. As a result, flying foxes are losing their nomadism and increasingly remaining in urban and agricultural areas where horse populations have a higher density, increasing spillover risk. Yet when remnant habitat flowered abundantly, providing pulses of natural nectar, flying foxes returned to their natural nomadic behaviors. Every time the increasingly rare flowering occurred during the study period, risk was mitigated and no spillovers occurred. Although spillover can be prevented by vaccinating horses with HeV vaccine, its adoption is declining. Therefore, the behavioral responses of flying foxes to flowering suggest that strategic habitat restoration of species that provide food during periods of resource limitation could sustainably mitigate spillover risk by supporting bat health and reducing contact with spillover hosts.
Conference presentations also built on complementary studies that have linked food shortages with viral shedding whereby viral excretion occurs at higher prevalences, at higher viral loads, and across a broader viral diversity following acute food shortages (19,20). Those observations suggest that bats live on an energetic edge, and those periods of shortage result in bats experiencing allostatic overload, i.e., when insufficient available energy (i.e., food) forces diversion of energy away from immunity, which reduces suppression of viral infections and increases viral shedding (21).
In Bangladesh, Indian flying foxes (Pteropus medius) have been found to prefer roosting in forest fragments near higher human population density in the country, rather than intact forest (22,23). Removal of those timber trees can disrupt bat colonies and lead to colony dispersal. According to a study presented at the conference, 75% of roosts were affected by tree cutting; bat population was an average of 1,700 bats per roost. Bat populations declined at 4 locations, by 63%, 70%, 54% and 20%, from 2021 to 2024. Raw date palm sap, the primary route of zoonotic transmission of NiV in Bangladesh, is not the bat's natural food source but is available during winter months when other food is scarce. NiV outbreaks were most likely to occur in winter months, coinciding with date palm sap cultivation (24).
Collectively, ecologic insights suggest that similar mechanisms might drive henipavirus spillover risk across multiple systems. Scientists in Bangladesh and India have emphasized the need to protect bat habitats and maintain consistent access to natural dietary sources to reduce NiV spillovers (25). Ecologic countermeasures that maintain and restore natural bat foraging habitats could also provide, pending local ecologic and socioeconomic context validation, a generalizable, sustainable approach to reducing henipavirus spillover risk in other areas at risk by supporting bat health and minimizing contact with domestic animals and humans.
## Closing Discussions
Both Nipah@20 and Hendra@30 included closing panel discussions that contextualized henipavirus R&D within a global public health perspective. The panel at Hendra@30 specifically addressed the 2024 update of the Nipah R&D roadmap (5). The original goal of the roadmap was to develop effective medical countermeasures, including diagnostics, therapeutics, and vaccines, by 2030. The panelists highlighted the enthusiasm and ambitious timelines of the expert group involved. They emphasized the importance of securing funding for R&D, addressing cross-cutting issues, overcoming access and implementation hurdles, and navigating regulatory challenges during outbreaks. Shortly after updating the roadmap, WHO shifted its R&D strategy; the shift in strategy resulted from extensive consultations during 2022-2024 and was influenced by the Coalition for Epidemic Preparedness and Innovations-driven 100-day mission (26), which emphasized rapid response to new epidemic threats through the proactive development of prototype vaccines for priority pathogens. Consequently, the WHO strategy prioritizes preparedness and response alongside prevention efforts (4). In contrast to earlier pathogen prioritization efforts, which listed 10 priority diseases, including henipaviruses, the new framework identified >30 priority pathogens, of which NiV remains notable within the Paramyxoviridae family as both a priority and a prototype pathogen. The designation highlighted NiV as both a representative threat and a basis for potential vaccine development.
During the Hendra@30 panel discussion, WHO's new approach raised questions, which were discussed. Panelists also introduced the Paramyxovirus Collaborative Open Research Consortium (CORC), cohosted by WHO and the Indian Council of Medical Research, and emphasized its openness to global researchers and its role in addressing knowledge gaps, ensuring equitable access to medical countermeasures, and strengthening community trust. However, several concerns arose; those included the complexity, speed, and governance of the CORC's operationalization, funding sources, and R&D decision mechanisms. Panelists agreed that aligning Nipah roadmap priorities with the CORC paramyxovirus initiative could benefit overall henipavirus research efforts, reiterating concerns about sustainable R&D funding. Details of the discussion are described in the conference proceedings (7).
## Considerations for the Next Henipavirus Conference
References to the intergovernmental negotiating body drafting and negotiating the pandemic agreement were absent from discussions at Hendra@30 (27). WHO member states recently concluded negotiations and voted to adopt the instrument during the World Health Assembly on May 19, 2025 (28). The agreement provides for targeted research over the next 5 years, benefiting henipavirus R&D and achievement of the WHO CORC paramyxovirus priorities. Articles 19 and 20 of the pandemic agreement address international cooperation and sustainable financing for pandemic prevention, preparedness and response (29); therefore, those are areas of opportunity for henipavirus researchers and international policymakers to establish dialogue and learn from each other.
Another area for improvement is ensuring timely access to medical countermeasures when needed by the affected populations. Because of time limitations, neither the Nipah@20 nor the Hendra@30 conference provided a specific forum to discuss access to medical countermeasures; nonetheless, the 2024 update to the Nipah disease R&D blueprint addressed several issues related to ensuring access (5). We propose that access be integrated into the discussions of the new Paramyxoviridae CORC, as well as in the planning of the next major henipavirus conference, possibly Nipah-Malaysia@30 in Malaysia (30 years after the first outbreak in 1998-1999) or Nipah-Bangladesh@25 in Bangladesh (25 years after the first outbreak in Meherpur, Bangladesh, in 2001).
Although Hendra@30 discussed ecologic countermeasures not contemplated in its Nipah@20 predecessor, it missed the opportunity to host representatives from national regulatory agencies in henipavirus-affected countries, who were invited but were unable to travel; having those experts will be considered in planning the next conference. To ensure participation from stakeholders involved in reviewing data and regulating licensure of, and ultimately access to, medical countermeasures for the persons who need them most, one possibility will be to conduct hybrid science and policy interface sessions, similar to sessions conducted during the 8th World OneHealth Congress (30).
In terms of basic science, 2 presentations addressed the topic of cell-mediated immunity (CMI), and a few more addressed innate immunity indirectly. Encouraging more presentations on immunity will be an area to consider for the next meeting; we anticipate a full session dedicated to CMI, including vaccine-elicited immunity and CMI-elicited immunity by natural infection among NiV and HeV infection survivors in Bangladesh, Malaysia, India, and Australia. One intention is to ensure greater representation from the Philippines; the henipavirus R&D community has much to learn from epidemiologic studies and from survivors of the Nipah-like outbreak in that country.
We anticipate that ex vivo models, including organoids and organs-on-a-chip, will become a booming area of research in the coming years, because that topic was of great interest during Hendra@30. As noted in the proceedings, 1 poster and 4 oral presentations at the conference featured the use of reconstituted airway epithelia as a model for the study of viral replication and pathogenesis; the development of human cerebral organoids, or 3-dimensional, selforganizing tissue-like structures derived from human induced pluripotent stem cells; organ-on-a-chip micro physiologic models to emulate the alveoli and capillaries of the human lung; the use of a 3-dimensional cortical organoid model of human cerebral cortex; and the use of primary normal human bronchial epithelial cells grown and cultured in transwells at air-liquid interfaces.
Last, we anticipate that major topics at the next meeting will be, again, bat ecology, human behavior, and the design of strategies to prevent spillover. Furthermore, given that many novel henipa-like viruses are increasingly being detected in shrews, a specific focus on evolutionary ecology of henipaviruses in bats versus other species could be informative. We recommend that ecologic countermeasures be at the forefront of henipavirus prevention. We will continue to advocate for the design and implementation of these ecologic countermeasures and the fundamental ecology studies that inform them as part of a One Health strategy (30). That approach will ensure that primary and secondary prevention remains in the equation of pandemic prevention, preparedness and response, as enshrined in the WHO Pandemic Agreement.
## References
1. Spengler, Lo, Welch et al. (2025) "Henipaviruses: epidemiology, ecology, disease, and the development of vaccines and therapeutics" *Clin Microbiol Rev*
2. Vasudevan, Subash, Mehta et al. (2024) "Global and regional mortality statistics of Nipah virus from 1994 to 2023: a comprehensive systematic review and meta-analysis"
3. Yuen, Fraser, Henning et al. (2021) "Hendra virus: epidemiology dynamics in relation to climate change, diagnostic tests and control measures" *One Health*
4. (2024) "World Health Organization Pathogens prioritization: a scientific framework for epidemic and pandemic research preparedness. Geneva: The Organization"
5. Moore, Mehr, Ostrowsky et al. (2024) "Measures to prevent and treat Nipah virus disease: research priorities for 2024-29" *Lancet Infect Dis*
6. (2020) "WHO South-East Asia Regional strategy for the prevention and control of Nipah virus infection 2023-2030. Geneva: The Organization"
7. (2024) "Proceedings from Hendra@30 Henipavirus International Conference"
8. Gómez Román, Wang, Lee et al. (2020) "Nipah@20: lessons learned from another virus with pandemic potential" *MSphere*
9. Gómez Román, Tornieporth, Cherian et al. (2022) "Medical countermeasures against henipaviruses: a review and public health perspective" *Lancet Infect Dis*
10. Geisbert, Bobb, Borisevich et al. "A single dose investigational subunit Hendra@30 Henipavirus Conference"
11. (2021) "vaccine for human use against Nipah virus and Hendra virus" *NPJ Vaccines*
12. Chan, Haeusler, Choy et al. (2025) "Therapeutics for Nipah virus disease: a systematic review to support prioritisation of drug candidates for clinical trials" *Lancet Microbe*
13. Tripp, Rawlinson, Edwards et al. (2025) "The intracellular virus-host interface of henipaviruses" *J Virol*
14. Annand, Horsburgh, Xu et al. (2022) "Novel Hendra virus variant detected by sentinel surveillance of horses in Australia" *Emerg Infect Dis*
15. Wang, Anderson, Halpin et al. (2021) "A new Hendra virus genotype found in Australian flying foxes" *Virol J*
16. (2024) "International Committee on Taxonomy of Viruses. ICTV taxon details" *Genus: Henipavirus*
17. Caruso, Edwards (2023) "Recently emerged novel henipa-like viruses: shining a spotlight on the shrew" *Viruses*
18. Madera, Kistler, Ranaivoson et al. (2022) "Discovery and genomic characterization of a novel henipavirus, Angavokely virus, from fruit bats in Madagascar" *J Virol*
19. Eby, Peel, Hoegh et al. (2023) "Pathogen spillover driven by rapid changes in bat ecology" *Nature*
20. Peel, Wells, Giles et al. (2019) "Synchronous shedding of multiple bat paramyxoviruses coincides with peak periods of Hendra virus spillover" *Emerg Microbes Infect*
21. Becker, Eby, Madden et al. (2023) "Ecological conditions predict the intensity of Hendra virus excretion over space and time from bat reservoir hosts" *Ecol Lett*
22. Plowright, Ahmed, Coulson et al. (2024) "Ecological countermeasures to prevent pathogen spillover and subsequent pandemics" *Nat Commun*
23. Hahn, Epstein, Gurley et al. (2014) "Roosting behaviour and habitat selection of Pteropus giganteus reveals potential links to Nipah virus epidemiology" *J Appl Ecol*
24. Hahn, Gurley, Epstein et al. (2014) "The role of landscape composition and configuration on Pteropus giganteus roosting ecology and Nipah virus spillover risk in Bangladesh" *Am J Trop Med Hyg*
25. Gurley, Hegde, Hossain et al. (2017) "Convergence of humans, bats, trees, and culture in Nipah virus transmission" *Bangladesh. Emerg Infect Dis*
26. Yadav, Baid, Patil et al. (2025) "A One Health approach to understanding and managing Nipah virus outbreaks" *Nat Microbiol*
27. (2024) "Coalition for Epidemic Preparedness and Innovations. CEPI 2.0 and the 100 days mission"
28. (2025) "WHO member states conclude negotiations and make significant progress on draft pandemic agreement"
29. (2025) "World Health Assembly adopts historic pandemic agreement to make the world more equitable and safer from future pandemics"
30. (2025) *Resolution WHA* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645992&blobtype=pdf | # Mutations of two amino acids in VP5 mediate the attenuation of human rotavirus vaccine: evidence from in vitro and in vivo studies
Theresa Bessey, Yuhuan Wang, Sung-Sil Moon, Liliana Sanchez-Tacuba, Philippe Jaïs, Harry Greenberg, Baoming Jiang
## Abstract
Various vaccines, like polio, measles, and rotavirus, have been developed by serial-passaging in cell culture. Live oral rotavirus vaccines have been shown to be generally safe, but mechanisms of attenuation are not known. We have used a new, entirely plasmid-based reverse genetics system to artificially generate the novel human rotavirus vaccine strain ) and analyze the effect of the mutations within the VP4 gene on adaptation in vitro and attenuation in vivo. We demonstrated that out of the 6 amino acid mutations that appeared after serial passaging in Vero cells, mutations of wild-type CDC-9 P11 at VP4 AA331 and AA385 each or in combination were associated with increased replication in vitro comparable to cell-culture adapted CDC-9 P45. Neonatal rats infected with the single AA331 or AA385 mutant had reduced viral shedding, comparable to cell-culture passaged CDC-9 P45. We observed additional reduced shedding in neonatal rats that were infected with combination mutants harboring mutations at position AA331_385_388, indicative of a slight additive effect. Our data indicate that mutations in the VP5* region of the VP4 gene, particularly at position AA331 and AA385, are the determining factor for in vitro replication adaptation and in vivo attenuation of a G1P[8] rotavirus vaccine. This information provides great potential for targeted mutation in rotavirus vaccine generation instead of labor-consum ing serial passaging in cell culture. IMPORTANCE Live oral rotavirus vaccines have been developed through serial passaging in cell culture and found to be generally safe and efficacious in children. Live vaccines are also found to be associated with rare but severe adverse events, such as intussusception, in vaccinated children. Mechanisms for vaccine attenuation and adverse effects are unknown. We have developed a novel human rotavirus vaccine strain (CDC-9) and demonstrated several amino acid mutations in the VP4 gene of cell-passaged virus.In the present study, we identified two key amino acid mutations via reverse genetics technology in VP4 that mediated enhanced growth in cell culture, including a human intestinal cell line, reduced virus shedding, and downregulated inflammatory response in neonatal rats. This study is the first to identify the molecular signatures that define attenuation of human rotavirus vaccine and should help provide guidance for develop ing new generations of safe and effective vaccines.KEYWORDS in vivo, in vitro, mutations, vaccine, rotavirus S erial-passaging in cell culture has been used to generate attenuated virus strains for measles, polio, mumps, rubella, varicella, and rotavirus vaccine development (1-6). Sequence changes within the genome can be correlated with viral adaptation in tissue culture and attenuation in vivo making serial-passaging an important tool for
of human rotavirus vaccine strains. In the present study, we took advantage of complete passage history and fully delineated nucleotide and amino acid sequences to examine associations between specific mutations in VP5* and documented phenotypes in vitro and in vivo of the human rotavirus vaccine strain CDC-9 utilizing reverse genetics technology. The findings from the present study would help expedite the development, skipping the time-consuming process of serial passaging in cell culture, and improve the safety of rotavirus vaccines.
## RESULTS
## Importance of rotavirus VP4 for infectivity in vitro and virulence in vivo
Rotavirus VP4 protein is essential for cell attachment, cell wall penetration, and entry into the cytoplasm. To understand the effect of VP4 on infectivity and pathogenicity, we generated recombinant mono-reassortment strains on a CDC-9 P11 background with different VP4 genes from simian (SA11), bovine (UK), and human (CDC-9) rotaviruses (rCDC-9 P11 VP4_RRV, rCDC-9 P11 VP4_SA11, rCDC-9 P11 VP4_UK, and rCDC-9 P11 VP4_P45) (Fig. 1a). A recombinant mono-reassortment between wild-type CDC-9 P11 with only the VP4 gene from attenuated CDC-9 P45 (rCDC-9 P11 VP4_P45) demonstrated a significant increase in titer in MA104 cells when compared with rCDC-9 P11, indicating that VP4 in CDC-9 P45 is primarily responsible for the enhanced cell culture replication capacity of this strain. Interestingly, the titers of mono-reassortment strains with simian (RRV or SA11) or bovine (UK) VP4 remained lower than rCDC-9 P11 VP4_P45 most likely due to the attenuated nature of CDC-9 P45 VP4. To determine the effect of VP4 origin on in vivo infectivity and pathogenesis, we initially infected 5-day-old neonatal rats with 1 × 10 7 FFU of rhesus rotavirus (RRV), rCDC-9 P11, rCDC-9 P11 harboring RRV VP4 (rCDC-9 P11 VP4_RRV), rCDC-9 P11 harboring P45 VP4 (rCDC-9 P11 VP4_P45), or mock control. Body weight in neonatal rats after virus infection reflected the general state and health of these animals (Fig. 1b). We observed comparable body weight gain in animals infected with rCDC-9 P11 VP4_P45 and mock control but reduced body weight gain in animals infected with RRV and rCDC-9 P11 VP4_RRV, and rCDC-9 P11. Of note, no diarrhea was observed in any infected animals. We observed that animals infected with RRV had the highest level of rotavirus shedding, followed by rCDC-9 P11 VP4_RRV on day 4 (Fig. 1c). Rats with rCDC-9 P11 infection had moderate levels of shedding but significantly reduced rotavirus shedding after infection with rCDC-9 P11 VP4_P45 (Fig. 1c). These results indicate that the VP4 gene of the attenuated vaccine strain CDC-9 P45 mediates reduced shedding in neonatal rats and the increased virus growth in cell culture and demonstrate that this vaccine strain is adapted in vitro and attenuated in vivo due to sequence changes within the VP4 gene.
## Generation of recombinant CDC-9 strains with mutations in VP4 gene
By creating recombinant viruses harboring CDC-9 P45 VP4 on a CDC-9 P11 background, we demonstrated the importance of mutations in VP4 for attenuation in vivo and adaptation in vitro (Fig. 1). During serial passaging in Vero cells, we observed a total of six functional AA mutations within VP4 (Table 1). To identify which mutation in the VP4 gene of the rotavirus vaccine candidate strain CDC-9 was responsible for growth adaptation in vitro and attenuation in vivo, we have generated multiple recombinant rotavirus strains with single or multiple point mutations in the VP4 gene. The structural view of VP4 indicated that 5 of the 6 present mutations are within the VP5* region of VP4, whereas only one mutation is in the VP8* region of VP4 (Fig. 2a). We success fully generated constructs of CDC-9 P11 (rCDC-9 P11) and CDC-9 P45 (rCDC-9 P45), single point mutations in VP4 on a P11 backbone (rCDC-9 P11 VP4_AA51, rCDC-9 P11 VP4_AA331, rCDC-9 P11 VP4_AA385, and rCDC-9 P11 VP4_AA499) as well as multiple point mutations in VP4 on a P11 background (rCDC-9 P11 VP4_385_388, rCDC-9 P11 VP4_AA331_385_388, and rCDC-9 P11 VP4_AA331_364_385_388) by reverse genetics and determined their titers at passage 4 in MA104 cells (Fig. 2b). We observed a significant titer difference of rCDC-9 P45, rCDC-9 P11 VP4_AA385_388, rCDC-9 P11 VP4_AA331_385_388, and rCDC-9 P11 VP4_AA331_364_385_388 when compared to rCDC-9 P11. Of note, rCDC-9 P11 VP4_AA51 and rCDC-9 P11 VP4_AA499 showed lower titers than rCDC-9 P45, but comparable titer to rCDC-9 P11. We were not able to generate single point mutants for AA364 and AA388. Sequences and mutations for all recombi nant CDC-9 strains generated were confirmed by Next Generation Sequencing (data not shown).
## Comparison of recombinant CDC-9 strains in vitro
In a previous study (22), we showed that native CDC-9 P45 grew to higher titers than CDC-9 P11 in a human intestinal cell line (Caco-2). Here, we used the recombinant CDC-9 strains and VP4 mutant strains to determine if there were differences in growth for these recombinant strains as well. As seen for the original CDC-9 P11 and CDC-9 P45 strains, we showed that recombinant CDC-9 P11 (rCDC-9 P11) grew to significantly lower titers compared to rCDC-9 P45 over the time course of 84 h in Caco-2 cells (Fig. 3). Recombinant viruses with single point mutations at position AA331 or AA385 were able to replicate to titers seen for rCDC-9 P45 indicating that these two mutations are mainly responsible for the adaptation of CDC-9 strain in vitro. Recombinant viruses with single point mutations at position 51 or 499 showed comparable growth to rCDC-9 P11 in Caco-2 cells, indicating that these two mutations do not play a role in cell culture adaptation (Fig. S1). Of note, infection with recombinant virus harboring both mutations at position 385 and 388 (rCDC-9 P11 VP4_AA385_388) did not appear to significantly increase the titer compared to rCDC-9 P11 VP4_AA385 alone. Infection with rCDC-9 P11 VP4_AA331_385_388 and rCDC-9 P11 VP4_AA331_364_385_388 showed a slight additive effect on replication. In addition to infectious titers, we examined the sizes of plaques from infections by the original CDC-9 P11 and its VP4 mutants. We observed clear large plaque sizes after infection with native CDC-9 P45 and its mutants rCDC-9 P11 VP4_AA331 and rCDC-9 P11 VP4_AA385 in MA104 cells. By contrast, we did not observe any plaques by CDC-9 P11 (data not shown). These data indicate that mutations at position AA331 and AA385 are critical for optimal adaptation and growth in vitro.
## Identification of molecular signatures involved in CDC-9 attenuation and pathogenesis in neonatal rats
Since we previously observed no effect of CDC-9 P45 VP4 on bodyweight gain but significantly reduced virus shedding, we next assessed which of these VP4 mutations were important for attenuation of the strain CDC-9 in vivo by using a neonatal rat model. Infection with rCDC-9 P11 led to significantly less gain in body weight from day 6 to 13 compared to infection with rCDC-9 P45 and mock-inoculated animals (Fig. 4a). Infection with rCDC_9 P11 VP4_AA331 and rCDC-9 P11 VP4_AA385 showed normal gain in body weight comparable to rCDC-9 P45 and mock control. Combination mutants rCDC-9 P11 VP4_AA331_385_388 and rCDC-9 P11 VP4_AA331_364_385_388 also showed significantly greater gain in body weight than rCDC-9 P11 (Fig. 4b).
To identify the effect of VP4 and its mutants on disease progression and rotavirus shedding, we measured rotavirus shedding over 14 days in neonatal rats. rCDC-9 P11 induced a high level of shedding with a peak shedding between day 4 and 5 (Fig. 5a). rCDC-9 P11 further induced diarrhea in 4 out of 10 animals 2-3 days post inoculation, no diarrhea was observed in neonatal rats infected with other recombinant viruses (Table 2). Infection with rCDC-9 P45 induced significantly less shedding compared to rCDC-9 P11 with peak shedding from day 4 to 6. A similar pattern was seen in animals infected with rCDC-9 P11 VP4_AA331 or rCDC-9 P11 VP4_AA385 with significantly reduced shedding compared to rCDC-9 P11. However, the peak shedding in animals infected with rCDC-9 P11 VP4_AA385 lasted from day 3 to day 7, whereas peak shedding after infection with rCDC_9 P11 VP4_AA331 had a shorter peak between day 4 and day 6. Similar to the in vitro data, mutations at position AA51 and AA499 did not appear to significantly change the virulence of the viruses compared to rCDC-9 P11 (Fig. S2). We observed elevated levels of shedding in animals infected with rCDC-9 P11 VP4_AA51 or rCDC-9 P11 VP4_AA499 and rCDC-9 P11 with a peak on days 4 and 5.
To examine whether multiple mutations in VP4 had an additive effect, rats were infected with recombinant CDC-9 with two, three, or four mutations in VP4 (Fig. 5b). rCDC-9 P11 double VP4 mutants AA385 and AA388 induced level of shedding compa rable to the single mutant AA385. rCDC-9 P11 triple VP4 mutants AA331, AA385, and AA388 and quadruple mutants AA331, AA364, AA385, and AA388 induced comparable levels of shedding with peak at day 4 or 5, which appeared to be slightly lower than that from the double mutant. These results showed that double mutations at VP4 AA331 and AA385 appeared to have an additive effect on the level of attenuation of this G1P [8] rotavirus, whereas mutation at VP4 AA388 seemed to have a small added effect on its further attenuation. Additional mutation at position AA364 (rCDC-9 P11 VP4_AA331_364_385_388) did not appear to significantly reduce the shedding further.
## Cytokine response after infection with recombinant CDC-9 strains and mutants
Cytokines and chemokines play a pivotal role in disease progression and rotavirus shedding (26). We analyzed cytokine and chemokine profiles in neonatal animals after infection with different recombinant rotaviruses (Fig. 6). We showed that anti-inflammatory cytokine Interleukin (IL)-10 and immune-modulatory granulocyte macrophage colony-stimulating factor (GM-CSF) were significantly upregulated after infection with rCDC-9 P45 compared to rCDC-9 P11 (Fig. 6a). Additionally, we observed significant upregulation of GM-CSF after infection with rCDC-9 P11 VP4_AA331_385_388 and rCDC-9 P11 VP4_AA331_364_385_388. We also observed a significant increase of IL-10 in rCDC-9 P11 VP4_AA331-infected animals compared to rCDC-9 P11 VP4_AA385 infected rats. On the other hand, pro-inflammatory cytokines IL-1β, IL-7, IL-18, and chemokines MCP-1 and MIP-1α were upregulated after infection with rCDC-9 P11 compared to rCDC-9 P45 (Fig. 6b). We showed that infection with rCDC-9 P45 led to reduced expression of pro-inflammatory cytokines IL-2 and significantly reduced expression of IL-12 compared to infection with rCDC-9 P11 (Fig. S3). We showed that rCDC-9 P11 VP4_AA385 induced comparable levels of IL-1β, IL-18, MCP-1, and MIP-1α to as rCDC-9 P11 indicating that this mutation was not involved in modulating these cytokines in vivo. Of note, no significant difference was found for type II interferon expression in any of the VP4 mutants tested (data not shown).
## DISCUSSION
In the present study, we have comprehensively analyzed the mutations within rotavirus VP4 gene that occurred during serial passaging of the human vaccine strain CDC-9 in Vero cells and identified for the first time, two molecular signatures in the membrane fusion domain of VP5 that facilitated adaptation in vitro and attenuation in vivo. It has been previously shown that VP4 is the key factor facilitating replication in cell culture when examining heterologous recombinant rotavirus strains with human or murine VP4 on a simian rotavirus backbone (25,27,28). We also showed that infection of neonatal rats with a simian VP4 (RRV) on a human backbone (rCDC-9 P11) leads to significantly higher shedding in neonatal rats compared to the parental human rCDC-9 P11 rotavirus infection (Fig. 1b). In addition, when using rCDC-9 P11 (wild-type) backbone with VP4 from CDC-9 P45 (attenuated), we were able to recreate findings we observed with a full recombinant and cell culture-derived attenuated CDC-9 P45. We further showed that rCDC-9 P11 VP4_P45 and rCDC-9 P45 grew to significantly higher titers in simian (MA104) and human (Caco-2) cell lines compared to rCDC-9 P11 (Fig. 2b and3). These data support a previously published study from our lab that original, cell culture-adap ted CDC-9 P45 grew to significantly higher titers than CDC-9 P11 (22). Our findings demonstrate the primary importance of VP4 in adaptation to high growth in cell culture despite a few mutations occurring in other rotavirus genes. One early study examined the effect of mutations within the NSP4 gene on the attenuation of human rotavirus vaccine Rotarix, but no correlation was identified (29).
Previous studies analyzed the effect of whole VP4 gene substitution and did not identify specific mutations within VP4 by reverse genetics (25,28). When analyzing mutations of VP4 in circulating or passaged strains, most mutations appeared within the VP5* region, especially between AA360 and AA400 (30)(31)(32). Region AA382-400 of VP4 has been defined as the membrane interaction loop in VP5 (33) with the fusion domain at AA385. Substitutions in this region may favor a growth advantage in cell culture and render the virus attenuated in humans. Five mutations appear in the vaccine strain Rotarix within VP4, namely, G51D, L167F, S331F, D385Y, and N695I (13). Mutations at positions 51, 331, and 385 are found conserved in Rotarix and CDC-9 after serial passaging in cell culture (21,22) (Table 1). Rotavirus entry is highly dependent on the fusion domain in VP5* (AA384-AA404) and AA changes within this region can result in conformational changes in VP4 (34). The position AA385 was identified by several other studies as a mutation following serial passaging (30,32,35,36) of human rotavirus, while high conservation of 385D was reported in wild-type P [8] rotavirus and not in cell culture passaged strains (37). However, these findings only identify associations between AA mutation at this one position and strain adaptation and attenuation.
Nevertheless, these published data with a single point mutation at AA385 were in agreement with increased replication capacity in cell culture and reduced infectivity in neonatal rats in the present study. We further observed an additive effect of the double mutations at AA331 and AA385 on attenuation, which agreed with the data that these two mutations appeared together at passage 13 (7 passages in MA104 cells and 6 in Vero cells) when the strain was adapted to grow in Vero cells. In addition, we observed mutations at positions AA388 and AA499 that appeared at passage 26 and 43 (19 and 36 in Vero cells), respectively (Table 1). The appearance of the mutation in Vero cell-adapted virus can be directly linked to an increased titer (~10 7 FFU/mL at P25) compared to MA104-passaged CDC-9 P11 (titer ~10 5 FFU/mL; Table 1). Infectious titer did not increase further following the appearance of these late mutations since the titer ranged from 1.3 × 10 7 FFU/mL to 4.5 × 10 7 FFU/mL between passages 26 and 45. We demonstrated a negative association between cell culture passages and virus virulence; CDC-9 P11 induced high levels of virus shedding and diarrhea in animals, while CDC-9 P45-infected animals had significantly reduced viral shedding and no diarrhea (Table 1), confirming findings in a previous study (22).
We were not able to generate recombinant CDC-9 strains with single point muta tions at position AA364 and AA388. Since those mutations occurred after the initial AA change at positions 51, 331, and 385 at passage 13, we speculate that there could be a structural change within the VP4 protein that induced additional mutations at AA364 and AA388 to produce a stable VP4 protein. For CDC-9 P45, we observed that the VP4 protein appeared in an upright conformation compared to VP4 from CDC-9 P11 (38). The appearance of the upright confirmation of VP4 was explained by the mutations mainly at AA331, AA385, and AA388 stabilizing the apex of VP4 spikes. These data were in accordance with a recently published study that compared AA mutations in various rotavirus genotypes and showed that mutations accumulate between AA380 and AA390 (31). No mutations in AA331 region were seen in that study for the analyzed rotavirus genotypes G1P [8], G3P [8], G9P [8], G12P [8], and G2P [4]. However, that study was done by passaging rotavirus genotypes in primary RhMK cells (up to 5 passages) followed by passaging in MA104 cells (up to 10 passages), whereas we saw mutations accumulating after adapting and passaging in Vero cells, a cell substrate commonly used for vaccine manufacturing. We showed no additive effect of mutations AA385 and AA388 (rCDC-9 P11 VP4_AA385_388); however, once we introduced AA331 (rCDC-9 P11 VP4_AA331_385_388) and AA364 (rCDC-9 P11 VP4_AA331_364_385_388), we were able to see comparable titers in vitro (Caco-2 cells, Fig. S1) as well as normal body weight gain and reduced shedding in neonatal rats (Fig. 4b and5b). Cytokines are an important tool to target and help clear viral infections. Various cytokines have been reported after infection with different rotaviruses. Pro-inflammatory cytokines and chemokines like IL-1β, IL-8, GRO-α, RANTES, MIP-1α, MCP-1, and IFN-γ have been detected in cell culture supernatants and plasma and mucosal surfaces after infection in humans (39)(40)(41)(42). We detected induction of pro-inflammatory cytokines and chemokines (e.g., IL-1β and MIP-1α; Fig. 6b) in neonatal rats after infection with rCDC-9 P11 and a shift to anti-inflammatory cytokines (e.g., IL-10, Fig. 6a) after infection with attenuated rCDC-9 P45 or VP4 mutants harboring AA331 mutations, either single or in combination (VP4_AA331, VP4_AA331_385_388, and VP4_AA331_364_385_388). These data highlight the association of mutations at AA331 and reduced viral shedding as well as down regulated inflammatory response. Of note, infection with rCDC-9 P11 VP4_AA385 did not downregulate the pro-inflammatory response for selected cytokines (e.g., IL-1β, IL-18 MIP-1α) indicating that AA385 might be involved in reduced viral shedding in vivo but did not modulate inflammatory responses as seen with rCDC-9 P11 VP4_AA331. Further studies will need to investigate the effects of AA385 and other mutations on host responses in vivo to fully understand the immunological properties of attenuated rotavirus vaccines. The present study examined cytokine profiles in sera of animals 21 days post infection. Due to the limited number of neonatal rats available, no cytokine analysis was conducted for early time points, which should be done in future studies.
For CDC-9, we have observed major changes in amino acid properties for position 331 (S331F, hydrophilic to hydrophobic) and a change from hydrophilic acidic to hydrophilic basic amino acid at position 385 (D385H) (Table 1), resulting in increased growth in cell culture (MA104 and Caco-2) and reduced pathogenicity in neonatal rats (Table 2). In a previously conducted study analyzing the protein structure of VP4 in CDC-9 P11 and CDC-9 P45, we demonstrated that the VP5* domain of CDC-9 P45 remained stable in an upright position, indicating a fully infectious virion, compared to CDC-9 P11 which was unstable and unable to mediate cell entry (38). These data were consistent with the lower infectivity in vitro of CDC-9 P11 as well as rCDC-9 P11. During this study, we found that a mutation in VP4 at position AA51, AA331, or AA385 was likely to stabilize the tip of the VP4 spike and, therefore, result in the formation of triple-layered CDC-9 P45 particles. In accordance with these previous findings, our results from rCDC-9 P11 strains with single point mutations at position AA331 or AA385 showed a significant increase in viral growth in vitro (Fig. 3) and reduced viral shedding in vivo (Fig. 5) compared to rCDC-9 P11. In contrast, rCDC-9 P11 VP4_AA51 did not result in an increased viral growth in vitro (Fig. S1) or decreased viral shedding in vivo (Fig. S2) compared to rCDC-9 P11. Conse quently, we concluded that mutation at either AA331 or AA385, or both were responsible for attenuation and adaptation. When analyzing recombinant variants with multiple point mutations (Fig. S1 for in vitro results and Fig. 5 for in vivo results), we demonstrated that the combination of mutations at position AA331, AA385, and AA388 appeared to result in comparable results to the attenuated rCDC-9 P45. Since mutations at position AA331 and AA385 appeared at the same time during serial passaging and the two mutants singularly or in combination induced in vitro and in vivo responses comparable to rCDC-P45 remained to be determined whether all three mutations at AA331, AA385, and AA388 are needed to work together to fully attenuate CDC-9. We showed that rCDC-9 P11 VP4 AA331 and rCDC-9 P11 VP4 AA385 mutants induced predominantly anti-inflammatory or suppressed pro-inflammatory response compared with a general pro-inflammatory state in rCDC-9 P11 infected animals. This shift from pro-inflammatory response against wild-type CDC-9 P11 to anti-inflammatory state or reduced pro-inflammatory response against late passaged CDC-9 P45 or VP4 mutants might explain dramatic reduction in shedding and their apparent attenuation in neonatal rats. Future studies will be conducted to determine the three-dimensional structures of VP4 mutants containing the single point mutations (rCDC-9 P11 VP4_AA331, rCDC-9 P11 VP4_AA385, or rCDC-9 P11 VP4_AA388) and multiple point mutations by cryo-electron microscopy and examine virus-host interactions to elucidate mechanisms of mutations and associated functional changes.
In summary, we were successful in generating recombinant human rotaviruses based on the sequence of the vaccine candidate CDC-9 wildtype (P11) and attenuated (P45) strain. We demonstrated that mutations at positions AA331 and AA385 within the VP5* region of the VP4 gene of this vaccine strain were crucial in mediating adaption in cell culture and attenuation as well as downregulated inflammatory response in neonatal rats. These findings together with conserved mutations at AA331 and AA385 from the licensed Rotarix vaccine allowed us to identify the molecular signatures that define the attenuation and safety of human rotavirus vaccine G1P [8] strains. Studies are in progress to examine specific mutations in VP4 and other genes of non G1P [8] strains (e.g., DS-1 like G9P [6]) and their effect on virus attenuation, pathogenesis, and safety in animals and children.
## MATERIALS AND METHODS
## Cells and viruses
African green monkey kidney epithelial cell line MA104 (ATCC CRL-2378.1) was passaged in 199 medium (Sigma Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS, one shot, Gibco), 100 IU penicillin/mL, 100 µg/mL streptomycin, and 0.292 mg/mL L-glutamine as previously described (25). A baby hamster kidney fibroblast (BHK) cell line stably expressing T7 RNA polymerase BHK-T7 was kindly gifted by Dr. Buchholz at the NIH and previously described (43). BHK-T7 cell line was cultivated in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% FBS, 100 IU penicillin/mL, 100 µg/mL streptomycin, 0.292 mg/mL L-glutamine, and 0.2 µg/mL G-418 (Promega, Madison, WI, USA). A human intestinal carcinoma cell line Caco-2 (ATCC HTB-37) was passaged in Eagle's minimum essential medium (EMEM) supplemented with 20% FBS.
Infection of Caco-2 cells for growth curve analysis was performed as previously described (22). In brief, Caco-2 cells were seeded in 24-well plates for 5 days and infected with the indicated viruses at MOI 0.1 for 1 h. After removal of inoculum, fresh serum-free EMEM containing 10 µg/mL porcine trypsin (Gibco) was added and samples collected at the indicated time points and used for virus titration.
Original CDC-9 virus was isolated in 2003 from a 5-month-old boy hospitalized in the United States that presented with acute diarrhea (21). CDC-9 was passaged for up to 7 passages on MA104 cells followed by passaging on Vero cells for up to 45 passages (P45) for vaccine seed virus preparation. To increase titer of early passage virus, the common progenitor of the vaccine seed virus CDC-9 P4 was passaged in MA104 cells up to P11/12 separately. For large-scale virus production, MA104 (for CDC-9 P11/12) or Vero (for CDC-9 P45) cells were seeded in roller bottles until confluency is reached and infected at MOI 0.1 for 2-3 h with the respective viruses. After the addition of fresh serum-free DMEM containing 30 µg/mL porcine trypsin, roller bottles were incubated for 3-5 days until cytopathic effect (CPE) was observed. Roller bottles were placed at -80°C, freeze-thawed three times and clarified by centrifugation at 8,000 rpm for 30 min. Clarified CDC-9 strains were aliquoted and stored at -65 to -95°C for subsequent use. Uninfected MA104 cells in roller bottles were processed in the same manner for use as mock infection.
## Virus titration
Titration of original and recombinant viruses was performed by using an Immunospot assay as previously described (22). In brief, MA104 cells were seeded in 96-well plates and cultivated until confluency was reached. Cells were washed with serum-free Iscove's modified Dulbecco medium (IMDM, Gibco) and serial dilution of viruses was performed in serum-free IMDM also. Cells were fixed 18 h after infection, and plates were incuba ted with a rabbit anti-Wa antibody, followed by incubation with a peroxidase-labeled goat-anti rabbit IgG antibody (KPL) and staining with True Blue Peroxidase Substrate (VWR, Radnor, PA, USA). Plates were scanned by using a CTL analyzer (Cellular Technol ogy, Kennesaw, GA, USA). Stained cells over two dilutions were counted and the average number of viruses per milliliter was calculated.
## Plasmids
Human CDC-9 plasmid collection (pT7-CDC-9 P11 VP1, pT7-CDC-9 P11 VP2, pT7-CDC-9 P11 VP3, pT7-CDC-9 P11 VP4, pT7-CDC-9 P11 VP6, pT7-CDC-9 P11 VP7, pT7-CDC-9 P11 NSP1, pT7-CDC-9 P11 NSP2, pT7-CDC-9 P11 NSP3, pT7-CDC-9 P11 NSP4, pT7-CDC-9 P11 NSP5, pT7 CDC-9 P45 VP2, pT7 CDC-9 P45 VP4, pT7 CDC-9 P45 VP7, pT7 CDC-9 P45 NSP1, pT7-CDC-9 P11 VP4_AA51, pT7-CDC-9 P11 VP4_AA331, pT7-CDC-9 P11 VP4_AA385, pT7-CDC-9 P11 VP4_AA499, pT7-CDC-9 P11 VP4_AA385_388, pT7-CDC-9 P11 VP4_AA331_385_388, and pT7-CDC-9 P11 VP4_AA331_364_385_388) was commercially synthesized (GenScript, Piscataway, NJ). Simian SA11 plasmid collection (pT7-SA11_VP1, pT7-SA11_VP2, pT7-SA11_VP3, pT7-SA11_VP4, pT7-SA11_VP6, pT7-SA11_VP7, pT7-SA11_NSP1, pT7-SA11_NSP2, pT7-SA11_NSP3, pT7-SA11_NSP4, and pT7-SA11_NSP5) was originally made by Takeshi Kobayashi (Research Institute for Microbial Diseases, Osaka University, Japan) and obtained from Addgene (44). Plasmid pCMVScript-NP868R-(G4S)4-T7RNAP (C3P3-G1), which autonomously synthesize viral mRNAs containing the main post-transcriptional modifications, was kindly provided by Dr. Jais (Eukarÿs, Evry, France) (45). Plasmid purification was performed using Qiagen EndoFree Plasmid Maxi Kit (Qiagen, Hilden, Germany) according to manufactur er's instructions. Recombinant mono-reassortment viruses on CDC-9 P11 background with different VP4 genes were received from Stanford University, namely, rCDC-9 P11 VP4_RRV, rCDC-9 P11 VP4_SA11, rCDC-9 P11 VP4_UK, and rCDC-9 P11 VP4_P45 (VP4 gene with all six mutations seen in CDC-9 P45 VP4 was used on a CDC-9 P11 back ground).
## Reverse genetics
Recombinant rotavirus particles were generated by an improved reverse genetics protocol as previously described (25). In brief, BHK-T7 cells were seeded in 12-well plates and cultivated until 80% confluency was reached. 0.4 µg of 9 rotavirus plasmids (all except NSP2 and NSP5), 1.2 µg of rotavirus plasmids NSP2 and NSP5, and 0.8 µg of helper plasmids C3P3-G1 were mixed in Gibco OPTI-MEM I Reduced Serum Medium (Fisher Scientific, Waltham, MA, USA) and transfected with TransIT-LT1 (Mirus Bio) into BHK-T7 cells. Twenty-four hours after transfection, medium was replaced with serum-free DMEM. Forty-eight hours after transfection, MA104 cells were added to transfected BHK-T7 cells and cultured in the presence of 0.5 µg/mL trypsin. Rescued viruses were amplified by passaging in MA104 cells up to four times. To generate VP4 mutant viruses, we replaced P11 VP4 or P45 VP4 with the appropriate VP4 plasmids containing single or multiple point mutations. Generated recombinant rotavirus sequences were confirmed by next-generation sequencing.
## Next-generation sequencing of virus strains
Viral RNA from rotavirus-infected cell culture was extracted using the Qiagen Viral RNA Kit according to manufacturer's instructions followed by a DNase I treatment (Qiagen, Hilden, Germany) and cleanup with QIAquick PCR purification kit. Total RNA (250 ng) was used as input for rRNA Depletion Kit (NEB, Ipswich, MA, USA) followed by the NEB Ultra II RNA library preparation kit according to the manufacturer's instructions. Sequencing was performed using an Illumina MiSeq V2 reagent kit 500 cycles (Illumina, San Diego, CA, USA). Data were analyzed by CLC Genomics Workbench (V21 and V22, Qiagen, Hilden, Germany).
## Infection of neonatal rats
Timed-pregnant Lewis rats that were seronegative for RV were purchased from Charles River (Wilmington, MA, USA). One day after birth, neonatal pups were randomly distributed to a litter size of 4-10 pups per mother depending on experiment. Animal numbers for each experiment are given in the respective figure legends. Five-day-old neonatal rats were infected by oral gavage with 1 × 10 7 FFU with the indicated viruses diluted in serum-free IMDM or given mock control (clarified supernatant of uninfected MA104 cells). Animals were checked daily for 14 days post infection for body weight and diarrhea score. Diarrhea scoring was as follows. 0: normal feces; 1: unusually loose, yellow stool; 2: mucus with liquid stool, some loose stool; 3: totally loose yellow-green feces; 4: high amount of watery feces. Rectal swabs were collected in 0.5 mL Premier Rotaclone Diluent daily for 14 days post infection and stored at -65 to -95°C before testing. Shedding was analyzed by testing rectal swabs with Premier Rotaclone EIA (Meridian, Cincinnati, OH, USA) according to the manufacturer's instructions. On day 21 post infection, all animals were euthanized, whole blood collected and allowed to clot at room temperature for 30 min before centrifuging for 10 min at 2,000 × g in a refrigerated centrifuge. Serum was aliquoted into clean tubes for subsequent assays and stored at -80°C.
## Cytokine assay
Cytokine and chemokine expression in rat serum was performed using Bioplex Pro Rat Cytokine 23-plex Assay (G-CSF, GM-CSF, GRO/KC, IFN-γ, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-7, IL-10, IL-12 (p70), IL-13, IL-17A, IL-18, M-CSF, MCP-1, MIP-1α, MIP-3α, RANTES, TNF-α, and VEGF) according to manufacturer's instructions. Samples were analyzed by using BioPlex 200 System (BioRad, Hercules, CA, USA) and Bioplex Manager Software (BioRad, Hercules, CA, USA).
## Statistical analysis
Graphs were generated, and statistical analyses were performed by using GraphPad Prism V10 (GraphPad Software, La Jolla, CA, USA). Two-way analysis of variance (ANOVA) was used to determine statistical significance of rotavirus shedding and body weight gain in neonatal rats. For all other analyses, two-tailed Student's t-test was performed. P values < 0.05 were considered statistically significant and are indicated with an asterisk.
## References
1. Sabin, Hennessen, Winsser (1954) "Studies on variants of poliomyelitis virus. I. Experimental segregation and properties of avirulent variants of three immunologic types" *J Exp Med*
2. Katz, Kempe, Black et al. (1960) "Studies on an attenuated measles-virus vaccine. VIII. General summary and evaluation of the results of vaccination" *Am J Dis Child*
3. Hilleman, Buynak, Weibel et al. (1968) "Live, attenuated mumps-virus vaccine" *N Engl J Med*
4. Plotkin, Farquhar, Katz et al. (1969) "Attenuation of RA 27-3 rubella virus in WI-38 human diploid cells" *Am J Dis Child*
5. Takahashi, Okuno, Otsuka et al. (1975) "Development of a live attenuated varicella vaccine" *Biken J*
6. Bernstein, Smith, Sherwood et al. (1998) "Safety and immunogenicity of live, attenuated human rotavirus vaccine 89-12" *Vaccine (Auckl)*
7. Bhandari, Rongsen-Chandola, Bavdekar et al. (2014) "Efficacy of a monovalent human-bovine (116E) rotavirus vaccine in Indian infants: a randomised, double-blind, placebo-controlled trial" *Lancet*
8. Dang, Nguyen, Vu et al. (2012) "A dose-escalation safety and immunogenicity study of a new live attenuated human rotavirus vaccine (Rotavin-M1) in Vietnamese children" *Vaccine (Auckl)*
9. Troeger, Khalil, Rao et al. (2018) "Rotavirus vaccination and the global burden of rotavirus diarrhea among children younger than 5 years" *JAMA Pediatr*
10. Buttery, Danchin, Lee et al. (2011) "PAEDS/APSU Study Group"
11. Haber, Patel, Pan et al. (2006) "Intussusception after rotavirus vaccines reported to US VAERS" *Pediatrics*
12. Patel, López-Collada, Bulhões et al. (2011) "Intussusception risk and health benefits of rotavirus vaccination in Mexico and Brazil" *N Engl J Med*
13. Ward, Kirkwood, Sander et al. (2006) "Reductions in cross-neutralizing antibody responses in infants after attenuation of the human rotavirus vaccine candidate 89-12" *J Infect Dis*
14. Mattion, Mitchell, Both et al. (1991) "Expression of rotavirus proteins encoded by alternative open reading frames of genome segment 11" *Virology (Auckl)*
15. Crawford, Mukherjee, Estes et al. (2001) "Trypsin cleavage stabilizes the rotavirus VP4 spike" *J Virol*
16. Groome, Fairlie, Morrison et al. (2020) "Safety and immunogenicity of a parenteral trivalent P2-VP8 subunit rotavirus vaccine: a multisite, randomised, double-blind, placebocontrolled trial" *Lancet Infect Dis*
17. Park, Yoon, Park et al. (2021) "Rotavirus spike protein ΔVP8* as a novel carrier protein for conjugate Full-Length Text Journal of Virology November"
18. "vaccine platform with demonstrated antigenic potential for use as bivalent vaccine" *Sci Rep*
19. Roier, Prasad, Mcneal et al. (2023) "mRNA-based VP8* nanoparticle vaccines against rotavirus are highly immunogenic in rodents" *NPJ Vaccines*
20. Xue, Yu, Jia et al. (2016) "Immunogenicity and protective efficacy of rotavirus VP8* fused to cholera toxin B subunit in a mouse model" *Hum Vaccin Immunother*
21. Denisova, Dowling, Shaw et al. (1999) "Rotavirus capsid protein VP5* permeabilizes membranes" *J Virol*
22. Esona, Foytich, Wang et al. (2010) "Molecular characterization of human rotavirus vaccine strain CDC-9 during sequential passages in Vero cells" *Hum Vaccin*
23. Resch, Wang, Moon et al. (2020) "Serial passaging of the Human rotavirus CDC-9 strain in cell culture leads to attenuation: characteriza tion from in vitro and in vivo studies" *J Virol*
24. Kanai, Onishi, Kawagishi et al. (2020) "Reverse genetics approach for develop ing Rotavirus vaccine candidates carrying VP4 and VP7 genes cloned from clinical isolates of human Rotavirus" *J Virol*
25. Kawagishi, Nurdin, Onishi et al. (2020) "Reverse genetics system for a human group A Rotavirus" *J Virol*
26. Sánchez-Tacuba, Feng, Meade et al. (2020) "An optimized reverse genetics system suitable for efficient recovery of simian, human, and murine-like Rotaviruses" *J Virol*
27. Amimo, Raev, Chepngeno et al. (2021) "Rotavirus interactions with host intestinal epithelial cells" *Front Immunol*
28. Kawagishi, Sánchez-Tacuba, Feng et al. (2024) "Reverse genetics of murine rotavirus: a comparative analysis of the wildtype and cell-culture-adapted murine Rotavirus VP4 in replication and virulence in neonatal mice" *Viruses*
29. Valusenko-Mehrkens, Schilling-Loeffler, Johne et al. (2024) "VP4 Mutation boosts replication of recombinant Human/Simian Rotavirus in cell culture" *Viruses*
30. Ward, Mason, Bernstein et al. (1997) "Attenuation of a human rotavirus vaccine candidate did not correlate with mutations in the NSP4 protein gene" *J Virol*
31. Simsek, Bloemen, Jansen et al. (2022) "Rotavirus vaccine-derived cases in Belgium: evidence for reversion of attenuating mutations and alternative causes of gastroenteritis" *Vaccine (Auckl)*
32. Carter, Gribble, Diller et al. (2024) "Human Rotaviruses of multiple genotypes acquire conserved VP4 mutations during serial passage" *Viruses*
33. Guo, Wentworth, Stucker et al. (2020) "Amino acid substitutions in positions 385 and 393 of the hydrophobic region of VP4 may be associated with Rotavirus attenuation and cell culture adaptation" *Viruses*
34. Kim, Trask, Babyonyshev et al. (2010) "Effect of mutations in VP5 hydrophobic loops on rotavirus cell entry" *J Virol*
35. López, Arias (2004) "Multistep entry of rotavirus into cells: a Versaillesque dance" *Trends Microbiol*
36. Dormitzer, Nason, Prasad et al. (2004) "Structural rearrangements in the membrane penetration protein of a nonenveloped virus" *Nature*
37. Tsugawa, Tatsumi, Tsutsumi (2014) "Virulence-associated genome mutations of murine rotavirus identified by alternating serial passages in mice and cell cultures" *J Virol*
38. Tsugawa, Tsutsumi (2016) "Genomic changes detected after serial passages in cell culture of virulent human G1P[8] rotaviruses" *Infect Genet Evol*
39. Jenni, Li, Wang et al. (2022) "Rotavirus VP4 Epitope of a broadly neutralizing human antibody defined by its structure bound with an attenuated-strain virion" *J Virol*
40. Jiang, Snipes-Magaldi, Dennehy et al. (2003) "Cytokines as mediators for or effectors against rotavirus disease in children" *Clin Diagn Lab Immunol*
41. Gómez-Rial, Curras-Tuala, Rivero-Calle et al. (2018) "Rotavirus intestinal infection induces an oral mucosa cytokine response" *PLoS One*
42. Casola, Estes, Crawford et al. (1998) "Rotavirus infection of cultured intestinal epithelial cells induces secretion of CXC and CC chemokines" *Gastroenterology*
43. Ge, Mansell, Ussher et al. (2013) "Rotavirus NSP4 triggers secretion of proinflammatory cytokines from macrophages via Toll-like receptor 2" *J Virol*
44. 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*
45. Kanai, Komoto, Kawagishi et al. (2017) "Entirely plasmid-based reverse genetics system for rotaviruses" *Proc Natl Acad Sci*
46. Jaïs, Decroly, Jacquet et al. (2019) "C3P3-G1: first generation of a eukaryotic artificial cytoplasmic expression system" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548469&blobtype=pdf | # HIV proviral transcription and infectivity are enhanced by neddylation
Cristina Vaca, Hannah Hudson, Isabelle Clerc, Chisu Song, Richard D'aquila
## Abstract
A spreading infection can quickly restart from the persistent reservoir of cells harboring HIV proviruses if antiretroviral therapy (ART) is stopped. HIV transcription can also be increased by latency-reversing agents (LRAs) during ART. Neddylation is a post-translational modification that activates Cullin-RING ligases (CRLs), which ubiqui tinate several proteins that regulate HIV transcription and infectivity, marking these regulators for proteasomal degradation. We studied how inhibiting neddylation affects HIV gene expression ex vivo after three LRAs: TNFα, PMA and ionomycin, and JQ1. In provirus-containing T cells, broad inhibition of CRL-mediated ubiquitination using MLN4924 (MLN) with the above LRAs reduced HIV transcriptional initiation, decreased virus production, and diminished virion infectivity by increasing A3G incorporation. Decreased degradation of inhibitor of kB alpha (IkBα) was implicated in reducing LRA-stimulated HIV transcription. MLN also decreased HIV reactivation after PMA and ionomycin treatment of CD4+ T cells from ART-suppressed people living with HIV. Results indicate that neddylation can enhance HIV proviral transcription and reactivated virion infectivity.IMPORTANCE Results indicate that neddylation contributes to reactivating HIV provirus transcription and antigen expression, as well as enhancing infectivity of resulting virions. This suggests hypotheses to test in the future that may inform a novel strategy for research to enable antiretroviral therapy (ART)-free remission of HIV infection, includ ing whether inhibiting neddylation when ART stops reduces spontaneous provirus reactivation and increases virus A3G content to help control HIV rebound from latent reservoirs post-ART.
ionomycin (PMAi), and BET-bromodomain inhibitor JQ1, which are each well documen ted to increase transcription and virus production from provirus-containing ACH2 and J-Lat cells (3).
Neddylation is a post-translational modification that attaches a ubiquitin-like protein, Nedd8, to lysine residues (4). This modification to any of the eight Cullins in the many Cullin-RING Ligases (CRLs) is necessary to activate the enzymatic function of virtually all CRLs. This enables a CRL to polyubiquitinate its substrate protein, directing it for proteasomal degradation (5). Several cellular proteins that are degraded after activation of CRL ubiquitination have been reported to affect HIV transcription, including inhibitors of NF-kB (6)(7)(8)(9)(10). Therefore, we studied the effects of broadly inhibiting CRL-mediated degradation on LRA-induced provirus transcription and HIV protein expression ex vivo in provirus-containing T cells.
In this work, we used a small molecule Nedd8-activating-enzyme inhibitor, MLN4924 (also known as pevonedistat, referred to here as MLN), to inhibit CRL activity. MLN was developed as an adjunctive therapy for certain cancers and was shown to be safe in multiple clinical trials (11)(12)(13)(14)(15)(16)(17). Among its broad effects, MLN has been shown to block degradation of several proteins critical for de novo HIV infections in T cells and mono cyte-derived macrophages, including those co-opted by HIV accessory gene products Vif, Vpr, and Vpu (18)(19)(20)(21)(22). For example, previous work showed that MLN reduced HIV virion infectivity and spread through a culture after de novo infection by protecting the anti-viral host factor APOBEC3G (A3G) from Cullin-5 (CUL5)-mediated degradation by HIV Vif (18,19). However, prior works were limited to de novo infections of HIV and did not study the effects of inhibiting neddylation on infectivity of viruses reactivated from cells persistently harboring proviruses, the effects on provirus transcription, nor the subsequent production of virus from provirus-containing cells.
Here, we investigated the effect of MLN on provirus-containing cells induced to express and release HIV following treatment with three LRAs: TNFα, PMAi, and JQ1. Given previous data from cancer studies that MLN can reduce NF-kB signaling (23)(24)(25)(26), we studied whether inhibiting neddylation reduced HIV transcription in addition to evaluating its effects on A3G. We characterized how MLN affected HIV transcription following these LRAs in ACH2 and J-Lat cells. Effects on HIV transcription in CD4+ T cells from ART-suppressed PLWH were also studied after treatment with PMAi ex vivo. The amount and infectivity of viruses produced from provirus-containing T-cell lines after each of the three LRAs was also assessed. Results show that inhibiting neddylation limited increases in provirus transcriptional initiation caused by each of these three different LRAs and suggested that MLN decreased the degradation of an inhibitor of NF-kB. MLN also decreased the production and infectivity of Vif+ viruses reactivated from ACH2 cells. These results indicate that neddylation enhances provirus transcription and virus spread caused by these LRAs.
## RESULTS
## Inhibiting neddylation reduced HIV expression in provirus-containing cells after LRAs
We first studied the impact of inhibiting neddylation on the effects of LRAs in two well-characterized HIV provirus-containing T-cell lines: ACH2 and J-Lat cells. ACH2 cells contain a replication-competent provirus, while both the J-Lat 6.3 and 11.1 clones studied here harbor a replication-incompetent provirus encoding a GFP reporter. It is important to note that ACH2 cells have a single-nucleotide mutation in the trans-acti vation response (TAR) region of the HIV long-terminal repeat (LTR). Nevertheless, they maintain functional Tat and responsiveness to TNFα (27). Because MLN caused cell death of cancer cells ex vivo at concentrations from 500 nM to 3 µM (23,28,29), we first studied whether lower MLN concentrations inhibited Cullin neddylation in these HIV provirus-containing cell lines. Cullin 2 (CUL2) neddylation was inhibited at 100 nM in J-Lat 6.3 cells and at 200 nM in both ACH2 and J-Lat 11.1 cells (Fig. 1). We therefore used 200 nM MLN for all experiments shown here.
We started by testing the hypothesis that MLN would decrease the number of cells expressing HIV after LRA stimulation. After pretreatment with 200 nM of MLN, cells were stimulated with TNFα, PMAi, or JQ1 in the continued presence of MLN for an additional 48 hours (Fig. 2A). The percentage of cells expressing intracellular HIV-1 capsid protein, p24, was measured by flow cytometry 48 hours after adding each LRA. In ACH2 cells, MLN reduced the percentage of p24+ cells from 91% with TNFα alone to 46% with MLN plus TNFα (Fig. 2B; P = 0.001). Similarly, in J-Lat 6.3 cells, MLN reduced the percentage of GFP+ cells from 12% following TNFα alone to 3% with both (Fig. 2C; P = 0.03). Immuno blots of ACH2 and J-Lat 6.3 cell lysates also confirmed decreased levels of TNFα-induced intracellular p24 protein with MLN in both cell lines (Fig. 2D).
This phenotype was also seen with the other two LRAs, PMAi and JQ1, using the same experimental workflow. After PMAi treatment, MLN reduced the percentage of p24+ ACH2 cells from 85% with PMAi alone to 69% with PMAi and MLN (Fig. 2E; P = 0.003). After JQ1 treatment, the percentage of p24+ ACH2 cells was decreased from 53% to 33% by adding MLN (Fig. 2F; P = 0.04). This trend was consistent across addi tional experiments, not shown here, using various other J-Lat clones (e.g., 11.1 and 5a8) stimulated with either TNFα or PMAi. Overall, we observed that inhibiting neddylation with MLN reduced the percentage of cells expressing intracellular p24 or GFP after LRA stimulation across multiple cell lines and following three different LRAs. This indicates that a reduction in provirus reactivation when neddylation is inhibited is not specific to a single cell line, integration site, or LRA.
Next, the amount of extracellular p24 in the cell culture supernatant of ACH2 cells was quantified using an HIV p24 ELISA to assess how much virus was produced from these cells (Fig. 3). Because J-Lat cells do not produce replication-competent virus, this and subsequent experiments assessing virus production and infectivity focused only on ACH2 cells. MLN significantly reduced the amount of extracellular p24 produced from ACH2 cells treated with TNFα (P = 0.0004), PMAi (P = 0.0025), or JQ1 (P < 0.0001), compared to the amount produced after each LRA alone (Fig. 3A through C). Together, these initial experiments showed that inhibiting neddylation reduced the number of ACH2 or J-Lat cells newly expressing HIV after these LRAs and also reduced the amount of virus released from ACH2 cells.
## Inhibiting neddylation reduced HIV gag RNA
To explore the mechanism underlying reduced p24 expression and release after treatment with the three LRAs studied here, we assessed the impact of inhibiting neddylation via MLN on provirus transcription. We measured HIV gag gene expression in ACH2 cells following each LRA, with and without MLN. In ACH2 cells treated with TNFα and MLN, MLN reduced TNFα-stimulated gag RNA expression sixfold (P = 0.0003) compared to TNFα alone (Fig. 4A). MLN reduced gag RNA expression by fivefold (P < 0.0001) with PMAi treatment (Fig. 4B), and by sevenfold (P < 0.0001) with JQ1 treatment (Fig. 4C).
To determine whether effects on subsequent rounds of infection (e.g., superinfection) contributed to the reduction in gag RNA by MLN (Fig. 4), we cultured ACH2 cells in the presence and absence of saquinavir (Fig. 5). This HIV protease inhibitor specifically blocks cleavage of the Gag polyprotein, which is necessary for the maturation of newly released virions into infectious virions. Extracellular p24 decreased in the presence of saquinavir, whether MLN was also present or not, indicating that saquinavir effectively blocked processing of Gag polyprotein into p24 in both conditions (Fig. 5A). In the presence of saquinavir, MLN still reduced TNFα-stimulated HIV gag RNA (Fig. 5B). Thus, MLN reduced HIV gag RNA expression in cells stimulated by TNFα even when spreading infection was blocked by saquinavir. Overall, these data show that inhibiting neddylation hinders the induction of HIV gag gene expression by these three LRAs in ACH2 provirus-containing T cells.
## Inhibiting neddylation reduced HIV RNA by decreasing transcriptional initiation following TNFα, PMAi, or JQ1
Next, we tested whether the initiation or elongation of HIV transcription was responsible for this reduction in HIV RNA by quantifying transcripts associated with these steps of HIV transcription. Transcription of full-length HIV RNA by RNA polymerase II begins at the 5′ LTR promoter and pauses at the TAR stem-loop region in what is known as promoter proximal pausing (31)(32)(33). Blocks in elongation can cause RNA polymerase II not to extend beyond the TAR region, preventing elongation of the remaining transcript. Thus, the level of TAR transcripts serves as a measure of HIV transcriptional initiation, whereas the level of transcripts that include the region just downstream of TAR ("Long LTR") quantifies elongation past the TAR region (34,35). Prior literature has established that a greater than twofold excess of TAR over Long LTR transcripts in cells treated with both an LRA and MLN would signify that MLN inhibits HIV transcriptional elongation. Reduced TAR RNA levels with MLN, relative to LRA alone, would indicate that MLN decreased initiation of HIV transcripts (31,34,35). With MLN treatment, TAR and Long LTR transcript levels were comparable to each other, with no statistically significant difference between the two transcripts in both ACH2 and J-Lat 6.3 cells treated with LRAs plus MLN (Fig. 6A through D). This indicates that there was no block to HIV transcript elongation with MLN. Conversely, when cells were treated with TNFα, the level of TAR RNA was significantly reduced with MLN compared to TNFα alone (P = 0.008 for ACH2, P = 0.004 for J-Lat 6.3), consistent with a block in initiation when neddylation is inhibited (Fig. 6A andB). This was also true in ACH2 cells treated with PMAi (P < 0.001) or JQ1 (P = 0.002) plus MLN, compared to the LRA-only control (Fig. 6C andD).
## Impact of neddylation on inhibitor of kB-alpha (IkBα) and LRA-driven HIV transcription
The results above suggested that these three LRAs share a common mechanistic pathway, at least in part, whose stimulation of HIV transcription was diminished by neddylation inhibition. We tested whether CRL-mediated degradation of a negative regulator of the canonical NF-kB pathway was impacted by MLN. In the canonical pathway, NF-kB is sequestered by inhibitor of kB-alpha (IkBα) in the cytoplasm (36,37). When IkBα is phosphorylated by IkB kinase (IKK), the resulting p-IkBα must be ubiquitinated by a neddylation-activated CUL1 ubiquitin ligase (CRL β-TrCP ), directing it to the proteasome (Fig. 7). After degradation of p-IkBα, NF-kB (RelA and p50) can then translocate to the nucleus to initiate transcription from the HIV LTR (6, 38, 39) (Fig. 7). Therefore, we tested the hypothesis that inhibiting neddylation would increase p-IkBα levels in ACH2 cells. Levels of p-IkBα were increased by 200 nM MLN, relative to the LRA-only control, after stimulation with TNFα (P < 0.08), PMAi (P < 0.08), or JQ1 (P < 0.6) (Fig. 8A through C). Given these suggestive results, two independent approaches were then taken to assess whether MLN impacted LRA-stimulated HIV transcription via the NF-kB pathway.
A well-characterized, highly selective IKK inhibitor, BMS-345541 (BMS), was used here at 5 µM to inhibit the phosphorylation of IkBα, which is necessary for CRL-mediated degradation of p-IkBα (IKK-1 IC 50 = 4 µM, IKK-2 IC 50 = 0.03 µM) (41). 5 µM BMS signif icantly reduced TNFα-induced HIV gag RNA, relative to control (Fig. 9A; P < 0.0001). HIV RNA induced by either PMAi or JQ1 was also reduced by BMS as a single agent, albeit to a lesser degree that did not reach statistical significance (Fig. 9B andC). Of note, TNFα is documented to act on HIV transcription primarily, if not entirely, through NF-kB (27,43). By contrast, PMAi and JQ1 are reported to stimulate HIV transcription via additional mechanisms as well as via NF-kB, as detailed in the discussion (44)(45)(46)(47). In the presence of BMS, HIV gag RNA stimulated by each of these 3 LRAs was not significantly further reduced by adding MLN to the LRA and BMS (Fig. 9A through C). That is to say, BMS inhibition of IKK, which blocks formation of p-IkBα, precluded a further reduction in LRA-stimulated HIV transcription by MLN. No significant difference was seen in supernatant p24 levels between 5 µM BMS alone versus 200 nM MLN alone after TNFα stimulation (Fig. 9D).
A different inhibitor of the NF-kB pathway, which acts downstream of BMS to prevent nuclear translocation of RelA/p65 released by IkBα degradation, JSH-23 (JSH), was also studied (42). Extracellular HIV p24 was also similarly reduced by sole treatment with either 200 nM MLN or 50 µM JSH (Fig. 9E). Taken together, these results support the hypothesis that MLN acted via inhibition of the NF-kB pathway to reduce HIV transcrip tion and subsequent p24 production.
## Inhibiting neddylation in CD4 T cells from ART-suppressed PLWH reduced HIV multiply spliced RNA
We next expanded upon the results we found using cell lines by assessing PMAi-driven HIV provirus transcription in primary CD4+ T cells from ART-suppressed PLWH. Due to the low frequency of provirus-infected cells in the blood of ART-suppressed PLWH (48,49), we used the highly sensitive Tat/rev Induced Limiting Dilution Assay (TILDA) to quantify PMAi-driven transcription ex vivo in cells from three donors (50,51). After negative selection of CD4+ T cells from cryopreserved peripheral blood mononuclear cells (PBMCs), CD4+ T cells were stimulated for a total of 12 hours with PMAi as is standard in this assay (50,51). Cells were treated with dimethyl sulfoxide (DMSO) or 200 nM of MLN starting 4 hours after PMAi stimulation. After 12 hours, cells were collected and multiply spliced tat/rev RNA (msRNA) was measured using a plate-based digital PCR platform (Fig. 10A).
MLN treatment reduced msRNA after PMAi stimulation of cells from all three donors (Fig. 10B). PMAi-driven provirus transcription was reduced with MLN by 35%, 67%, and 43%, respectively, in each donor's cells compared to DMSO control. As expected, variation in HIV reservoir size across individuals resulted in differing magnitudes of msRNA between donors. More information on each donor is found in Table 1. In summary, inhibiting neddylation limited increases in provirus transcription after PMAi stimulation in this primary cell model, supporting the relevance of results in cell lines.
## Inhibiting neddylation protected APOBEC3G from degradation in ACH2 cells and decreased infectivity of LRA-stimulated virions
We have shown thus far that inhibiting neddylation via MLN curtailed increases in the number of HIV expressing cells, virus production, HIV gag RNA, and HIV transcrip tional initiation after LRA treatment in provirus-containing T-cell lines. MLN also reduced PMAi-driven provirus transcription in CD4+ T cells from ART-suppressed PLWH. Given that MLN has been demonstrated to reduce infectivity of virions in de novo HIV infections (18)(19)(20), we next studied the infectivity of viruses produced from LRA-stimulated ACH2 cells containing Vif-positive provirus.
If A3G escapes degradation in HIV-producing cells and is packaged into virions, it potently reduces virion infectivity (53)(54)(55). We tested whether Vif-positive virions produced from ACH2 cells after LRA stimulation, in the presence of MLN that inhibits activation of CRL Vif , would be less infectious due to increased A3G incorporation (18,19). We pelleted virus from the supernatant of ACH2 cells that were pretreated with MLN or DMSO prior to TNFα stimulation (as in Fig. 2A). MLN resulted in higher A3G incorporation in virions with and without TNFα treatment (Fig. 11A). The infectivity of virions was studied by infecting TZM-bl cells with equal amounts of supernatant virus from ACH2 cells with and without MLN treatment and assessing the relative luminescence units (RLU) (56). Vif-positive virions produced from MLN and LRA-treated ACH2 cells had reduced virion infectivity compared to virions from LRAtreated cells without MLN exposure (Fig. 11B andD). Infectivity of virions from TNFα-trea ted ACH2 cells was reduced by 67% (P = 0.002) with 100 nM MLN and 85% (P < 0.0004) with 200 nM MLN (Fig. 11B). MLN reduced the infectivity of virions induced by PMAi by 61% (P = 0.007) with 100 nM MLN and 59% (P = 0.008) with 200 nM (Fig. 11C). With JQ1 treatment, MLN reduced infectivity by 39% (P = 0.007) with 100 nM MLN and 45% (P = 0.003) with 200 nM MLN (Fig. 11D). Taken together, MLN increased A3G incorpora tion into virions released from ACH2 cells after each of these three LRAs, decreasing infectivity. Thus, inhibiting neddylation also limited infection spread in provirus-contain ing cell cultures.
## DISCUSSION
This study identified that neddylation increased LRA-stimulated proviral HIV transcription by broadly inhibiting neddylation using MLN. With that same approach, neddylation was also found to decrease A3G incorporation into LRA-reactivated HIV virions. The ex vivo results will be summarized. Hypotheses suggested by this study that could be tested in future animal model research using MLN at the time ART is stopped, rather than with an LRA during ART, will be discussed. Rationale for future discovery of targeted alternatives to broadly acting MLN will also be described below.
First, MLN limited LRA-driven HIV transcription and virus production by TNFα, PMAi, and JQ1 in provirus-containing ACH2 and J-Lat 6.3 T-cell lines, diminishing the inten ded effects of an LRA (Fig. 2 to 4). MLN likely impacted the cells initially reactivating provirus because LRA-stimulated HIV RNA and p24 antigen production were each also decreased when saquinavir blocked HIV spread in ACH2 cultures (Fig. 5). MLN also decreased PMAi-induced transcription in CD4+ T cells from ART-suppressed PLWH (Fig. 10), indicating relevance to primary cells. Provirus transcription stimulated by each of these LRAs was decreased at the initiation step by MLN (Fig. 6), raising the hypothesis (20,40). BMS inhibits phosphorylation of IkBα by IKK (41). JSH inhibits RelA/p65 translocation (42) Several lines of evidence presented in this study support that a common mechanism among the three LRAs was an MLN-mediated decrease in degradation of an inhibitor of NF-kB, p-IkBα (Fig. 7). p-IkBα levels were non-significantly increased by MLN (Fig. 8). BMS, which inhibits phosphorylation of IkBα by IKK (41), and JSH, which inhibits nuclear translocation of RelA/p65 (42), each individually diminished HIV transcription and precluded a further reduction in LRA-stimulated HIV transcription by adding MLN (Fig. 9). Additional support came from the differing magnitudes by which MLN decreased HIV expression stimulated by each of the three LRAs. MLN decreased the number of p24+ cells and extracellular p24 levels more when given with TNFα treatment than the other two LRAs (Fig. 9A through C). These results are consistent with prior reports that PMAi and JQ1 each act to increase HIV transcription via separate mechanisms in addition to NF-kB. PMA acts via NF-kB (44), while ionomycin increases NFAT signaling to further activate transcription at the HIV LTR (45). JQ1 has been previously reported to increase HIV transcriptional activation via several mechanisms: epigenetic effects, enhancement of pTEFb recruitment to TAR by HIV Tat, as well as indirectly increasing NF-kB tran scription (46,47,(57)(58)(59)(60)(61). Taken together, these results suggest that MLN is decreasing HIV expression after TNFα, PMAi, or JQ1 via inhibition of canonical NF-kB signaling.
Determining whether NF-kB binding to the HIV LTR is decreased when neddylation is inhibited in the presence of each of these single LRAs, relative to the LRA alone, will help validate this hypothesis in the future. CRL β-TrCP also degrades a protein with a similar function as p-IkBα in the non-canonical NF-kB pathway, p100 (62,63). This suggests future study of MLN's effects on non-canonical NF-kB signaling. a CD4+ T cells came from ART-suppressed, de-identified PLWH participating in the RADAR cohort (52). Undetectable plasma viral load was defined as <40 copies/mL. ND = not determined. Seroconverted refers to RADAR study participants who seroconverted during the RADAR study and were subsequently put on ART treatment.
In addition to its effects on HIV transcription, broadly inhibiting neddylation with MLN was also found to block A3G degradation in LRA-stimulated ACH2 cells, reactivating Vif-positive virus production. MLN increased A3G packaging into virions and potently diminished infectivity of those virions (Fig. 11). This extended previous findings from previous studies of de novo HIV infections (18,19). This is relevant because others have reported that CD4+ T cells harboring A3G-hypermutated, replication-defective proviruses caused by infection with virions containing increased A3G can more potently activate HIV-specific CD8+ cytolytic T cells (CTL) than do CD4+ T-cell proviruses lacking this magnitude of A3G hypermutation. This has been seen in both ex vivo and in vivo experiments (64,65). Other effects reported in previous studies also impacted anti-HIV immunity. DNA repair mechanisms induced by A3G-hypermutated HIV genomes upregulated expression of a ligand on CD4 T cells that activates natural killer (NK) cells, specifically increasing NK cell-mediated killing of HIV-infected CD4+ T cells containing such A3G-hypermutated proviruses (66). Dendritic cells infected with lethally A3G-hyper mutated HIV have also been reported to more effectively present HIV antigens to CD8+ CTL to enhance their specific recognition/cytolysis of those A3G-hypermutated HIV-expressing CD4+ T cells (67). Others also verified that antigens are expressed from defective HIV proviruses and recognized by CD8+ CTLs to shape the dynamic changes in the proviral landscape in CD4+ T cells in vivo over time (68).
Several new research directions are suggested by our results, including ex vivo studies if the broad effects of MLN also decrease Vpu-mediated degradation of tetherin and/or Vpr-mediated degradation of cellular factors impacting cell cycle, apoptosis, and/or transcription (69)(70)(71)(72). The combined effects of MLN on both HIV transcription and A3G virion incorporation, and the HIV-specific immunity-enhancing effects of increased A3G, summarized above (64)(65)(66)(67)(68), suggest animal model research evaluating whether these effects can help control HIV rebound from latent reservoirs off-ART. A current paradigm in HIV cure research, often called "shock and kill, " adds LRA(s) to ART. An LRA aims to expose newly HIV-expressing cells to clearance by immune effector cells, while ART protects against reactivated HIV spread. However, administering LRA(s) during ART has not yet led to substantial reactivation of HIV antigen expression or detectable clearance of newly HIV-expressing cells by immunity. In contrast to adding an LRA to ART, stopping ART reliably leads to robustly rebounding viremia, reflecting a spreading infection. We hypothesize that MLN given when ART stops in an animal model of HIV infection may add to lessons learned here about leveraging mechanisms for a different paradigm than using LRA(s) during ART: attempting to use MLN to curtail HIV rebound from latency when ART stops. Studying MLN when ART stops in animal models can test whether and how uninfected CD4+ T cells may be protected from infection by increased levels of virion A3G and/or if CD4+ T or dendritic cells expressing A3G-hyper mutated HIV genomes may be better recognized and eliminated by immune effector cells (CD8+ T and NK cells), as described previously (64)(65)(66)(67)(68). This hypothesized effect of A3G to increase anti-HIV immunity is supported by reports that adding an adjuvant that increases cellular A3G improved innate and adaptive immune responses to SIV vaccines in rhesus macaques (73)(74)(75). Such animal model research using MLN is envisioned as only a first step in developing such a new research paradigm, if it validates these hypotheses.
Although useful for characterizing the above mechanisms in animal models, MLN's broad effects on inhibiting neddylation-activated ubiquitination of many cellular proteins (23-25, 40, 76) and its potential cytotoxicity on uninfected cells limit its potential for clinical application in HIV. A risk of cytotoxicity that is acceptable in those with cancer does not seem appropriate for ART-treated persons with HIV. Development of more specific, less cytotoxic inhibitors of one or a few CRLs is preferable and may become possible. New knowledge explicates how, specifically and notably, the NEDD8 modification of CUL1 promotes the transfer of ubiquitin to p-IkBα by CRL β-TrCP (77), potentially enabling more specific targeting of that CRL to either increase or decrease its ubiquitinating activity. Molecular understanding of how neddylation activates CUL1 in CRL β-TrCP may also inform targeting of some of the other eight Cullins, which scaffold regulation of many additional protein substrates. This, and efforts to improve broad neddylation inhibitors for cancer treatment (4), could lead to new compounds targeting specific CRLs and/or their targets. Lead compounds that can either increase or decrease the degradative capacities of specific CRLs by modulating neddylation or its effects on CRL ubiquitinating activity may emerge as candidates with more specificity than MLN and with less potential for uninfected cell cytotoxicity. We suggest that more specific agents than MLN be available for study before refining operational protocols for this proposed new approach for sustaining HIV remission that we hypothesize will enhance anti-HIV immunity when ART stops, first in animal models and then in any potential clinical research.
This hypothesis-generating study is also limited in not providing enough pre-clinical data with the lower concentration of MLN used here to yet enable advancing such concepts. Longer-term evaluations of MLN in both ex vivo and animal model studies of HIV infection will be needed to enable animal experiments. Another limitation is that the separation and study of different T-cell subtypes, or other immune cells implicated as HIV reservoirs, was not possible given the small number of blood cells available from ART-suppressed PLWH. In summary, this work highlights neddylation as a targetable mechanism increasing HIV transcription. MLN reduced latency reversal, including in blood CD4+ T cells from ART-suppressed PLWH. Results suggest that MLN prevented degradation of p-IkBα to limit NF-kB-driven HIV transcription. MLN also increased reactivated virion A3G. This result, and earlier studies showing that increased virion A3G enhanced anti-HIV immunity (64)(65)(66)(67)(68), suggests testing whether using MLN when latent HIV spontaneously reactivates after stopping ART can help control HIV rebound off-ART. Studying this in an animal model of HIV infection can begin to validate whether specific proteins and pathways affected by MLN in this study may contribute to the development of a new research strategy to sustain HIV remission off-ART. While broad inhibition of neddylation is unlikely to be clinically applicable to HIV management, the discovery of an agent specifically impacting one or a few proteins identified by studying MLN, such as IkBa or A3G, may help advance such an approach.
## MATERIALS AND METHODS
## Cell culture and cell lines
Suspension cell lines were cultured in RPMI with L-glutamine (Corning) plus 10% fetal bovine serum (FBS), 50 IU/mL penicillin, and 50 μg/mL streptomycin (Corning, #30-002-CI) and maintained at 37°C and 5% CO 2 . ACH2 cells (ARP-349) and J-Lat 6.3 (ARP-9846) were obtained from BEI Resources (formerly the NIH HIV Reagent Program). J-Lat 11.1 cells, originally developed by Dr. Eric Verdin (78), were a generous gift from Dr. Steven Wolinsky at Northwestern University. TZM-bl cells (ARP-8129) were obtained from BEI Resources and cultured in DMEM (Corning) with 10% FBS, 50 IU/mL penicillin, and 50 μg/mL streptomycin as above.
## Reactivation of provirus-containing T-cell lines and LRAs
Provirus-containing T-cell lines (ACH2 or J-Lat) were cultured at 2 × 10 6 cells/mL and pretreated for 6 hours with 100 or 200 nM of MLN (Cayman Chemical, #15217) or equivalent amounts of DMSO. MLN was replenished every 24 hours throughout the course of the experiment (three times total, including the initial pretreatment) (see Fig. 2A). After 6 hours of pretreatment, cells were stimulated with one of the following LRAs: 10 ng/mL TNFα (PeproTech, #300-01A), a 1:125,000 dilution of PMA and ionomy cin (PMAi) (eBioscience Cell Stimulation Cocktail, #00-4970-93), or 100 nM JQ1 (Tocris, #4499). Cells treated with TNFα were cultured in RPMI with only 1% FBS (plus penicillin and streptomycin as above). Controls were treated with equivalent amounts of DMSO (PMA/i, JQ1) or water (TNFα). Cells and supernatants were collected for downstream assays 48 hours after LRA treatment, unless otherwise specified.
## Flow cytometry
All experiments were conducted on the BD Biosciences LRS Fortessa Analyzer. Cells were collected 48 hours after LRA stimulation and stained with either Ghost Dye Red 710 Viability Dye (Tonbo Bioscience, #13-0871-T100) or LIVE/DEAD Fixable Blue Dead Cell Stain Kit (Invitrogen, #L34962) to measure cell viability. ACH2 cells were fixed with 4% paraformaldehyde for 30 minutes, then permeabilized and stained with anti-p24 KC57-FITC (Beckman Coulter, #6604665). Analysis was performed by gating on live cells followed by gating on GFP+ cells (J-Lats) or p24+ cells (ACH2). The percentage of p24+ or GFP+ cells in each condition was plotted. Statistical analysis was done in Graphpad Prism. Conditions were compared using Repeated Measures (RM) one-way ANOVA with the Geisser-Greenhouse Correction. Turkey's multiple comparisons test, with individual variances, was computed for each comparison.
## Immunoblotting
Cell pellets containing the same number of cells across conditions were collected 48 hours after treatment with LRAs and were lysed in a 1% NP-40-based lysis buffer with protease inhibitor (Roche, #11836145001) and subsequently sonicated to lyse the nuclear membrane as we have previously described (79). Cells were centrifuged at 10,000 × g for 12 min at 4°C and denatured using 10% DTT (Sigma, D0632) and 4× Bolt LDS sample buffer (Thermo Fisher Scientific, #B0008).
To pellet viruses, culture supernatant was filtered through a 0.45 µm filter, layered over a 20% sucrose cushion, and ultracentrifuged at 32,000 rpm at 4°C for 1.5 hours (Beckman Coulter SW41Ti). Pellets were resuspended in buffer (20 µL Bolt LDS sample buffer, 8 µL DTT, and 12 µL lysis buffer) and transferred to Eppendorf tubes.
All samples were then boiled for 7 minutes at 92°C and underwent electrophore sis through a 4%-12% Bolt Bis-Tris gel (Invitrogen). Proteins were transferred to a PVDF membrane (Thermo Scientific, #88518) and probed with primary antibodies. The following antibodies were used: anti-CUL2 (Santa Cruz, #166506), anti-phospho-IkB alpha (Proteintech, #82349-1-RR), anti-p24 (183-H12-5C, Tennessee Center for AIDS Research Virology Core), anti-lamin B1 (Proteintech, #66095-1) or anti-lamin B1 (GeneTex, #GTX103292), anti-beta-actin (Proteintech, #66009), and anti-A3G C17 (ARP-10082, polyclonal anti-human APOBEC3G, obtained from BEI Resources, formerly NIH HIV Reagent Program). Fluorescently labeled secondary antibodies (Li-Cor #926-68071 and #926-32210) were used in conjunction with the Li-Cor Odyssey CLx imaging system to visualize membranes. Quantification of immunoblots was done using Li-Cor Empiria Studio 3.2. Quantification of bands was done by dividing the signal of the band of interest by the signal of the corresponding loading control. Values were then normalized to the DMSO/no treatment control.
## Gene expression analysis
Total RNA was isolated from cells using the Qiagen RNeasy Mini Kit (#74106) according to the manufacturer's protocol. cDNA was immediately synthesized from extracted RNA using Invitrogen SuperScript III first-strand synthesis kit (#18080-051) according to the manufacturer's instructions. Following cDNA synthesis, viral transcripts were assessed by RT-qPCR. For HIV gag, the following primers were used: F: 5′ TGCTATGTCAGTTCCCCTTG GTTC TCT 3′, R: 5′ AGTTGGAGGACATCAAGCAGCCATCGAAAT 3′. The cycling conditions were as follows: 95°C for 15 min followed by 40 cycles of 95°C for 15 s, 57°C for 30 s, and 72°C for 30 s. HIV TAR and Long LTR were assessed based on this previously described method (35). TAR was assessed using F: 5′ GTCTCTCTGGTTAGACCAG 3′, R: 5′ TGGGTTCC CTAGYTAGCC 3′, and Long LTR using F: 5′ GCCTCAATAAAGCTTG CCTTGA 3′, R: 5′ GGGC GCCACTGCTAGAGA 3′. The cycling conditions were 95°C for 5 min followed by 40 cycles of 95°C for 15 s, 60°C for 30 s, 72°C for 1 min. RT-qPCR was performed using SYBR Green Low ROX qPCR Mix (Thermo Scientific #AB-1323A) and analyzed using a QuantStudio 6 Flex Real-Time PCR system. Gene normalization was performed using 18S, and ROX was the internal control. The Relative Quantification (RQ) of gene expression was calculated using the following formula: 2 (-ddCT) , where ddCT is the difference between the dCT(gene of interest) and the dCT(control gene). The dCT(gene of interest) is the average Ct value (of three technical replicates) of the gene of interest minus the average Ct value of 18S in that sample. 18S was used as the control gene for all experiments except for those in which cells were treated with saquinavir. In these experiments, GAPDH was used as the control. The following primers were used as described in (80): F: 5′ CTCTGCTCCTCCTGTT CGAC 3′; R: 5′ AGTTAAAAGCAGCCCTGGTGA 3′. An ordinary one-way ANOVA was used to compare conditions. Turkey's multiple comparisons test, with a single pooled variance, was used to derive P values.
## Additional drug treatments
Saquinavir treatment of ACH2 cells was conducted as described above (Fig. 2A) plus the addition of 5 µM saquinavir (MedChemExpress, HY-17007) at time zero, along with the initial MLN treatment. Saquinavir was not replenished after the initial dose. When measuring RNA expression in cells treated with Saquinavir, GAPDH was used as the control in place of 18S for RT-qPCR. JSH-23 (TargetMol, #749886-87-1) and BMS-345541 (TargetMol, #445430-58-0) were added at the concentration noted in figure legends at time zero, along with initial MLN treatment, and not replenished.
## Evaluation of HIV msRNA in CD4+ T cells from ART-suppressed PLWH (TILDA)
Frozen, banked PBMCs were selected from the RADAR cohort at Northwestern University (52). Donors who had been on ART for >3 years or who seroconverted during the study (and thus were immediately put on medication) were undetectable at the time of sample collection (<40 copies/mL), and self-reported good adherence to the ART regimen and lacked substance use were selected.
PBMCs were thawed and rested at 37°C for 2 hours prior to negative selection of CD4+ T cells using the EasySep Human CD4+ T-cell Isolation Kit (StemCell Technologies, #17952). Cells were seeded at a density of 2 × 10 6 cells/mL and rested in complete media for an additional 3 hours prior to cellular activation with 500× PMA and ionomycin (eBioscience Cell Stimulation Cocktail, #00-4970-93) or DMSO control for a total of 12 hours. After 4 hours, non-stimulated and stimulated cells were divided in half and treated with either DMSO (control) or 200 nM of MLN.
After 12 hours of stimulation, cells were washed and pelleted in a benchtop centrifuge, and tat/rev msRNA was measured according to the previously defined TILDA v2.0 protocol (51). In brief, cells were diluted to 3 × 10 5 -1×10 5 cells/mL in RPMI. From each aliquot, 10 µL of the cell suspension was distributed to 24 wells of a 96-well plate containing 2 µL One-Step RT-PCR enzyme (Qiagen, #210212), 10 µL 5× One-Step buffer, 10 µL 0.3% Triton-x (Fisher BP151-500), 0.25 µL 40 U/µL RNAsin (Promega N2111), 2 µL dNTPs (10 mm each), 1 µL 20 µM forward primers (tat1.4 5′ TGGCAGGAAGAAGCGGAG 3′, 1 µL 20 µM reverse primer (rev 5′ GGATCTGTCTCTGTCTCTCTCTCCACC 3′), and nucleasefree water (50). To increase sensitivity, for donors B and C, RNA was extracted first using the Qiagen RNeasy Mini Kit (#74106) from each of the 24 wells individually, and 10 µL of RNA was added to the one-step PCR. One-step RT-PCR was run on a thermocycler using the following conditions: 50°C for 30 min, 95°C for 15 min, 25 cycles of 95°C for 30 s, 55°C for 1 min, 72°C for 2 min followed by 72°C for 5 min. After reverse transcription, 2 µL from each of the 24 wells was pooled across each condition and run on the QuantStudio Absolute Q Digital PCR System to give an absolute quantification of msRNA. 1 µL of sample was added to 2 µL Absolute Q 5X Master Mix (Applied Biosciences), 5 µM Probe HIV, 20 µM tat2.0, 20 µM reverse primer, and dH 2 O, and run with the following cycling conditions: 45 cycles of 95°C for 23 s, followed by 30 s at 60°C (50).
## p24 ELISA
Supernatant from ACH2 cells was collected and clarified by either spinning down and discarding the pellet or by straining through a 0.45 µm filter. Extracellular p24 capsid protein was measured in triplicate via sandwich ELISA using commercially available Alliance HIV-1 P24 Antigen Kit (NEK050B001KT) according to the manufactur er's protocol. An ordinary one-way ANOVA was used to compare conditions. Turkey's multiple comparisons test, with a single pooled variance, was used to derive P values.
## Infectivity assays
ACH2 cells were treated with LRAs and MLN as described in Fig. 2A. ACH2 cell superna tant was collected 48 hours after LRA reactivation and normalized for virus content (as determined by p24 ELISA). To test the infectivity of these viruses, the TZM-bl reporter cell line was plated at a density of 10,000 cells/well in a 96-well plate. 24 hours later, the media was replaced, and cells were infected with equivalent amounts of virus from ACH2 cell supernatant in triplicate wells. The following morning, the media was replaced, and cells were incubated for another 24 hours. 48 hours post-infection, cell luminescence was measured using the Britelite plus Reporter Gene Assay System (Perkin Elmer, #50-209-9084) according to the manufacturer's protocol. Infectivity was determined based on the relative luciferase activity downstream of the HIV LTR in TZM-bl cells. Infectivity was calculated by normalizing the average relative luminescence units (RLU) to the DMSO control. Relative infectivity was calculated by setting the LRA-only control to 100% infectivity and calculating the percent change in infectivity compared to this condition. An ordinary one-way ANOVA was used to compare conditions. Turkey's multiple comparisons test, with a single pooled variance, was used to derive P values.
## References
1. Finzi, Blankson, Siliciano et al. (1999) "Latent infection of CD4+ T cells provides a mechanism for lifelong persistence of HIV-1, even in patients on effective combination therapy" *Nat Med*
2. Wong, Hezareh, Günthard et al. (1997) "Recovery of replication-competent HIV despite prolonged suppression of plasma viremia" *Science*
3. Spina, Anderson, Archin et al. (2013) "An in-depth comparison of latent HIV-1 reactivation in multiple cell model systems and resting CD4+ T cells from aviremic patients" *PLoS Pathog*
4. Zhang, Yu, Li et al. (2024) "Protein neddylation and its role in health and diseases" *Signal Transduct Target Ther*
5. Duda, Borg, Scott et al. (2008) "Structural insights into NEDD8 activation of cullin-RING ligases: conformational control of conjugation" *Cell*
6. Fernandez, Zaikos, Khan et al. (2013) "Targeting IκB proteins for HIV latency activation: the role of individual IκB and NF-κB proteins" *J Virol*
8. Deshmane, Amini, Sen et al. (2011) "Regulation of the HIV-1 promoter by HIF-1α and Vpr proteins" *Virol J*
9. Duette, Gerber, Rubione et al. (2018) "Induction of HIF-1α by HIV-1 infection in CD4 + T cells promotes viral replication and drives extracellular vesicle-mediated inflammation" *mBio*
10. Zhuang, Pedroza-Pacheco, Nawroth et al. (2020) "Hypoxic microenvironment shapes HIV-1 replication and latency" *Commun Biol*
11. Jiang, Espeseth, Hazuda et al. (2007) "C-Myc and Sp1 contribute to proviral latency by recruiting histone deacetylase 1 to the human immunodeficiency virus type 1 promoter" *J Virol*
12. Shah, Jakubowiak, Connor et al. (2016) "Phase I study of the novel investigational NEDD8-activating enzyme inhibitor pevonedistat (MLN4924) in patients with relapsed/refractory multiple myeloma or lymphoma" *Clin Cancer Res*
14. Murthy, Saliba, Szabo et al. (2024) "A phase I study of pevonedistat, azacitidine, and venetoclax in patients with relapsed/refractory acute myeloid leukemia" *Haematologica*
15. Foster, Reid, Minard et al. (2024) "Phase 1 study of NEDD8 activating enzyme inhibitor pevonedistat in combination with chemotherapy in pediatric patients with recurrent or refractory solid tumors (ADVL1615)" *Eur J Cancer*
16. Sekeres, Watts, Radinoff et al. (2021) "Correction to: randomized phase 2 trial of pevonedistat plus azacitidine versus azacitidine for higher-risk MDS/CMML or low-blast AML" *Leukemia*
17. Wong, Micel, Selby et al. (2017) "Targeting the protein ubiquitination machinery in melanoma by the NEDD8-activating enzyme inhibitor pevonedistat (MLN4924)" *Invest New Drugs*
18. Bhatia, Pavlick, Boasberg et al. (2016) "A phase I study of the investiga tional NEDD8-activating enzyme inhibitor pevonedistat (TAK-924/ MLN4924) in patients with metastatic melanoma" *Invest New Drugs*
19. Qin, Wells, Malhotra et al. (2024) "A phase II trial of pevonedistat and docetaxel in patients with previously" *Clin Lung Cancer*
20. Stanley, Bartholomeeusen, Crosby et al. (2012) "Inhibition of a NEDD8 cascade restores restriction of HIV by APOBEC3G" *PLoS Pathog*
21. (2025) *Full-Length Text Journal of Virology*
22. Nekorchuk, Sharifi, Furuya et al. (2013) "HIV relies on neddylation for ubiquitin ligase-mediated functions" *Retrovirology (Auckl)*
23. Ramirez, Depaula-Silva, Szaniawski et al. (2015) "HIV-1 Vpu utilizes both cullin-RING ligase (CRL) dependent and independent mechanisms to downmodulate host proteins" *Retrovirology (Auckl)*
24. Greenwood, Williamson, Sienkiewicz et al. (2019) "Promiscuous targeting of cellular proteins by Vpr drives systems-level proteomic remodeling in HIV-1 infection" *Cell Rep*
25. Tokarev, Stoneham, Lewinski et al. (2015) "Pharmacologic inhibition of Nedd8 activation enzyme exposes CD4-induced epitopes within env on cells expressing HIV-1" *J Virol*
26. Milhollen, Traore, Adams-Duffy et al. (2010) "MLN4924, a NEDD8-activating enzyme inhibitor, is active in diffuse large B-cell lymphoma models: rationale for treatment of NF-{kappa}Bdependent lymphoma" *Blood*
27. Godbersen, Humphries, Danilova et al. (2014) "The Nedd8-activating enzyme inhibitor MLN4924 thwarts microenvironment-driven NF-κB activation and induces apoptosis in chronic lymphocytic leukemia B Cells" *Clin Cancer Res*
28. Ehrentraut, Curtis, Wang et al. (2016) "Perturbation of neddylation-dependent NF-κB responses in the intestinal epithelium drives apoptosis and inhibits resolution of mucosal inflammation" *Mol Biol Cell*
29. Barth, Mavis, Hernandez-Ilizaliturri et al. (2012) "MLN4924, an investigational small molecule inhibitor of NEDD8activating enzyme (NAE), inhibits NF-κB activity and induces G1 cell cycle arrest and apoptosis in pre-clinical models of Hodgkin lymphoma (HL)" *Blood*
30. Folks, Clouse, Justement et al. (1989) "Tumor necrosis factor alpha induces expression of human immunodeficiency virus in a chronically infected T-cell clone" *Proc Natl Acad Sci*
31. Blank, Liu, Cosmopoulos et al. (2013) "Novel DNA damage checkpoints mediating cell death induced by the NEDD8activating enzyme inhibitor MLN4924" *Cancer Res*
32. Soucy, Smith, Milhollen et al. (2009) "An inhibitor of NEDD8-activating enzyme as a new approach to treat cancer" *Nature*
33. Wei, Guo, Liu et al. (2014) "A first-in-class NAE inhibitor, MLN4924, blocks lentiviral infection in myeloid cells by disrupting neddylation-dependent Vpx-mediated SAMHD1 degradation" *J Virol*
34. Kao, Calman, Luciw et al. (1987) "Anti-termination of transcription within the long terminal repeat of HIV-1 by tat gene product" *Nature*
35. Adams, Sharmeen, Kimpton et al. (1994) "Cellular latency in human immunodeficiency virus-infected individuals with high CD4 levels can be detected by the presence of promoter-proximal transcripts" *Proc Natl Acad Sci*
36. Krumm, Hickey, Groudine (1995) "Promoter-proximal pausing of RNA polymerase II defines a general rate-limiting step after transcription initiation" *Genes Dev*
37. Kaiser, Joshi, Kim et al. (2017) "Assays for precise quantification of total (including short) and elongated HIV-1 transcripts" *J Virol Methods*
38. Yukl, Kaiser, Kim et al. (2018) "HIV latency in isolated patient CD4 + T cells may be due to blocks in HIV transcriptional elongation, completion, and splicing" *Sci Transl Med*
39. Coiras, López-Huertas, Rullas et al. (2007) "Basal shuttle of NF-κB/IκBα in resting T lymphocytes regulates HIV-1 LTR dependent expression" *Retrovirology (Auckl)*
40. Wong, Jiang (2021) "NF-κB sub-pathways and HIV cure: a revisit" *EBioMedicine*
41. Read, Brownell, Gladysheva et al. (2000) "Nedd8 modification of Cul-1 activates SCF βTrCP -dependent ubiquitination of IκBα" *Mol Cell Biol*
42. Khan, Gasperino, Zeichner (2019) "Nuclear transit and HIV LTR binding of NF-κB subunits held by IκB proteins: implications for HIV-1 activation" *Viruses*
43. Xiong, Zheng, Liu et al. (2022) "Activation of the β-TrCP/IκBα/inflammation axis limits the sensitivity of liver cancer cells to neddylation inhibition" *Oncol Rep*
44. Burke, Pattoli, Gregor et al. (2003) "BMS-345541 Is a highly selective inhibitor of IκB kinase that binds at an allosteric site of the enzyme and blocks NF-κB-dependent transcription in mice" *J Biol Chem*
45. Shin, Kim, Kim et al. (2004) "Inhibitory action of novel aromatic diamine compound on lipopolysaccharide-induced nuclear translocation of NF-κB without affecting IκB degradation" *FEBS Lett*
46. Duh, Maury, Folks et al. (1989) "Tumor necrosis factor alpha activates human immunodeficiency virus type 1 through induction of nuclear factor binding to the NF-κB sites in the long terminal repeat" *Proc Natl Acad Sci*
47. Jiang, Dandekar (2015) "Targeting NF-κB signaling with protein kinase C agonists as an emerging strategy for combating HIV latency" *AIDS Res Hum Retroviruses*
48. Acchioni, Remoli, Marsili et al. (2019) "Alternate NF-κBindependent signaling reactivation of latent HIV-1 provirus" *J Virol*
49. Boehm, Calvanese, Dar et al. (2013) "BET bromodomain-targeting compounds reactivate HIV from latency via a Tat-independent mechanism" *Cell Cycle*
50. Karn (2013) "A new BET on the control of HIV latency" *Cell Cycle*
51. Bachmann, Siebenthal, Vongrad et al. (2019) "Determinants of HIV-1 reservoir size and long-term dynamics during suppressive ART" *Nat Commun*
52. Enick, Brooker, Tumiotto et al. (2021) "Comparison of methods to quantify inducible HIV-1 outgrowth" *J Virus Erad*
53. Procopio, Fromentin, Kulpa et al. (2015) "A novel assay to measure the magnitude of the inducible viral reservoir in HIVinfected Individuals"
54. Lungu, Procopio, Overmars et al. (2020) "Inter-laboratory reproducibility of inducible HIV-1 reservoir quantification by TILDA" *Viruses*
55. Mustanski, Ryan, Newcomb et al. (2020) "Very high HIV incidence and associated risk factors in a longitudinal cohort Full-Length Text Journal of Virology October"
56. "study of diverse adolescent and young adult men who have sex with men and transgender women" *AIDS Behav*
57. Iwatani, Chan, Wang et al. (2007) "Deaminase-independent inhibition of HIV-1 reverse transcription by APOBEC3G" *Nucleic Acids Res*
58. Stopak, De Noronha, Yonemoto et al. (2003) "HIV-1 Vif blocks the antiviral activity of APOBEC3G by impairing both its translation and intracellular stability" *Mol Cell*
59. Marin, Rose, Kozak et al. (2003) "HIV-1 Vif protein binds the editing enzyme APOBEC3G and induces its degradation" *Nat Med*
60. Montefiori (2009) "Measuring HIV neutralization in a luciferase reporter gene assay" *Methods Mol Biol*
61. Banerjee, Archin, Michaels et al. (2012) "BET bromodomain inhibition as a novel strategy for reactivation of HIV-1" *J Leukoc Biol*
62. Bartholomeeusen, Xiang, Fujinaga et al. (2012) "Bromodo main and extra-terminal (BET) bromodomain inhibition activate transcription via transient release of positive transcription elongation factor b (P-TEFb) from 7SK small nuclear ribonucleoprotein" *J Biol Chem*
63. Li, Guo, Wu et al. (2013) "The BET bromodomain inhibitor JQ1 activates HIV latency through antagonizing Brd4 inhibition of Tattransactivation" *Nucleic Acids Res*
64. Zhu, Gaiha, John et al. (2012) "Reactivation of latent HIV-1 by inhibition of BRD4" *Cell Rep*
65. Falcinelli, Peterson, Turner et al. (2022) "Combined noncanonical NF-κB agonism and targeted BET bromodomain inhibition reverse HIV latency ex vivo" *J Clin Invest*
66. Fong, Sun (2002) "Genetic evidence for the essential role of betatransducin repeat-containing protein in the inducible processing of NF-κB2/p100" *J Biol Chem*
67. Melikian, Eluard, Bertho et al. (2017) "Model of the interaction between the NF-κB inhibitory protein p100 and the E3 ubiquitin ligase β-TrCP based on NMR and docking experiments" *J Chem Inf Model*
68. Casartelli, Guivel-Benhassine, Bouziat et al. (2010) "The antiviral factor APOBEC3G improves CTL recognition of cultured HIV-infected T cells" *J Exp Med*
69. Squires Kdbs, Mms, Cfms et al. (2015) "Impact of APOBEC mutations on CD8+ T cell recognition of HIV epitopes varies depending on the restricting HLA" *J Acquir Immune Defic Syndr*
71. Norman, Mashiba, Mcnamara et al. (2011) "The antiviral factor APOBEC3G enhances the recognition of HIV-infected primary T cells by natural killer cells" *Nat Immunol*
72. Cardinaud, Urrutia, Rouers et al. (2017) "Triggering of TLR-3, -4, NOD2, and DC-SIGN reduces viral replication and increases T-cell activation capacity of HIV-infected human dendritic cells" *Eur J Immunol*
73. Pollack, Jones, Pertea et al. (2017) "Defective HIV-1 proviruses are expressed and can be recognized by cytotoxic T lymphocytes, which shape the proviral landscape" *Cell Host Microbe*
74. Zhou, Delucia, Hao et al. (2017) "HIV-1 Vpr protein directly loads helicase-like transcription factor (HLTF) onto the CRL4-DCAF1 E3 ubiquitin ligase" *J Biol Chem*
75. Wei, Guo, Han et al. (2012) "A novel DCAF1-binding motif required for Vpx-mediated degradation of nuclear SAMHD1 and Vpr-induced G2 arrest" *Cell Microbiol*
76. Hauser, Lopez, Yang et al. (2010) "HIV-1 Vpu and HIV-2 Env counteract BST-2/tetherin by sequestration in a perinuclear compartment" *Retrovirology (Auckl)*
77. Wallet, Rohr, Schwartz (2020) "Evolution of a concept: from accessory protein to key virulence factor, the case of HIV-1 Vpr" *Biochem Pharmacol*
78. Sui, Zhu, Gagnon et al. (2010) "Innate and adaptive immune correlates of vaccine and adjuvantinduced control of mucosal transmission of SIV in macaques" *Proc Natl Acad Sci*
79. Wang, Bergmeier, Stebbings et al. (2009) "Mucosal immunization in macaques upregulates the innate APOBEC 3G anti-viral factor in CD4 + memory T cells" *Vaccine (Auckl)*
80. Wang, Whittall, Rahman et al. (2012) "The role of innate APOBEC3G and adaptive AID immune responses in HLA-HIV/SIV immunized SHIV infected macaques" *PLoS One*
81. Amir, Iwai, Ciechanover (2002) "The NEDD8 pathway is essential for SCF β-TrCP -mediated ubiquitination and processing of the NF-κB precursor p105" *J Biol Chem*
82. Baek, Krist, Prabu et al. (2020) "NEDD8 nucleates a multivalent cullin-RING-UBE2D ubiquitin ligation assembly" *Nature*
83. Jordan, Bisgrove, Verdin (2003) "HIV reproducibly establishes a latent infection after acute infection of T cells in vitro" *EMBO J*
84. Scholtes, Sawyer, Vaca et al. (2021) "The von Hippel-Lindau Cullin-RING E3 ubiquitin ligase regulates APOBEC3 cytidine deaminases" *Transl Res*
85. Zheng, Chen, Huang et al. (2021) "CPI-637 as a potential bifunctional latency-reversing agent that targets both the BRD4 and TIP60 proteins" *Front Cell Infect Microbiol*
86. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12164193&blobtype=pdf | # Correction: Lumpy skin disease virus suppresses the antiviral response of bovine peripheral blood mononuclear cells that support viral dissemination
Manoj Kumar, Ohad Frid, Asaf Sol, Alexander Rouvinski, Sharon Karniely
## References
1. Kumar, Frid, Sol et al. (2025) "Lumpy skin disease virus suppresses the antiviral response of bovine peripheral blood mononuclear cells that support viral dissemination" *Vet Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724137&blobtype=pdf | # Emergence and characterization of historically extinct virulent genotype IV Newcastle disease virus in wild and domestic birds: genetic insights, pathogenicity, and vaccine efficacy
Weiwen Yan, Xinxin Liu, Shanshan Jiang, Hongjin Li, Weiwei Chi, Rui Luo, Jiahuizi Peng, Feng Jiang, Hongli Li, Tobias Stoeger, Abdul Wajid, Aleksandar Dodovski, Chao Gao, Claro Mingala, Dmitry Andreychuk, Renfu Yin, Hongjin Li
## Abstract
Class II genotype IV Newcastle disease virus (NDV), a historically virulent strain responsible for the first Newcastle disease (ND) panzootic from the 1940s to 1960s, was presumed extinct after its last reported isolate in India in 2000. Here, we report the emergence of four virulent genotype IV NDV isolates from 6,731 wild birds and domestic poultry across nine provinces of China between 2021 and 2023, representing the first genetically confirmed isolation of this ancestral genotype in over two decades. These isolates exhibit remarkably high genetic similarity to ancestral strains, showing minimal divergence despite a temporal span of 50-90 years, yet they differ from the most recently reported isolate from India. Infection with the representative isolate, KS02, caused severe lethality and higher transmissibility in specific-pathogen-free chicks than a genotype VII reference virus. Notably, LaSota vaccination provided only limited protection at the conventional hemagglutination inhibition (HI) threshold and achieved complete protection only when HI titers were at least twofold higher, in contrast to the protection observed against genotype VII challenge. This unusual genetic stability raises concerns about the origin and evolutionary history of these viruses as well as highlights the urgent need to update vaccination strategies, including approaches to elicit higher HI titers through intensified immunization schedules. This study underscores the critical importance of sustained global surveillance of genotype IV NDV in domestic poultry and migratory birds to monitor its spread and evolution. IMPORTANCE Virulent Newcastle disease virus (NDV), particularly emerging isolates, poses a major threat to poultry health and production, causing severe morbidity, high mortality, and economic losses. While ancestral class II genotypes II and IX have persisted globally across various bird species, the status of genotype IV NDV-last reported in India in 2000-had been uncertain. This study documents the emergence of genotype IV isolates in wild and domestic birds across China from 2021 to 2023, marking their return after more than two decades of presumed extinction. The representative isolate, KS02, showed severe lethality and high transmissibility in chicks compared to the circulating virus. The LaSota vaccine conferred complete protection against this isolate only when HI titers were at least twofold above the conventional protective threshold. These findings underscore the significant risk posed by reemerging genotype IV NDV, highlighting the urgent need for surveillance and updated vaccination strategies.
## RESULTS
## Emergence of genotype IV NDV in migratory birds and domestic poultry along the East Asia-Australasia flyway during 2021 to 2023
Genotype IV NDV was first identified in England in 1933 and has since been documented across Europe, Asia, the Middle East, the Americas, and parts of Africa (Fig. 1A andB). The most recent report of this ancestral genotype IV virus was from India in 2000 (13), after which its status remained uncertain. Through comprehensive molecular epidemiological surveillance of NDV in 6,731 wild and domestic birds across nine provinces in China from 2021 to 2023 (Fig. 2), this study confirms the reemergence of ancestral genotype IV NDV in both wild and domestic bird populations over two decades.
The first isolate was isolated in 2021 from fresh fecal samples of a wild migratory gray goose (Anser anser) at Poyang Lake, Jiangxi Province (longitude 116.266, latitude 28.502), China's largest freshwater lake and a critical stopover along the East Asia-Australasia Flyway in southeastern China. The second isolate was obtained in 2021 from the fresh feces of a pheasant (Gallus gallus) near Poyang Lake (longitude 116.207, latitude 29.186). The third isolate was isolated in 2022 from fecal and oropharyngeal swabs of a domestic duck (Anas platyrhynchos) at a live bird market (LBM) in Changchun, Jilin Province viruses in wild and domestic birds (10,15,16). Metadata for the four isolates are presen ted in Fig. 2. All genotype IV NDV isolates were confirmed by conventional RT-PCR targeting the F gene of AOAV-1 using established protocols, followed by Sanger sequenc ing and BLAST similarity search for identification. The isolates were designated as AOAV-1/gray goose/China/KS02/2021 (KS02), AOAV-1/pheasant/China/2526/2021 (2526), AOAV-1/duck/China/HL69/2022 (HL69), and AOAV-1/Chicken/China/DL85/2023 (DL85). Avian influenza virus and other avian avulaviruses were ruled out using the hemag glutination inhibition (HI) assay and RT-PCR, with primer sequences available upon request from the corresponding author. All isolates were successfully propagated in 9-to 10-day-old specific-pathogen-free (SPF) embryonated chicken eggs, and the harvested infectious allantoic fluid tested positive by HA assay, with titers ranging from 256 to 1024 reciprocal dilutions.
To our knowledge, this study represents the genetically confirmed reemergence of ancestral class II genotype IV NDV in wild birds and domestic poultry along the East Asia-Australasia Flyway in China, with no detections in other major migratory routes within the country. This significant finding, following a more than two-decade absence since its last report in India in 2000, highlights the reemergence and persistence of this ancestral NDV genotype, previously presumed extinct, in a critical migratory corridor. These findings provide critical insights into the epidemiological dynamics and potential transmission pathways of genotype IV NDV, emphasizing its importance for regional and global avian disease surveillance.
## Current genotype IV NDV isolates exhibit unexpectedly high genetic similarity to the presumed extinct ancestral strains
Phylogenetic analyses of the full-length fusion (F) gene nucleotide sequences of NDV clearly distinguish historical class II genotypes I, II, III, and IV-responsible for the first NDV panzootic in the late 1920s and persisted through the 1950s-from currently circulating genotypes V, VI, VII, and XII to XXI (17,18) (Fig. 3A). Within the IV geno type, three subgenotypes (IVa, IVb, and IVc) were defined based on full-length F gene sequences, with inter-subgenotype distances ranging from 0.0616 to 0.0856, exceeding the 0.05 divergence threshold (Fig. 3A) (19). All four NDV isolates in this study clustered within subgenotype IVa, showing close genetic similarity to virulent NDV strains isolated in Europe (1930s-1970s), Western Africa (1970s), and the Americas (pre-1980s). Notably, no subgenotype IVb or IVc NDV strains were detected, consistent with their exclusive reports from Nigeria and India, respectively (13,20). These findings, supported by limited literature on genotype IV NDV, suggest that subgenotype IVa NDV has potential for intercontinental transmission via wild birds, unlike IVb and IVc, which do not exhibit this behavior.
The genotype IV NDV isolates in this study exhibited remarkable nucleotide identity in the F and HN genes, exceeding 94.6% and 95.5%, respectively, with multiple historical subgenotype IVa strains (Fig. 3B andC). For example, divergence from the chicken/VRD Ibadan/1973 was less than 0.2% for both genes, and from the fowl/Herts/1933 strain, it ranged from 0.2% to 0.4%, despite a 50-to 90-year temporal gap. The average evolution ary distance of F gene sequences between current isolates and historical subgenotype IVa strains was 0.0249, significantly lower than distances to IVb (0.0580) and IVc (0.8160) subgenotypes. Bayesian analyses of F gene sequences from 13 genotype IV viruses estimated a mean substitution rate of 6.185 × 10⁻⁵ (standard error: 8.505 × 10⁻⁶). Wholegenome sequencing of two representative isolates, KS02 and HL69, corroborated these findings, confirming a close phylogenetic relationship with historical virulent subgeno type IVa viruses (Fig. 4A). The minimal nucleotide divergence (0.3%-5.7%) across wholegenome sequences within the subgenotype IVa (Fig. 4B), combined with the low F gene substitution rate, suggests an unnatural origin. Despite NDV's expected evolution rate as an RNA virus, the unusually close genetic similarity between recent subgenotype IVa isolates and ancestral viruses from before the 1970s raises questions about their natural transmission dynamics in domestic poultry or wild bird populations. IV strains from this study and selected reference strains from other genotypes. In both panels, the color gradient indicates sequence identity percentages, with warmer colors denoting higher similarity. The genotype IVa, IVb, and IVc strains are outlined in blue, purple, and green dashed boxes, respectively. The newly isolated genotype IV strains from China exhibit > 99% identity to genotype IV reference strains from the mid-20th century, highlighting their close genetic relationship and the limited sequence divergence over time.
## Current genotype IV NDV isolates exhibit high lethality and transmissibility in chicks
All current genotype IV NDV isolates exhibited a characteristic virulent multibasic amino acid sequence (112RRQRRF117) at the fusion (F0) protein cleavage site, consistent with other genotype IV viruses. Their intracerebral pathogenicity index (ICPI) scores exceeded 1.4, and the mean death time (MDT) in eggs was less than 60 hours, further confirming their classification as typical virulent viruses. Among the isolates, KS02 was chosen for further studies due to its status as the first isolate obtained and its high genetic similarity to the other three isolates (Fig. 3B andC and Fig. 4B). In parallel, the genotype VII NDV strain NA-1, a virulent reference strain and the predominant genotype currently circulating in several countries, including China, was used as a control.
Twenty SPF chicks were challenged with a single dose of 10⁵ EID 50 of the KS02 virus via intraocular-nasal drops, resulting in 100% mortality within 2-3 days post-challenge (dpc) (Fig. 5C). Clinical signs of ND infection, including depression, drooping wings, diarrhea, and nasal discharge, appeared approximately 36 hours post-challenge (hpc). Necropsy revealed extensive multi-organ hemorrhages, particularly in the respiratory system, gastrointestinal tract, brain, liver, and heart, along with viscous fluid in the trachea and cloaca in challenged chicks (Fig. 5A). Histopathological analysis identified severe congestion, hemorrhage, and edema in alveolar spaces; infiltration of inflammatory cells (neutrophils and macrophages) in the mucosal layer; non-suppurative encephalitis with neuronal degeneration, perivascular cuffing, and gliosis; lymphoid depletion and congestion in the spleen's white and red pulp; and tubular necrosis, vacuolar degeneration, and glomerular hypercellularity in the kidney of challenged chicks (Fig. 5B). By comparison, chicks infected with NA-1 died between 2 and 5 dpc and developed clinical signs of ND at approximately 42 hpc-about 6 h later than those infected with KS02-but also exhibited extensive multi-organ hemorrhages (Fig. 5A through C). These findings confirm the systemic spread of the virulent KS02 virus, comparable to the currently predominant genotype VII strains circulating globally, resulting in rapid lethality and severe multi-organ damage.
To evaluate the transmissibility of the KS02 isolate, two SPF chicks were challenged intraocularly and nasally with 10⁵ EID₅₀ of the virus. Twenty naïve SPF day-old chicks were then randomly assigned to two exposure groups: cohabitation (direct contact in the same cage) and cohousing (indirect contact in the same room), with 10 chicks in each group. Clinical signs, oropharyngeal and cloacal swab viral RNA positivity, and survival rates were monitored daily for 15 days post-challenge (dpc). The two challenged chicks died on 3 dpc, serving as the source of infection for both contact groups. Cohabited and cohoused chicks began to exhibit clinical symptoms around 36 hours after exposure, with mortality occurring between 6 and 9 dpc in both groups. Although the cohabited group showed a higher overall mortality rate (40%) compared to the cohoused group (30%), statistical analysis indicated no significant difference between the two groups (Fig. 5C). Viral RNA detection in oropharyngeal and cloacal swabs reached 100% in the cohabited group between 4 and 5 dpc, then declined to 66.7% by 8 dpc and remained stable thereafter. In contrast, the cohoused group showed a lower peak RNA positivity rate of 40% during the same period, which consistently remained below that of the cohabited group throughout the observation period (Table 1). Although naïve chicks cohoused or cohabited with NA-1-infected birds displayed comparable mortality and viral shedding, the onset of mortality and viral RNA detection in swabs was delayed by more than 24 h compared to KS02 (Fig. 5C, Table 2). These findings demonstrate that genotype IV NDV (KS02) exhibits higher virulence than the virulent genotype VII NA-1 strain and is highly transmissible, causing lethal infection not only in directly inoculated birds but also in those exposed through both direct and indirect contact.
## Limited protection of LaSota vaccine-induced HI titers at the minimal protective threshold against the current genotype IV NDV isolate in chicks
The LaSota vaccine, widely used in poultry for its safety, efficacy, and ease of administra tion via drinking water, injection, or spraying, has historically provided robust protection against ND clinical signs, including respiratory and digestive symptoms, while reducing mortality (5). HI titers are critical for neutralizing NDV and preventing infection, with levels ≥4 log 2 generally considered sufficient to protect against virulent NDV (5,14). However, the efficacy of the LaSota vaccine-induced HI titers at this minimal protective threshold against the current genotype IV NDV isolates remains uncertain. To assess this, 30 SPF day-old chicks were vaccinated with LaSota vaccine via intraocular-nasal drops, following the manufacturer's guidelines. Once HI titers against LaSota reached a protective level of at least 4 log₂ in all vaccinated chicks, they were randomly divided into two groups: 20 chicks were challenged with 10⁵ EID₅₀ of the KS02 isolate via intraocularnasal drops, while 10 chicks received an equivalent dose of phosphate-buffered saline (PBS) as a nonchallenged control. Parallel groups included vaccinated and unvaccinated chicks challenged with the virulent genotype VII NDV strain NA-1 (circulating virus control) as well as unvaccinated, nonchallenged negative controls. Consistent with KS02 pathogenicity results (Fig. 5C), all unvaccinated chicks challenged with either KS02 or NA-1 died within 5 dpc, whereas the nonchallenged negative and vaccinated-nonchallenged controls remained healthy throughout the observation period (Fig. 6A). Among vaccinated chicks challenged with NA-1, no obvious clinical signs were observed, and only 1 of 20 (5 %) died at 7 dpc. In contrast, vacci nation conferred only partial protection against KS02: although no clinical signs were observed within the first 3 dpc, 9 of 20 vaccinated chicks died between 5 and 9 dpc (2, 4, and 3 deaths on 5, 7, and 9 dpc, respectively), resulting in a 55% survival rate (Fig. 6A). Surviving vaccinated chicks challenged with KS02 exhibited significantly slower growth rates from 4 dpc onward (Fig. 6B) and displayed persistent ND-like clinical signs, including head shaking, lethargy, reduced appetite, and respiratory distress, throughout the 14-day observation period. Pathological examination of surviving KS02-challenged chicks revealed mild tissue damage, including slight congestion and inflammatory cell infiltration in the lung bronchi, mild neuronal degeneration in the brain (characterized by cell shrinkage and occasional loss without significant structure disruption), slight vacuolization in kidney tubular epithelial cells, and moderate lymphocyte depletion in the spleen's white pulp (Fig. 6C). These changes were less severe than those observed in naïve challenged chicks and non-surviving vaccinated and KS02-challenged chicks (Fig. 6C and5B). No significant lesions were detected in the vaccinated-nonchallenged, vaccinated-NA-1-challenged, and negative control groups over the 14-day period. Despite 55% survival after a single LaSota dose, oropharyngeal virus shedding was detected in 60% of vaccinated chicks challenged with KS02 (12/20) at 3 dpc, gradually declining and becoming undetectable by 14 dpi (0/11). Cloacal shedding was observed in 15% of chicks (3/20) at 3 dpc and was not detected thereafter (Table 3). In contrast, vaccinated chicks challenged with NA-1 showed significantly lower shedding rates and shorter duration (Table 3).
Because HI titers are pivotal for NDV neutralization, sera from vaccinated nonchal lenged and vaccinated challenged chicks were tested with LaSota and KS02 viruses at 1, 3, 5, 7, 10, and 14 dpc. Vaccinated nonchallenged chicks showed robust LaSota-specific HI titers but minimal or undetectable KS02-specific titers (Fig. 6D). In vaccinated chal lenged chicks, LaSota-specific HI titers initially declined from a peak between 3 and 7 dpc, then sharply increased, reaching a second peak by 14 dpc, while KS02-and NA-1specific HI titers rose steadily throughout the observation period (Fig. 6E andF). Despite LaSota-specific HI titers remaining above the minimal protective threshold of 4 log 2 , the LaSota vaccine provided limited protection, with only a 55% survival rate against the current genotype IV KS02, compared with its stronger protection against NA-1 over the 14-day period (Fig. 6A). Collectively, these results indicate that the LaSota vaccineinduced HI titers at minimal protective threshold offer only limited protection against clinical signs, tissue damage, growth retardation, and virus shedding caused by the current genotype IV NDV isolate, in contrast to the stronger protection observed against the genotype VII strain NA-1. LaSota-specific HI titers remained consistently high and showed a modest anamnestic rise after challenge. NA-1-specific HI titers were initially undetectable, began increasing from 3 dpc, and rose progressively by 14 dpc, although overall levels remained lower than LaSota-specific titers. Data are presented as mean ± SD; the dashed line indicates the 4-log₂ threshold.
## LaSota vaccine confers complete protection against the current genotype IV NDV isolate when HI titers exceed the protective threshold twofold
Because HI titers are pivotal for NDV neutralization, and LaSota-induced HI titers at the minimal protective threshold provide only limited protection against the current genotype IV NDV isolate, we next examined whether the LaSota vaccine can confer complete protection when the HI titers are twofold (≥5 log₂) or even fourfold (≥6 log₂) above the protective threshold. Once vaccinated chicks achieved HI titers either ≥5 to <6 log 2 or ≥6 log₂, they were randomly allocated to four groups. In all, 20 chicks (10 from each titer category) were challenged with 10⁵ EID₅₀ of the KS02 isolate via intraocularnasal inoculation, while 10 chicks (five from each titer category) received an equivalent dose of PBS as nonchallenged controls. Parallel groups included unvaccinated chicks challenged with KS02 and unvaccinated nonchallenged negative controls. In the experimental setup, all unvaccinated, challenged birds began showing clinical signs-including depression, dyspnea, anorexia, and watery greenish-white diarrheaby 36 hpc and died within 2-3 dpc. In contrast, no obvious clinical signs or mortality were observed in any vaccinated birds with an HI titer of either ≥5 to <6 log ₂ or ≥6 log ₂ through 14 dpc observation period, nor in the unvaccinated, nonchallenged controls (Fig. 7A andB). Notably, virus shedding in vaccinated birds was largely restricted to the first 7 days dpc, with the exception of four birds with HI titer ≥5 to <6 log ₂ and two birds with titer ≥6 log ₂ (Table 4).
Moreover, there was no significant difference in either the number of shedders or the duration of shedding between the two KS02-challenged strata (HI ≥6 log₂ vs HI ≥5 to <6 log₂; P > 0.05). In contrast, both higher-titer strata shed significantly less virus and cleared earlier than the minimal-threshold vaccinated group challenged with KS02 (HI ≥4 log₂; P < 0.05). Collectively, these findings indicate that the LaSota vaccine provides complete protection against the current genotype IV NDV isolate when HI titers exceed the protective threshold by at least twofold.
## DISCUSSION
The confirmed emergence of a died-out historical virulent genotype IV NDV in wild and domestic birds over two decades raises serious concerns for poultry health. This study documents the isolation of four genotype IV NDV isolates from migratory geese, pheasants, ducks, and chickens between 2021 and 2023, marking the emergence of a lineage presumed extinct since its last documentation in India in 2000. Phylogenetic analyses place these isolates within subgenotype IVa, revealing an unexpectedly high genetic similarity to virulent strains from the 1930s to 1940s, with nucleotide divergence as low as 0.3%-2.9% (Fig. 4B). This minimal genetic divergence over a 50-to 90-year span is remarkable for an RNA virus like NDV, which typically evolves rapidly (10 -3 to 10 -4 substitutions per site per year) due to the error-prone nature of its RNA-depend ent RNA polymerase (21)(22)(23). Bayesian analysis estimated a mean substitution rate All isolates exhibited a virulent multibasic cleavage site (112RRQRRF117) in the F protein, an ICPI score above 1.4, and an MDT of less than 60 hours, confirming their velogenic nature. The pathogenicity and transmissibility of the representative isolate, KS02, further underscore the threat posed by these genotype IV NDV strains. Experi mental infection with KS02 in SPF chicks caused 100% mortality within 2-3 days, with severe multi-organ damage, including hemorrhages in the respiratory and gastrointesti nal systems, non-suppurative encephalitis, and lymphoid depletion in the spleen (Fig. 5). These findings align with the systemic spread typical of velogenic NDV strains (24)(25)(26)(27). Additionally, KS02 demonstrated high transmissibility, with a 40% mortality rate in cohabitated chicks and 100% viral RNA detection in oropharyngeal and cloacal swabs within 4-5 days post-cohabitation (Fig. 5C and Table 1). The delayed onset of clinical signs and reduced viral shedding in cohoused chicks compared with directly cohabitated ones suggests that direct contact is a more significant transmission route than environ mental exposure, consistent with known NDV dynamics (27). In contrast, chicks either directly challenged with the virulent reference strain NA-1 or cohoused/cohabited with NA-1-infected birds displayed similar mortality and morbidity, but both the onset of death and the appearance of viral shedding were delayed compared with KS02 (Fig. 5C, Table 2). Collectively, these findings demonstrate that genotype IV NDV (KS02) is more virulent than the virulent genotype VII NA-1 strain and is highly transmissible, causing lethal infection not only in directly inoculated birds but also in those exposed through both direct and indirect contact. Notably, these results emphasize the need for enhanced biosecurity measures in LBM and farms along migratory flyways, particularly the East Asia-Australasia Flyway, which facilitates viral spread through migratory waterfowl.
The LaSota strain (Class II, Genotype II) is one of the most widely used poultry vaccines, and the vaccine-induced HI titers of ≥4 log 2 are generally considered sufficient to protect against virulent NDV (5,14). However, the effectiveness of LaSota against current genotype IV NDV isolates remains uncertain. Despite maintaining HI titers at the conventional protective threshold of 4 log2, the vaccine achieved only a 55% survival rate in KS02-challenged chicks, with survivors exhibiting persistent clinical signs, growth retardation, and mild tissue damage, including lung congestion, neuronal degenera tion, and spleen lymphocyte depletion (Fig. 6). All vaccinated and challenged chicks exhibited viral shedding, with oropharyngeal shedding persisting throughout the 14-day observation period. In the KS02-challenged group, cloacal shedding was detected in 40% of chicks at 1 dpc (Table 3). Complete protection against the current genotype IV NDV isolate was achieved only when LaSota-specific HI titers exceeded the protective threshold by at least twofold. Under these conditions, both the rate and duration of viral shedding reduced in an HI titer-dependent manner (Fig. 7C andD, and Table 4). These observations are consistent with previous reports that higher HI titers are required to fully block infection or virus shedding by virulent heterologous NDV strains (14,(28)(29)(30). In contrast, vaccinated chicks with minimal protective HI titers of 4 log 2 that were challenged with the NA-1 strain showed no obvious clinical signs or significant lesions, shed markedly less virus for a shorter duration, and experienced only 1 of 20 deaths (5%) by 7 dpc (Fig. 6A and Table 3). Collectively, these findings highlight the critical role of robust humoral immunity in cross-protection against emergent virulent NDV lineage and underscore the need to maintain HI titers at least twofold above the conventional protective threshold to effectively control outbreaks of potential virulent isolates such as the genotype IV NDV described in this study.
Migratory waterfowl, such as the gray goose (Anser anser), likely play a key role in maintaining and disseminating these viruses, as evidenced by the initial isolation from fecal samples at Poyang Lake, a critical stopover site. The absence of genotype IV NDV in other flyways during 2021-2023 indicates that host susceptibility, environmen tal conditions, or viral fitness may influence its distribution. Whether the genotype IV NDV described in this study occupies a distinct ecological niche remains uncertain and warrants further investigation. Notably, bidirectional spillover between wild birds and domestic poultry, demonstrated by isolates from wild geese as well as domestic ducks and chickens, complicates control efforts. Therefore, a One Health approach, integrating surveillance of wild and domestic bird populations, is essential to monitor and mitigate NDV spread across ecological interfaces. The involvement of migratory birds also raises concerns about the international spread of genotype IV NDV, as these birds can disseminate the virus across continents via established flyways, posing a global risk to poultry industries and food security.
In conclusion, the emergence of genotype IV NDV in wild and domestic birds after over two decades presents a significant challenge to poultry health and production due to its high pathogenicity, efficient transmissibility, and the limited protection provided by LaSota-induced HI titers at the conventional protective threshold. The unusual genetic stability of these isolates raises critical questions about their origin and evolutionary history, necessitating further genomic and epidemiological studies. Notably, the observation that the LaSota vaccine confers complete protection only when HI titers reach at least twice the conventional protective level highlights the urgency of updating vaccination strategies, including efforts to elicit higher HI titers via intensified immunization schedules. Enhanced surveillance along migratory flyways, strengthened biosecurity measures, and a One Health framework are essential to prevent future ND outbreaks and address the broader implications of emerging ancestral strains for global poultry industries. Together, these findings highlight the continuing risk posed by divergent NDV strains and the necessity of reassessing vaccine and surveillance strategies for newly emerged genotypes.
## MATERIALS AND METHODS
## Sample collection
From 2021 to 2023, a total of 6,731 samples, comprising fresh fecal droppings from wild migratory birds at stopover points and oropharyngeal and cloacal swabs from domestic poultry in live bird markets (LBM), were collected across nine provinces in China (Hubei, Hunan, Henan, Hebei, Shandong, Anhui, Jilin, Liaoning, and Heilongjiang) (Fig. 1). Samples were collected under sterile conditions using swabs placed in 2 mL EP tubes containing 1.5 mL of viral transport medium (40% glycerol, 2,000 U/mL penicillin, 2 mg/mL streptomycin, 50 µg/mL gentamycin, 50 U/mL nystatin, and 0.5% bovine serum albumin). During field collection, samples were kept on dry ice and subsequently stored at -80°C upon arrival at the laboratory.
## Virus isolation
Samples were inoculated into the allantoic cavities of 9-to 10-day-old SPF chicken embryos (Jinan SAIS Poultry Company, Shandong, China) in accordance with the WOAH standard manual for ND detection. Allantoic fluids were harvested upon embryo death or at the end of the incubation period and tested for NDV using hemagglutination (HA) assays and RT-PCR to detect NDV-specific nucleic acid signatures. Hemagglutinated allantoic fluids were stored at -80°C for subsequent analysis.
## RNA extraction, RT-PCR, whole-genome sequencing, and phylogenetic analysis
RNA extraction, RT-PCR, and whole-genome sequencing were performed following the protocols and conditions described in our previous studies (15,31,32). Briefly, viral RNA was extracted from allantoic fluid using TRIzol Reagent (Sigma, Shanghai, China) according to the manufacturer's instructions. Reverse transcription and PCR were subsequently conducted using a reverse transcription kit (Novoprotein, Suzhou, China) and a 2× M5 SuperLong Taq MasterMix kit (Mei5Bio, Beijing, China), respectively, to amplify the NDV F gene, adhering to previously established protocols (15,31,32). Purified RT-PCR products were sequenced using an ABI 3730XL automated DNA analyzer (Applied Biosystems, Massachusetts, USA). A BLAST similarity search confirmed NDV identity, and NDV-positive samples were subjected to further sequencing to obtain F gene or whole-genome sequences.
Whole-genome sequencing of the KS02 and HL69 isolates was performed as previously described (31,32), while the full-length F and HN gene sequences of the 2526 and DL85 isolates were obtained through RT-PCR and Sanger sequencing (30). Phylo genetic trees were constructed using MEGA 11 software (33,34) based on full-length F gene and whole-genome sequences of NDV across various genotypes. Maximum Likelihood (ML) trees were constructed using the Kimura 2-parameter model, incorporat ing a discrete Gamma distribution (+G) and invariant sites (+I). Bootstrap analysis with 1000 replicates was used to assess node support. All trees were drawn to scale, with branch lengths representing the number of substitutions per site. Analyses included codon positions 1st, 2nd, and 3rd, as well as noncoding regions; gaps and missing data were excluded. The final data set for the F gene alignment comprised 1,662 nucleotide positions.
## Pathogenicity tests and EID50 assays
The virulence of the Chinese genotype IV NDV isolates was determined using the ICPI per WOAH standards for ND. Briefly, 10-day-old SPF chicks were inoculated intracerebrally with 0.05 mL of allantoic fluid serially diluted 10-fold in sterile PBS. Birds were examined once daily for 8 consecutive days and scored 0 (normal), 1 (sick), or 2 (dead); the ICPI was calculated as the mean daily score per bird over the 8-day observation period, with higher values indicating greater virulence. MDT was assessed by inoculating 9-to 10-day-old SPF embryonated chicken eggs with serially diluted allantoic fluid (10-fold in PBS). A volume of 0.1 mL was inoculated into the allantoic cavity of 5-6 embryos per dilution, and MDT was defined as the mean time to death (in hours) among embryos receiving the minimal lethal dose. Eggs were candled twice daily to monitor embryonic death. Viruses causing death within 60 hours were classified as velogenic, 61-90 hours as mesogenic, or beyond 90 hours as lentogenic (35). The 50% embryo infectious dose (EID50) for the KS02 was determined using the Reed-Muench method. EID50 titers were expressed as log 10 EID50/mL, and back-titrations were performed for all challenge stocks.
## Challenge and transmission studies of genotype IV (KS02) and Genotype VII (NA-1) NDV in SPF chicks
Sixty-five SPF day-old chicks were divided into five groups. Two challenge groups (20 chicks each) were inoculated via ocular and nasal routes with 10⁵ EID₅₀/100 µL of KS02 or NA-1, respectively. To evaluate transmission capacity, two SPF chicks from each challenge group were randomly selected and used as seeders for subsequent exposure experi ments. For each virus, two contact groups of uninfected chicks (n = 20) were established: one group (n = 10) was cohoused in the same room for indirect contact, and the other group (n = 10) was placed in the same cage for direct contact with a seeder chick. The remaining five chicks served as an unchallenged negative control group. All chicks were monitored daily over a 15-day period post-challenge. Oropharyngeal and cloacal swabs were collected daily to assess viral shedding by RT-PCR, and clinical signs and survival were recorded.
## Evaluation of LaSota vaccine protective efficacy at the conventional HI protective threshold against KS02 or NA-1 isolates in chicks
For each virus, 30 chicks were vaccinated with a single dose of the LaSota vaccine per manufacturer's instructions. Once HI titers against LaSota virus reached a minimum of 4 log₂, the chicks were divided into two groups. The challenge group (20 chicks) received 100 µL of 10⁵ EID50 of either KS02 or NA-1 via ocular and nasal routes, while 10 chicks served as vaccinated, unchallenged controls. Chicks were observed for 14 dpc, with daily oropharyngeal and cloacal swabs collected to monitor viral shedding. HI antibody titers against LaSota, NA-1, and KS02 were measured throughout the observation period. An additional group of 10 unvaccinated, unchallenged chicks served as mock controls.
## Evaluation of LaSota-mediated protection against KS02 at elevated HI titers
Although LaSota-induced HI titers at the minimal protective threshold provide only limited protection against the current genotype IV NDV isolate, we next assessed whether higher HI titers could confer complete protection. Specifically, we examined protection when the LaSota-specific HI titer was twofold (≥5 log₂) or fourfold (≥6 log₂) above the conventional threshold. Once vaccinated chicks reached pre-challenge HI titers of either ≥5 to <6 log₂ or ≥6 log₂, they were randomly allocated into four groups. For each titer level, 10 chicks were challenged with 100 µl of 10⁵ EID50 of KS02 via ocular and nasal routes, while 5 chicks served as vaccinated, unchallenged controls. In parallel, 5 unvaccinated chicks challenged with KS02 and 10 unvaccinated, unchallenged chicks served as additional controls. Clinical signs, survival, body weight, HI titers, and virus shedding were monitored as described above.
## Hemagglutination (HA) and HI assays
HA and HI assays were performed using the WOAH-recommended standard microtiter plate method without any modifications. For the HI test, 4 HA units of either the LaSota strain or the KS02 isolate were used to measure HI titers against each virus.
## Histopathological analysis
Tissue samples from chicks that died due to infection or were selected for analysis were fixed, sectioned, and subjected to histopathological examination. Lesions were documented to characterize pathological changes associated with NDV infection.
## Statistical analysis
Statistical analyses were conducted using GraphPad Prism version 10.1.2. Data from three independent experiments are expressed as mean ± SD. Statistical significance was determined using an unpaired two-tailed Student's t-test, with significance levels denoted as * P < 0.05, ** P < 0.01, *** P < 0.001.
## References
1. Cheng, Liu, Mu et al. (2022) "Intense innate immune responses and severe metabolic disorders in chicken embryonic visceral tissues caused by infection with highly virulent newcastle disease virus compared to the avirulent virus: a bioinformatics analysis" *Viruses*
2. Gao, Ding, Qian et al. (2019) "The virulence of NDV NA-1 strain regulated by the 3' leader or 5' trailer sequences" *Microb Pathog*
3. Sha, Liu, Wang et al. (2024) "Long non-coding RNA analysis: severe pathogenicity in chicken embryonic visceral tissues infected with highly virulent newcastle disease virus-A comparison to the avirulent vaccine virus" *Microorgan isms*
4. Yin, Ding, Liu et al. (2010) "Inhibition of Newcastle disease virus replication by RNA interference targeting the matrix protein gene in chicken embryo fibroblasts" *J Virol Methods*
5. Ganar, Das, Sinha et al. (2014) "Newcastle disease virus: current status and our understanding" *Virus Res*
6. Sánchez-Felipe, Villar, Muñoz-Barroso (2013) "Entry of newcastle disease virus into the host cell: role of acidic pH and endocytosis" *Biochim Biophys Acta*
7. Dimitrov, Ramey, Qiu et al. (2016) "Temporal, geographic, and host distribution of avian paramyxovirus 1 (Newcastle disease virus)" *Infect Genet Evol*
8. Cao, Bo, Ruan et al. (2022) "Construction of novel thermostable chimeric vaccine candidates for genotype VII newcastle disease virus" *Viruses*
9. Ni, Deng, Chen et al. (2023) "Role of macrophages in the pathogenesis of genotype VII newcastle disease virus in chickens" *Animals (Basel)*
10. Xue, Cong, Yin et al. (2017) "Genetic diversity of the genotype VII Newcastle disease virus: identification of a novel VIIj sub-genotype" *Virus Genes*
11. Alexander, Manvell, Parsons (2006) "Newcastle disease virus (strain Herts 33/56) in tissues and organs of chickens infected experimentally" *Avian Pathol*
12. Jakhesara, Prasad, Pal et al. (2016) "Pathotypic and sequence characterization of newcastle disease viruses from vaccinated chickens reveals circulation of genotype II, IV and XIII and in India" *Transbound Emerg Dis*
13. Tirumurugaan, Kapgate, Vinupriya et al. (2011) "Genotypic and pathotypic characterization of Newcastle disease viruses from India" *PLoS One*
14. Liu, Shen, Li et al. (2023) "Efficacy of Newcastle disease LaSota vaccine-induced hemagglutination inhibition antibodies against challenges with heterologous virulent strains of genotypes VII and IX" *Vet Immunol Immunopathol*
15. Zhang, Xie, Liu et al. (2015) "High genetic diversity of newcastle disease virus in wild and domestic birds in Northeastern China from 2013 to 2015 reveals potential epidemic trends" *Appl Environ Microbiol*
16. (2025) *Full-Length Text Journal of Virology*
17. Xue, Xu, Yin et al. (2017) "Identification and pathotypical analysis of a novel VIk sub-genotype Newcastle disease virus obtained from pigeon in China" *Virus Res*
18. Dimitrov, Lee, Williams-Coplin et al. (2016) "Newcastle disease viruses causing recent outbreaks worldwide show unexpectedly high genetic similarity to historical virulent isolates from the 1940s" *J Clin Microbiol*
19. Ballagi-Pordány, Wehmann, Herczeg et al. (1996) "Identification and grouping of Newcastle disease virus strains by restriction site analysis of a region from the F gene" *Arch Virol*
20. Dimitrov, Abolnik, Afonso et al. (2019) "Updated unified phylogenetic classification system and revised nomenclature for Newcastle disease virus" *Infect Genet Evol*
21. Welch, Shittu, Abolnik et al. (2019) "Genomic comparison of Newcastle disease viruses isolated in Nigeria between 2002 and 2015 reveals circulation of highly diverse genotypes and spillover into wild birds" *Arch Virol*
22. Fan, Xu, Zhang et al. (2017) "Analysis of molecular evolution of nucleocapsid protein in Newcastle disease virus" *Oncotarget*
23. Jia, Liang, Wu et al. (2022) "Circulation, genomic characteristics, and evolutionary dynamics of class I Newcastle disease virus in China" *Virulence*
24. Chong, Padhi, Hudson et al. (2010) "The effect of vaccination on the evolution and population dynamics of avian paramyxovirus-1" *PLoS Pathog*
25. Susta, Jones, Cattoli et al. (2015) "Pathologic characterization of genotypes XIV and XVII Newcastle disease viruses and efficacy of classical vaccination on specific pathogen-free birds" *Vet Pathol*
26. Kabiraj, Mumu, Chowdhury et al. (2020) "Sequential pathology of a genotype XIII newcastle disease virus from Bangladesh in chickens on experimental infection" *Pathogens*
27. Dhaygude, Sawale, Chawak et al. (2017) "Molecular characterization of velogenic viscerotropic Ranikhet (Newcastle) disease virus from different outbreaks in desi chickens" *Vet World*
28. Dimitrov, Ferreira, Mj et al. (2019) "Pathogenicity and transmission of virulent Newcastle disease virus from the 2018-2019 California outbreak and related viruses in young and adult chickens" *Virology (Auckl)*
29. Izquierdo-Lara, Chumbe, Calderón et al. (2019) "Genotype-matched Newcastle disease virus vaccine confers improved protection against genotype XII challenge: The importance of cytoplasmic tails in viral replication and vaccine design" *PLoS One*
30. Bello, Mahamud, Yusoff et al. (2020) "Development of an effective and stable genotypematched live attenuated newcastle disease virus vaccine based on a novel naturally recombinant malaysian isolate using reverse genetics" *Vaccines (Basel)*
31. Sultan, Elfeil, Nour et al. (2021) "Efficacy of the newcastle disease virus genotype VII.1.1matched vaccines in commercial broilers" *Vaccines (Basel)*
32. Sha, Yan, Liu et al. (2025) "The first report and biological characteriza tion of Avian Orthoavulavirus 16 in wild migratory waterfowl and domestic poultry in China reveal a potential threat to birds" *Avian Pathol*
33. Wang, Liu, Chen et al. (2024) "First detection and biological characterization of an avian metaavulavirus 8 isolated from a migratory swan goose in Qinghai Lake" *Microbiol Immunol*
34. Stecher, Tamura, Kumar (2020) "Molecular Evolutionary Genetics Analysis (MEGA) for macOS" *Mol Biol Evol*
35. Tamura, Stecher, Kumar (2021) "MEGA11: Molecular Evolutionary Genetics Analysis Version 11" *Mol Biol Evol*
36. Woah (2021) "Newcastle disease (infection with Newcastle disease virus)" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12229004&blobtype=pdf | # Analysis of the overall development trends and hotspots in the research field of the human gut virome
Yidi Zhang, Ting Mei, Minhui Wang, Chunxia Huang, Hanzhi Yi, Yu Zhan, Sen Yang, Han Wang, Qiulong Yan, Ruochun Guo, Changming Chen
## Abstract
Background Dysbiosis of the human gut virome is associated with a variety of factors, yet the underlying mechanisms remain poorly understood. This study aims to map the current research trajectory of the human gut virome and propose a strategic framework for future scientific research.Methods A bibliometric analysis was performed on articles retrieved from the Web of Science (WoS) Core Collection database covering the period 2000 to 2024, utilizing VOSviewer, CiteSpace, and the R software environment.
ResultsOver the past 20 years, the number of published papers and citations in the field of enterovirus research has shown a significant growth trend. This trend is attributed to the breakthrough progress of high-throughput sequencing technology and the iterative upgrade of viral genome databases such as CheckV, which has made the classification resolution of the enterovirus group more accurate and discovered a large number of unknown bacteriophages. Technological innovation has led to a fundamental transformation in the research model, evolving from the traditional singlevirus species identification to the multi-omics integrated analysis of virus-host interaction networks. It is worth noting that the existing research shows a distinct feature of "imbalance between dry and wet experiments". Most of the achievements are based on bioinformatics analysis, while the translational medicine research involving virus isolation, culture and functional verification is still in its infancy (accounting for only 25%), especially the mechanism research evidence chain for key scientific issues such as the viral-host molecular interaction mechanism has not yet been completed.
ConclusionThe progress of enterovirus group research is developing rapidly. However, the species-level taxonomy of a large proportion of newly discovered bacteriophages remains unresolved. Future research must give priority to establishing a comprehensive functional database, and at the same time combine the systematic characterization of viral functions with the study of the mechanism of host-virus interaction to enable bacteriophages to play a role in maintaining human health.
## Introduction
The human gut microbiome represents a vast and intricate ecosystem of microorganisms colonizing the digestive tract, comprising bacteria, fungi, viruses, archaea, and other microbial entities. These microbes play a pivotal role in maintaining intestinal homeostasis, regulating host immunity, and modulating host metabolism [1,2]. Previous studies on the gut microbiome predominantly concentrated on bacteria as the primary microbial domain of interest [3,4]. As a critical component of the gut microbiome, the gut virome has received growing scientific attention. The gut virome encompasses all viruses and their genomes residing in the human gut, including bacteriophages that infect bacteria, eukaryotic viruses, and free viral nucleic acids [5][6][7]. Studies have demonstrated that the gut virome plays a regulatory role in shaping the structure, composition, and function of the gut microbiome through phage-mediated bacterial lysis, ultimately influencing host health [8,9].
With the rapid development of high-throughput sequencing and bioinformatics technologies, research on the human gut virome has undergone rapid expansion over the past decade. Studies have revealed that the human gut virome is dominated by bacteriophages [10]. In obese individuals, phage diversity is reduced, and the abundance of specific phages shows a negative correlation with body mass index (BMI). Among patients with Crohn's disease, lytic bacteriophages (e.g., Myoviridae) are enriched, which correlates with the activation of host antiviral immune pathways, including interferon-gamma (IFN-γ) [11]. These findings highlight the pivotal regulatory role of the gut virome in disease pathogenesis.
Historically, methodological limitations and a limited understanding of viral diversity have impeded the comprehensive exploration of the human virome [12]. In recent years, technological advancements, particularly breakthroughs in high-throughput sequencing, have enabled researchers to achieve a more comprehensive characterization of the human virome's complexity [13,14]. However, despite substantial progress in gut virome research, numerous challenges and unresolved questions persist. Current viral databases and annotation tools possess limited capacity to identify the vast repertoire of unknown gut viruses, predict their functional roles, or elucidate the complexity of virus-host interactions [15], thereby hindering the establishment of causal relationships in most studies.
Given the rapid advancement and growing importance of gut virome research, it is necessary to systematically analyze its developmental trends, research hotspots, collaboration networks, and key literature to better understand current progress and inform future research directions.
Bibliometrics is widely used for quantitative analysis and visualization of scientific literature. This method enables examination and visualization of multiple dimensions, such as co-authorship, countries, institutions, co-citation and co-occurrence networks, emerging research themes, and thematic evolution [16,17]. Thus, bibliometrics is a crucial tool for research evaluation, highly valued globally by scholars and researchers.
This study conducts a comprehensive bibliometric analysis of human gut virome research literature from the Web of Science Core Collection. Using an integrated framework of CiteSpace, VOSviewer, and R-bibliometrix, it quantifies knowledge production and diffusion patterns, uncovering the domain's knowledge structure, evolutionary trajectories, and emerging directions. This approach addresses limitations of traditional reviews, providing a framework to identify research priorities, promote interdisciplinary collaboration, and translate findings into clinical applications.
## Materials and methods
## Search strategy
A comprehensive search was conducted in the Web of Science Core Collection, targeting English-language articles related to various aspects of the human gut virome. The publication period was set from 2000 to 2024. The search strategy was as follows: TS = (("human gut virome"OR"human intestinal virome"OR"human enteric virome") OR (human AND (gut OR intestinal OR enteric) AND (virome OR virus* (community OR diversity OR ecology OR metagenom*) OR phage* OR bacteriophage* OR virusome OR"viruslike particles"OR"viral-like particles"OR VLP))).
Only English-language articles were included using the database's built-in filters. The primary publication types were research articles and reviews, while conference abstracts, editorial materials, conference papers, duplicate records, and unrelated literature were excluded. The research team conducted multiple rounds of discussion to remove irrelevant and duplicate records, incorporate diverse perspectives, and finalize the dataset under expert guidance.
A total of 3,356 curated articles were included in the bibliometric analysis. Relevant metadata were extracted, including publication and citation counts, contributing countries and institutions, journals, authors, and keywords.
## Bibliometric analysis
This study employed VOSviewer (v.1.6.20), CiteSpace (v.6.1.R6), and the R-Bibliometrix package to conduct the bibliometric analysis. VOSviewer is a widely used software tool for constructing and visualizing bibliometric networks [18]. In this study, VOSviewer was applied to construct and explore collaboration networks among institutions, authors, and keywords. CiteSpace, developed by Professor Chaomei Chen, is a bibliometric analysis tool designed to generate interactive visualizations of scientific knowledge structures, temporal trends, and emerging patterns [19]. CiteSpace was used to detect keyword citation bursts, generate keyword clustering and timeline visualizations, and to produce a dual-map overlay of journal citation landscapes in enterovirus-related research.
The R programming environment was employed to generate annual publication and citation trends from 2000 to 2024The Biblioshiny interface was employed to generate a global collaboration map, analyze international co-authorship patterns, and assess author productivity. The detailed parameter Settings of Cit-eSpace and VOSviewer are provided in Supplement material. R scripts are deposited in https:// github. com/ 1DZha ng/ Bibli ometr ic-/ tree/ main.
## Results
Using a bibliometric research strategy, this study initially retrieved 3,515 articles. After excluding 129 documents that were neither research articles nor reviews, and 30 non-English publications, a total of 3,356 refined records were included in the final bibliometric analysis (Fig. 1).
## Annual publication growth trend
An analysis of publication trends revealed a general upward trajectory in gut virome research over the past two decades, with the number of publications increasing from 13 in 2000 to a peak of 334 in 2021, despite minor fluctuations (Fig. 2). During this period, citation frequency also exhibited a consistent upward trend, reflecting heightened research activity and growing scholarly attention to the gut virome field.
## Regional collaboration analysis
A visual network analysis was conducted on the countries represented in the 3,356 included publications. For the national collaboration analysis, a minimum threshold of five publications per country was applied. Of the 130 countries identified, 70 met this criterion. The top five countries by publication count were the United States (n = 896), China (n = 396), France (n = 163), the United Kingdom (n = 140), and Italy (n = 134) (Fig. 3a).
Although a wide range of countries had contributed to gut virome research, the majority of publications were concentrated in a limited number of nations. Collaborative networks wre primarily centered on the United States and China, while connections among other countries appeared comparatively weaker (Fig. 3b). Most countries engaged in international collaborative research. The United States accounted for the highest proportion of internationally co-authored publications. Notably, although the United Kingdom published more articles than both Italy and Germany, its rate of international collaboration was lower than that of these two countries (Fig. 3c; Table 1).
## Institutional analysis
A total of 4,090 institutions contributed to publications in the field of gut virome research. The top 10 institutions collectively published 473 articles, representing approximately 14% of the total (Table 2).
Prior to 2015, research published by most institutions predominantly focused on fundamental virology, such as elucidating viral molecular structures. However, a significant shift occurred post-2015, with research focus moving towards exploring disease-virus associations. Within the domain of systemic lupus erythematosus (SLE) research, the University of Chinese Academy of Sciences (UCAS) demonstrated the highest publication output from 2016 to 2020. The University of California System displayed a distinct evolution in its research themes. Publications prior to 2015 centered on double-stranded RNA viruses, while those post-2015 gradually transitioned towards investigations concerning waterborne diseases. The French National Institute of Health and Medical Research (INSERM) and the French National Centre for Scientific Research (CNRS) maintained a sustained collaborative relationship from 2000 to 2024. Their collaborative research themes reflected a phased shift, with investigations prior to 2015 focusing on microsporidia, transitioning to enterobacterial virus as the primary focus post-2015 (Fig. 4).
## Journal analysis
To identify the most influential journals in the field, a minimum threshold of five publications was applied, and VOSviewer was used to visualize the distribution of journals related to gut virome research (Fig. 5a). A total of 768 journals were identified, of which 137 published more than five articles. In terms of publication volume, the top three journals were Frontiers in Microbiology, Viruses-Basel, and PLOS ONE (Table 3). The concentrated darker blue blocks in recent years for Viruses-Basel reflect its agility in addressing emerging virology frontiers (e.g., meta-viromics, environmental virology), aligning with increasing global demand for viral ecology researcha rapidly evolving field-thereby attracting heightened manuscript submissions. The dispersed pattern with light blue coloring in the Journal of Virology reflects its status as a premier journal for fundamental virology; it sustains consistent publication output but necessitates innovative themes (e.g., advanced viromics technologies) to maintain its competitive edge (Fig. 5b). Although Applied and Environmental Microbiology ranked fourth in publication count, it received the highest number of citations among all journals, possibly due to its openaccess status, which facilitates unrestricted access for researchers worldwide.The dual-map overlay enabled the analysis, comparison, and contrast of publication characteristics across disciplines (Fig. 5c). In this study, publications in the fields of Veterinary Science, Animal Science,
## Author analysis
Across publications in this field, a total of 17,103 authors contributed. Only 39 authors published more than 10 articles each. These highly productive authors cumulatively contributed 557 papers, accounting for approximately 17% of all publications in this domain (Fig. 6a, Supplement Table 1). Further analysis revealed that the The larger the node, the more publications it has. Lines represent the intensity of collaboration between countries. Thicker lines indicate closer collaborations. b Global map of international collaborations. Darker colors of countries represent more publications, and denser lines represent deeper collaborations between countries. c Ranking chart of the number of collaborative articles by country. The X-axis represents the number of articles, the Y-axis represents countries, red bar charts represent the number of articles co-authored with other countries, and green represents the number of articles published independently by the country distribution of author productivity conforms to Lotka's Law. Its parameter b was calculated as 2.907, exceeding the theoretical value of 2 defined by classical Lotka's Law. This indicates a highly right-skewed distribution characterized by"left-side density and right-side sparsity": the majority of authors have low publication counts, while a small cohort of highly productive authors accounts for the predominant share of output. This pattern prominently exhibits the classic"Pareto Principle"(80/20 rule) prevalent in scientific research output (Fig. 6b). Analysis of the thematic evolution pathways (Author Paths) among core authors showed that the research directions of most authors displayed significant continuity over time. For instance, GIRONES R's research trajectory evolved from early work on PCR technology to subsequent studies on the virome. SAIF LJ consistently maintained a focus on porcine enteric viruses, demonstrating thematic stability. Notably, a minority of authors, exemplified by HILL C, underwent a discernible disciplinary pivot, with their research emphasis shifting from virology towards phage metagenomics. Additionally, some authors, such as ZHANG Y and SHKOPOROV AN, were active only within a single period, failing to exhibit continuity across the different research phases (Fig. 6c).
Analysis of the shifting research foci among the top 10 most productive authors revealed that their research themes exhibited distinct, phase-specific characteristics over time. In the early phase (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008), research primarily centered on molecular detection techniques and single viral pathogens. During the middle phase (2009-2018), metagenomics and gut microbiome research emerged as dominant directions. Progressing into the late phase (2019-2024), the research emphasis gradually extended into applied and environmental-ecological domains (Fig. 6d). This signifies an overarching trend of transition from basic research toward practical applications and environmental contexts.
## Keyword analysis
Keywords serve as concise representations of an article's core themes and research focus. By constructing a cooccurrence network of keywords, it is possible to uncover thematic associations, knowledge structures, and emerging research trends within the field. A visual analysis was performed on keywords that appeared at least five times, yielding a total of 386 terms. Among them, microbiome, bacteriophage, norovirus, metagenomics, and virome were the most frequently occurring (Fig. 7a). Keywords such as gut virome, microbiome, dysbiosis, and SARS-CoV-2 have emerged frequently in recent years (Fig. 7b), indicating the growing relevance of gut virome-related research.
Keyword clustering analysis using CiteSpace identified eight major clusters (Fig. 7c), including: #0 Microbiome, #1 Microbial Source Tracking, #2 molecular characterization, #3 Phage Therapy, #4 Protein, #5 HIV, #6 Virulence plasmid, #7 genome. Notably, Microbiome ranked first among these clusters, highlighting its central role in gut virome research.
A keyword timeline graph was constructed to visualize the evolution of research themes over time (Fig. 8a). Early studies in the field primarily focused on Cluster #1 Microbial Source Tracking, #2 molecular characterization. With the advancement of 16SrRNA and metagenomic sequencing technologies, recent research has shifted toward core topics such as microbiome and phage therapy.
Additionally, the top 25 keywords with the strongest citation bursts were identified (Fig. 8b). The timeline is illustrated in blue, with red segments denoting periods of burst activity. The keyword with the highest burst intensity was Norwalk-like Virus (36.94), followed by Round Structured Virus (22.95) and Reverse Transcription PCR (21.44). Norwalk-like Virus ceased its burst in 2013, Notably, the keywords that have exhibited burst activity within the past two years include Virome, Human Gut, Alignment, and Gut Virome. This suggests that these topics are currently active areas of research and may represent future research frontiers.
## Disease coexistence analysis
Using R, we extracted the top 20 high-frequency keywords for virus-bacterium and virus-disease interactions (Supplement Table 2, 3, Supplement Fig. 1). The analysis revealed a concentrated research focus on intestinal virology, viral pathogenesis-notably enteric infections such as norovirus-induced diarrheal diseases-alongside bacteriophage-driven modulation of gut microbiota, exemplified by Escherichia coli dynamics. Importantly, noroviruses and bacteriophages emerged as predominant research targets, while gastroenteritis and diarrheal diseases represented key clinical manifestations. Collectively, these findings underscore the gut ecosystem's centrality in contemporary virome research.
Further co-occurrence analysis using VOSviewer identified the top diseases co-mentioned with gut virome studies: HIV, gastroenteritis, diarrhea, COVID-19, and inflammatory bowel disease (Table 5). This demonstrates sustained research emphasis on gut virus-disease relationships, particularly involving autoimmune and gastrointestinal disorders, with notable associations extending to cancer, respiratory diseases, and metabolic syndromes.
To gain deeper insight into the clinical dimensions of human gut virome research, we conducted a subset analysis focusing on autoimmune diseases, digestive disorders, and cancer. A total of 168 articles focused on associations between the gut virome and diseases such as HIV, inflammatory bowel disease (IBD), as well as other autoimmune disorders. Twelve authors contributed four or more articles, indicating their prominent roles in advancing this research area (Supplement Table 4). Publication output peaked at 27 articles in 2021, potentially driven by advancements in third-generation sequencing technologies (Supplement Fig. 2). Frontiers in Immunology, Frontiers in Microbiology, and Microorganisms each published five articles, underscoring their contributions to gut virome research (Supplement Table 5). Notably, within autoimmune disease studies (e.g., IBD and type 1 diabetes), over 75% of articles relied primarily on bioinformatics approaches (e.g., viral genome annotation and co-occurrence network analysis), while experimental validations (e.g., phage-host interaction assays or animal models) accounted for only 25% of methodologies. This disparity highlights critical gaps in mechanistic exploration of gut virome-disease causality (Supplement Fig. 3).
Sixty clinical studies on cancer were identified, primarily investigating the relationship between the virome and colorectal cancer (CRC) or gastric cancer. Key research themes included viral biomarker screening and bacteriophage-host interaction mechanisms. The most prolific authors were Allgayer, H. and Marongiu, L., with three publications each (Supplement Table 6). Frontiers in Immunology, Frontiers in Microbiology, and Microorganisms were also the most active journals in this domain, each publishing five articles (Supplement Table 7), reflecting growing interest in the virome's role in modulating the tumor microenvironment.
A total of 155 clinical studies focused on digestive system disorders, with particular emphasis on viral gastroenteritis, liver diseases, and intestinal infectionsespecially among pediatric and immunocompromised populations. Khamrin, P. was the most prolific author in this category, with 10 publications, followed by Maneekarn, N. and Ushijima, H., each with nine (Supplement Table 8). Archives of Virology (n = 11) and Journal of Medical Virology (n = 20) emerged as core journals in gut virome-related digestive research (Supplement Table 9), contributing to innovation in clinical diagnostics and therapeutic strategies. 6b visualizes the Lottka Law of the author's scientific research productivity in this field. After logarithmic transformation of the horizontal axis (Documents) and the vertical axis (Authors), the red fitting line reflects the power-law distribution relationship between the number of published papers by authors and the corresponding number of authors. c The chart type is a path map. The horizontal axis represents the period (T1, T2, T3), and the vertical axis represents the author's name (arranged in descending order of the number of published articles). The two keywords of each author in different periods are displayed in the form of tags, and the nodes of different periods are connected by lines. d Based on the TF-IDF algorithm, the evolution of the popularity of research topics of the top 10 high-yield authors in different time periods (2000-2008, 2009-2018, 2019-2024) was presented. The TF-IDF values of the key words are presented in the form of heat maps (the higher the value, the stronger the importance of the key words in the corresponding author-period combination) (See figure on next page.) Fig. 6 (See legend on previous page.)
## Discussion
Through the analysis of publication trends, geographical distribution, institutions, journals, authorship patterns, interdisciplinary subjects, and thematic evolution, a comprehensive knowledge map of the human gut virome research field was constructed. These findings not only reflect the current research landscape. Over the past two decades, publications and citations related to gut viruses have shown a steady upward trend. The United States contributed the highest number of publications and participated in the most extensive international collaborations. This can be attributed to its central role in the global research network, fostering dense collaborative clusters with countries across Europe and Asia. Its international collaborations span multiple cutting-edge areas, including virome functional annotation and disease biomarker discovery. While the majority of publications are concentrated in a limited number of countries, international collaborations are primarily centered around the United States and China. However, the concentration of high productivity in specific countries and institutions remains notable.
Among the top 10 most productive institutions, Ohio State University produced the highest number of publications, while the Chinese Academy of Sciences ranked The field demonstrates a clear stratification in academic productivity: 60% of researchers have published only one paper, whereas less than 1% have authored more than ten. This distribution is similar to the pattern described by Lotka's law [20].
Through keyword clustering, timeline maps and emergent word analysis, high-frequency terms such as"microbiome","phage therapy"and"virome"were central themes identified in this field. A notable increase in the use of keywords such as"gut virome"and"dysbiosis" was observed, suggesting growing research focus on the interaction between the virome and host health. This trend coincides with advancements in metagenomic technologies. With the wider application of 16S rRNA sequencing and long-read sequencing technology [21], the relative representation of viral data in microbiome studies has likely increased, while the accuracy and efficiency of virome annotation methods have reportedly improved [22]. Meanwhile, the integration of bioinformatics tools such as CheckV and VIBRANT has further optimized the annotation integrity of the viral genome.
Research interest in phage therapy has grown substantially. In earlier periods, research predominantly focused on basic mechanisms such as the lysis cycle, while clinical translation appeared limited [23]. However, coinciding with growing concerns over antibiotic resistance, the number of published papers in this area soared globally between 2011 and 2020, alongside a shift in research emphasis towards therapeutic applications [24].
Research focus has broadened from the detection of single pathogens (e.g.,"microbial origin tracing") toward more systemic studies of microbiome-phage-host interactions. Recent studies demonstrate progress in understanding the gut virome's role. At a basic science level, researchers have constructed reference databases for the Asian population enteric virome using metagenomic assembly [25], addressing a prior geographical bias in data availability. Recent breakthrough research highlights the dual progress in this field.
In the exploration of the mechanism, the Danish team found that after transmigrating the enterovirus group of lean donor mice into a metabolic syndrome model induced by a high-fat diet, the glucose metabolism of the recipient mice was significantly improved and their weight gain slowed down. Analysis indicated that phages could regulate host metabolism by enriching short-chain fatty acid-producing bacteria (such as Faecalibacterium) [26]. These findings support the concept of the virome's therapeutic potential and provide a theoretical rationale for exploring phage-based interventions. For instance, such insights could inform the development of strategies like combining phages with probiotics for metabolic diseases or exploring virome transplantation to modulate microbiota balance. However, translating these findings into safe and effective human therapies remains a significant challenge.
The keyword bursts of "Norwalk-like virus" and "round structured virus" were concentrated between 2008 and 2013, indicating a high research focus on norovirus detection technology development during that period. Research mainly focused on the standardization of pathogen identification methods, including PCR-based detection, reflecting efforts to enhance diagnostic accuracy and reliability [27,28]. The term "round structured virus" emerged from the morphological classification needs of small, round-structured viruses during the era when electron microscopy was a primary diagnostic tool [29]. The subsequent decline in the use of these keywords indicates a maturation of foundational detection technologies. Consequently, research appears to have shifted towards advanced topics, including virus ecology, hostvirus interactions, and pathogenesis.
In contrast, the recent keyword bursts of "virome" and "gut virome" underscore a growing recognition of the central role of the gut virome in human health and disease [30,31]. Existing studies have found that changes in the composition and function of the gut virome are strongly associated with the onset and progression of various diseases. Gut bacteriophages influence metabolic diseases through modulation of microbial community structure [5], while virome imbalance is associated with various autoimmune diseases [32][33][34], such as autism spectrum disorders, Parkinson's disease, among others [35,36]. This shift aligns with broader research trends exploring microbiome complexity, driven by advancements in metagenomics and bioinformatics. The burst of the keyword"alignment" signals the critical and growing importance of bioinformatics methodologies in virome research. The optimization of virus annotation tools such as VirFinder and CheckV, along with advancements in machine learning algorithms, has enhanced the resolution and accuracy of virushost interaction networks [37,38]. Research initiatives have revealed aspects of the Chinese population enteric virome composition by developing resources like cnGVC, a database containing > 93,000 non-redundant viral sequences with high completeness, over 70% of which were novel at the time of publication [39]. In addition, the application of single-virus sequencing and spatial transcriptomics has facilitated the multi-omics integration of virome data with metabolomic and immunologic profiles [40,41]. For a large number of newly discovered unknown bacteriophages, future research needs to break through the species classification bottleneck and establish functional databases. By conducting whole-genome conserved region sequence alignment and annotating host interaction metabolic pathways, a multi-dimensional classification system can be constructed to provide data support for precision medicine and bacteriophage therapy.
With the continuous development of sequencing technology and bioinformatics tools, these technologies have provided strong support for in-depth research on the relationship between the enterovirus group and human diseases.
Subset analysis has revealed patterns of disease cooccurrence within human gut virome studies, reflecting its diverse applications in immunology, gastroenterology, and oncology.
Notably, dysbiosis of the gut virome exhibits a multifaceted yet pivotal role in autoimmune pathogenesis, involving dysregulated cross-kingdom interactions. In conditions such as Crohn's disease (CD), rheumatoid arthritis (RA), and systemic lupus erythematosus (SLE), virome perturbations are consistently documented, particularly through altered abundance of Caudovirales phages. For instance, crAssphage abundance is markedly elevated in CD patients and demonstrates a significant inverse correlation with anti-inflammatory genera such as Faecalibacterium, Blautia [42][43][44]. These findings collectively indicate that bacteriophages disrupt immune homeostasis-specifically Th17/Treg balanceby remodeling gut bacterial ecology and disrupting mucosal barrier integrity [42,45]. This mechanistic insight provides a novel framework for understanding autoimmune etiology and suggests therapeutic potential in phage-targeted interventions to recalibrate immune responses.
In gastrointestinal diseases, rotavirus is recognized as a leading viral pathogen responsible for diarrhea worldwide. Diarrhea remains a major contributor to morbidity and mortality among children globally, particularly those under the age of five [46]. A team from Nanjing Medical University in China investigated the molecular basis underlying β-diversity variation using virome metagenomic approaches. They found that rotavirus A infection resulted in a significant decline in the abundance of gut bacteriophages from the Microviridae and Caudovirales families in children with diarrhea, while the proportions of adenovirus and calicivirus increased [47].This is not only a change in the viral community, but also reveals the possible antagonistic or synergistic effects among specific viral species, providing new clues for understanding the pathological mechanism of diarrhea and suggesting the potential for developing novel intervention measures for specific viral combinations.
In the tumor microenvironment, viruses function as both symbionts and pathogenic agents, capable of directly damaging DNA, modulating the immune system, and potentially triggering chronic inflammation, thereby contributing to carcinogenesis [48]. CRC as an example, the gut virome of CRC patients is marked by an expansion of Caudovirales and a reduction in Microviridae. T These bacteriophages may mediate through mechanisms like horizontal gene transfer (HGT), the transfer of bacterial virulence genes, potentially fostering a pro-inflammatory tumor microenvironment [49]. Preclinical models suggest Helicobacter pylori infection induces alterations in the gut virome early in tumorigenesis, specifically among temperate bacteriophages, which might subsequently impact the gut microbiota and CRC progression [50]. These findings support the possibility of the enterovirome as a biomarker for tumor development, especially as a sensitive biomarker for the early diagnosis of cancer or a new intervention target.
Among obese people, the richness, diversity and dominance of phage communities have significantly increased [51]. Phage-mediated lysis of Bacteroidetes reduces SCFA-producing bacteria [52]. This process may impair dietary fiber fermentation capacity and reduce production of metabolically active SCFAs (e.g., acetate and propionate), potentially disrupting host energy balance and satiety signaling [53]. Similarly, the gut viromes of type 2 diabetes mellitus (T2DM) patients exhibit abnormalities, with altered temperate phage abundance correlating with microbiota-driven bile acid dysmetabolism and reduced insulin sensitivity [54]. However, the precise mechanisms by which viromes directly or indirectly regulate lipid metabolism through microbiota-host axes remain poorly understood, representing critical knowledge gaps in the field.
While mechanistic insights grow, translational progress in gut virome therapy has been both promising and challenging. The widely reported case in 2016 where Tom Patterson was successfully treated for multidrug-resistant Acinetobacter baumannii infection using phage therapy [55] highlighted the potential of this approach and stimulated developments such as regulatory pathways like the FDA's'Compassionate Use'(eIND). The new-generation oncolytic adenovirus (ONCOS-102) combined with PD-1 inhibitors triggers a"super-progressive response"in melanoma, with an objective response rate of 43% [56], Topical application of phage gel in the treatment of diabetic foot ulcers increases the wound healing rate to 75% [57]. However, these advancements are limited by technological gaps. Due to viral variations among individuals, some patients do not respond to standardized phage cocktails. The high cost of personalized phage treatment and the conflicting standards for phage purity set by the FDA and EMA have increased global development costs [58]. These gaps prominently indicate the need for interdisciplinary solutions, ranging from AI-driven phage design to international regulatory coordination.
Based on our subgroup analysis, it is found that there is a serious imbalance between dry and wet experiments in the current research field of enteroviruses. At present, most of the research relies on bioinformatics analysis, while the proportion of wet experiments is only about 25%. Take CrAssphage phage as an example. Although it was predicted to exist in 2014, it was not isolated and cultured for the first time until 2018 [59], and only three new isolates were added in the following three years [60,61]. This is due to the fact that > 90% of gut virome components remain unidentified; current methods poorly enrich low-abundance viruses [62]. Bacteriophages require specific host bacteria for proliferation, yet most gut bacteria (e.g., butyrate-producing species) are unculturable. Furthermore, DNA/RNA viruses and temperate/ lytic phages exhibit varying sensitivities to environmental factors such as temperature, pH, and osmolarity. The absence of unified cultivation systems exacerbates these challenges, leading to the failure of isolating most bacteriophages [63,64]. Future efforts must prioritize the development of host-independent viral cultivation methods and the establishment of humanized animal models to advance phage GMP production standards and address formulation bottlenecks.
In view of the problems existing in the current research, such as technical bottlenecks, imbalance between dry and wet experiments, and difficulties in clinical transformation, the research methods of a single discipline have been difficult to meet the needs of in-depth exploration of the complex relationship between the enterovirus group and human health. This requires the integration of knowledge from multiple disciplines such as biology, medicine, and bioinformatics to solve. From a broader perspective, the study of the enterovirus group involves multiple disciplinary fields such as microbiology, immunology, gastroenterology, and oncology. The cross-integration among these disciplines can provide a new perspective for a comprehensive understanding of the functions and mechanisms of action of the enterovirus group.
Currently, our research has several limitations. All literature data were sourced exclusively from the WOS Core Collection. While comprehensive for peer-reviewed journal publications indexed in major international databases, this approach omits critical literature types. These include:
(1) studies published in regional journals not indexed in the Web of Science (e.g., Chinese Medical Journal, Indian Journal of Microbiology); (2) book chapters, dissertations, and conference proceedings (e.g., gut virome abstracts presented at DDW or UEG Week); and (3) publications within local or national databases (e.g., CNKI for Chinese literature, SciELO for Latin America). Consequently, contributions from specific regions or emerging research fields may be systematically underrepresented. Furthermore, the exclusion of research deemed unrelated to the study's predefined goals by the researchers introduces potential selection bias. This may compromise the representativeness and comprehensiveness of the mapped research landscape. For instance, research focusing on non-human viruses or technical approaches lacking direct health applications could be inappropriately excluded. Finally, while our bibliometric analysis captures publication and citation trends, it does not directly measure scientific impact beyond these quantitative metrics, nor does it evaluate research quality. Citation counts in bibliometrics do not take into account research quality, such as mechanical rigor or clinical relevance, which limits the interpretation of scientific impact. Despite these limitations, the findings provide an overview of publication and research activity concerning the human gut virome from 2000 to 2024.Identifying these trends can inform discussions about future research priorities and resource allocation.
## Conclusion
The human gut virome constitutes a pivotal constituent of the gut microbiota and exerts a fundamental influence on host health. This study leverages bibliometric tools to delineate the rapid advancement of this research domain, wherein research priorities have transitioned from PCR-based pathogen detection to mechanistic studies of viral ecology and pathogenesis. In the future, it is imperative to conduct further investigations into the cross-disease impacts of the gut virome, formulate effective therapeutic strategies, elucidate the intricate host-virus interaction mechanisms, and intensify international cooperation. Such efforts are essential for integrating multi-national data and constructing large-scale databases, which will ultimately facilitate the progress of precision medicine.
## References
1. Chioma, Hesse, Chapman et al. (2021) "Role of the Microbiome in Interstitial Lung Diseases" *Front Med (Lausanne)*
2. Jazayeri, Daghighi, Rezaee (2017) "Lifestyle alters GUT-bacteria function: Linking immune response and host" *Best Pract Res Clin Gastroenterol*
3. Wang, Chen, Han et al. (2025) "Comparative study of gut microbiota and metabolite variations between severe and mild acute pancreatitis patients at different stages" *Microb Pathog*
4. Wang, Guo, Li et al. (2023) "Composition and functional profiles of gut microbiota reflect the treatment stage, severity, and etiology of acute pancreatitis" *Microbiol Spectr*
5. Shkoporov, Hill (2019) "Bacteriophages of the Human Gut: The "Known Unknown" of the Microbiome" *Cell Host Microbe*
6. Guerin, Shkoporov, Stockdale et al. (2018) "Biology and Taxonomy of crAss-like Bacteriophages, the Most Abundant Virus in the Human Gut" *CELL HOST & MICROBE*
7. Kim, Park, Roh et al. (2011) "Diversity and Abundance of Single-Stranded DNA Viruses in Human Feces" *APPLIED AND ENVIRONMENTAL MICROBIOLOGY*
8. Clooney, Sutton, Shkoporov et al. (2019) "Whole-Virome Analysis Sheds Light on Viral Dark Matter in Inflammatory Bowel Disease" *Cell Host Microbe*
9. Wang, Li, Ren (1873) "Microbiota-Immune Interaction in the Pathogenesis of Gut-Derived Infection" *Front Immunol*
10. Gao, Sohn, Wang (2022) "Editorial: Gut virome and human health" *Front Cell Infect Microbiol*
11. Ungaro, Massimino, Furfaro et al. (2019) "Metagenomic analysis of intestinal mucosa revealed a specific eukaryotic gut virome signature in early-diagnosed inflammatory bowel disease" *Gut Microbes*
12. Fancello, Raoult, Desnues (2012) "Computational tools for viral metagenomics and their application in clinical research" *Virology*
13. Gholamzad, Khakpour, Hashemi et al. (2024) "Exploring the virome: An integral part of human health and disease" *Pathol Res Pract*
14. Yang, Shan, Liu et al. (2024) "VirID: Beyond Virus Discovery-An Integrated Platform for Comprehensive RNA Virus Characterization" *Mol Biol Evol*
15. Coclet, Camargo, Roux (2024) "MVP: a modular viromics pipeline to identify, filter, cluster, annotate, and bin viruses from metagenomes" *Msystems*
16. Waltman, Van Raan, Smart (2014) "Exploring the relationship between the engineering and physical sciences and the health and life sciences by advanced bibliometric methods" *PLoS ONE*
17. Liu, Wang, Wen (2024) "Bibliometric analysis of the top 100 most cited articles on Th17/Treg balance and rheumatoid arthritis" *Heliyon*
18. Van Eck, Waltman (2010) "Software survey: VOSviewer, a computer program for bibliometric mapping" *Scientometrics*
19. Chen (2004) "Searching for intellectual turning points: progressive knowledge domain visualization" *Proc Natl Acad Sci U S A*
20. Ioannidis, Baas, Klavans et al. (2019) "A standardized citation metrics author database annotated for scientific field" *PLoS Biol*
21. Zhao, Shi, Lau et al. (2022) "Uncovering 1058 Novel Human Enteric DNA Viruses Through Deep Long-Read Third-Generation Sequencing and Their Clinical Impact" *Gastroenterology*
22. Wei, Hung, Kao et al. (2020) "Characterization of Fecal Microbiota with Clinical Specimen Using Long-Read and Short-Read Sequencing Platform" *Int J Mol Sci*
23. Nobrega, Costa, Kluskens et al. (2015) "Revisiting phage therapy: new applications for old resources" *Trends Microbiol*
24. Walter, Mirzaei, Deng et al. (2024) "The Potential of Bacteriophage Therapy as an Alternative Treatment Approach for Antibiotic-Resistant Infections" *Med Princ Pract*
25. Nishijima, Nagata, Kiguchi et al. (2022) "Extensive gut virome variation and its associations with host and environmental factors in a populationlevel cohort" *Nat Commun*
26. Leal Rodríguez, Shah, Rasmussen et al. (2024) "The infant gut virome is associated with preschool asthma risk independently of bacteria" *Nat Med*
27. Scipioni, Bourgot, Mauroy et al. (2008) "Detection and quantification of human and bovine noroviruses by a TaqMan RT-PCR assay with a control for inhibition" *Mol Cell Probes*
28. Qu, Vongpunsawad, Atmar et al. (2014) "Development of a Gaussia luciferase-based human norovirus protease reporter system: cell type-specific profile of Norwalk virus protease precursors and evaluation of inhibitors" *J Virol*
29. Lambden, Caul, Ashley et al. (1993) "Sequence and genome organization of a human small round-structured (Norwalk-like) virus" *Science*
30. Avellaneda-Franco, Dahlman, Barr (2023) "The gut virome and the relevance of temperate phages in human health" *Front Cell Infect Microbiol*
31. Chen, Cao, Liu et al. (2023) "Disturbed gut virome with potent interferonogenic property in systemic lupus erythematosus" *SCIENCE BULLETIN*
32. Chen, Zhang, Yao et al. (2023) "Characterizations of the multi-kingdom gut microbiota in Chinese patients with gouty arthritis" *BMC Microbiol*
33. Li, Zhang, Yan et al. (2023) "Alterations in the gut virome in patients with ankylosing spondylitis" *Front Immunol*
34. Kim, Horton, Pang et al. (2019) "Higher abundance of enterovirus A species in the gut of children with islet autoimmunity" *Sci Rep*
35. Wan, Zhang, Xu et al. (2024) "Alterations in fecal virome and bacteriome virome interplay in children with autism spectrum disorder" *Cell Rep Med*
36. Wang, Guo, Xing et al. (2024) "Alterations of the gut virome in patients with Parkinson's disease"
37. Nayfach, Camargo, Schulz et al. (2021) "CheckV assesses the quality and completeness of metagenome-assembled viral genomes" *Nat Biotechnol*
38. Ren, Song, Deng et al. (2020) "Identifying viruses from metagenomic data using deep learning" *Quant Biol*
39. Yan, Huang, Li et al. (2025) "The Chinese gut virus catalogue reveals gut virome diversity and disease-related viral signatures" *Genome Med*
40. Kuchina, Brettner, Paleologu et al. (2021) "Microbial single-cell RNA sequencing by split-pool barcoding" *Science*
41. Ma, Amemiya, He et al. (2023) "Smillie CS, Hung DT. Bacterial droplet-based single-cell RNA-seq reveals antibiotic-associated heterogeneous cellular states" *Cell*
42. Tomofuji, Kishikawa, Maeda et al. (2022) "Whole gut virome analysis of 476 Japanese revealed a link between phage and autoimmune disease" *Ann Rheum Dis*
43. Cao, Sun, Huang et al. (2024) "The gut ileal mucosal virome is disturbed in patients with Crohn's disease and exacerbates intestinal inflammation in mice" *Nat Commun*
44. Su, Cao, Feng et al. (2025) "Dietary whey protein protects against Crohn's disease by orchestrating cross-kingdom interaction between the gut phageome and bacteriome" *Gut*
45. Tun, Peng, Massimino et al. (2024) "Gut virome in inflammatory bowel disease and beyond" *Gut*
46. Wagner, Handley, Donato et al. (2025) "Early-life gut microbiome associates with positive vaccine take and shedding in neonatal schedule of the human neonatal rotavirus vaccine RV3-BB" *Nat Commun*
47. Bao, Wang, Li et al. (2023) "Viral metagenomics of the gut virome of diarrheal children with Rotavirus A infection" *Gut Microbes*
48. Read, Douglas (2014) "Virus induced inflammation and cancer development" *Cancer Lett*
49. Tito, Verbandt, Vazquez et al. (2024) "Microbiome confounders and quantitative profiling challenge predicted microbial targets in colorectal cancer development" *Nat Med*
50. Luo, Ru, Mirzaei et al. (2023) "Gut virome profiling identifies an association between temperate phages and colorectal cancer promoted by Helicobacter pylori infection" *Gut Microbes*
51. Bikel, López-Leal, Cornejo-Granados et al. (2021) "Gut dsDNA virome shows diversity and richness alterations associated with childhood obesity and metabolic syndrome" *Iscience*
52. Zhang, Wang (2023) "The human gut phageome: composition, development, and alterations in disease" *Front microbiol*
53. Vaiserman, Romanenko, Piven et al. (2020) "Differences in the gut Firmicutes to Bacteroidetes ratio across age groups in healthy Ukrainian population" *BMC Microbiol*
54. Yang, Niu, Zuo et al. (1213) "Alterations in the Gut Virome in Obesity and Type 2 Diabetes Mellitus" *Gastroenterology*
55. Schooley, Biswas, Gill et al. (2017) "Development and Use of Personalized Bacteriophage-Based Therapeutic Cocktails To Treat a Patient with a Disseminated Resistant Acinetobacter baumannii Infection" *Antimicrob Agents Chemother*
56. Shoushtari, Olszanski, Nyakas et al. (2023) "Pilot Study of ONCOS-102 and Pembrolizumab: Remodeling of the Tumor Microenvironment and Clinical Outcomes in Anti-PD-1-Resistant Advanced Melanoma" *Clin Cancer Res*
57. Nir-Paz, Onallah, Dekel et al. (2025) "Randomized double-blind study on safety and tolerability of TP-102 phage cocktail in patients with infected and non-infected diabetic foot ulcers" *Med*
58. (2024) "Advocating for phage therapy" *Nat Microbiol*
59. Shkoporov, Khokhlova, Fitzgerald et al. (2018) "ΦCrAss001 represents the most abundant bacteriophage family in the human gut and infects Bacteroides intestinalis" *Nat Commun*
60. Hryckowian, Merrill, Porter et al. (2020) "Bacteroides thetaiotaomicron-Infecting Bacteriophage Isolates Inform Sequence-Based Host Range Predictions" *CELL HOST & MICROBE*
61. Guerin, Shkoporov, Stockdale et al. (2021) "Isolation and characterisation of φcrAss002, a crAss-like phage from the human gut that infects Bacteroides xylanisolvens" *Microbiome*
62. Li, Wang, Hu et al. (2022) "Metagenomic analysis reveals unexplored diversity of archaeal virome in the human gut" *NATURE COM-MUNICATIONS*
63. Rand, Quinones-Olvera, Jean et al. (2025) "Phage DisCo: targeted discovery of bacteriophages by coculture. mSystems"
64. He, Li, Zong et al. (2024) "Biological and genome characteristics of a novel broad host-range phage ΦEP1 infecting enteroinvasive Escherichia coli" *Sheng Wu Gong Cheng Xue Bao* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12570474&blobtype=pdf | # Viral apoptotic mimicry and the role of phosphatidylserine receptors
| Virology, | Minireview, Melinda Brindley, Michael Pisciotta, Melinda R01ai139238, Brindley
## Abstract
Building on the observation that enveloped viral lipids can mediate viral entry, subsequent studies have expanded these observations and further defined the lipids that can initiate infection, the cellular receptors involved, as well as started to explore how viruses obtain their membrane and potentially alter the lipid environment to enhance infection. Exploring the role viral apoptotic mimicry plays in the host has proven more difficult, as many of the receptors that interact with viral lipids also play key roles in immune signaling. This Full Circle review summarizes the lipids and receptors that are involved with viral apoptotic mimicry, the viruses that use them, as well as examines the studies that attempt to explore the role apoptotic mimicry plays in a host.
and TIM4 are long enough to breach the cell glycocalyx, presenting the PS-binding domain above most of the other plasma membrane proteins. TIM3 is much shorter, resulting in its IgV domain sitting within the glycocalyx (22). While both TIM1 and TIM4 readily enhance enveloped virus entry, TIM3 did not, despite it containing a PS-binding site (23,24). Chimeric TIM proteins containing TIM3 PS-binding domain/TIM1 mucin domains enhanced entry, while TIM1 PS-binding domain/TIM3 mucin domain did not (25), suggesting the IgV PS-binding domain needs to be above the glycocalyx to readily interact with virion PS. TIM3 can enhance apoptotic clearance (26), but in a cell-type-spe cific manner. This could suggest that in cells with a shorter glycocalyx layer, TIM3 is capable of binding to PS debris, but in cells with a thicker glycocalyx layer, TIM3 cannot mediate PS-binding. Both TIM1 and TIM3 contain tyrosine phosphorylation motifs in their cytoplasmic tails, yet TIM4 does not, suggesting TIM4 can only tether to PS while TIM1 and TIM3 can mediate signaling when bound to PS (27). It remains unclear whether TIM binding to virions is sufficient to initiate internalization. TIM1 lacking its cytoplas mic tail continued to enhance entry of vesicular stomatitis virus (VSV) pseudovirions coated in the Ebola glycoprotein lacking its mucin domain (ppVSV∆G + EBOV-GP∆O) particles (25). However, residues in the cytoplasmic tail were found to be critical for TIM1 enhancement of dengue entry (28).
TIM family members are associated with a variety of immune regulatory roles. TIM1 was first known as kidney injury molecule 1 (KIM1) and was found to be expressed on luminal epithelial cells within the kidney, especially after renal injury (29). TIM1 is expressed on T helper 2 (Th2) cells and is involved with T-cell activation (30). It is also found on other immune cells including B cells, dendritic cells (DCs), macrophages, invariant NKT cells, and mast cells (31,32). TIM4 is produced in antigen-presenting cells (APCs) including macrophages and DCs (33,34). When TIM4 on APCs and TIM1 on T cells both bind to PS-containing exosomes, it simulates T-cell proliferation (20,35). Expression of TIM4 alone is not sufficient to clear apoptotic debris in several tissue macrophage lines and requires TAM expression to efficiently efferocytose (36,37).
TIM1 was the first TIM family member associated with enhancing virus entry. In 1996, Kaplan et al. generated antibodies against African green monkey kidney cells that were susceptible to hepatitis A virus (HAV) (38). They identified three antibodies capable of inhibiting HAV infection that reacted with a mucin-like class 1 integral membrane protein that they named HAV cellular receptor 1 (HAVcr-1), now known as TIM1. Even though HAV is a picornavirus and lacks viral membrane proteins, the PS-binding domain of TIM1 is required for HAV entry (39). HAV and other "naked" viruses can acquire an envelope and travel from cell to cell in exosome-like vesicles (40,41). TIM1 is needed for Vero cell uptake of these quasi-enveloped HAV particles (42).
In 2011, TIM1 expression was found to correlate with Ebola virus glycoprotein (EBOV-GP) mediated entry in a panel of human tumor lines (43). TIM1 monoclonals blocked both EBOV-GP and Marburg glycoprotein-mediated entry into several epithelial cell lines, suggesting TIM1 could serve as a receptor for filoviruses (43). However, TIM1 was not found in all susceptible cells; therefore, it was not required, and other factors could also mediate uptake. A gain-of-function cDNA library approach in 2012 found TIM3, TYRO3, and AXL enhanced dengue virus entry into 293T cells. Confirmatory studies found TIM3 modestly increased 293T susceptibility to dengue, while TIM1 and TIM4 were highly effective, increasing infection rates 500-fold (23). Interactions with both TIMs and TAMs occurred via PS in the viral particle (23). The following year, Jemielity et al. demonstrated entry of many enveloped viruses could be enhanced by both TIM and TAM family members and confirmed the enhancement is mediated by PS in the viral membrane (24). Subsequently, several studies have expanded the viruses that can utilize TIMs in entry (Table 1). Interestingly, human TIM1 is more efficient in mediating viral entry when compared to murine or rhesus TIM1 (44). Human TIM1 can bind to both PS and PE (45,46), while murine and rhesus TIM1 only bind to PS (44), and the added ability to interact with PE may enable more efficient virus interactions.
Most studies conclude that TIM1 and TIM4 facilitate virus entry by binding to the PS within the virion envelope (28,69,70). However, some studies suggest TIM1 directly binds to viral glycoproteins (71,72). The results become difficult to interpret as the mutations in TIM1 identified to reduce binding to the glycoproteins are the same residues associated with PS binding (19,71,72). Furthermore, the same TIM1 residues associated with binding to Ebola and Japanese encephalitis glycoproteins, when mutated, also reduced VSV, Junin, and Lujo glycoprotein entry (73). Potentially, the PS-binding pocket also readily binds with proteins non-specifically. Additional studies, including control proteins, are needed to demonstrate that the binding is lipid-mediated.
## TAMs
Tyro3, Axl, and Mer (TAM) are a family of receptor tyrosine kinases primarily found on macrophages and DCs. Structurally, they contain two N-terminal Ig-like domains that mediate ligand binding, followed by two fibronectin type III repeats (Fig. 1B) (74,75). The ectodomain is tethered by a single-transmembrane region, and the cytoplasmic tail contains a protein-tyrosine kinase domain. TAMs are involved in PS-mediated apoptotic clearance, but they do not bind PS directly; they bind proteins called growth arrest-spe cific-6 (Gas6) and protein S (ProS) (76). Gas6 and ProS are both secreted proteins that require vitamin K carboxylation to bridge TAMs and PS. The C-terminal domains of Gas6 Asfarviruses (ASFV) PAM ( 62)
Vero, A549, H1650, HCC1944, HCC2303
Retroviruses 293T ( 17) 293T ( 17) HMVEC ( 17)
A549 ( 64)
Picornavirus (HAV) AGMK (66)
Huh7.5 (67) LtK (38) HEV Huh7.5 (68) a Studies that exogenously produce the PSRs in cells are in plain text and include the cell lines used. Studies that either knock down a PSR or mask with PSR antibodies are underlined and include the cell lines used.
and ProS bind to the TAM Ig-like domains, while the γ-carboxylated N-termini bind to PS (77). Gas6 and ProS's affinity for PS is increased when PE is also present in the membrane (78). Gas6 and ProS do not interact with all TAMs equally. For example, Gas6 binds to Axl preferentially while ProS binds with Tyro3 (79). TAMs are needed for apoptotic clearance in adult cells (80). In humans, it is estimated that more than a billion cells die every day, yet they are difficult to detect because macrophages and other phagocytes are so efficient in clearance (81,82). Deletion of TAM receptors in mice results in the accumulation of dead cells due to inefficient clearance, which leads to male sterility and retinal diseases as well as autoimmune phenotypes from the build-up of self-antigens (83)(84)(85)(86)(87).
In addition to their role in apoptotic clearance, TAMs also play a role in innate immune signaling. Following engagement with PS-containing Gas6/ProS ligands, TAM receptor tyrosine kinases can dampen the innate immune response (88). For example, in DCs, Axl activation induces the downregulation of pro-inflammatory cytokines. Axl can form a complex with the type I interferon receptor (IFNAR) and switches signaling by activating suppressor of cytokine signaling (SOCS) 1 and 3 (88). TAMs are critical during innate immune signaling because if type 1 interferon binds to IFNAR, a pro-inflammatory signal is produced; however, if it binds to a TAM-IFNAR complex, an immunosuppressive response is produced. Since many viruses induce an interferon response, activation of TAM receptors will suppress the IFN signaling and may enable enhanced replication (89). TAM receptor activation can lead to cell survival and virus activation of TAMs could prolong the survival of viral-infected cells (90).
TAMs were first found to enhance entry of pseudotyped retroviral particles carrying the EBOV and MARV glycoproteins using a cDNA library approach (51). Follow-up studies confirmed TAMs enhanced viral uptake (91, 92), yet the mechanism did not involve interaction with the glycoprotein (93). Morizono et al. determined that Gas6 and ProS bind to virion PS and TAMs, which mediates enveloped virus internalization (21). TAMs' role in virus entry was confirmed by Meertens et al. with several additional enveloped viruses (23).
$$Coronaviruses (SARS-CoV-2) 293T (63) 293T(63)$$
$$Bunyaviruses (Hantan) 293H(64)$$
$$Jurkat (65) A549(64)$$
## Additional PS-binding proteins
Once the PS/PE in the viral membrane was demonstrated to be the ligand in apoptotic mimicry, studies to examine the role of other PS-binding partners were completed to determine if all PSRs worked in a similar manner. Moller-Tank et al. determined that not all PS-binding receptors mediated enveloped virus entry, as the receptor for advanced glycation end products (RAGE) did not enhance entry (47). Furthermore, Morizono and Chen tested whether MFG-E8, CD300a, BAI1, and stabilin 1 and 2 could facilitate PS-mediated viral uptake (17). MFG-E8 was found to work similarly to Gas6/Axl uptake, while CD300a could modestly increase uptake. However, not all PSRs enhanced virus uptake, neither BAI1 nor stabilins facilitated entry (17). While additional cellular proteins can interact with PS (12), little evidence suggests these other PSRs enhance viral uptake.
## CD300
CD300a binds to the aminophospholipids PS and PE. Human CD300a binds to PE with higher affinity than PS (94,95), while CD300f (or mouse CD300lf ) binds to ceramides and sphingomyelin in addition to PS (96). CD300 family members are found on many immune cells, including natural killer cells, DCs, and T cells. While CD300a binds to lipids exposed on apoptotic cells, it does not enhance apoptotic clearance (97). PS/PE binding to CD300a will negatively regulate the phagocytosis of apoptotic cells via an immunoreceptor tyrosine-based inhibitory motif (ITIM) within its cytoplasmic tail (Fig. 1C) (97). Despite reducing phagocytosis of apoptotic materials, CD300a enhances entry of dengue in 293T cells, and CD300a antibodies could partially inhibit dengue uptake into macrophages (54). CD300lf is the receptor for murine norovirus (98). While CD300lf binds to several lipids, including PS, TIM1 was unable to mediate norovirus entry, suggesting PS binding was not required (98).
## MFG-E8
Milk fat globule-EGF8 (MFG-E8) is a soluble protein that binds to PS on the surface of cells and can initiate apoptotic clearance by linking to phagocytes through alpha(v)beta (1) integrin (Fig. 1D) (99). MFG-E8, also known as lactadherin, is abundantly found in milk but also produced in various other tissues. MFG-E8 was the first factor identified that could link apoptotic cells to phagocytes and contains 2 PS-binding domains (100). While MFG-E8 levels in milk can bind to rotavirus and prevent infection (101,102), it was not associated with enhanced viral entry until Morizono and Chen found that it enhanced pseudoparticle entry into 293T cells (17). The integrin/MFG-E8/PS bridge helps HIV-1 virus-like particles spread in DCs (103).
## VIRAL APOPTOTIC MIMICRY
Once the first few studies showed that select enveloped virus entry can occur through PSRs, many additional reports were published demonstrating that the pathway can be used by numerous viruses, including Ebola, chikungunya, dengue, Zika, Tacaribe, and many others (Table 1). Most studies exploring the role of apoptotic mimicry and virus entry focus on taking poorly susceptible cells, overexpressing a PSR, and demonstrating viral entry is enhanced. For instance, 293T cells do not naturally produce PSRs, yet they are highly transfectable cells that can readily produce PSRs from exogenous plasmids (47). These enhanced entry observations have been documented for many viruses or viral glycoproteins (Table 1). While these studies demonstrate that overexpressed PSRs can bind and internalize PS-containing particles, they do not confirm that these pathways are used in naturally susceptible cells. A subset of studies knocked out PSRs from susceptible cells and demonstrated a large reduction in virus entry. For example, African green monkey Vero cells produce both TIM1 and Axl (104). A decrease or removal of TIM1 from these cells dramatically reduced entry of many viruses (28,43,48,50,52,53,55,57,59,61,(64)(65)(66)(67)(68). The abundant levels of TIM1 and Axl may partially explain why Vero cells are frequently used for viral isolation. Vero cells readily uptake virions through apoptotic mimicry, and their lack of interferon production improves permissivity to several viruses (105).
Most studies characterizing viral apoptotic mimicry utilize viruses produced in various tissue culture cells. Moller-Tank harvested rVSV∆O/EBOV particles from several mouse tissues and found they readily infected cells producing TIM1 and virus entry was blocked by TIM1 antibodies or PS-containing liposomes, confirming host-derived viruses also become enriched in PS which can mediate entry (47).
Select studies demonstrate PSR requirements in disease-relevant cell types. Ebola virus is thought to target immune cells including macrophages during infection (106). While infection of various cell culture lines is attributed to TIM or TAM expression, monocyte-derived macrophages did not produce appreciable TIM1 or Axl, and they were not needed for infection, while integrin alphaV and Mer could mediate EBOV infection (50).
Bhattacharyya et al. hypothesized that the enveloped viruses may interact with TAMs not for entry, but because activated TAM receptors inhibit interferon signaling, and thus would be pro-viral (107). Using primary bone marrow-derived DCs from wild-type or TAM gene knockout mice, they determined that Gas6-bound virus rapidly activated Axl. Axl activation led to a dampened innate immune response, including type I interferons (107). They also found that blocking interferon with antibodies restores virus replication in TAM triple knockout cells, suggesting TAM's role in dampening the immune response is more important than its role in viral entry (107).
SARS-CoV-2 entry into cells requires ACE2 (108), but in cells with low levels of ACE2 PSRs including TIM1, TIM4, and Axl can enhance entry (63). ACE2 expression in the respiratory tract is low, while many lung cells produce Axl (109). SARS-CoV-2 entry into several human lung cell lines was reduced by bemcentinib, an inhibitor that blocks Axl signaling (63), suggesting Axl may facilitate entry into disease-relevant cell types.
## HOW ENVELOPED VIRUSES OBTAIN THEIR MEMBRANE RICH IN PS
Cellular membranes are composed of phospholipid bilayers. The plasma membrane of a healthy cell is asymmetric, with distinct lipid compositions in each layer. The exoplas mic/outer leaflet is highly ordered and rigid, enriched in neutral phosphatidylcholine, sphingolipids, ceramides, and cholesterol. The tight packing is important to maintain the barrier role the plasma membrane plays, reducing permeability to extracellular material (110). In contrast, the inner leaflet is more fluid and contains amino-phospholipids PS and PE, as well as phosphatidylinositol (PI) (111). The net negative charge found on the inner leaflet serves as a platform to recruit many membrane-binding proteins to the inner leaflet of the plasma membrane (112,113).
Plasma membrane asymmetry is critical for cellular function. To maintain distinct lipid organizations between the leaflets, cells produce families of lipid transport proteins, termed flippases and scramblases, that readily move lipids between leaflets (114)(115)(116)(117). Flippases maintain bilayer asymmetry by actively flipping their lipid substrates found in the outer leaflet to the inner leaflet in an energy-requiring process (118). During specific signaling events, the asymmetric lipid organization can be briefly reversed, and inner leaflet lipids are exposed to the extracellular environment for short periods (119). Brief PS translocation is mediated by lipid scramblases, such as TMEM16F (120), which moves PS to the outer leaflet during myoblast fusion and exosome budding. Once the signaling event is complete, scramblases are turned off, and flippases restore membrane asymmetry. When cells are programmed for apoptosis, caspases cleave flippases, permanently inactivating them from lipid flipping. In addition, caspases activate scramblases such as XKR8, resulting in PS accumulation in the outer leaflet (121). Apoptosis induction results in an irreversible change to membrane asymmetry and will eventually lead to clearance by phagocytes. Some viruses manipulate cellular scramblases to alter the lipid distribution and enhance their infectivity, while others rely on the apoptotic changes that occur in the cell to increase PS on their membrane (8).
For enveloped viruses to utilize apoptotic mimicry, they must incorporate PS and/or PE in their outer leaflet. As discussed above, cellular membranes are typically asymmetric, and PS/PE are only found in the inner leaflet of the plasma membrane. Therefore, virions that bud from the plasma membrane must alter the localization of lipids in the healthy bilayer in order to effectively produce virions that can utilize apoptotic mimicry. Many viral infections will eventually induce apoptosis, which will enhance the levels of PS in virions budding from the plasma membrane. Some viral infections activate scramblases, altering the membrane asymmetry and enhancing particle infectivity via PSRs. Ebola infection in Vero and Huh7 cells resulted in PS externalization via TMEM16F enhancing particle infectivity, while downregulation of the scramblase reduced particle infectivity (122). If cells lack the apoptotic-induced scramblase XKR8, the produced EBOV virions contain lower levels of external PS and are less infectious (53,123,124). Conversely, if the producing cells lack flippase complexes, viral particles become highly enriched in outer leaflet PS and can interact with PSRs more efficiently (53).
Viruses that bud from internal membranes can also incorporate PS and PE within their membrane to utilize apoptotic mimicry. Although PS is synthesized within the endoplasmic reticulum (ER), the ER membrane contains lower levels of PS than the plasma membrane. PS is also asymmetrically distributed with PS located in the luminal leaflet of the ER (125). While PS levels are low, PE is more abundant in the ER and Golgi, and viruses that bud from internal membranes may rely on PE interactions with various PSRs to help mediate entry. For example, flaviviruses including Zika (ZIKV), dengue (DENV), and West Nile (WNV) viruses bud within the ER (126). These viruses also infect and replicate in both mammalian and mosquito cells which have very different lipid levels (127,128). Richard et al. determined Axl-dependent infection for flaviviruses could change depending on the cell type used to produce the virus (129). When the virus was produced in mammalian Vero 76 cells, ZIKV was able to readily infect human umbilical vein epithelial cells (HUVECs), while DENV and WNV were not (129). ZIKV entry into HUVECs was mediated by Axl, yet the closely related DENV and WNV viruses could not use Axl for entry. This was surprising, as previous studies had found DENV and WNV can use Axl for entry (23,107). It was noted that previous studies produced DENV and WNV virus in mosquito C6/36 cells. When the experiment was repeated with C6/36 produced virus, DENV and WNV could utilize Axl. Lipid differences in ZIKV stocks also play a role in infecting immortalized villous trophoblasts and human placental explants; however, the role of specific receptors was not assessed (128). Mosquito-produced virus may contain altered lipid levels that facilitate Axl interaction.
Entry of several non-enveloped viruses is enhanced by PSRs (40,130,131). While these viruses lack viral membrane proteins, they can utilize cellular exosome machinery to travel between cells (132). Exosome formation is enhanced in cells with high PS levels in the outer leaflet (133,134). These quasi-enveloped viruses are enriched in PS/PE lipids that can facilitate uptake via PSRs. TIM1 expression of Vero cells was first identified as a viral entry receptor for the naked HAV (38).
Virus binding to PSRs can facilitate entry, but cells producing PSRs on their surface can bind to newly budded virions and prevent their release, limiting virus spread. Several reports suggest PSRs reduce the release of newly formed HIV, Chikungunya virus (CHIKV), and Japanese encephalitis virus (JEV) (104,135,136). To efficiently spread and dissemi nate these viruses, actively reduce the levels of PSRs on the cell surface (104,(136)(137)(138). HIV's nef protein reduces Axl levels on the cell surface, Japanese encephalitis virus NS2B-3 protein promotes the degradation of Axl, while CHIKV nsP2 protein reduces TIM1 expression (104,(136)(137)(138).
Few studies have examined how arboviruses such as dengue and chikungunya acquire PE or PS in the outer leaflet of their membranes within mosquito vectors. Perera et al. demonstrated that dengue virus infection of C6/36 cells leads to an altered lipidome, with significantly increased PE levels at sites of viral replication (139). In parallel, studies showing that dengue virus and West Nile virus (WNV) could utilize Axl in HUVEC cells only after passage through C6/36 cells (140) suggest that the altered lipid profile of insect cells is critical for PSR utilization and possibly for vector transmission. While homologs of known PSRs are not found in mosquitos, other PSRs are produced in insect cells and aid in apoptotic clearance (9), raising questions about species-specific roles of lipid-mediated entry mechanisms. The role of apoptotic mimicry in insect cells has not been robustly explored. However, CHIKV-which can use PSR for entry in mammalian cells-was found not to rely on PSRs in mosquito C6/36 cells, as entry could not be competitively inhibited by PC:PE:PS liposomes (53). This suggests that PSRs are not needed for CHIKV entry into C6/36 cells, but additional studies are needed to assess viral apoptotic mimicry in mosquito cells.
## THE COMPLICATED ROLE OF PSR IN ANIMAL MODELS OF DISEASE
Demonstrating that PSRs enhance viral uptake in tissue culture is clear; however, linking these interactions with outcomes of viral disease is complicated. The redundancy of PS-binding proteins makes it difficult to explore their roles in disease (141). For example, Zika infection is primarily mediated by Axl and Tyro3 in human skin fibroblasts, with TIM-1 playing a minor role (58,140). However, mice deficient in Axl or Mer, or even double knockouts of both Axl/Mer or Axl/Tyro3, remained readily infectable with Zika virus after a subcutaneous inoculation (142,143). Zika replicated to similar titers as control mice, suggesting removal of two associated receptors in the skin is not sufficient to alter mouse infection (142,143). Additionally, similar results were observed after transplacental, vaginal, and intracranial routes of infection (143). Presumably, other PSR receptors like TIMs and/or CD300a (54,78), or other non-PS pathways, facilitated mouse infection (143). In mice deficient for both interferon alpha receptor (IFNAR) and Axl, Axl played an age-dependent role in Zika-induced pathology (144). While all 3-week-old mice died from infection, 6-week-old mice lacking Axl were more likely to survive Zika infection due to reduced apoptosis in the brain (144). However, Mertk-deficient mice were more resistant to VSV because Mertk dampens the interferon response permitting more VSV replication (145).
TAMs play important roles in innate immune signaling, and while removal of a viral receptor may be predicted to reduce viral entry and spread, removal of important immune signaling pathways can play a more important role. Mice lacking Mer or Axl were more vulnerable to neuroinvasive West Nile and La Crosse viruses due to increased blood-brain barrier permeability (146). Axl-deficient mice were more susceptible to JEV, and more virus was observed in neuronal tissue, suggesting Axl's role in altering immune pathways is more important than its ability to mediate entry within the mouse (147).
While the role of TAMs in animal models of disease suggests their immunologic functions are more important than their ability to enhance virus uptake, mice knocked out for TIM1 survived mouse-adapted EBOV significantly more than parental controls (148). Surprisingly, survival did not correlate with reduced viremia, as viral RNA copies in the blood were similar in TIM1 -/-and wild-type controls. However, viral levels were only noted in the blood on day 6 and no other tissues were reported. The lack of TIM1 reduced EBOV-stimulated cytokine production, and the authors suggest TIM1's role in the T cell immune response to infection altered disease severity more than its role in virus entry (148).
TIM1's role in viral entry was also monitored in a mouse model using a recombinant VSV encoding the EBOV glycoprotein (rVSV/EBOV). While Ifnar -/-control and Ifnar -/-/ TIM1 -/-mice both succumbed to rVSV/G infection, Ifnar -/-/TIM1 -/ -mice were significantly more likely to survive infection with rVSV/EBOV (149). Titers were examined over time and in various tissues following infection. While no differences were observed early in infection, by day 5, the TIM1 -/-mice had significantly lower titers in the serum, liver, kidneys, and adrenal glands (149), suggesting TIM1 may play a role in virus load. To examine the role of T cells in TIM1-deficient mice, mice were depleted of T cells and infected with rVSV/EBOV. T cell depletion did not alter mouse survival, suggesting that in this model of disease, T cells do not play a significant role (149).
TIM1 plays a significant role in mouse models of tick-borne encephalitis virus disease (60). Ifnar -/-mice all succumbed to TBEV infection while 80% of Ifnar -/-/TIM1 -/-mice survived infection. Surviving mice showed little weight loss and lower viral titers than Ifnar -/-that produced TIM1. Even in wild-type C57BL/6 mice lacking TIM1, they were significantly more likely to survive TBEV-Neudoerfl infection than wild-type mice. TIM1's presence correlated with significantly more virus in the serum, brain, kidney, and lungs (60).
HAV results in mild or asymptomatic disease in most patients; however, a subset of patients develops serious hepatitis. Interestingly, HAV-induced severe liver disease is associated with an insertion in TIM1 which improves TIM1-HAV binding (150). The insertion adds six amino acids within the mucin domain of TIM1, and the insertion is also associated with HIV disease progression (151). While this association was observed in human cohorts, TIM1 did not play a role in entry into Huh-7.5 human hepatoma cells and mice lacking IFNAR1/TIM1 or IFNAR1/TIM4 remained susceptible to HAV infection and disease (42), suggesting TIM1 may play a limited role in HAV-induced disease.
## TARGETING APOPTOTIC MIMICRY FOR THERAPEUTICS
Rather than trying to reduce viral entry through genetic ablation of various PSRs, Soares et al. used an antibody against PS, bavituximab, to mask virion PS, preventing entry (152). Fifty percent of the guinea pigs infected with a lethal dose of Pichinde virus were protected if provided bavituximab 7 days following infection, and survival was improved if given in combination with the nucleoside analog ribavirin (152). Bavituximab also binds to the Ebola virus and EBOV-infected cells (153).
Rather than blocking the PS with an antibody, Song et al. engineered a soluble TIM domain (sTIMEdMLDR801). It was capable of binding to virion PS/PE and reduced viral infection in cell culture systems (154). The molecule slightly reduces Zika viral levels in a mouse model in the serum and spleen but did not alter levels in the brain, suggesting additional modifications will be needed for use as a therapeutic.
Bodily fluids such as saliva and semen are rich in extracellular vesicles (EVs), which can be antiviral (155)(156)(157). EVs are rich in PS, and EV production is enhanced in cells with high outer leaflet PS, suggesting EVs contain exposed PS (158,159). Similar to lab-produced PS-containing liposomes, PS-rich EVs purified from human semen, saliva, urine, and milk were able to block Zika infection in Vero cells (160). The EVs from semen contained the highest levels of PS and were most effective in blocking viruses that utilize apoptotic mimicry including WNV, CHIKV, and EBOV-VLPs. Groß et al. noted viruses that readily transmit in bodily fluids such as HIV, herpesvirus, and hepatitis C virus do not use apoptotic mimicry. Many of the viruses that use apoptotic mimicry, DENV, ZIKV, CHIKV, EBOV, spread throughout the host in the blood, where EVs incorporate little PS (160). They suggest PS-rich EV-like particles could be developed into a therapeutic. Drugs or gene therapy formulated into lipid nanoparticles are seen to have increased effectiveness when the lipid species composition includes PS (161)(162)(163)(164).
## FUTURE QUESTIONS AND PERSPECTIVE
Several outstanding questions remain regarding viral apoptotic mimicry. For example, some viruses utilize select PSRs, but not all. For example, SARS-CoV-2 entry is enhanced by Axl more than TIM1, while DENV entry is more efficient with TIM1 compared to Axl (23,165). Are all PSRs interacting with the viral membrane in a similar manner, or are there virion differences that result in the selective use of PSRs? Are other cellular factors needed for different PSRs to efficiently internalize viruses? Does virion structure play a role in PSR usage? What is the role of apoptotic mimicry in the host? Can an effective, broadly active therapeutic be developed that effectively blocks viral entry via apoptotic mimicry?
## References
1. Chen (2019) "Molecular mechanism of HIV-1 entry" *Trends Microbiol*
2. Kaplan, Roitburd-Berman, Lewis et al. (2016) "Range of CD4-bound conformations of HIV-1 gp120, as defined using condi tional CD4-induced antibodies" *J Virol*
3. Li, Deng, Yang et al. (2019) "Effects of CD4 binding on conformational dynamics, molecular motions, and thermodynamics of HIV-1 gp120" *IJMS*
4. Mercer, Helenius (2008) "Vaccinia virus uses macropinocytosis and apoptotic mimicry to enter host cells" *Science*
5. Vanlandschoot, Leroux-Roels (2003) "Viral apoptotic mimicry: an immune evasion strategy developed by the hepatitis B virus?" *Trends Immunol*
6. Segawa, Nagata (2015) "An apoptotic "eat me" signal: phosphatidyl serine exposure" *Trends Cell Biol*
7. Brindley (2021) "mSphere of influence: apoptotic mimicry and virus entry"
8. Acciani, Brindley (2022) "Scrambled or flipped: 5 facts about how cellular phosphatidylserine localization can mediate viral replication" *PLoS Pathog*
9. Van Den Eijnde, Boshart, Baehrecke et al. (1998) "Cell surface exposure of phosphati dylserine during apoptosis is phylogenetically conserved" *Apoptosis*
10. Poon, Lucas, Rossi et al. (2014) "Apoptotic cell clearance: basic biology and therapeutic potential" *Nat Rev Immunol*
11. Segawa, Suzuki, Nagata (2011) "Constitutive exposure of phosphatidylserine on viable cells" *Proc Natl Acad Sci U S A*
12. Naeini, Bianconi, Pirro et al. (2020) "The role of phosphati dylserine recognition receptors in multiple biological functions" *Cell Mol Biol Lett*
13. Meesmann, Fehr, Kierschke et al. (2010) "Decrease of sialic acid residues as an eat-me signal on the surface of apoptotic lymphocytes" *J Cell Sci*
14. Quinn, Brindley, Weller et al. (2009) "Rho GTPases modulate entry of Ebola virus and vesicular stomatitis virus pseudotyped vectors" *J Virol*
15. Roy (2020) "TAM receptors: a phosphatidylserine receptor family and its implications in viral infections" *Int Rev Cell Mol Biol*
16. Evans, Liu (2019) "Multifaceted roles of TIM-family proteins in virushost interactions" *Trends Microbiol*
17. Morizono, Chen (2014) "Role of phosphatidylserine receptors in enveloped virus infection" *J Virol*
18. Moller-Tank (2014) "Phosphatidylserine receptors: enhancers of enveloped virus entry and infection" *Virology (Auckland)*
19. Kobayashi, Karisola, Peña-Cruz et al. (2007) "TIM-1 and TIM-4 glycoproteins bind phosphatidylserine and mediate uptake of apoptotic cells" *Immunity*
20. Miyanishi, Tada, Koike et al. (2007) "Identification of Tim4 as a phosphatidylserine receptor" *Nature*
21. Morizono, Xie, Olafsen et al. (2011) "The soluble serum protein Gas6 bridges virion envelope phosphatidyl serine to the TAM receptor tyrosine kinase Axl to mediate viral entry" *Cell Host Microbe*
22. Chongsaritsinsuk, Steigmeyer, Mahoney et al. (2023) "Glycoproteomic landscape and structural dynamics of TIM family immune checkpoints enabled by mucinase SmE" *Nat Commun*
23. Meertens, Carnec, Lecoin et al. (2012) "The TIM and TAM families of phosphatidylserine receptors mediate dengue virus entry" *Cell Host Microbe*
24. Jemielity, Wang, Chan et al. (2013) "TIMfamily proteins promote infection of multiple enveloped viruses through virion-associated phosphatidylserine" *PLoS Pathog*
25. Moller-Tank, Albritton, Rennert (2014) "Characterizing functional domains for TIM-mediated enveloped virus entry" *J Virol*
26. Dekruyff, Bu, Ballesteros et al. (2010) "T cell/transmembrane, Ig, and mucin-3 allelic variants differentially recognize phosphatidylserine and mediate phagocytosis of apoptotic cells" *J Immunol*
27. Freeman, Casasnovas, Umetsu et al. (2010) "TIM genes: a family of cell surface phosphatidylserine receptors that regulate innate and adaptive immunity" *Immunol Rev*
28. Dejarnac, Hafirassou, Chazal et al. (2018) "TIM-1 ubiquitination mediates dengue virus entry" *Cell Rep*
29. Ichimura, Bonventre, Bailly et al. (1998) "Kidney injury molecule-1 (KIM-1), a putative epithelial cell adhesion molecule containing a novel immunoglobulin domain, is upregulated in renal cells after injury" *J Biol Chem*
30. De Souza, Oriss, Kj et al. (2005) "TIM-1) is expressed on in vivo-activated T cells and provides a costimulatory signal for T cell activation" *T cell Ig and mucin*
31. Lee, Meyer, Goya et al. (2010) "Apoptotic cells activate NKT cells through T cell Ig-like mucin-like-1 resulting in airway hyperreactivity" *J Immunol*
32. Li, Ju, Frieri (2013) "The T-cell immunoglobulin and mucin domain (Tim) gene family in asthma, allergy, and autoimmunity" *Allergy Asthma Proc*
33. Wong, Valdez, Tan et al. (2010) "Phosphatidylserine receptor Tim-4 is essential for the maintenance of the homeostatic state of resident peritoneal macrophages" *Proc Natl Acad Sci*
34. Nurtanio, Yang (2011) "Role of TIM-4 in innate or adaptive immune response" *N Am J Med Sci*
35. Meyers, Chakravarti, Schlesinger et al. (2005) "TIM-4 is the ligand for TIM-1, and the TIM-1-TIM-4 interaction regulates T cell proliferation" *Nat Immunol*
36. Yanagihashi, Segawa, Maeda et al. (2017) "Mouse macrophages show different requirements for phosphatidylser ine receptor Tim4 in efferocytosis" *Proc Natl Acad Sci*
37. (2025) "Minireview mSphere October"
38. Flannagan, Canton, Furuya et al. (2014) "The phosphatidylserine receptor TIM4 utilizes integrins as coreceptors to effect phagocytosis" *Mol Biol Cell*
39. Kaplan, Totsuka, Thompson et al. (1996) "Identification of a surface glycoprotein on African green monkey kidney cells as a receptor for hepatitis A virus" *EMBO J*
40. Thompson, Lu, Kaplan (1998) "The Cys-rich region of hepatitis A virus cellular receptor 1 is required for binding of hepatitis A virus and protective monoclonal antibody 190/4" *J Virol*
41. Feng, Hensley, Mcknight et al. (2013) "A pathogenic picornavirus acquires an envelope by hijacking cellular membranes" *Nature*
42. Kachko, Costafreda, Zubkova et al. (2018) "Determinants in the Ig variable domain of human HAVCR1 (TIM-1) are required to enhance hepatitis C virus entry" *J Virol*
43. Das, Hirai-Yuki, González-López et al. (2017) "TIM1 (HAVCR1) is not essential for cellular entry of either quasi-enveloped or naked hepatitis A virions" *mBio*
44. 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*
45. Zhang, Kitzmiller, Richard et al. (2024) "The ability of human TIM1 to bind phosphatidylethanolamine enhances viral uptake and efferocytosis compared to rhesus and mouse orthologs" *J Virol*
46. Ichimura, Asseldonk, Humphreys et al. (2008) "Kidney injury molecule-1 is a phosphatidylserine receptor that confers a phagocytic phenotype on epithelial cells" *J Clin Invest*
47. Richard, Zhang, Park et al. (2015) "Virionassociated phosphatidylethanolamine promotes TIM1-mediated infection by Ebola, dengue, and West Nile viruses" *Proc Natl Acad Sci U S A*
48. Moller-Tank, Kondratowicz, Davey et al. (2013) "Role of the phosphatidylserine receptor TIM-1 in enveloped-virus entry" *J Virol*
49. Kuroda, Fujikura, Noyori et al. (2014) "A polymorphism of the TIM-1 IgV domain: Implications for the susceptibility to filovirus infection" *Biochem Biophys Res Commun*
50. Rhein, Brouillette, Schaack et al. (2016) "Characterization of human and murine T-cell immunoglobulin mucin domain 4 (TIM-4) IgV domain residues critical for Ebola virus entry" *J Virol*
51. Dahlmann, Biedenkopf, Babler et al. (2015) "Analysis of Ebola virus entry into macrophages" *J Infect Dis*
52. Shimojima, Takada, Ebihara et al. (2006) "Tyro3 family-mediated cell entry of Ebola and Marburg viruses" *J Virol*
53. Kirui, Abidine, Lenman et al. (2021) "The phosphatidylserine receptor TIM-1 enhances authentic chikungunya virus cell entry" *Cells*
54. Ballista, Miazgowicz, Acciani et al. (2023) "Chikungunya virus entry and infectivity is primarily facilitated through cell line dependent attachment factors in mammalian and mosquito cells" *Front Cell Dev Biol*
55. Carnec, Meertens, Dejarnac et al. (2016) "The phosphatidylser ine and phosphatidylethanolamine receptor CD300A binds dengue virus and enhances infection" *J Virol*
56. Brouillette, Phillips, Patel et al. (2018) "TIM-1 mediates dystroglycan-independent entry of lassa virus" *J Virol*
57. Fedeli, Torriani, Galan-Navarro et al. (2018) "Axl can serve as entry factor for lassa virus depending on the functional glycosylation of dystroglycan" *J Virol*
59. Niu, Jiang, Xu et al. (2018) "TIM-1 promotes Japanese encephalitis virus entry and infection" *Viruses*
60. Hamel, Dejarnac, Wichit et al. (2015) "Biology of zika virus infection in human skin cells" *J Virol*
61. Chu, Yang, Peng et al. (2019) "TIM-1 as a signal receptor triggers dengue virus-induced autophagy" *IJMS*
62. Zhang, Liang, Wang et al. (2022) "T-cell immunoglobulin and mucin domain 1 (TIM-1) is a functional entry factor for tick-borne encephalitis virus" *mBio*
63. Wei, Li, Qiao et al. (2020) "Porcine reproduc tive and respiratory syndrome virus utilizes viral apoptotic mimicry as an alternative pathway to infect host cells" *J Virol*
64. Chen, Zheng, Li et al. (2023) "Coreceptor AXL facilitates African swine fever virus entry via apoptotic mimicry" *J Virol*
65. Bohan, Van Ert, Ruggio et al. (2021) "Phosphatidylserine receptors enhance SARS-CoV-2 infection" *PLoS Pathog*
66. Mayor, Torriani, Rothenberger et al. (2020) "T-cell immunoglo bulin and mucin (TIM) contributes to the infection of human airway epithelial cells by pseudotype viruses containing Hantaan virus glycoproteins" *Virology (Auckl)*
67. Ma, Zheng, Gu et al. (2024) "T-cell immunoglobulin and mucin 1 (TIM-1) mediates infection of Hantaan virus in Jurkat T cells" *Virus Res*
68. Costafreda, Kaplan (2018) "HAVCR1 (CD365) and its mouse ortholog are functional hepatitis a virus (hav) cellular receptors that mediate HAV infection" *J Virol*
69. Wang, Qiao, Hou et al. (2017) "TIM-1 promotes hepatitis C virus cell attachment and infection" *J Virol*
70. Corneillie, Lemmens, Montpellier et al. (2023) "The phosphatidylserine receptor TIM1 promotes infection of enveloped hepatitis E virus" *Cell Mol Life Sci*
71. Zhang, Wang, Hu et al. (2022) "TIM-1 augments cellular entry of ebola virus species and mutants, which is blocked by recombinant TIM-1 protein" *Microbiol Spectr*
72. Yu, Zheng, Sheridan et al. (2025) "Autophagy-mediated downregulation of AXL and TIM-1 promotes sustained Zika virus infection" *Proc Natl Acad Sci U S A*
73. Liang, Pan, Xie et al. (2023) "Interaction between hTIM-1 and envelope protein is important for JEV infection" *Viruses*
74. (2025) "Minireview mSphere October"
75. Yuan, Cao, Ling et al. (2015) "TIM-1 acts a dual-attachment receptor for Ebolavirus by interacting directly with viral GP and the PS on the viral envelope" *Protein Cell*
76. 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*
77. Sasaki, Knyazev, Clout et al. (2006) "Structural basis for Gas6-Axl signalling" *EMBO J*
78. Rescigno, Mansukhani, Basilico (1991) "A putative receptor tyrosine kinase with unique structural topology" *Oncogene*
79. Stitt, Conn, Goret et al. (1995) "The anticoagulation factor protein S and its relative, Gas6, are ligands for the Tyro 3/Axl family of receptor tyrosine kinases" *Cell*
80. Rajotte, Hasanbasic, Blostein (2008) "Gas6-mediated signaling is dependent on the engagement of its gamma-carboxyglutamic acid domain with phosphatidylserine" *Biochem Biophys Res Commun*
81. Zhang, Richard, Jackson et al. (2020) "Phosphatidyle thanolamine and phosphatidylserine synergize to enhance GAS6/AXLmediated virus infection and efferocytosis" *J Virol*
82. Tsou, Nguyen, Calarese et al. (2014) "Receptor tyrosine kinases, TYRO3, AXL, and MER, demonstrate distinct patterns and complex regulation of ligand-induced activation" *J Biol Chem*
83. Lemke, Burstyn-Cohen (2010) "TAM receptors and the clearance of apoptotic cells" *Ann N Y Acad Sci*
84. Henson, Hume (2006) "Apoptotic cell removal in development and tissue homeostasis" *Trends Immunol*
85. Ravichandran (2010) "Find-me and eat-me signals in apoptotic cell clearance: progress and conundrums" *J Exp Med*
86. Suh, Hilliard, Li et al. (2010) "TAM receptor ligands in lupus: protein S but not Gas6 levels reflect disease activity in systemic lupus erythematosus" *Arthritis Res Ther*
87. Ye, Han, Lu et al. (2011) "Retinal self-antigen induces a predominantly Th1 effector response in Axl and Mertk double-knockout mice" *J Immunol*
88. Wium, Paccez, Zerbini (2018) "The dual role of TAM receptors in autoimmune diseases and cancer: an overview" *Cells*
89. Chen, Wang, Qi et al. (2009) "Functions of TAM RTKs in regulating spermatogenesis and male fertility in mice" *Reproduction*
90. Lu, Gore, Zhang et al. (1999) "Tyro-3 family receptors are essential regulators of mammalian spermatogenesis" *Nature*
91. Rothlin, Ghosh, Zuniga et al. (2007) "TAM receptors are pleiotropic inhibitors of the innate immune response" *Cell*
92. Lemke (2013) "Biology of the TAM receptors" *Cold Spring Harb Perspect Biol*
93. Wang, Zhen, Fan et al. (2007) "Axl deficiency promotes the neuroinvasion of Japanese encephalitis virus by enhancing IL-1α production from pyroptotic macrophages" *J Infect Dis*
94. Hunt, Kolokoltsov, Davey (2011) "The Tyro3 receptor kinase Axl enhances macropinocytosis of Zaire ebolavirus" *J Virol*
95. Brindley, Hunt, Kondratowicz et al. (2011) "Tyrosine kinase receptor Axl enhances entry of Zaire ebolavirus without direct interactions with the viral glycoprotein" *Virology (Auckl)*
96. Dimitrova, Zenarruzabeitia, Borrego et al. (2016) "CD300c is uniquely expressed on CD56bright Natural Killer Cells and differs from CD300a upon ligand recognition" *Sci Rep*
97. Simhadri, Andersen, Calvo et al. (2012) "Human CD300a binds to phosphatidylethanolamine and phosphatidyl serine, and modulates the phagocytosis of dead cells" *Blood*
98. Izawa, Isobe, Matsukawa et al. (2014) "Sphingomyelin and ceramide are physiological ligands for human LMIR3/CD300f, inhibiting FcεRI-mediated mast cell activation" *J Allergy Clin Immunol*
99. Zenarruzabeitia, Vitallé, Eguizabal et al. (2015) "The biology and disease relevance of CD300a, an inhibitory receptor for phosphatidylserine and phosphatidylethanolamine" *J Immunol*
100. Orchard, Wilen, Doench et al. (2016) "Discovery of a proteinaceous cellular receptor for a norovirus" *Science*
101. Asano, Miwa, Miwa et al. (2004) "Masking of phosphatidylserine inhibits apoptotic cell engulfment and induces autoantibody production in mice" *J Exp Med*
102. Hanayama, Tanaka, Miwa et al. (2002) "Identification of a factor that links apoptotic cells to phagocytes" *Nature*
103. Newburg, Peterson, Ruiz-Palacios et al. (1998) "Role of human-milk lactadherin in protection against symptomatic rotavirus infection" *Lancet*
104. Kvistgaard, Pallesen, Arias et al. (2004) "Inhibitory effects of human and bovine milk constituents on rotavirus infections" *J Dairy Sci*
105. Park, Kehrl (2019) "An integrin/MFG-E8 shuttle loads HIV-1 viral-like particles onto follicular dendritic cells in mouse lymph node" *Elife*
106. Ballista, Hoover, Noble et al. (2024) "Chikungunya virus release is reduced by TIM-1 receptors through binding of envelope phosphatidylserine" *J Virol*
107. Emeny, Morgan (1979) "Regulation of the interferon system: evidence that Vero cells have a genetic defect in interferon production" *J Gen Virol*
108. Wanninger, Millian, Saldarriaga et al. (2022) "Macrophage infection, activation, and histopathological findings in ebolavirus infection" *Front Cell Infect Microbiol*
109. Bhattacharyya, Zagórska, Lew et al. (2013) "Enveloped viruses disable innate immune responses in dendritic cells by direct activation of TAM receptors" *Cell Host Microbe*
110. Minami, Kotaki, Sakai et al. (2025) "Maeda Minireview mSphere October"
111. Okamoto, Tachibana, Matsuura et al. (2024) "Vero celladapted SARS-CoV-2 strain shows increased viral growth through furinmediated efficient spike cleavage" *Microbiol Spectr*
112. Wang, Qiu, Hou et al. (2021) "AXL is a candidate receptor for SARS-CoV-2 that promotes infection of pulmonary and bronchial epithelial cells" *Cell Res*
113. Hill, Zeidel (2000) "Reconstituting the barrier properties of a water-tight epithelial membrane by design of leaflet-specific liposomes" *J Biol Chem*
114. Gupta, Korte, Herrmann et al. (2020) "Plasma membrane asymmetry of lipid organization: fluorescence lifetime microscopy and correlation spectroscopy analysis" *J Lipid Res*
115. Li, Shi, Guo et al. (2014) "Ionic protein-lipid interaction at the plasma membrane: what can the charge do?" *Trends Biochem Sci*
116. Lemmon (2008) "Membrane recognition by phospholipid-binding domains" *Nat Rev Mol Cell Biol*
117. Dolis, Moreau, Zachowski et al. (1997) "Aminophospholipid translocase and proteins involved in transmembrane phospholipid traffic" *Biophys Chem*
118. Sebastian, Baldridge, Xu et al. (2012) "Phospholipid flippases: building asymmetric membranes and transport vesicles" *Biochim Biophys Acta*
119. Norris, Mansueto, Jimenez et al. (2024) "Flipping the script: advances in understanding how and why P4-ATPases flip lipid across membranes" *Biochim Biophys Acta Mol Cell Res*
120. Nagata, Sakuragi, Segawa (2020) "Flippase and scramblase for phosphatidylserine exposure" *Curr Opin Immunol*
121. Tanaka, Fujimura-Kamada, Yamamoto (2011) "Functions of phospholipid flippases" *J Biochem*
122. Jeong, Conboy (2011) "Phosphatidylserine directly and positively regulates fusion of myoblasts into myotubes" *Biochem Biophys Res Commun*
123. Falzone, Malvezzi, Lee et al. (2018) "Known structures and unknown mechanisms of TMEM16 scramblases and channels" *J Gen Physiol*
124. Suzuki, Denning, Imanishi et al. (2013) "Xk-related protein 8 and CED-8 promote phosphatidylserine exposure in apoptotic cells" *Science*
125. Younan, Iampietro, Santos et al. (2018) "Role of transmembrane protein 16F in the incorporation of phosphatidylserine into budding Ebola virus virions" *J Infect Dis*
126. Nanbo, Maruyama, Imai et al. (2018) "Ebola virus requires a host scramblase for externalization of phosphatidylserine on the surface of viral particles" *PLoS Pathog*
127. Acciani, Mendoza, Havranek et al. (2021) "Ebola virus requires phosphatidylserine scrambling activity for efficient budding and optimal infectivity" *J Virol*
128. Holthuis, Menon (2014) "Lipid landscapes and pipelines in membrane homeostasis" *Nature*
129. Martín-Acebes, Merino-Ramos, Blázquez et al. (2014) "The composition of West Nile virus lipid envelope unveils a role of sphingolipid metabolism in flavivirus biogenesis" *J Virol*
130. Hitakarun, Williamson, Yimpring et al. (2022) "Cell type variability in the incorporation of lipids in the dengue virus virion" *Viruses*
131. Mazzetto, Bortolami, Bovo et al. (2025) "Infectivity in full-term placenta of Zika viruses with different lipid profiles" *Virus Res*
132. Richard, Shim, Kwon et al. (2017) "AXL-dependent infection of human fetal endothelial cells distinguishes Zika virus from other pathogenic flaviviruses" *Proc Natl Acad Sci U S A*
133. Yin, Ambardekar, Lu et al. (2016) "Distinct entry mechanisms for nonenveloped and quasi-enveloped hepatitis E viruses" *J Virol*
134. Chen, Du, Hagemeijer et al. (2015) "Phosphatidylserine vesicles enable efficient en bloc transmission of enteroviruses" *Cell*
135. Costafreda, Abbasi, Lu et al. (2020) "Exosome mimicry by a HAVCR1-NPC1 pathway of endosomal fusion mediates hepatitis A virus infection" *Nat Microbiol*
136. Wang, Rumin, Doktorova et al. (2025) "Loss of lipid asymmetry facilitates plasma membrane blebbing by decreasing membrane lipid packing" *Proc Natl Acad Sci U S A*
137. Dixson, Dawson, Vizio et al. (2023) "Context-specific regulation of extracellular vesicle biogenesis and cargo selection" *Nat Rev Mol Cell Biol*
138. Li, Ablan, Miao et al. (2014) "TIM-family proteins inhibit HIV-1 release" *Proc Natl Acad Sci U S A*
139. Li, Waheed, Yu et al. (2019) "TIM-mediated inhibition of HIV-1 release is antagonized by Nef but potentiated by SERINC proteins" *Proc Natl Acad Sci*
140. Yu, Zheng, Sheridan et al. (2025) "Autophagy-mediated downregulation of AXL and TIM-1 promotes sustained zika virus infection" *bioRxiv*
141. Xie, Liang, Yang et al. (2021) "Japanese encephalitis virus NS2B-3 protein complex promotes cell apoptosis and viral particle release by down-regulating the expression of AXL" *Virol Sin*
142. Perera, Riley, Isaac et al. (2012) "Dengue virus infection perturbs lipid homeostasis in infected mosquito cells" *PLoS Pathog*
143. Liu, Delalio, Isakson et al. (2016) "AXL-mediated productive infection of human endothelial cells by Zika virus" *Circ Res*
144. Zapatero-Belinchón, Dietzel, Dolnik et al. (2019) "Characterization of the filovirus-resistant cell line SH-SY5Y reveals redundant role of cell surface entry factors" *Viruses*
145. Miner, Sene, Richner et al. (2016) "Zika virus infection in mice causes panuveitis with shedding of virus in tears" *Cell Rep*
146. Hastings, Yockey, Jagger et al. (2017) "TAM receptors are not required for Zika virus infection in mice" *Cell Rep*
147. Hastings, Hastings, Uraki et al. (2019) "Loss of the TAM receptor AXL ameliorates severe Zika virus pathogenesis and reduces apoptosis in microglia"
148. Adomati, Cham, Hamdan et al. (2020) "Dead cells induce innate anergy via mertk after acute viral infection" *Cell Rep*
149. Miner, Daniels, Shrestha et al. (2015) "The TAM receptor Mertk protects against neuroinvasive viral infection by maintaining blood-brain barrier integrity" *Nat Med*
150. Yang, Li, Yuan et al. (2022) "Axl-/-neurons promote JEV infection by dampening the innate immunity" *Virus Res*
151. Younan, Iampietro, Nishida et al. (2017) "Ebola virus binding to Tim-1 on T lymphocytes induces a cytokine storm" *mBio*
152. Brunton, Rogers, Phillips et al. (2019) "TIM-1 serves as a receptor for Ebola virus in vivo, enhancing viremia and pathogenesis" *PLoS Negl Trop Dis*
153. Kim, Eyheramonho, Pichavant et al. (2011) "A polymorphism in TIM1 is associated with susceptibility to severe hepatitis A virus infection in humans" *J Clin Invest*
154. Biasin, Sironi, Saulle et al. (2017) "A 6-amino acid insertion/deletion polymorphism in the mucin domain of TIM-1 confers protections against HIV-1 infection" *Microbes Infect*
155. Soares, King, Thorpe (2008) "Targeting inside-out phosphati dylserine as a therapeutic strategy for viral diseases" *Nat Med*
156. Dowall, Graham, Corbin-Lickfett et al. (2015) "Effective binding of a phosphatidylserine-targeting antibody to Ebola virus infected cells and purified virions" *J Immunol Res*
157. Song, Garcia, Situ et al. (2021) "Development of a blocker of the universal phosphatidylserine-and phosphatidylethanolaminedependent viral entry pathways" *Virology (Auckl)*
158. Müller, Harms, Krüger et al. (2018) "Semen inhibits Zika virus infection of cells and tissues from the anogenital region" *Nat Commun*
159. Conzelmann, Groß, Zou et al. (2020) "Salivary extracellular vesicles inhibit Zika virus but not SARS-CoV-2 infection" *J Extracell Vesicles*
160. Yáñez-Mó, Siljander, Andreu et al. (2015) "Biological properties of extracellular vesicles and their physiological functions" *J Extracell Vesicles*
161. Llorente, Skotland, Sylvänne et al. (2013) "Molecular lipidomics of exosomes released by PC-3 prostate cancer cells" *Biochim Biophys Acta*
162. Minh, Star, Stupak et al. (2021) "Characterization of extracellular vesicles secreted in lentiviral producing HEK293SF cell cultures" *Viruses*
163. Groß, Reßin, Maltitz et al. (2024) "Phosphatidylserineexposing extracellular vesicles in body fluids are an innate defence against apoptotic mimicry viral pathogens" *Nat Microbiol*
164. Lotter, Alter, Bolten et al. (2022) "Incorporation of phosphatidylserine improves efficiency of lipid based gene delivery systems" *Eur J Pharm Biopharm*
165. Ren, Fan, Yao et al. (2026) "A phosphatidylserine-targeted self-amplifying nanosystem improves tumor accumulation and enables efficient tumor therapy by modulat ing anticancer immunity" *Biomaterials*
166. Xu, Li, Yan et al. (2024) "Phosphatidylserine and/or sialic acid modified liposomes increase uptake by tumor-associated macrophages and enhance the anti-tumor effect" *AAPS PharmSciTech*
167. Klein, Rieckmann, Sedding et al. (2021) "Towards the development of long circulating phosphatidylser ine (PS)-and phosphatidylglycerol (PG)-enriched anti-inflammatory liposomes: Is PEGylation effective?" *Pharmaceutics*
168. Bohan, Ert, Ruggio et al. (2021) "Phosphatidylserine receptors enhance SARS-CoV-2 infection: AXL as a therapeutic target for" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12791831&blobtype=pdf | # Effect of Recipient Age and Cytomegalovirus Sero-status on Patient and Graft Survival in Liver Transplant Recipients
Abhay Dhand, Seigo Nishida, Kenji Okumura
Background. Cytomegalovirus (CMV) may impact allograft function and overall survival in solid organ transplant recipients in various direct or indirect mechanisms. The association between recipient age and donor-recipient CMV status and outcomes of liver transplantation (LT) are unclear.
Methods. Retrospective analysis of the Scientific Registry of Transplant Recipients for LT recipients ≥18 years old between 2000-2020 (N=89,753) was conducted. Recipient age-based stratification analysis was performed to compare the association of CMV serostatus and recipient age on overall and graft survival after LT. Survival curves were generated using the Kaplan-Meier method. All-cause mortality and graft failure hazard ratios were calculated at one-and fiveyears post-LT.
Poster Abstracts • OFID 2026:13 (Suppl 1) • S1109 |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12368227&blobtype=pdf | # Metagenomic analysis and genomic characterization of enterovirus A76 and Norovirus GI.6[P11] coinfection in a patient with acute gastroenteritis in Thailand
Watchaporn Chuchaona, Ray Izquierdo-Lara, Claudia Schapendonk, Sarawut Khongwichit, Marion Koopmans, Miranda De Graaf, Yong Poovorawan
## Abstract
Acute gastroenteritis (AGE) remains a significant global health concern, with noroviruses among the most prevalent viral pathogens. However, other enteric viruses also contribute substantially to the public health burden. This study provides the first molecular characterization of a co-infection involving a rarely reported enterovirus A76 (EV-A76) and a norovirus GI.6[P11] in a patient from Thailand. Metagenomic sequencing successfully identified complete viral genomes, revealing unique genetic variations. Phylogenetic analysis demonstrated that the EV-A76 strain shares high nucleotide similarity with a recently reported strain from Nepal, distinguishing it from previously identified recombinant strains. The amino acid sequence alignment of the complete EV-A76 genome revealed several distinctive amino acid substitutions compared to the most closely related strains. Notably, variations in the VP1 C-terminus and VP2 EF loop, known for high variability, were observed. These regions, crucial for epitope formation, are particularly susceptible to high-frequency mutations. This study reports the first documented co-infection of EV-A76 and norovirus GI.6[P11] in a single sample, identified through metagenomic sequencing in an AGE case in Thailand in 2023. The observed genetic variations highlight the necessity for ongoing monitoring of viral diversity to strengthen genomic surveillance and inform prevention strategies, especially for emerging pathogens with significant public health implications.
sporadic cases and outbreaks, highlighting the importance of its molecular characterization for epidemiological surveillance, particularly GI. 6[P11] which has been detected as a predominant genotype in recent outbreaks in China 15 . However, genetic data for this genotype remains limited in Thailand.
Enteroviruses, which belong to the Picornaviridae family 16 are also common causes of viral gastroenteritis and are associated with a broad range of diseases in both children and adults. Enteroviruses have a non-segmented, positive-sense, single-stranded RNA genome of approximately 7.5 kb, encoding a polyprotein that is cleaved into four structural proteins (VP1-VP4) and seven non-structural proteins (2A-2C and 3A-3D) by two viral proteases (2A pro and 3C pro ) 17 . The VP1, VP2, and VP3 proteins constitute the capsid surface of the virus, while VP4 is located on the inside. The arrangement of these capsid proteins plays a crucial role in viral antigenicity by forming conformational epitopes that are important for antigenicity and host immune recognition 18 . The viral capsid proteins are divided into three sites: site 1, site 2, and site 3. Site 1 is situated in the "north rim" area of the canyon near the fivefold axis. Site 2 is positioned in the "puff " region close to the twofold axis. Site 3 comprised two sections-one in the "knob" region on the south rim of the canyon and the other in the threefold axis region 19,20 . Conformational epitopes exhibit specificity in binding to antibodies that inhibit the interaction between cellular receptors and the virus. Furthermore, they are basic determinants of viral antigenicity and serve as critical markers for the classification of enterovirus serotypes 21 . Among Enterovirus species A, Enterovirus A71 (EV-A71), Coxsackievirus A6 (CV-A6), and Coxsackievirus A16 (CV-A16) are commonly associated with hand, foot, and mouth disease (HFMD) and herpangina, while EV-D68 (a species D enterovirus) is primarily associated with respiratory infections. Co-infections with multiple enterovirus species are frequently detected in patients presenting with gastrointestinal, neurological, and respiratory illnesses 22 .
Although many enterovirus strains are well studied, EV-A76 is a relatively rare type within the enterovirus A species, with a worldwide prevalence estimated to be less than 1% of reported enterovirus strains 22 . Due to its limited global epidemiological data, further genetic analysis of EV-A76 is crucial to better understand its genetic characteristics.
In this study, we report a co-infection in a patient with both EV-A76 and norovirus GI.6[P11] strains, and we further analyze the genetic variation of these two enteric viruses.
## Results
## Metagenomic sequencing of Norovirus GI.6 ([P6] and [P11]) samples
There were five GI.6-infected patient samples sequenced by metagenomics, which revealed co-infections with other enteric viruses in two of the samples (B9955 and B10503), as shown in Table 1. In sample B9955, a rare EV-A76 strain was co-detected in a norovirus GI. 6[P11] positive sample. This co-infection was identified in a 58-year-old woman from Bangkok, with a recent history of international travel; however, the country visited was not reported. On the third day after returning to Thailand, she consumed papaya salad and developed diarrhea. She was diagnosed with gastroenteritis of presumed infectious origin. Over three days, her symptoms included approximately five episodes of watery diarrhea per day, nausea, and abdominal cramping, without fever or vomiting.
## Phylogenetic analysis of Norovirus GI.6[P6] and GI.6[P11] strains
The phylogenetic analysis was conducted to investigate the genetic relationships among norovirus GI.6 strains based on the complete RdRp and VP1 genes. Phylogenetic analysis of both two genes of norovirus GI.6 strains revealed that the GI.6[P6] strain from this study clustered with a previously reported strain from Japan (MW305501) for the complete RdRp gene, sharing 98.7% nucleotide identity. For the complete VP1 gene, this strain closely clustered with reference strains from Japan, Taiwan, and the USA, presenting nucleotide identities ranging from 97.4 to 98.7% (Fig. 1). The GI.6[P11] strain (B9471), collected from Chaiyaphum province in 2023, was also most closely related to reference strains from Japan (LC646339) and China (MW243609) for both genes. It showed nucleotide identities of 99.2-99.5% for the RdRp gene and 99.1-99.2% for the VP1 gene. Notably, three GI.6[P11] strains collected in Bangkok formed a distinct cluster, separate from the reference strains.
Simplot nucleotide pairwise analysis of the complete genome sequence of norovirus GI.6[P6] and reference strains was performed using the prototype of GI.6[P6] strain (AF093797) as the query, revealing nucleotide sequence identity percentages consistent with the prototype strain and showing a similar trend in similarity percentages to closely related strains, ranging from 81 to 99% (Fig. 2A) with the most diversity observed in p48 gene. Next, the GI.6[P11] strains were investigated to identify prominent amino acid residues (Fig. 2B), that differentiate these strains from previous strains, including the most closely related strains from China that caused outbreaks between 2016 and 2019 (OR267445, OR267447, OR268208, and OR268635). We aligned complete genome sequences from different lineages, which revealed numerous unique residue changes, particularly in the ORF1 and 2 regions. The amino acid sequences of the histo-blood group antigens (HBGAs) binding sites were consistent across all strains. However, ten amino acid substitutions in the complete VP1 gene (10 A, 296I, 311 S, 338 T, 343 S, 356 V, 357 T, 378 N, 391 S, and 424 M) were identified, potentially affecting other viral characteristics.
## Phylogenetic analysis of Thai EV-76 strain
Phylogenetic analysis of the complete genome sequences (Fig. 3A) revealed that the EV-A76 strain from Thailand is genetically closest to a strain identified from a stool sample collected in Nepal in 2023 (PP621603). This strain shares 92.2% nucleotide identity with the complete genome sequence of the Nepal strain, while comparisons with other strains showed nucleotide identities ranging from 84.8 to 90.7%. Alignment of the amino acid sequences revealed several distinctive amino acid substitutions in the complete genome when compared with the most closely related strains (Fig. 3B).
## Similarity plot and BootScan analyses of Thai EV-76 strain
To investigate the genetic relationships among EV-A76 strains, we performed Similarity Plot and BootScan analyses, using the SimPlot program (v 3.5.1). The analysis revealed that the nucleotide sequence identities of the Thai strain showed the highest similarity percentage (81%-99%) with the Nepal strain (PP621603) over the entire genome (Fig. 4A). The nucleotide identity in the structural protein-coding region (P1 region, including VP1-4) of the Thai EV-A76 strain was notably lower (61-96%) when compared with prototype (AY697458) and reference EV-A76 strains (ON646245, MH118030, and PP621603). This contrasts with the non-structural regions (P2 and P3 regions), which displayed higher conservation (79-97%). The BootScan analysis was conducted to investigate potential recombination events in the EV-A76 genome (Fig. 4B). Our results showed an overall consistent clustering of the Thai EV-A76 strain with the selected reference EV-A76 strains, indicating no strong evidence of recombination across most of the genome between EV-A76 and EV-A89. However, a minor recombination signal was observed at the 5′ end of the P1 region.
To further explore possible epidemiological links between the Thai EV-A76 strain and environmental isolates, we conducted a similarity plot analysis using reference strains derived from wastewater and sewage samples. The analysis revealed that the partial VP1 and partial 2C regions of the Thai EV-A76 strain showed the highest nucleotide identity with the sewage-derived strain from Nigeria (2018) compared with the wastewater strain from the USA (2020) or the sewage strain from Cameroon (2018) (Supplementary Fig. 1). In the VP1 region, nucleotide identities ranged from 94 to 98%, which were higher than those observed for the Nepalese strain (91.5-95%), whereas the environmental strains from Cameroon and the USA showed only 67.5-84.5% identity. In contrast, in the nonstructural protein region (late 2A, 2B, and partial 2C regions), the Thai strain shared higher similarity (94-96%) with the wastewater strain from the USA.
## Specific amino acid differences of the capsid gene in EV-A76 strains compared with other enterovirus A strains
The epitope alignment of EV-A76 strains was compared with previously analyzed EV-A71, CV-A16, and CV-A10 strains, as shown in Table 2. Interestingly, two unique amino acid changes were observed only in the Thai EV-A76 strain in this study: 140D in the VP2 EF loop at site 2b and 60 A in the VP3 N-terminus at site 3a.
## Discussion
Enteric viruses are a leading cause of AGE, with norovirus recognized as a major etiological agent in both sporadic cases and outbreaks among children and adults worldwide include Thailand 1,23 . Other enteric viruses including rotavirus, adenovirus, astrovirus, sapovirus, and enterovirus, are also implicated in AGE 24,25 . In this study, we provide significant insights into the molecular characteristics of norovirus GI.6[P11] strains and report the detection of the rare EV-A76 in Thailand.
Through comprehensive genomic and phylogenetic analyses, we characterized five stool samples: one norovirus GI.6[P6] and four GI.6[P11] strains, detected in AGE cases in Thailand from 2022 to 2024. The GI.6[P6] strain showed close genetic relationships with reference strains from Japan, Taiwan, and the USA. One GI.6[P11] strain from Chaiyaphum province clustered with strains from China and Japan, while three GI.6[P11] strains from Bangkok clustered separately from reference strains. Notably, no amino acid changes were observed at the HBGA binding sites I, II, and III (positions 331, 333, 347, 384, 386, and 389 in the GI.6 VP1 sequence, accession number MZ227264) 26 . However, distinct amino acid substitutions in the VP1 gene of GI.6[P11] strains were identified, which may contribute to their persistence or pathogenicity.
EV-A76 was first identified in a strain obtained from a French military recruit with AGE during a 1991 outbreak of astrovirus-associated gastroenteritis among military recruits in Caen, France 27 . Subsequent studies have linked EV-A76 with neurological conditions such as acute encephalitis in India (2006) 28 and acute flaccid paralysis (AFP) in China (2011) 29 . However, the Thai patient in this study did not present neurological symptoms. Interestingly, EV-A76 has also been detected in patients with oyster-associated gastroenteritis 30 and sporadically in wastewater samples in France (2014-2015) 31 and Argentina (2017-2018) 32 . In this study, SimPlot analysis covering the partial VP1 through partial 2C regions of our EV-A76 strain revealed the highest nucleotide identity with a sewage-derived strain from Nigeria (2018) 33 whereas the nonstructural region showed greater similarity to a wastewater-derived strain from the USA 34 . These contrasting patterns suggest that different genomic regions of the detected strain may have divergent evolutionary origins. Fig. 3. Genetic analysis of the Thai EV-A76 strain. (A) A Maximum likelihood tree of EV-A76 strains was inferred for complete nucleotide sequences using the GTR best-fit model, with 1,000 bootstrap replications for branch support as implemented in MEGA X. The scale bar indicates nucleotide substitutions per site, and selected bootstrap values greater than 80 are shown. The rare EV-A76 strain identified in Thailand is denoted with bold black and a red circle. The genome of strain Enterovirus_A76_FRA_FRA91-10369_1991 (AY697458) as prototype EV-A76 strain is highlighted in green. (B) Amino acid residues of the complete genome sequence of the Thai EV-A76 strain (red bold) was compared with other full genome EV-A76 strains, with unique amino acid changes labeled in red. Amino acids are colored by the properties of their side-chains: purple for basic -R, H, K; pink for acidic -D, E; Green for nonpolar -G, A, V, L, I, M, F, W, and P; blue for polar -S, T, C, Y, N, and Q.
Our findings suggest a broader clinical spectrum as evidenced by its detection in a patient with AGE symptoms. The nucleotide identity in the structural protein of Thai EV-A76 strain was lower when compared with reference EV-76 strains. This pattern is consistent with the high genetic variability typical of capsid-coding regions in enteroviruses 18 . Alignment of amino acid sequences between the Thai and other EV-A76 strains revealed unique mutations in several genomic regions that may have functional implications, such as altered viral replication, host adaptation, or immune evasion 18 . Notably, BootScan analysis revealed a minor recombination signal at the 5' end of the structural protein-coding region of the Thai EV-A76 strain in this study. This observation is consistent with previous large-scale genomic analyses indicating that the junction between the 5' UTR and P1 region represents a recombination hotspot in enteroviruses [35][36][37] . Moreover, the structured nature of the 5' UTR-in particular the cloverleaf and IRES domains connected by intrinsically disordered RNA segments-likely predisposes this region to template switching events during replication, facilitating recombination 38 .
The Thai EV-A76 strain exhibited the closest genetic similarity to strains from Nepal. Notably, it presented unique amino acid mutations and showed no evidence of recombination with EV-A89, unlike certain EV-A76 strains from China that contain recombinant fragments in the 2C and P3 regions 29 . The predominant enterovirus serotypes worldwide, including EV-A76, CV-A16, and EV-A71, were the most common epidemic strains in Africa from 2007 to 2017. In Asia, particularly in Bangladesh, enterovirus epidemics exhibit high serotype diversity, with EV-A76 being the most prevalent 39 . However, in other countries, EV-A76 remains a relatively rare and uncommonly reported enterovirus type 22 . In Thailand, Enterovirus species A, including CV-A6, CV-A4, EV-A71, and CV-A16, are major pathogens associated with HFMD 40 . The identification of EV-A76 is particularly significant due to its rarity and limited genomic data available in public databases as well as the association with severe disease.
The conformational epitope alignment analysis of EV-A76 compared with EV-A71, CV-A16, and EV-A10, which are the predominantly causative agents for several outbreaks of HFMD 41 . The genetic diversity of previously predicted conformational epitopes analysis identified the BC and HI loops, the C-terminus of VP1, the EF and HI loops of VP2, and the C-terminus of VP3 as regions with high variability 20 . Our study revealed multiple amino acid changes in the capsid gene. Notably, VP1 C-terminus and VP2 EF loops variations which are highly diverse regions, were also observed. These findings associated with the epitope formation are particularly susceptible to high-frequency mutations 20 . Furthermore, previous studies have indicated that the VP1 BC and EF loops play a crucial role in the immunogenicity of EV-A71 and in providing natural crossprotection against CV-A16 42 . Specifically, the VP1 BC and HI loops are involved in neural cell tropism 43 . For other enteroviruses, such as EV-D68, it has been reported that the dominant amino acid composition of the BC-loop and DE-loop continued to change ongoing viral evolution and may lead to enteroviruses outbreaks 19,44 . Moreover, VP2 and VP3 genes of EV-A71 and CV-A16 which constitute antigenic sites 2 and 3, play a crucial role in antigenic evolution 20 while the exchange of VP3 GH loop may contribute to partial neutralization escape 42 . However, research on the conformational epitopes of the EV-A76 capsid gene remains limited. The observed variability underscores the importance of continued surveillance of antigenic drift to ensure effective immunization strategies, which will be crucial for anticipating emerging variants and maintaining long-term control of viral transmission.
These findings highlight the utility of advanced molecular techniques, such as metagenomic sequencing, for identifying uncommon viral pathogens and co-infections in clinical samples. Although our results provide valuable insights into the genetic diversity, evolution, and potential public health implications of these viruses, the small sample size limits our ability to accurately determine the prevalence of EV-A76 and its co-infection with norovirus. Larger-scale surveillance and more comprehensive clinical data collection are needed to fully elucidate the epidemiological and clinical significance of EV-A76.
In conclusion, this study reports the detection of an EV-A76 strain in a patient presenting in Thailand, which may represent an imported case. Further studies are warranted to determine the source and transmission dynamics of this virus, and to support the development of broad-spectrum HFMD detection methods and panenterovirus vaccines, while ongoing viral surveillance will provide valuable insights to inform public health strategies and guide future vaccine development and antiviral interventions.
## Materials and methods
## Ethical approval statement
The research protocol for this study was approved by the Ethical Committee of the Faculty of Medicine, Chulalongkorn University, Thailand, under the institutional review board (approval number IRB 549/62). All patient information and identifiers were anonymized to safeguard patient confidentiality. The institutional review board of the Ethics Committee for Human Research granted a waiver for written informed consent because all clinical specimens were anonymized. All experiments conducted in this study adhered to the relevant guidelines and regulations. CV-A16 (JQ354992), and c CV-A10 (KP009574). 1 Amino acid change observed in the Thai EV-A76 strain and a reference strain from Nepal (PP621603). 2 Amino acid change found in the Thai EV-A76 strain, reference strains from India (MH118028, MH118029, and MH11803) as well as Nepal (PP621603). *Unique amino acid changes specific to the Thai EV-A76 strain identified in this study.
## Study samples
In this study, fecal samples were collected from patients with AGE (e.g., nausea, vomiting, watery diarrhea) between 2022 and 2024. Samples that tested positive for norovirus GI with a Ct of < 30 by quantitative realtime reverse-transcription polymerase chain reaction (qRT-PCR) assay 45 were included. Norovirus genotyping was confirmed using conventional RT-PCR 46 Sanger sequencing, and genotyping via the norovirus genotyping tool 47 . We included one GI.6[P6], and four GI.6[P11] sequences for further analysis, these sequences included one from Chaiyaphum province and four from Bangkok. Feces samples were stored at -80 °C until processing.
## RNA extraction, library preparation, sequencing, and metagenomic analysis
RNA extraction and Illumina sequencing were performed as previously described 48 . Firstly, viral RNA was isolated from 200 uL of the supernatants, which were diluted 1:10 (w/v) in PBS, using the High Pure RNA Isolation Kit (Roche) according to the manufacturer's instructions. Secondly, first-strand cDNA synthesis was performed using random hexamers and the Superscript IV reverse transcriptase (ThermoFisher Scientific). Thirdly, second-strand cDNA synthesis was performed using the Klenow Fragment DNA Polymerase I (3′→5′ exo-) (New England Biolabs) to convert the first-strand cDNA into double-stranded DNA. Paired-end libraries were generated using the KAPA HyperPlus kit (Roche), followed by enrichment with an in-house set of probes, named GastroCap (designed and synthesized by Roche), targeting at least seven vertebrate viral families known to cause gastrointestinal disease (Adenoviridae, Astroviridae, Caliciviridae, Hepeviridae, Parvoviridae, Picornaviridae, and Reoviridae) 48 . Hybridization-based capture of viral genomic fragments was performed according to Roche's standard protocol using the KAPA HyperCapture Bead Kit (Roche). Following capture, 14 cycles of post-capture PCR amplification were carried out with the KAPA HiFi HotStart ReadyMix (Roche) to enrich the target DNA prior to sequencing. Libraries were then sequenced using the 2 × 250 bp paired-end protocol on an Illumina MiSeq sequencer, following the manufacturer's instructions.
Raw reads generated from the Illumina MiSeq platform were processed with fastp (version 1.0.1) 49 to remove short reads (< 50 nts) and low-quality bases, and to trim low-quality ends. A de novo assembly was performed using metaSPAdes (version 4.2.0) 50 for each sample. The retrieved near-complete genomes from the assembly were used as sample-specific references genome for subsequent analyses, conducted using a customized workflow in Galaxy EU 51 . Consensus genomes were generated by aligning the metaSPAdes-assembled scaffolds to the references. A virus was considered "detected" if the assembled contig(s) exhibited ≥ 90% nucleotide identity to reference genome in the NCBI RefSeq database over ≥ 500 nt, or if multiple non-overlapping contigs collectively aligned ≥ 1,000 nt to the same viral species. These thresholds were selected to minimize false positives from short or low-complexity sequences and are consistent with criteria used in previous metagenomic studies of enteric viruses 48 . GenBank accession numbers for the assembled genomes are listed in Table 1.
## Genome typing and phylogenetic analysis
The nucleotide sequences of norovirus GI.6 ([P6] and [P11]) and EV-A76 were analyzed using the Basic Local Alignment Search Tool (BLAST), and genotyped using the Enterovirus Genotyping Tool ( h t t p s : / / w w w . r i v m . n l / m p f / t y p i n g t o o l / e n t e r o v i r u s / ) and the Norovirus Typing Tool v2.0 ( h t t p s : / / w w w . r i v m . n l / m p f / t y p i n g t o o l / n o r o v i r u s / ) 4 7 , both provided by the National Institute for Public Health and the Environment (RIVM), Netherlands. Sequence alignment was performed using ClustalW as implemented in BioEdit (v 7.2) 52 . Maximum-likelihood phylogenetic trees with 1,000 bootstrap replicates were constructed for the norovirus GI.6 ([P6] and [P11]) and EV-A76 strains using MEGA X 53 . Bootstrap values greater than 80 were considered significant.
## Nucleotide analysis
Pairwise nucleotide sequence similarity analysis was performed on norovirus GI.6[P6] and the Thai EV-A76 strain, in comparison with reference strains available in the GenBank database, using SimPlot version 3.5.1 54 . A similarity plot was generated with a window size of 200 and a step size of 20. The Kimura 2-parameter distance model was used and the ratio of transitions and transversions was determined empirically.
## Amino acid variations in the viral capsid proteins of EV-A76
Comparison of conformational epitope positions among three strains, EV-A71 (EU703812), CV-A16 (JQ354992), and CV-A10 (KP009574), with the full-length VP1-VP3 sequences of a Thai EV-A76 strain and 11 EV-A76 reference strains. The reference strains include the EV-A76 prototype strain from France (AY697458, 1991), six strains from China (JF905564, 2004; ON646232, ON646239, ON646240, ON646243, and ON646245, 2011), three strains from India (MH118028, MH118029, and MH118030, 2010-2011), and one strain from Nepal (PP621603, 2023). Only positions with amino acid changes were highlighted when compared within the EV-A76 group.
## References
1. Troeger (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. Aliabadi, Tate, Haynes et al. (2015) "Sustained decrease in laboratory detection of rotavirus after implementation of routine vaccination-United states, 2000-2014" *MMWR Morb Mortal. Wkly. Rep*
3. Mans "Norovirus infections and disease in Lower-MiddleandLow-Income countries, 1997⁻2018" *Viruses*
4. Zhang (2022) "Global burden and trends of Norovirus-Associated diseases from 1990 to 2019: an observational trend study" *Front. Public. Health*
5. Jiang, Wang, Wang et al. (1993) "Sequence and genomic organization of Norwalk virus" *Virology*
6. Chhabra "Updated classification of Norovirus genogroups and genotypes" *J. Gen. Virol*
7. Cannon (2021) "Global trends in Norovirus genotype distribution among children with acute gastroenteritis" *Emerg. Infect. Dis*
8. Kendra, Tohma, Parra (2022) "Global and regional circulation trends of Norovirus genotypes and recombinants, 1995-2019: A comprehensive review of sequences from public databases" *Rev. Med. Virol*
9. Kroneman (0705) "Proposal for a unified Norovirus nomenclature and genotyping" *Arch. Virol*
10. Chiu (2015) "Molecular epidemiology of GI.3 Norovirus outbreaks from acute gastroenteritis surveillance system in taiwan" *Biomed. Res*
11. Wulandari (2019) "Predominance of Norovirus GI.4 from children with acute gastroenteritis in jambi, indonesia" *J. Med. Virol*
12. Navarro-Lleó (2022) "Recombinant Noroviruses Circulating in Spain from 2016 to 2020 and proposal of two novel genotypes within genogroup I" *Microbiol. Spectr*
13. Alarcón-Linares (2023) "Rare Recombinant GI.5[P4] Norovirus that caused a large foodborne outbreak of gastroenteritis in a hotel in Spain in 2021" *Microbiol. Spectr*
14. Khamrin (2015) "Genetic recombination and genotype diversity of Norovirus GI in children with acute gastroenteritis in thailand" *J. Infect. Public. Health*
15. Fu (2016) "Genotypic diversity and recombination of Norovirus GI.6[P11] associated acute gastroenteritis outbreaks in beijing" *Infect. Genet. Evol*
16. Zell (2017) "ICTV virus taxonomy profile: Picornaviridae" *J. Gen. Virol*
17. Jiang, Liu, Ma et al. (2014) *Picornavirus morphogenesis. Microbiol. Mol. Biol. Rev*
18. Huang (2009) "Reemergence of enterovirus 71 in 2008 in taiwan: dynamics of genetic and antigenic evolution from 1998 to 2008" *J. Clin. Microbiol*
19. Fang (2021) "The role of conformational epitopes in the evolutionary divergence of enterovirus D68 clades: A bioinformaticsbased study" *Infect. Genet. Evol*
20. Wang (2021) "Bioinformatics-based prediction of conformational epitopes for enterovirus A71 and coxsackievirus A16" *Sci. Rep*
21. Yee, Laa Poh (2017) "Impact of genetic changes, pathogenicity and antigenicity on Enterovirus-A71 vaccine development" *Virology*
22. Brouwer, Moreni, Wolthers et al. (2021) "World-Wide prevalence and genotype distribution of enteroviruses" *Viruses*
23. Okada (2020) "Etiologic features of diarrheagenic microbes in stool specimens from patients with acute diarrhea in Thailand" *Sci. Rep*
24. Wang (2021) "Etiological, epidemiological, and clinical features of acute diarrhea in China" *Nat. Commun*
25. Chio, Chien, Chan et al. (2024) "Overview of the Trending Enteric Viruses and Their Pathogenesis in Intestinal" *Epithelial Cell Infection. Biomedicines*
26. Tan, Jiang, Gastroenteritis et al.
27. Oberste (2005) "Enteroviruses 76, 89, 90 and 91 represent a novel group within the species human enterovirus A" *J. Gen. Virol*
28. Sapkal (2009) "Enteroviruses in patients with acute encephalitis, Uttar pradesh" *India. Emerg. Infect. Dis*
29. Li (2023) "Transmission of Recombinant enterovirus A76 (EV-A76) in Xinjiang Uighur autonomous region of China" *Emerg. Microbes Infect*
30. Iritani (2014) "Detection and genetic characterization of human enteric viruses in oyster-associated gastroenteritis outbreaks between 2001 and 2012 in Osaka city" *Japan. J. Med. Virol*
31. Bisseux (2020) "Monitoring of enterovirus diversity in wastewater by ultra-deep sequencing: an effective complementary tool for clinical enterovirus surveillance" *Water Res*
32. Lizasoain (2021) "Environmental surveillance through Next-Generation sequencing to unveil the diversity of human enteroviruses beyond the reported clinical cases" *Viruses*
33. Majumdar (2021) "High diversity of human Non-Polio enterovirus serotypes identified in contaminated water in Nigeria" *Viruses*
34. Faleye (1803) "Wastewater-Based epidemiology and Long-Read sequencing to identify enterovirus circulation in three municipalities in Maricopa county, arizona, Southwest united States between June and October" *Viruses*
35. Lukashev (2005) "Role of recombination in evolution of enteroviruses" *Rev. Med. Virol*
36. Simmonds (2006) "Recombination and selection in the evolution of picornaviruses and other mammalian positive-stranded RNA viruses" *J. Virol*
37. Nikolaidis (2019) "Large-scale genomic analysis reveals recurrent patterns of intertypic recombination in human enteroviruses" *Virology*
38. Mahmud "Structure of the 5' untranslated region of enteroviral genomic RNA" *J. Virol*
39. Qiao (2023) "Analysis of the epidemiological trends of enterovirus A in Asia and Europe" *J. Infect. Chemother*
40. Puenpa (2019) "Evolutionary and genetic recombination analyses of coxsackievirus A6 variants associated with hand, foot, and mouth disease outbreaks in Thailand between" *Viruses*
41. Esposito, Principi, Hand (2018) "foot and mouth disease: current knowledge on clinical manifestations, epidemiology, aetiology and prevention" *Eur. J. Clin. Microbiol. Infect. Dis*
42. Tan (2023) "Substitution of coxsackievirus A16 VP1 BC and EF loop altered the protective immune responses in chimera enterovirus A71. Vaccines (Basel)"
43. Cordey (2012) "Identification of Site-Specific adaptations conferring increased neural cell tropism during human enterovirus 71 infection" *PLoS Pathog*
44. Shi, Liu, Wu et al. (2023) "Molecular epidemiology and recombination of enterovirus D68 in China" *Infect. Genet. Evol*
45. Debbink (2013) "Human Norovirus detection and production, quantification, and storage of virus-like particles" *Curr. Protoc. Microbiol*
46. Chhabra "Single-step RT-PCR assay for dual genotyping of GI and GII Norovirus strains" *J. Clin. Virol*
47. (2021)
48. Kroneman (2011) "An automated genotyping tool for enteroviruses and Noroviruses" *J. Clin. Virol*
49. Izquierdo-Lara (2024) "Patterns of the within-host evolution of human Norovirus in immunocompromised individuals and implications for treatment"
50. Chen, Zhou, Chen et al. (2018) "Fastp: an ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
51. Nurk, Meleshko, Korobeynikov et al. (2017) "MetaSPAdes: a new versatile metagenomic assembler" *Genome Res*
52. Afgan (2018) "The galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update" *Nucleic Acids Res*
53. Hall (1999) "& BioEdit A user-friendly biological sequence alignment editor and analysis program for windows 95/98/NT" *Nucl. Acids Symp. Ser*
54. Kumar, Stecher, Li et al. (2018) "MEGA X: molecular evolutionary genetics analysis across computing platforms" *Mol. Biol. Evol*
55. Lole (1999) "Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in india, with evidence of intersubtype recombination" *J. Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584639&blobtype=pdf | # TUBB1 promoter methylation is a promising biomarker for predicting HBeAg seroconversion in chronic hepatitis B
Tong Zhao, Yuna Tang, Yu Sun, Jihui Li, Yuchen Fan, Chao Cui, Shuai Gao, Kai Wang, Victoria Delpino
## Abstract
The identification of predictive indices for hepatitis B e antigen seroconver sion (HBeAg SC) in patients with chronic hepatitis B (CHB) remains a challenge. We aimed to investigate whether the TUBB1 promoter methylation in peripheral blood mononuclear cells (PBMCs) can predict HBeAg SC. A total of 271 participants were recruited, comprising 145 patients with HBeAg-positive CHB, 94 with HBeAg-negative CHB, and 32 healthy controls (HCs). The patients with HBeAg-positive CHB were followed up for 72 weeks. The TUBB1 promoter methylation and the corresponding mRNA levels in PBMCs were detected using MethyLight and quantitative real-time PCR, respectively. The methylation levels of the TUBB1 promoter were remarkably elevated in patients with positive HBeAg, in comparison to those with negative HBeAg and HCs. Conversely, the relative mRNA expression levels of TUBB1 were significantly downregulated in patients with positive HBeAg, when compared to those with negative HBeAg and HCs. The TUBB1 promoter methylation levels showed a gradual decrease across the four phases of CHB. Patients with HBeAg SC had lower baseline methylation levels of the TUBB1 promoter than those without HBeAg SC. The TUBB1 promoter methylation was an independent predictor of HBeAg SC (odds ratio [OR] = 0.683, 95% CI 0.553-0.845, P < 0.001). The methylation of the TUBB1 promoter showed good predictive value for HBeAg SC in patients with positive HBeAg (area under the curve [AUC] = 0.805, 95% CI 0.704-0.907, P < 0.001). The methylation level of the TUBB1 promoter might be a potent biomarker for predicting HBeAg SC. IMPORTANCE Previous studies emphasized hepatitis B e antigen seroconversion (HBeAg SC) as a milestone for chronic hepatitis B (CHB) remission associated with reduced disease progression risks. While the significance of HBeAg SC is widely recognized, reliable non-invasive predictors for achieving this endpoint remain limited. Additionally, in our previous studies, DNA methylation of key regulatory genes has been linked to CHB progression. However, the association between TUBB1 promoter methylation and HBeAg SC, as well as its potential as a biomarker for clinical application, has not been fully elucidated. We demonstrated that TUBB1 promoter methylation levels were significantly higher in HBeAg-positive patients and that decreased methylation levels were independ ently associated with subsequent HBeAg SC during a 72-week follow-up. Our findings underscore the potential clinical utility of TUBB1 promoter methylation as a non-invasive biomarker for predicting HBeAg SC. This study provides strong evidence supporting the role of TUBB1 promoter methylation in predicting HBeAg SC, offering a novel biomarker for monitoring CHB. KEYWORDS TUBB1, promoter methylation, chronic hepatitis B, HBeAg seroconversion G lobally, an estimated 296 million individuals are chronically infected by hepati tis B virus (HBV), and in 2022, hepatitis B was responsible for approximately 1.1 million deaths (1). Patients who are hepatitis B e antigen (HBeAg) positive with
persistently elevated alanine aminotransferase (ALT) levels generally exhibit high HBV DNA concentrations, putting them at an increased risk of hepatocellular carcinoma and cirrhosis, necessitating antiviral therapy (2)(3)(4)(5)(6)(7). HBeAg seroconversion (SC), defined as the clearance of HBeAg accompanied by the appearance of anti-HBe antibodies, is a key therapeutic objective. Early HBeAg SC may indicate disease remission and is associated with a favorable prognosis (8)(9)(10). Accordingly, HBeAg SC is an important goal in antiviral strategies (3)(4)(5)(6). Therefore, a reliable and accurate non-invasive marker is urgently needed to predict early HBeAg SC, assess patient status, and guide clinical decision-making.
Methylation of cytosine phosphate-guanine (CpG) islands in deoxyribonucleic acid (DNA) is a highly prevalent epigenetic phenomenon in mammalian genomes, playing a crucial role in gene regulation. This process has been demonstrated to exert a myriad of biological effects, encompassing normal developmental processes, ribonucleic acid (RNA) metabolism, X-chromosome inactivation, genomic imprinting, and even the development of tumors (11)(12)(13)(14)(15). Previous studies showed that DNA methylation of key regulatory regions might be a biomarker for the progression of chronic hepatitis B (CHB), underscoring its potential role in monitoring and predicting disease advancement (16)(17)(18)(19).
Microtubules, assembled by heterodimers of α-tubulin and β-tubulin, constitute one of the primary cytoskeletal structures in cells, playing crucial roles in maintaining cell morphology, intracellular transport, signal transduction, cell motility, and mitosis. Specifically encoded by the TUBB1 gene, β1-tubulin serves as the most prevalent isoform of β-tubulin (20)(21)(22)(23)(24). It was reported that TUBB1 was involved in the pathophysiology of various diseases, including thrombocytopenia, thyroid dysgenesis, neurodevelopmen tal defects, and so forth (25)(26)(27)(28). TUBB1 was reported to be a potential therapeutic compound and druggable target for hepatocellular carcinoma patients (29). Besides, in the study proposed by Wang H et al., single-cell RNA sequencing revealed that TUBB1+ monocyte in peripheral blood mononuclear cells (PBMCs) might be associated with decreased antiviral activity in patients with CHB (30). However, it remains uncertain whether TUBB1 and the methylation of its promoter take part in the natural history of CHB and HBeAg SC.
In this study, we evaluated the mRNA expression levels of TUBB1 and the methylation levels of the TUBB1 promoter in PBMCs among patients with CHB and healthy controls (HCs). We observe varying levels of methylation in the TUBB1 promoter across the four phases of CHB. Our findings indicate that TUBB1 promoter methylation levels in HBeAg-positive patients independently predict HBeAg SC.
## MATERIALS AND METHODS
## Patients' selection
A total of 271 participants, including 239 patients with CHB and 32 HCs, were recruited at Qilu Hospital of Shandong University from January 2022 to June 2023. All the patients with CHB were identified as HBsAg-positive for a minimum duration of 6 months. Exclusion criteria included: (i) coinfection with hepatitis C virus, hepatitis D virus, hepatitis E virus, or human immunodeficiency virus; (ii) combined with other liver disease (alcoholic liver disease, non-alcoholic fatty liver disease, autoimmune liver disease, drug-induced liver injury, Wilson disease); (iii) presence of hepatocellular carcinoma or other malignant disease; and (iv) pregnancy. All the subjects gave their written informed consent to participate in the study. The research was approved by the local Research and Ethics Committee at Qilu Hospital of Shandong University, in accordance with the guidelines of the 1975 Declaration of Helsinki. All experiments involving human blood samples were performed in a Biosafety Level 2 facility in accordance with institutional guidelines and regulations.
## Study design
We collected the PBMCs from all enrolled patients and HCs at baseline and extracted the DNA and mRNA to detect the TUBB1 promoter methylation and the corresponding mRNA expression. Baseline clinical and laboratory data were collected and analyzed. Patients with positive HBeAg were followed up for 72 weeks to see whether they could achieve HBeAg SC. Seroconversion was confirmed by standard assays as the simultane ous loss of HBeAg and the emergence of anti-HBe positivity.
## DNA extraction and sodium bisulfite modification
PBMCs were separated by density gradient centrifugation with Ficoll-Paque (Pharma cia Diagnostics, Uppsala, Sweden) and stored at -80°C until use. Genomic DNA was extracted from PBMCs using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). DNA bisulfite modification was performed with an EZ DNA Methylation-Gold Kit (Zymo Research, Orange, CA, USA) according to the manufacturer's instructions. The modified DNA was used as a template for MethyLight.
## TaqMan probe-based quantitative methylation-specific PCR (MethyLight)
The methylation levels of TUBB1 promoter were detected using MethyLight in all participants. The promoter of TUBB1 was delineated with the website (https:// www.ncbi.nlm.nih.gov/) and the sequence was transformed on another website (https://www.urogene.org/methprimer/). The genome coordinates of TUBB1 are hg38, chr20:60802540-60809767. The promoter region was considered to be the upstream 2,000 bp region of its transcription start site, where one CpG island was found from 1,025 to 1,131 bp (Fig. S1). The primers and probes were designed at the CpG island region using oligo7 (OLIGO 1267 Vondelpark ColoradoSprings, CO 80907, USA). The specific primers and probe sequences for gene promoters are listed in Table 1. We used a 10 µL volume MethyLight reaction system, including 5 µL MethyLight Master Mix (consisting of HotStarTaq Plus DNA Polymerase, EpiTect Probe PCR Buffer, and dNTP mix [dATP, dCTP, dGTP, dTTP]), 0.4 µL forward primer, 0.4 µL reverse primer, 0.2 µL probe, 2 µL nuclease-free water, and 2 µL modified DNA. The reaction was cycled using the following conditions: 95°C for 15 min and 45 cycles of 95°C for 15 s and 60°C for 60 s. SSSI methylase and bisulfite-modified human control DNA (QIAGEN, Hilden, Germany) served as the reference for methylation and β-actin was used as a normalization control. The results of MethyLight data were indicated by percentage of methylated reference (PMR) (31). PMR = 100% × 2^(-[Delta Ct (target gene in sample -control gene in sample) -Delta Ct (100% methylated target in reference sample -control gene in reference sample)]) (32).
## RNA extraction and quantitative real-time polymerase chain reaction
Total RNA of PBMCs was extracted using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). We used a reverse transcription kit to convert RNA into cDNA following the manufac turer's instructions (ThermoFisher, Waltham, USA). The expression levels of TUBB1 and β-actin mRNA were quantified using real-time PCR. The total reaction volume was 10 µL, consisting of 5 µL of TB Green premix (Takara, Shiga, Japan), 3 µL of nuclease-free water, 0.5 µL of forward primer, 0.5 µL of reverse primer, and 1 µL of cDNA. The PCR cycling was performed using a thermocycler from Analytik Jena (Germany), with conditions of denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 5 s, 55°C for 30 s, and 72°C for 60 s. The primer sequences are shown in Table 1. The comparative method (2 -ΔΔCt ) was utilized.
## Statistical analysis
Quantitative variables are expressed as the median (centile 25 and centile 75) and categorical variables are expressed as frequency (percentage). We used the Mann-Whit ney U-test and the Kruskal-Wallis H-test to compare quantitative variables and the chi-square test to analyze categorical variables. The Spearman's rank correlation test was used to determine the correlation between the TUBB1 promoter methylation level and clinical data. Independent risk factors for HBeAg SC were analyzed by binary logistic regression analysis with multivariate stepwise regression. The receiver operat ing characteristics (ROC) curve was constructed to obtain the area under the curve (AUC) and the best cut-off value was calculated, corresponding to the highest Youden index. Statistical analyses were performed using SPSS (version 26.0) and GraphPad Prism (version 9.5.1). Two-tailed P < 0.05 was considered statistically significant.
## RESULTS
## Study population
A total of 271 participants were enrolled in this study, including 239 patients with CHB and 32 HCs. Among the patients with CHB, 145 patients were HBeAg-positive, and 94 patients were HBeAg-negative.
Compared to patients with negative HBeAg, patients with positive HBeAg were younger and had higher ALT, AST, GGT, AFP levels and lower ALB levels. Serum HBV DNA was more likely to be detected in patients with positive HBeAg, while HBV DNA and HBsAg levels in patients with positive HBeAg were significantly higher than those with negative HBeAg results (Table 2).
## Hypermethylation of the TUBB1 promoter and low mRNA expression of TUBB1 in patients with positive HBeAg
The methylation status of the TUBB1 promoter in PBMC was evaluated using MethyLight and expressed as PMR. Figure 1A depicts the methylation level of the TUBB1 promoter in HCs, HBeAg-negative, and HBeAg-positive groups, respectively. The TUBB1 methyla tion levels in patients with positive HBeAg (median 15.83, interquartile range 13.26-18.33) were significantly higher than that in those with negative HBeAg (median 13.96, interquartile range 11.41-16.17; P < 0.001) and HCs (median 11.01, interquartile range 8.85-13.43; P < 0.001). In addition, the TUBB1 methylation levels of HBeAg-negative participants were significantly higher than HCs (P < 0.001).
Since methylation is a prevalent mechanism that regulates transcription, we examined the expression pattern of TUBB1 mRNA in PBMCs from HCs and patients with HBeAg-negative, as well as HBeAg-positive (Fig. 1B). The mRNA expression level of TUBB1 in the HBeAg-positive group was significantly lower than that in the HBeAg-negative (P = 0.004) and HC (P = 0.005) groups. There was no difference in the mRNA level of the TUBB1 between the HBeAg-negative group and the HC group (P = 0.550).
To further elucidate the association between the methylation level of the TUBB1 promoter and its mRNA expression level, we conducted a Spearman's rank correlation analysis. Our results indicated a significant, albeit weak, negative correlation between the methylation status of the TUBB1 promoter and its mRNA expression (Spearman's r = -0.157, P = 0.033; Fig. 1C).
## The TUBB1 promoter methylation level was related to the phases of CHB
All the patients with CHB were selected into the different phases according to the 2017 EASL guidelines (5). There were 85 cases in HBeAg-positive chronic infection (immune tolerant), 53 in HBeAg-positive chronic hepatitis (immune [re]active), 70 in HBeAg-negative chronic infection (inactive carrier state), and 22 in HBeAg-negative chronic hepatitis (immune-active or reactivation). Besides, nine patients were in the gray zone. The TUBB1 promoter methylation levels showed a sequential decrease in the order of HBeAg-positive chronic infection, HBeAg-positive chronic hepatitis, HBeAg-negative chronic infection, and HBeAg-negative chronic hepatitis. Among the participants with positive HBeAg, those who had raised ALT during the HBeAg-positive chronic hepatitis phase exhibited lower TUBB1 promoter methylation levels than those with normal ALT, who were in the HBeAg-positive chronic infection phase. Among patients with normal ALT, those in the HBeAg-positive chronic infection phase had higher methylation levels compared to those in the HBeAg-negative chronic infection phase (Fig. 2A).
On the other hand, the relationship between the TUBB1 promoter methylation and clinicopathology was analyzed in patients with CHB. As shown in Table 3 and Fig. 2B, we found that PMR was significantly and positively correlated to HBeAg. Furthermore, HBeAg status (positive or negative) was significantly correlated with PMR values, as demonstrated by Spearman's correlation analysis (Spearman's r = 0.243, P < 0.001; Fig. 2C). However, there was no statistically significant correlation between PMR and other clinical indices.
## Relatively lower TUBB1 promoter methylation level as a predictor for HBeAg SC in patients with positive HBeAg
Of the 93 patients with positive HBeAg who completed the 72-week follow-up, 28 developed HBeAg SC, while 65 did not. As shown in Table 4 and Fig. 3A, the TUBB1 methylation levels were significantly lower in patients with HBeAg SC than in those without HBeAg SC. Besides, compared with patients without HBeAg SC, patients with HBeAg SC contained more males and had significantly higher ALT and AST. There was no significant difference between the two groups in terms of either HBV-related indices or treatment.
To identify the significant factors that may affect the HBeAg SC, binary logistic regression analysis was performed (Table 5). A univariate analysis showed that PMR, gender, ALT, AST, and GGT were significant factors for predicting HBeAg SC with P < 0.05 and they were selected into the multivariate binary logistic regression model with stepwise analysis. As a result, PMR representing the methylation level of the TUBB1 promoter was a significant independent predictor for HBeAg SC (odds ratio [OR] = 0.683, 95% CI 0.553-0.845, P < 0.001). In addition, elevated ALT levels (≥40 U/L) were also demonstrated to be independent factors influencing HBeAg SC. Besides, based on the ROC curve analysis, the optimal cut-off value of PMR was <14.80% for predicting the
a TDF, tenofovir disoproxil fumarate; ETV, entecavir; PEG-IFN, peginterferon; NAs, nucleos(t)ide analogs.
occurrence of HBeAg SC (AUC = 0.805, 95% CI 0.704-0.907, sensitivity 80.00%, specificity 71.43%, P < 0.001; Fig. 3B). Specifically, the TUBB1 promoter methylation exhibited statistically significant predictive ability for HBeAg SC in HBeAg-positive patients with elevated ALT (AUC = 0.773, 95% CI 0.609-0.937, sensitivity 62.50%, specificity 88.67%, P = 0.001) (Fig. 3C).
## DISCUSSION
Our study showed that the TUBB1 promoter methylation levels were significantly higher in HBeAg-positive patients than in HBeAg-negative patients or in HCs. The methylation degree of the TUBB1 promoter gradually declined during the four phases of CHB.
There was a weak but significant positive correlation between the TUBB1 promoter methylation level and the quantitative value of HBeAg. We found that TUBB1 promoter methylation and elevated ALT were independent factors for the achievement of HBeAg SC. Besides, the mRNA levels of TUBB1 were obviously lower in the HBeAg-positive group than in the HBeAg-negative group and HCs. The methylation status of the TUBB1 promoter exhibited a weak but significant negative correlation with the mRNA expression.
In our study, we observed hypermethylation of the TUBB1 promoter and, on the contrary, lower TUBB1 mRNA expression in the HBeAg-positive group compared to the HBeAg-negative group and HCs. Meanwhile, the methylation levels of the TUBB1 promoter were significantly inversely associated with the mRNA levels of TUBB1 (Spearman's r = -0.157, P = 0.033). Previous studies have shown that CpG islands function as molecular switches, suppressing gene expression when methylated (33)(34)(35). It is reasonable to speculate that the methylation of the TUBB1 promoter inhibits the expression of TUBB1 mRNA to some degree. The weak correlation between promoter methylation and mRNA expression levels is likely attributable to the multifaceted nature of gene regulation. While promoter methylation is a recognized mechanism of gene silencing, mRNA expression is additionally modulated by multiple other factors, such as histone modifications, transcription factor activity, non-coding RNAs, chroma tin architecture, and nuclear organization. Therefore, we did not delve into the poten tial molecular mechanisms in this research. Alternatively, we focused on the clinical significance of quantified methylation.
The PBMCs are mainly composed of different kinds of immune cells, including lymphocytes (T cells, B cells, and NK cells), monocytes, and dendritic cells. Several studies have already demonstrated that immune-metabolism disorder of liver tissue triggered by HBV exacerbation might result in the alteration of PBMCs (36,37). Meanwhile, several previous studies also proved that aberrant DNA methylation status of PBMCs existed in patients with hepatitis B virus infection (32,38). While PBMC methylation doesn't directly match liver methylation, it's a valuable indicator of the body's overall response to chronic HBV and the immune-metabolic stress from the liver. Therefore, in our study, TUBB1 promoter methylation in PBMCs serves as a promising non-invasive biomarker for assessing disease status and predicting clinical outcomes, such as HBeAg seroconversion.
Given the intricate pathophysiology of HBV, the history and phases of infection continue to be under ongoing investigation (39). According to the 2017 EASL guidelines (5), the natural progression of chronic HBV infection can be broadly categorized into four phases: HBeAg-positive chronic infection (immune-tolerant phase), HBeAg-positive chronic hepatitis (immune [re]active phase), HBeAg-negative chronic infection (inac tive carrier state phase), and HBeAg-negative chronic hepatitis (immune-active phase or HBeAg-negative disease). Specifically, patients within the HBeAg-positive chronic infection phase and the HBeAg-negative chronic infection phase maintain normal ALT levels. In contrast, those in the HBeAg-positive chronic hepatitis phase and the HBeAgnegative chronic hepatitis phase exhibit elevated ALT levels. CHB exhibits a non-linear clinical course and not all patients go through every phase (4,5). Therefore, the phases of CHB and TUBB1 promoter methylation cannot be used for correlation analysis and ordinal logistic regression analysis. In our study, the methylation levels of the TUBB1 promoter tended to differ across the four phases, associated with different clinical genotype. Among HBeAg-positive patients, those who experienced an elevation in ALT levels during the HBeAg-positive chronic hepatitis phase exhibited lower levels of TUBB1 promoter methylation compared to patients with normal ALT values who were in the HBeAg-positive chronic infection phase. This phenomenon is consistent with our subsequent observation that patients who achieved HBeAg SC had higher ALT levels and lower TUBB1 promoter methylation compared to those who did not lose HBeAg. Meantime, TUBB1 promoter methylation was significantly positively correlated with the quantitative value of HBeAg. It could be inferred that PMR, representing the methylation level of the TUBB1 promoter, could reflect the clinical progression of CHB patients.
Huang et al. found that quantitative HBeAg and detectable baseline HBV DNA could predict the clearance of HBeAg in patients treated with pegylated interferon (40). Buster's study showed that a higher ALT, low HBV DNA levels, female gender, older age, and an absence of previous interferon therapy were independent predictors of HBeAg SC in patients treated with peginterferon-alfa (41). In the studies conducted by Huang and Buster, patients were primarily treated with peginterferon, which raises concerns regarding selection bias due to the potential side effects and the limited applicability of this treatment to certain populations when investigating the predictive factors associated with HBeAg SC. However, our study is more closely aligned with actual clinical scenarios, and elevated ALT levels have been identified as independent predictive factors associated with HBeAg SC as well. In our study, however, there was no difference in HBV-related indicators between patients who had achieved HBeAg SC and those who had not. Besides, in our study, the HBeAg SC group comprised more male patients. Many other studies are consistent with our results (42)(43)(44).
However, there are still several limitations in our study. Firstly, it is a single-center study with a small sample size. Compared with lost patients, those who are willing to be reviewed are likely to have stronger treatment intention and follow the doctor's advice more regularly, potentially introducing selection bias. Secondly, due to the difficulty in obtaining biopsy samples, we were unable to analyze the intrahepatic methylation status of TUBB1 in the studied patients. Furthermore, the precise molecular mechanisms by which TUBB1 is involved in the natural history of CHB and its role in HBeAg SC have not been further investigated.
## Conclusion
In conclusion, elevated TUBB1 promoter methylation levels in PBMCs are significantly associated with HBeAg-positive status and may serve as a promising non-invasive biomarker for predicting HBeAg SC.
## References
1. (2024) "Global hepatitis report 2024: action for access in low-and middle-income countries"
2. Huang, Tran, Yeh et al. (2023) "Antiviral therapy substantially reduces HCC risk in patients with chronic hepatitis B infection in the indeterminate phase" *Hepatology*
3. (2024) "Guidelines for the prevention, diagnosis, care and treatment for people with chronic hepatitis B infection"
4. Terrault, Lok, Mcmahon et al. (2018) "Update on prevention, diagnosis, and treatment of chronic hepatitis B: AASLD 2018 hepatitis B guidance" *Hepatology*
5. Lampertico, Agarwal, Berg et al. (2017) "EASL 2017 Clinical Practice Guidelines on the management of hepatitis B virus infection" *J Hepatol*
6. Sarin, Kumar, Lau et al. (2016) "Asian-Pacific clinical practice guidelines on the management of hepatitis B: a 2015 update" *Hepatol Int*
7. Huang, Li, Le et al. (2022) "Natural history and hepatocellular carcinoma risk in untreated chronic hepatitis B patients with indeterminate phase" *Clin Gastroenterol Hepatol*
8. Liaw (2009) "HBeAg seroconversion as an important end point in the treatment of chronic hepatitis B" *Hepatol Int*
10. Ning, Han, Sun et al. (2014) "Switching from entecavir to PegIFN alfa-2a in patients with HBeAg-positive chronic hepatitis B: a randomised open-label trial (OSST trial)" *J Hepatol*
11. Van Zonneveld, Honkoop, Hansen et al. (2004) "Long-term follow-up of alpha-interferon treatment of patients with chronic hepatitis B" *Hepatology*
12. Li, Bestor, Jaenisch (1992) "Targeted mutation of the DNA methyltransferase gene results in embryonic lethality" *Cell*
13. Li, Beard, Jaenisch (1993) "Role for DNA methylation in genomic imprinting" *Nature*
14. Panning, Jaenisch (1998) "RNA and the epigenetic regulation of X chromosome inactivation" *Cell*
15. Koch, Joosten, Feng et al. (2018) "Analysis of DNA methylation in cancer: location revisited" *Nat Rev Clin Oncol*
16. Zhang, Wang, Su et al. (2023) "F-box protein 43 promoter methylation as a novel biomarker for hepatitis B virus-associated hepatocellular carcinoma" *Front Microbiol*
17. Zeybel, Vatansever, Hardy et al. (2016) "DNA methylation profiling identifies novel markers of progression in hepatitis B-related chronic liver disease" *Clin Epigenetics*
18. Li, Qin, Jiang et al. (2020) "The signature of HBV-related liver disease in peripheral blood mononuclear cell DNA methylation" *Clin Epigenetics*
19. Wang, Gao, Li et al. (2015) "Demethylation of tumor necrosis factor-α converting enzyme promoter associated with high hepatitis B e antigen level in chronic hepatitis B" *World J Gastroen terol*
20. Su, Wang, Li et al. (2024) "Hypermethylation of the glutathione peroxidase 4 gene promoter is associated with the occurrence of immune tolerance phase in chronic hepatitis B" *Virol J*
21. Gadadhar, Bodakuntla, Natarajan et al. (2017) "The tubulin code at a glance" *J Cell Sci*
22. Mohri (1968) "Amino-acid composition of "Tubulin" constituting microtubules of sperm flagella" *Nature*
23. Robertis (1953) "The submicroscopic organization of axon material isolated from myelin nerve fibers" *J Exp Med*
24. Stephens (1970) "Thermal fractionation of outer fiber doublet microtubules into A-and B-subfiber components. A-and B-tubulin" *J Mol Biol*
25. Schaedel, John, Gaillard et al. (2015) "Microtubules self-repair in response to mechanical stress" *Nat Mater*
26. Stefanucci, Collins, Sims et al. (2023) "The effects of pathogenic and likely pathogenic variants for inherited hemostasis disorders in 140 214 UK Biobank participants" *Blood*
27. Palma-Barqueros, Bury, Kunishima et al. (2021) "Expanding the genetic spectrum of TUBB1-related thrombocytopenia" *Blood Adv*
28. Stoupa, Adam, Kariyawasam et al. (2018) "TUBB1 mutations cause thyroid dysgenesis associated with abnormal platelet physiology" *EMBO Mol Med*
29. Tantry, Santhakumar (2023) "Insights on the role of α-and βtubulin isotypes in early brain development" *Mol Neurobiol*
30. Tian, Xiao, Yang et al. (2023) "Crosstalk between 5-methylcytosine and N 6 -methylade nosine machinery defines disease progression, therapeutic response and pharmacogenomic landscape in hepatocellular carcinoma" *Mol Cancer*
31. Wang, Li, Yang et al. (2023) "Single-cell RNA sequencing reveals transcriptional profiles of monocytes in HBV-infected pregnant women during mid-pregnancy" *J Cell Mol Med*
32. Um, Kim, Oh et al. (2011) "Aberrant CpG island hypermethylation in dysplastic nodules and early HCC of hepatitis B virus-related human multistep hepatocarcinogenesis" *J Hepatol*
33. Gao, Sun, Fan et al. (2015) "Aberrant GSTP1 promoter methylation predicts short-term prognosis in acute-on-chronic hepatitis B liver failure" *Aliment Pharmacol Ther*
34. Van Guelpen, Hultdin, Johansson et al. (2006) "Low folate levels may protect against colorectal cancer" *Gut*
35. Yoshikawa, Kuribayashi, Minami et al. (2020) "Epigenetic alterations and biomarkers for immune checkpoint inhibitors-current standards and future perspectives in malignant pleural mesothelioma treatment" *Front Oncol*
36. Wang, Wang, Zhu et al. (2024) "Promoter hypermethylation-induced downregulation of ITGA7 promotes colorectal cancer proliferation and migration by activating the PI3K/AKT/NF-κB pathway" *Biochim Biophys Acta Mol Cell Res*
37. Liang, Li, Jiang et al. "Chinese Group on the Study of Severe Hepatitis B (COSSH). 2023. Transcriptomics unveils immune metabolic disruption and a novel biomarker of mortality in patients with HBV-related acute-on-chronic liver failure" *JHEP Rep*
38. Gehring, Protzer (2019) "Targeting innate and adaptive immune responses to cure chronic HBV infection" *Gastroenterology*
39. Sun, Lu, Li et al. (2022) "Diagnostic and prognostic value of STAP1 and AHNAK methylation in peripheral blood immune cells for HBV-related hepatopathy" *Front Immunol*
40. Yim, Lok (2006) "Natural history of chronic hepatitis B virus infection: what we knew in 1981 and what we know in 2005"
41. Huang, Shen, Phyo et al. (2024) "Quantitative HBeAg is a strong predictor of HBeAg loss among patients receiving pegylated interferon" *Antiviral Res*
42. Buster, Hansen, Lau et al. (2009) "Factors that predict response of patients with hepatitis B e antigen-positive chronic hepatitis B to peginterferon-alfa" *Gastroenterology*
43. Yin, Wan, Issa et al. (2023) "The presence of baseline HBsAb-Specific B cells can predict HBsAg or HBeAg seroconversion of chronic hepatitis B on treatment" *Emerg Microbes Infect*
44. Song, Yang, Liu et al. (2023) "HBV pregenome RNA as a predictor of spontanous HBeAg seroconversion in HBeAgpositive chronic hepatitis B patients" *BMC Gastroenterol*
45. Yang, Xu, Cheng et al. (2024) "CXCL10 and its receptor in patients with chronic hepatitis B and their ability to predict HBeAg seroconversion during antiviral treatment with TDF" *J Med Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793372&blobtype=pdf | # Abstract citation ID: ofaf695.2236 P-2072. Sociodemographics and Viral Load Outcomes of People Newly Diagnosed with HIV or Returning to HIV Care Receiving RapidTx Cards at NYC Health +Hospitals/Bellevue
Ian Maynor, Abigail Smith, Cooper Urban, Emma Boockvar, Robert Pitts, Shelly Blumenthal, Kathryn Jano, Shree Sundaresh, Ofole Mgbako
Background. Immediate antiretroviral therapy (iART) is ART initiation immediately after HIV diagnosis or upon care linkage. In New York City (NYC), RapidTx cards cover the cost of 30 days of ART for un-or underinsured patients via the New York State (NYS) AIDS Institute HIV Uninsured Care Program. This study characterizes sociodemographics and viral load suppression (VLS) of patients initiated on iART via RapidTx cards at NYC Health+Hospitals/Bellevue.
Methods. We conducted a single-center, retrospective study of 115 patients issued RapidTx cards from January 1, 2019 to July 31, 2024. The primary outcome was VLS at 3 months after care linkage at the subsequent Bellevue HIV/virology appointment for patients receiving RapidTx cards. Descriptive statistics were used for socio-demographics and clinical characteristics. We compared the Rapid Tx cohort's sociodemographics to those of all Bellevue patients with HIV from 2020-2023. We then compared the primary outcome to VLS for Bellevue patients newly diagnosed with HIV in 2023.
Results. Out of 115 participants, 37.39% (N=43) were newly diagnosed with HIV at Bellevue. 62.61% (N=72) were diagnosed in the past and returning to care. Figure 1 summarizes the RapidTx card workflow. Table 1 summarizes sociodemographics and Table 2 summarizes clinical characteristics. RapidTx patients were more often Hispanic, cisgender male, uninsured or on Medicaid than the general Bellevue population with HIV. 83.48% (N=96) had viral load testing 3 months after care linkage; 14.78% (N=17) were lost to follow-up. 81.39% (N=35) of newly diagnosed patients and 79.16% (N=57) of patients returning to care had VLS at 3 months. For patients with new HIV diagnoses in 2023, 54% had VLS at 3 months at Bellevue while 51% had VLS at 3 months in NYC.
Conclusion. RapidTx cards were an effective tool for iART initiation with better initial VLS outcomes for RapidTx patients than for newly diagnosed patients at Bellevue and in NYC. There was no difference in VLS between new and previously diagnosed RapidTx patients. Future studies should prioritize more support for RapidTx patients prior to them becoming lost to follow-up.
Disclosures. Robert Pitts, MD MPH, Gilead Inc: Advisor/Consultant|ViiV: Advisor/Consultant Ofole Mgbako, MD, MS, Gilead Sciences: Advisor/ Consultant 1) captured at the near neighborhood level (9-digit ZIP for MORE 2 and 5-digit ZIP for Medicare FFS) were linked at the pt level to the Inovalon SDOH Data Warehouse. TF was defined as having a second oral antibiotic (ABX), intravenous ABX, or Poster Abstracts • OFID 2026:13 (Suppl 1) • S1261 |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12833909&blobtype=pdf | # Large-Scale Psychometric Assessment and Validation of the Modified COVID-19 Yorkshire Rehabilitation Scale Patient-Reported Outcome Measure for Long COVID or Post-COVID Syndrome
Mike Horton, Adam Smith, Ruairidh Milne, Darren Winch, Clare Rayner, Stephen Halpin, Rory O'connor, Roman Rocha, Darren Greenwood, Nawar Bakerly, Rachael Evans, Joseph Kwon, Helen Dawes, | Conor Wood, Paul Williams, | Harsha Master, Mae Mansoubi, Johannes De Kock, Jordan Mullard, Mike Ormerod, | Ghazala Mir, Stavros Petrou, Daryl O'connor, | Sivan, | Consortium
## Abstract
The C19-YRS was the first condition-specific for long COVID/post-COVID syndrome. Although the original C19-YRS evolved to the modified version (C19-YRSm) based on psychometric evidence, clinical content relevance, as well as feedback from patients and healthcare professionals, it has not been validated through Rasch analysis. The study aim was to psychometrically assess and validate the C19-YRSm using newly collected data from a large-scale, multicenter study (LOCOMOTION). In total, 1278 patients (67% Female; mean age = 48.6, SD 12.7) digitally completed the C19-YRSm. The psychometric properties of the C19-YRSm Symptom Severity (SS) and Functional Disability (FD) subscales were assessed using a Rasch Measurement Theory framework, assessing for individual item model fit, targeting, internal consistency reliability, unidimensionality, local dependency (LD), response category functioning and differential item functioning (DIF) by age group, sex and ethnicity. Rasch analysis revealed robust psychometric properties of both subscales, with each demonstrating unidimensionality, appropriate response category structuring, no floor or ceiling effects, and minimal LD and DIF. Both subscales also displayed good targeting and reliability (SS: Person Separation Index (PSI) = 0.81, Cronbach's α = 0.82; FD: PSI = 0.76, Cronbach's α = 0.81). Although some minor anomalies are apparent, the modifications to the original C19-YRS have strengthened its measurement characteristics and its clinical and conceptual relevance.
Trial Registration: NCT05057260, ISRCTN15022307This 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.
## 1 | Introduction
Long Covid (LC) or post COVID-19 syndrome (PCS) is a multisystem clinical syndrome where symptoms persist for more than 3 months after acute infection with SARS-CoV-2 [1]. The prevalence of the condition is estimated to be ~3.3% (2 million cases) in the United Kingdom alone [2] and up to 36% worldwide [3].
Common symptoms include fatigue, shortness of breath, cognitive impairment, muscle and joint pain, chest pain, palpitations, persistent loss of smell and taste, gastrointestinal upset, and headache [4]. Symptoms may fluctuate [5] and ~20% of those suffering with LC describe their symptoms as severe [2]. For some individuals, symptoms may persist for more than 4 years (persistent LC) following the initial COVID-19 infection [6]. This protracted course of LC leads to a significant compromise on individuals' ability to work and conduct day-to-day tasks and can result in severely reduced health-related quality of life (HRQoL) [7][8][9].
Measuring symptom burden, functional disability (or ability) and HRQoL through patient-reported outcome measures (PROMs) is therefore crucial to understand the impacts on health, condition trajectories, and the cost-effectiveness of interventions. There have been a number of condition-specific PROMs developed for the condition during or after the pandemic [10][11][12].
The COVID-19 Yorkshire Rehabilitation Scale (C19-YRS) is the first LC condition-specific PROM reported in the literature [13]. The scale was developed to cover all of the 2001 WHO International Classification of Functioning, Disability and Health (ICF) framework [14] and was designed to capture the symptoms and functional problems associated with LC [13]. The instrument has been widely employed in primary care and community settings [15][16][17], rehabilitation interventions [18], and post-COVID epidemiological studies [19,20]. Following an initial psychometric analysis of the C19-YRS [21], psychometric and clinical evidence, and feedback from patients and healthcare professionals was subsequently integrated, culminating in a modified version of the instrument (C19-YRSm) [22]. The C19-YRSm has since undergone further classical psychometric validation and has been shown, for instance, to have good internal reliability and convergent validity in a Croatian patient population [23]. More recent validation has demonstrated the C19-YRSm to have good psychometric properties [24], in terms of internal consistency and test-retest reliability, as well as discriminant and convergent validity. Factor analysis supported the instrument's factor structure. Furthermore, an exploratory minimal important difference (MID) and minimal clinically important difference (MCID) were determined for the subscale scores [24].
The psychometric properties of the modified C19-YRSm have not yet been assessed using Rasch measurement theory (RMT). In contrast to classical psychometrics, which are focused on the test-level, that is, the instrument as a whole, RMT allows for item-level analysis. This enables the identification of individual PROM items that may potentially require modification (or removal) in order to improve the measurement properties of a PROM. The aim of this study was therefore to use Rasch measurement methodology to psychometrically assess and further validate the C19-YRSm using data collected from a large-scale, multicenter study (LOng COvid Multidisciplinary consortium Optimising Treatments and servIces, LOCOMO-TION) [25].
## 2 | Materials and Methods
## 2.1 | C19-YRSm
The C19-YRSm [22] consists of four separate subscales: Symptom Severity (SS), Functional Disability (FD), Other Symptoms (OS), and Overall Health (OH). The OH subscale is a single item, scored on a 0-10 numeric rating subscale, with a score of 0 representing "worst health" and 10 being "best health." Given the OH is a single-item subscale it cannot be analyzed using the Rasch model. The OS subscale consists of a checklist of 25 additional symptoms, where respondents select the symptoms that they have experienced over the last 7 days based on yes/no options. The analysis of the OS subscale is not presented here.
For the remaining subscales, both the SS (26 items condensed to 10 core items, see below) and FD (5 items) are summed individually to form total scores for each subscale. All items on the SS and FD subscales are scored on a 4-point subscale (0 = No problem; 1 = Mild problem; 2 = Moderate problem; 3 = Severe problem) where a higher score represents a higher severity of the problem, that is, worse symptoms or worse functional disability. Some of the SS subscale core items are grouped within subsets. For these items, the maximum value observed within the subset is selected as the representative value. This scoring step is taken due to the inherent (local) dependency between the items within a subset, as observed during the modification of the original C19-YRS [20]. Taking the maximum score from within a set avoids local dependency (LD), whilst maintaining the clinical utility of the individual component items. The sections concerning breathlessness, cough/throat sensitivity, smell/taste, pain/discomfort, cognition, palpitation/dizziness, and anxiety/mood each have a maximum score that is taken from across multiple items in the section.
## 2.2 | Data Collection
Data collection was carried out as part of the LOCOMOTION study [25] (NIHR Ref: COV-LT2-0016), with the C19-YRSm data collected routinely within 10 participating LC services across the United Kingdom between December 2021 and October 2023. Participants with a clinical diagnosis of LC by a qualified healthcare professional were National Institute for Health and Care Excellence (NICE) eligible for inclusion in one of the ten participating centers. Participants had to meet the UK NICE case definition, that is, one or more persistent symptoms developed during or postinfection that are consistent with COVID-19 and not explained by alternative diagnoses [26]. Consent and clinical data were collected using the ELAROS digital PROMs platform [27]. Ethics approval for the LOCO-MOTION study was obtained from the Bradford and Leeds Research Ethics Committee on behalf of Health Research Authority and Health and Care Research Wales (Reference: 21/ YH/0276).
## 2.3 | Rasch Analysis
Rasch analysis of the data [26,28] was undertaken using RUMM2030 software [29], and carried out separately for the SS subscale (10 items) and the FD subscale (5 items). Key criteria for the RMT are (1) Unidimensionality-whether the items represent a single factor, (2) Item fit-whether the items fit the Rasch model, (3) Local dependence-the absence of any further association between items beyond that explained by the underlying trait, (4) Response category functioning (or threshold disordering)requiring the latent trait to increase monotonically across item response categories, and (5) Item invariance (or absence of item bias or differential item functioning (DIF))-requiring item properties to be invariant to subgroup characteristics (e.g., gender, ethnicity) where latent trait levels are equivalent.
1. Unidimensionality was evaluated by a series of t-tests [30],
with multidimensionality indicated if independent subsets of items delivered significantly different person estimates, and the lower bound 95% CI percentage of significantly different t-tests was > 5%.
2. The Rasch analytic process included several standard tests of fit, covering both the overall subscale and item-level fit. All items were assessed individually for fit to the Rasch model (Partial Credit Model) [31] within the subscale item set to assess whether each item contributes to the underlying construct. Item misfit was indicated where the Bonferroni-adjusted χ 2 p value was statistically significant for an item and the standardized (z-score) fit-residuals fell outside ±2.5 [32,33].
3. Tests of LD were carried out to determine whether any items in the subscale were more closely related than is explained by the underlying construct; LD was indicated using a residual correlation (Q3 value) criterion cut point of 0.2 above average residual correlation [34].
4. Response category functioning was assessed to determine whether the response structure of the items was operating in the intended manner. A functional 0-3 response category structure for each item would be indicated by sequential response thresholds (the crossover points between subsequent response categories) on the underlying (logit) subscale [35].
## 5. Item bias was assessed through uniform and nonuniform
DIF testing by sex (male/female), age group (16-49; 50+ years), and ethnicity, with significant DIF indicated at a Bonferroni-adjusted ANOVA p value.
Furthermore, reliability indices were taken as the Person Separation Index (PSI), and the Cronbach's α values, and the scalesample targeting was assessed graphically through the relative distribution of item and person locations, along with the calculation of floor and ceiling effects. When the Rasch model assumptions are satisfied, the sufficiency of the raw score allows for the transformation into a linear, interval-level transformation [33].
## 2.4 | Cross-Validation
In order to assess the replication of results across independent samples, the complete sample was randomly split into three equally sized subsamples which were examined separately. This strengthens the analysis through replication, whilst avoiding the overpowering of RUMM fit statistics and misinterpretation that can occur with sample sizes > 500 [36,37]. The subsamples were used to assess Rasch-based individual item fit and DIF. However, the complete sample was utilized to assess response category functioning, targeting, LD, and the reliability indices, as these tests operate better with the precision afforded by larger sample sizes. To allow for brevity of reporting, only the results of the first subsample are presented within the manuscript.
## 2.5 | Working Group/Patient Advisory Group
All empirical results were reported back to a working group made up of clinicians, patients, and social scientists with expertise in PROMs and psychometrics, and additionally to the wider LOCOMOTION team, for sense-checking from both the patient and clinical context. Results were also reported back to a patient advisory group (PAG), and any potential further modifications were discussed within the working group and the PAG, with an emphasis on the practical implications of any potential change in the instrument.
## 3 | Results
## 3.1 | Sample
Data from 1278 patients were included in the study. The mean age was 48.6 (standard deviation, SD: 12.7) years, predominantly female (67%) and White (79%). The mean time since infection was 418 (SD: 268.7) days. Levels of pre-COVID comorbidity were low with the most commonly reported being mental health (17%) and respiratory conditions (10%). Vaccination status was recorded for only 45% of the sample. The key demographic characteristics of the sample are presented in Table 1. The data for the presented cross-validation subsample were from 423 patients.
## 3.2 | Rasch Analysis
## 3.2.1 | Symptom Severity Subscale: Unidimensionality, Reliability, Floor/Ceiling, and Targeting
A summary of the psychometric properties of both subscales is presented in Table 2, including results for the full sample and for the cross-validation subsample. The results indicated that the SS subscale was unidimensional, although the percentage of significant t-tests fell just outside the lower bound 95% CI, 5% criterion (5.1%) on the full sample. A good level of internalconsistency reliability (0.82) was displayed with no floor or ceiling effect (see Table 2 and Figure 1).
## 3.2.2 | Symptom Severity Subscale: Local Dependency
The average residual correlation was -0.10 and therefore the criterion value to indicate LD was taken as 0.10 (-0.10 + 0.2). Two local dependencies out of the 45 (4%) pairs were identified (Table 3). The largest LD was observed between the "Fatigue" and "Post-exertional malaise" items (0.17); the other between the "Breathlessness" and "Cough" (0.13). Both of these dependencies appear to be conceptually logical, suggesting that these findings are real and not just a chance finding.
## 3.2.3 | Symptom Severity Subscale: Individual Item Fit
Most items were within the acceptable fit ranges, although misfit was identified for some items (Table 3). The "Cough" item (covering "cough" and "throat sensitivity") displayed the largest chi-square and fit residual misfit anomalies, with the high positive fit residual value indicating an underdiscrimination. The "Smell/Taste" and "Cognition" items also demonstrated some misfit, although this was borderline in both instances, and inconsistent among the different subsamples.
## 3.2.4 | Symptom Severity Subscale: Category Response Structure
The modified 4-response category format mostly displayed an ordered, functional response structure across all items, evidencing a marked improvement from the response functioning of the original C19-YRS. However, the "Smell/Taste" item was consistently disordered among the full sample and all subsamples, with a nonborderline response structure suggesting that a binary response format may be more appropriate. Additionally, the full sample displayed three further items as borderline disordered ("Fatigue," "PEM," and "Sleep"), with these same items either ordered or borderline disordered among the three smaller subsamples, suggesting that this is not problematic.
## 3.2.5 | Symptom Severity Subscale: Item Location Ordering (Easiest/Most Difficult Items to Affirm)
The item location ordering can be observed in Table 3. The "Fatigue" item marked the lowest location on the subscale, meaning that fatigue is observed as the most frequently problematic issue on the SS subscale, that is, the most easily endorsed item by people with LC. Conversely, the "Smell/Taste" item had the highest location on the subscale, meaning that smell/taste is observed as the least frequently problematic issue on the SS subscale.
## 3.2.6 | Symptom Severity Subscale: Differential Item Functioning
No items displayed any consistently significant DIF by sex, age group, or ethnicity grouping. However, the limited sample for minority ethnic groups was insufficient to make the findings on ethnicity reliable.
## 3.2.7 | Symptom Severity Subscale: Post Hoc Analysis Addressing the Issues Found
In order to determine whether amendments could be made to address and thereby potentially resolve the individual item issues that had been identified, the first subsample (N = 423) was taken as an experimental data set and a number of analysis iterations were run. The first analysis focused on retaining all items in the SS subscale and involved rescoring the "Smell/Taste" and "Cough" items into a binary response format. This resolved the associated response structure and item fit issues. Furthermore, the "Breathlessness" and "Cough" items were subtested, that is, added together into a single item, rather than contributing as two separate items, to account for local dependency. This resolved the dependency issues, although the subtested item displayed some misfit (Fit Residual: 3.8).
## Symptom Severity (SS)
The second analysis focused on model fit and involved removing "Cough" whilst retaining the rescored (binary) "Smell/ Taste" item. This resolved all fit and dependency issues, although the PEM item continued to display a borderline disordered response structure. Finally, in order to examine the impact of these amendments on person estimates ("scores"), the person estimates from the complete subscale (full sample) were correlated against the fullsample person estimates from both analyses. This indicated that there was a strong (Spearman's) correlation between the complete subscale person estimates and both post hoc analyses estimates (0.99 and 0.98, respectively), indicating that the subscale amendments have very little effect on person ordering.
Taken together, this suggests that, despite the issues identified, it is perhaps optimal to retain the complete SS subscale in its original format, in order to retain maximum information and allow for continuity of data collection and comparison. Given that Rasch model assumptions have been satisfied, the transformation of the raw ordinal scale scores into interval-level equivalent scores is appropriate, and these transformed scores are available in Table 5. Please note that this transformation is only valid for complete data, where all items have been included in the total score.
## 3.2.8 | Functional Disability Subscale: Unidimensionality, Reliability, Floor/Ceiling, and Targeting
A summary of the psychometric properties for the FD subscale is presented in Table 2. The FD subscale was unidimensional (only 2.6% of unidimensionality t-tests were statistically significant), displayed a good level of internal-consistency reliability (0.82), and had good subscale-sample targeting with no floor or ceiling effect (see Table 2 and Figure 1).
## 3.2.9 | Functional Disability Subscale: Local Dependency
The average correlation was -0.23 and therefore the criterion value to indicate LD was taken as -0.03 (-0.23 + 0.2). One local dependency (Table 4) was observed between the "Walking or moving around" and "Personal care" items. Again, there appears to be a conceptual connection between these two items, suggesting that this is a real dependency rather than a chance finding.
## 3.2.10 | Functional Disability Subscale: Individual Item Fit
The "Communication" item consistently displayed a fit residual misfit anomaly, with the high positive fit residual value indicating an underdiscrimination (Table 4). No other items indicated any evidence of misfit.
## 3.2.11 | Functional Disability Subscale: Category Response Structure
The response categories displayed an ordered, functional response structure across all items. None of the items were disordered.
## 3.2.12 | Functional Disability Subscale: Item Location Ordering (Easiest/Most Difficult Items to Affirm)
The "Other activities of daily living" item marked the lowest location on the subscale, that is, the most frequently problematic (or easily endorsed) issue on the FD subscale. Conversely, the "Personal care" item had the highest location on the subscale, representing the least frequently problematic issue on the FD subscale. The item location ordering can be observed in Table 4.
## 3.2.13 | Functional Disability Subscale: Differential Item Functioning
There was no significant DIF by sex, age group, or ethnicity grouping indicated for any of the items across any of the samples. As highlighted earlier, however, the ethnic minority sample was insufficiently powered to make the finding on ethnicity reliable. DIF by age group was observed for the "Walking or moving around" item, this was, however, consistent with expectation, as the older group are more likely to report issues with walking.
## 3.2.14 | Functional Disability Subscale: Post Hoc Analysis Addressing the Issues Found
Although there were relatively few issues found in the FD subscale, the first subsample (N = 423) was again taken as an experimental data set and amendments were made to address the issues that had been identified. This involved subtesting the "Walking or moving around" and "Personal care" items in order to resolve the dependency issues, although this resulted in a further borderline dependency between the "Other activities of daily living" and "Social role" in the full sample. However, no further subtesting was carried out due to the borderline nature of the apparent dependency.
In order to examine the impact of this amendment on person estimates, the person estimates from the complete scale (full sample) were correlated against the full-sample person estimates from the resolved analysis. This indicated a very strong (Spearman's) correlation of 0.999, indicating that the subscale amendment had very little effect on the ordering of persons. Again, this suggests that the parsimonious retention of the complete FD subscale in its original format would retain maximum information and allow for continuity of data collection and comparison.
Again, given that Rasch model assumptions have been satisfied, the transformation of the raw ordinal scale scores into intervallevel equivalent scores is appropriate, and these transformed scores are available in Table 5 (for complete data).
## 3.3 | Working Group/PAG
A benefit of the C19-YRSm is that it is a relatively short measure with a simple response structure. The feedback from the PAG suggested that it was not burdensome for patients to complete, and that the simplified response format was more appropriate than the previous 11-point numeric rating scale on the original C19-YRS. The PAG suggested that the C19-YRSm was comprehensible, easy to understand, and that the range of symptoms covered across the SS and OS scales (not presented here) was broadly comprehensive, whilst remaining manageable.
## 4 | Discussion
This study provides the first Rasch analysis of the latest version of the C19-YRSm following the initial development [13,21] and subsequent psychometric analysis [22]. In line with the classical psychometric analysis [24], the results demonstrated evidence of a two-factor structure, comprising unidimensional SS and FD subscales, with both displaying good internal reliability.
Within the subscale analysis, the few item anomalies observed concerned the "Cough" and "Smell/Taste" items, both within the SS subscale. It is uncertain why these items were inconsistent with the other subscale items, but there are some potential explanations. For instance, these items also marked the "most difficult" end of the subscale, meaning that they were generally reported to be problematic less frequently than the other items. This positioning means that there is less certainty in respect of the item fit characteristics of the items (given that less sample information is available). Furthermore, it also means that these items are an important demarcation of the upper measurement range of the subscale. The post hoc removal and amendment of these items had very little effect on the ordering of the person estimates that were generated, therefore the added clinical and measurement information provided by the retention of these items would seem to outweigh any potential improvement in subscale fit [37].
Although the cough, throat sensitivity, and anosmia items were recognized by the Working Group as common and important symptoms of COVID, the analysis results indicate that their contribution to the impact of LC on a person's daily life is less clear. It is also possible that a binary (no problem/problem) response format may be more appropriate for these items than the 4-point response structure, especially for the "Smell/Taste" item. Feedback from the Working Group suggested that a binary response would align with the manifestation of these symptoms that present in LC/PCS clinics, especially for the "Smell/Taste" item.
The input from the Working Group/PAG also suggested that the C19-YRSm was not burdensome for patients to complete, and that the simplified response format was more appropriate than the previous 11-point numeric rating subscale. The benefit of the C19-YRSm is that it is a relatively short measure with a simple response structure. If outcome measures are to be repeatedly used in clinical or research settings, they should be relevant, clinically useful, and nonburdensome to patients. Long PROMs may lead to questionnaire response burden, which is recognized as a threat to subscale completion and adherence in trials [38].
This study is the first, to the authors knowledge, to investigate item bias or DIF for the C19-YRSm and demonstrated little or no DIF across age groups and gender, thus reinforcing the instrument usability across a wide population of people living with LC. However, one potential limitation to be noted here is that minority ethnic groups were considerably underrepresented within our patient sample. Despite the fact there was no evident item bias, there is still a need for further research that is sufficiently powered to explore any variation by ethnicity in LC symptoms. Furthermore, although a recent systematic review [39] determined that the content validity of the C19-YRSm was sufficient, there is, in general, further research required involving additional input from people with LC as part of the instrument's ongoing validation process.
Nevertheless, the C19-YRS scored well as one of the suggested LC condition-specific instruments, covering a number of the core outcomes identified in an international consensus study [40] when evaluated against reporting a core outcome set of 12 outcomes that should be measured in all future clinical studies and in clinical care for people with LC [41]. In addition to the current study, previous validation studies [22,24] have shown the C19-YRSm to have robust psychometric properties and this is reflected in the growing evidence supporting the clinical utility of the C19-YRSm [42,43].
## 5 | Conclusion
The modified C19-YRSm has been demonstrated to have significant advantages over the original C19-YRS. The content coverage is much improved, including a number of common symptoms that were not included in the original version. This study and preceding validation studies have shown the SS and FD subscales to be far more robust than the original C19-YRS, thereby strengthening the measurement characteristics of the C19-YRSm and enhancing its clinical and conceptual relevance for use in both clinical and community settings. For research purposes, an interval-level score transformation is available, allowing for the calculation of parametric statistics on C19-YRSm scores. Note: This conversion is only valid in the case of complete data.
for their PROMs platform. The study was funded through a National Institute for Health and Care Research award (COV-LT2-0016). The views expressed in this publication are those of the authors and not necessarily those of NIHR or the Department of Health and Social Care.
## References
1. Davis, Mccorkell, Vogel et al. (2023) "Long COVID: Major Findings, Mechanisms and Recommendations" *Nature Reviews Microbiology*
2. (2024) "Self-Reported Coronavirus (COVID-19) Infections and Associated Symptoms"
3. Hou, Gu, Ni (2025) "Global Prevalence of Long COVID, Its Subtypes, and Risk Factors: An Updated Systematic Review and Meta-Analysis" *Open Forum Infectious Diseases*
4. Bowe, Xie, Al-Aly (2023) "Postacute Sequelae of COVID-19 at 2 Years" *Nature Medicine*
5. Greenwood, Mansoubi, Bakerly (2024) "Physical, Cognitive, and Social Triggers of Symptom Fluctuations in People Living With Long COVID: An Intensive Longitudinal Cohort Study" *Lancet Regional Health -Europe*
6. Sivan, Smith, Osborne (2024) "Long COVID Clinical Severity Types Based on Symptoms and Functional Disability: A Longitudinal Evaluation" *Journal of Clinical Medicine*
7. Davis, Assaf, Mccorkell (2021) "Characterizing Long COVID in an International Cohort: 7 Months of Symptoms and Their Impact" *EClinicalMedicine*
8. Kwon, Milne, Rayner (2024) "Impact of Long COVID on Productivity and Informal Caregiving" *European Journal of Health Economics*
9. Walker, Goodfellow, Pookarnjanamorakot (2023) "Impact of Fatigue as the Primary Determinant of Functional Limitations Among Patients With Post-COVID-19 Syndrome: A Cross-Sectional Observational Study" *BMJ Open*
10. Hughes, Haroon, Subramanian "Development and Validation of the Symptom Burden Questionnaire for Long Covid"
11. (2022) "Rasch Analysis"
12. Rofail, Somersan-Karakaya, Mylonakis (2024) "The Symptoms Evolution of Long COVID-19 (SE-LC19): A New Patient-Reported Content Valid Instrument" *Journal of Patient-Reported Outcomes*
13. Jandhyala (2021) "Design, Validation and Implementation of the Post-Acute (Long) COVID-19 Quality of Life (PAC-19QoL) Instrument" *Health and Quality of Life Outcomes*
14. Sivan, Halpin, Gees (2021) "The Self-Report Version and Digital Format of the COVID-19 Yorkshire Rehabilitation Scale (C19-YRS) for Long Covid or Post-COVID Syndrome Assessment and Monitoring" *Advances in Clinical Neuroscience & Rehabilitation*
15. Patel, Straudi, Yee Sien et al. (2020) "Applying the WHO ICF Framework to the Outcome Measures Used in the Evaluation of Long-Term Clinical Outcomes in Coronavirus Outbreaks" *International Journal of Environmental Research and Public Health*
16. López-López, Calvache-Mateo, Ortiz-Rubio (2023) "Differences of Disabling Symptoms Between Previously Hospitalized or Non-Hospitalized Currently Working Long-COVID Survivors One Year After Infection: A Descriptive Study" *Healthcare*
17. Partiprajak, Krongthaeo, Piaseu et al. (2023) "The Thai Version of the COVID-19 Yorkshire Rehabilitation Scale: A Valid Instrument for the Psychometric Assessment of the Community Members" *BMC Public Health*
18. Zhang, Chung, .-C. Chan (2023) "Determinants of Post-COVID-19 Symptoms Among Adults Aged 55 or Above With Chronic Conditions in Primary Care: Data From a Prospective Cohort in Hong Kong" *Frontiers in Public Health*
19. Kabir, Yin, Jeffree (2023) "Profile of Long COVID Symptoms Needing Rehabilitation: A Cross-Sectional Household Survey of 12,925 SARS-CoV-2 Cases Between July and December 2021 in Bangladesh" *Archives of Public Health*
20. García, Besteiro, Balado et al. (2023) "Assessment of Post-COVID Symptoms Using the C19-YRS Tool in a Cohort of Patients From the First Pandemic Wave in Northwestern Spain" *Telemedicine and e-Health*
21. Ho, Xu, Kwong (2023) "Prevalence, Patterns, and Clinical Severity of Long COVID Among Chinese Medicine Telemedicine Service Users: Preliminary Results From a Cross-Sectional Study" *International Journal of Environmental Research and Public Health*
22. O'connor, Preston, Parkin (2022) "The COVID-19 Yorkshire Rehabilitation Scale (C19-YRS): Application and Psychometric Analysis in a Post-COVID-19 Syndrome Cohort" *Journal of Medical Virology*
23. Sivan, Preston, Parkin "The Modified"
24. (2022) "Yorkshire Rehabilitation Scale (C19-YRSm) Patient-Reported Outcome Measure for Long Covid or Post-COVID-19 Syndrome" *Journal of Medical Virology*
25. Kustura, Bobek, Poljičanin (2024) "Psychometric Properties and Observational Data for COVID-19 Yorkshire Rehabilitation Scale (C19-YRSm) for Post-COVID-19 Syndrome" *QJM*
26. Smith, Greenwood, Horton (2024) "Psychometric Analysis of the Modified COVID-19 Yorkshire Rehabilitation Scale (C19-YRSm) in a Prospective Multicentre Study" *BMJ Open Respiratory Research*
27. Sivan, Greenhalgh, Darbyshire (2022) "LOng COvid Multidisciplinary Consortium Optimising Treatments and Services AcrOss the NHS (LOCOMOTION): Protocol for a Mixed-Methods Study in the UK" *BMJ Open*
28. (2024) "COVID-19 Rapid Guideline: Managing the Long-Term Effects of COVID-19"
29. Sivan, Rocha Lawrence, O'brien (2023) "Digital Patient Reported Outcome Measures Platform for Post-COVID-19 Condition and Other Long-Term Conditions: User-Centered Development and Technical Description" *JMIR Human Factors*
30. Rasch (1993) "Probabilistic Models for Some Intelligence and Attainment Tests"
31. Andrich, Sheridan, Luo (2010) "RUMM 2030 (RUMM Laboratory"
32. Smith (2002) "Detecting and Evaluating the Impact of Multidimensionality Using Item Fit Statistics and Principal Component Analysis of Residuals" *Journal of Applied Measurement*
33. Masters (1982) "A Rasch Model for Partial Credit Scoring" *Psychometrika*
34. Hagquist, Bruce, Gustavsson (2009) "Using the Rasch Model in Nursing Research: An Introduction and Illustrative Example" *International Journal of Nursing Studies*
35. Tennant, Küçükdeveci (2023) "Application of the Rasch Measurement Model in Rehabilitation Research and Practice: Early Developments, Current Practice, and Future Challenges" *Frontiers in Rehabilitation Sciences*
36. Christensen, Makransky, Horton (2017) "Critical Values for Yen's Q 3: Identification of Local Dependence in the Rasch Model Using Residual Correlations" *Applied Psychological Measurement*
37. Andrich (2011) "Rating Scales and Rasch Measurement" *Expert Review of Pharmacoeconomics & Outcomes Research*
38. Hagell, Westergren (2016) "Sample Size and Statistical Conclusions From Tests of Fit to the Rasch Model According to the Rasch Unidimensional Measurement Model (RUMM) Program in Health Outcome Measurement" *Journal of Applied Measurement*
39. Hagell (2014) "Testing Rating Scale Unidimensionality Using the Principal Component Analysis (PCA)/t-Test Protocol With the Rasch Model: The Primacy of Theory Over Statistics" *Open Journal of Statistics*
40. Basch, Abernethy, Mullins et al. (2011) "EV1 Development of a Guidance for Including Patient-Reported Outcomes (PROS) in Post-Approval Clinical Trials of Oncology Drugs for Comparative Effectiveness Research (CER)" *Value in Health*
41. Baalmann, Blome, Stoletzki et al. (2024) "Patient-Reported Outcome Measures for Post-COVID-19 Condition: A Systematic Review of Instruments and Measurement Properties" *BMJ Open*
42. Gorst, Seylanova, Dodd (2023) "Core Outcome Measurement Instruments for Use in Clinical and Research Settings for Adults With Post-COVID-19 Condition: An International Delphi Consensus Study" *Lancet Respiratory Medicine*
43. Munblit, Nicholson, Akrami (2022) "A Core Outcome Set for Post-COVID-19 Condition in Adults for Use in Clinical Practice and Research: An International Delphi Consensus Study" *Lancet Respiratory Medicine*
44. Germonpré, Rinia, Caeyers et al. (2025) "Effect of Normobaric and Hyperbaric Hyperoxia Treatment on Symptoms and Cognitive Capacities in Long COVID Patients: A Randomised Placebo-Controlled, Prospective, Double-Blind Trial" *Diving and Hyperbaric Medicine Journal*
45. Corrado, Iftekhar, Halpin (2024) "HEART Rate Variability Biofeedback for LOng COVID Dysautonomia (HEARTLOC): Results of a Feasibility Study" *Advances in Rehabilitation Science and Practice* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12587552&blobtype=pdf | # A nitrogen-containing diphyllin derivative C156-P1 exhibited broad-spectrum antiviral activity against Flaviviridae viruses by preventing endosomal acidification
Guoquan Chen, Wanfei Li, Ka Lam, Mingyue Hu, Qian Wu, Xiangyu Xu, Yunzhu Huang, Fei Tang, Guohui Cui, Ping Cui, Jianping Zuo, Linna Liu, Jun Qian, Hong-Jie Zhang, Yi-Ping Li
## Abstract
Dengue virus (DENV) represents a significant public health threat, with its four serotypes estimated to account for approximately 96 million symptomatic infections annually. Currently, there are no antiviral agents available for the prevention or treatment of DENV infection. Here, we initially screened 12 diphyllin derivatives and identified C156-P1, a nitrogen-containing compound, as a potent agent against DENV infection. Further, C156-P1 exhibited inhibitory effects against the viruses of the Flaviviridae family, including four serotypes of DENV (DENV-1 to DENV-4) in multiple human and monkey cell lines, as well as Zika virus, Japanese encephalitis virus, yellow fever virus, and hepatitis C virus. In addition, C156-P1 also showed inhibition of the infections of herpes simplex virus type 1 and vesicular stomatitis virus, but not adenovirus and Sendai virus. Mechanistic studies demonstrated that C156-P1 inhibited DENV-2 after cell entry but before the endosomal membrane fusion step. C156-P1 inhibited vacuo lar-type ATPase activity by perturbing the expression of ATP6V0A2 subunit, thereby suppressing endosomal acidification. Consequently, DENV was restricted in the late endosome, inhibiting virus fusion with endosomal membranes and resulting in infection inhibition. C156-P1 treatment also suppressed both IFN-I responses and endosomal TLR3 activation induced by DENV-2 infection. Furthermore, administration of C156-P1 in AG129 mice significantly reduced DENV-2 infection and effectively increased the survival rate of the mice. Taken together, our study demonstrates that the novel nitrogen-con taining diphyllin derivative C156-P1 functions as a broad-spectrum antiviral agent by inhibiting endosomal acidification, thus representing a promising host-targeting antiviral candidate for future development.
D engue virus (DENV) belongs to the Orthoflavivirus (formerly Flavivirus) genus of the Flaviviridae family, which also includes significant pathogenic viruses such as Zika virus (ZIKV), Japanese encephalitis virus (JEV), West Nile virus (WNV), and yellow fever virus (YFV) (1). DENV infections lead to a spectrum of diseases ranging from self-limit ing dengue fever to severe dengue, a potentially lethal condition with hemorrhagic manifestations and capillary leak, formerly known as dengue hemorrhagic fever and dengue shock syndrome (2). Four DENV serotypes (DENV-1 to DENV-4) sustain the transmission cycle in humans through Aedes mosquitoes as vectors, leading to epidemics across 129 countries. This results in approximately 390 million infections annually, with around 96 million symptomatic cases and an estimated 10,000 fatalities (3)(4)(5). Further more, dengue was announced as one of the top 10 threats to global health by the World Health Organization in 2019. To date, no antiviral drug has been approved for treating DENV infection, and both DENV and other flaviviruses continuously pose significant threats to global health (6). Therefore, there is an urgent need for the development of anti-flavivirus drugs to treat flavivirus infections.
DENV is an enveloped virus with a single-stranded, positive-sense RNA genome of approximately 10.7 kilobases, comprising a single open reading frame (ORF) flanked by structured 5′ and 3′ untranslated regions (7). The ORF is translated into a polyprotein that is cleaved by cellular and viral proteases to produce three structural proteins (capsid, C; precursor membrane, prM; and envelope, E) and seven nonstructural proteins (NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5) (8). The structural proteins constitute the virus particles and are critical for cellular entry, while the nonstructural proteins primarily facilitate viral genome replication and translation, as well as interactions with host factors essential for the viral life cycle (9). DENV is believed to enter host cells through clathrin-dependent receptor-mediated endocytosis (10). Internalized virions undergo low pH-dependent fusion of viral and endosomal membranes, releasing viral RNA into the cytoplasm, followed by translation in the endoplasmic reticulum (ER) and replication within invaginated membrane vesicles (11). After viral RNA is associated with C protein, a nucleocapsid is formed and coated with prM and E proteins before budding into the ER lumen, resulting in immature viral particles. These particles are transported through the Golgi to the trans-Golgi network, where they mature via prM cleavage by furin and are released (8,12). Targeting distinct stages of the viral life cycle represents a fundamental strategy in drug development. Furthermore, the similar life cycles of flaviviruses enhance the potential for developing broad-spectrum anti-flavivirus drugs.
Numerous chemical compounds with therapeutic activities have been derived from natural products, predominantly from plants (13). Previously, we identified diphyllin (DP), an arylnaphthalene lignan (ANL), from the traditional Chinese medicinal plant Justicia gendarussa (14,15). DP has been determined as a potent vacuolar-type ATPase (V-ATPase) inhibitor, which inhibits lysosomal acidification (16,17). Studies have provided evidence supporting the broad-spectrum antiviral activity of DP against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (18), human immunodeficiency virus type 1 (HIV-1) (19), influenza virus (20), and Rift Valley fever virus (RVFV) (21). Addition ally, DP exhibits multiple biological activities against tumors (22), inflammation (23), oxidation (24), and bacteria (25). However, the further development of DP as an antiviral therapeutic agent has been hampered by its low aqueous solubility and inadequate antiviral activity, highlighting the pressing need for structural modification. Recent studies have shown that nitrogen-containing diphyllin derivatives exhibit stronger anti-Ebola virus infection activity than DP (26). Our recent study showed that nitro gen-containing DP derivatives generally provide superior inhibitory activities against DENV-3 replicon and infections of DENV-1-4 (27). However, these findings require further validation using authentic flaviviruses, and the molecular mechanisms underlying their antiviral activities remain to be elucidated.
In the current study, we evaluated 12 DP derivatives for their potential anti-DENV activities using authentic virus infection assays. Among these, the nitrogen-containing DP derivative C156-P1 was identified and demonstrated potent antiviral activity against DENV-1-4 infections in vitro and in vivo. Furthermore, C156-P1 exhibited antiviral efficacy against Flaviviridae viruses, including ZIKV, JEV, YFV, and hepatitis C virus (HCV), as well as other enveloped viruses such as HSV-1 and VSV. Mechanistic studies revealed that C156-P1 inhibited viral infection by targeting the ATP6V0A2 subunit of V-ATPase, thereby preventing endosomal acidification.
## RESULTS
## The DP derivative C156-P1 efficiently inhibited DENV-2 infection in vitro
The natural compounds DP and patentiflorin A (PTA, also known as 6-deoxyglucosediphyllin) are ANL lead molecules isolated from the medicinal plant J. gendarussa, and they have shown promise as broad-spectrum antiviral agents (28,29). We recently screened a library of 83 chemically synthesized ANL derivatives using a DENV-3 replicon and discovered that nitrogen-containing ANLs (N-ANLs) exhibited the highest antiviral potency (27). We selected 12 ANL derivatives with pronounced inhibitory effects on the DENV-3 replicon and evaluated their antiviral activities against the authentic DENV-2 strain 16681 (Fig. 1). The results showed that DP derivatives (C176-P1 and C180A-P1), nitrogen-containing DP derivatives (C156-P1, C165-P1, C169-P1, and C201-P1), and nitrogen-containing PTA derivatives (C45-P1, C58-P1, C59-P1, and C70-P1) exhibited lower 50% effective concentration (EC 50 ) values against DENV-2 than their parent compounds (DP or PTA). However, nitrogen-containing derivatives of both DP and PTA also exhibited lower 50% cytotoxic concentration (CC 50 ) in A549 cells compared to their parent compounds. Interestingly, the nitrogen-containing DP derivative C156-P1 showed the lowest EC 50 value (1.36 nM) against DENV-2 and the highest selective index (SI = 292.65) value (Fig. 1G). The chemical structure of C156-P1 is shown in Fig. 2H, and the chemical structures of the other compounds are presented in Fig. S1. In summary, C156-P1 was identified as the most promising antiviral compound among ANL derivatives tested and was further explored in this study.
## C156-P1 inhibited the infection of four serotypes of DENV (DENV-1-4) in different cell lines
Subsequently, we assessed the antiviral efficacy of C156-P1 against DENV-2 in different cell lines. We found that C156-P1 inhibited DENV-2 infection in Vero, human umbilical vein endothelial cell line (HUVEC), and Huh7 cells, with EC 50 <5 nM and SIs > 100 (Fig. 2A). The level of DENV-2 NS3 protein decreased with increasing concentrations of C156-P1 in Vero, A549, HUVEC, and Huh7 cells (Fig. 2B). A concentration of 50 nM C156-P1, which is tolerable for cytotoxicity (<62.5 nM in HUVEC cells), was used for subsequent experiments. We infected A549 cells with DENV-2 strain 16681 and observed that the viral RNA level and NS3 protein were downregulated in the C156-P1-treated groups in comparison with the DMSO-treated groups at 12, 24, 36, and 48 h postinfection (hpi). Concurrently, the viral titers were significantly decreased upon C156-P1 treatment at 24, 36, and 48 hpi (Fig. 2C). Similar inhibitory effects were observed in DENV-2-infected HUVEC and Huh7 cells (Fig. 2D andE). Immunofluorescence staining also showed a significant reduction in the number of DENV-2 NS3-positive A549 cells after C156-P1 treatment (Fig. 2F). Furthermore, C156-P1 effectively inhibited infections of DENV-1, DENV-3, and DENV-4 in A549 and Vero cells, with EC 50 values ranging from 4.84 to 8.38 nM (Fig. 2G). Collectively, these results demonstrate that C156-P1 robustly suppresses DENV infection in multiple cultured cell lines in a dose-dependent manner.
## C156-P1 inhibited the infections of ZIKV, JEV, YFV, and HCV in vitro
To determine whether C156-P1 could inhibit other viruses of the Flaviviridae family, we examined the antiviral effects of C156-P1 on ZIKV (SZ01 strain), JEV (SA14-14-2 strain), YFV (17D strain), and HCV (JFH1 strain). In Vero and A549 cells, C156-P1 dose-depend ently inhibited the infections of ZIKV, JEV, and YFV, with EC 50 <6 nM for all three viruses, as determined by plaque assays (Fig. 3A through C). Consistently, the levels of ZIKV NS3, JEV E, and YFV E proteins decreased with increasing concentrations of C156-P1 in both Vero and A549 cells (Fig. 3D). Similar to its effects on DENV-2, C156-P1 significantly suppressed ZIKV (Fig. 3E), JEV (Fig. 3F), and YFV (Fig. 3G) infections, as evidenced by reductions in viral RNA, protein levels, and infectious virus titers in A549 cells at 12, 24, 36, and 48 hpi. In addition, we assessed the antiviral activity of C156-P1 against HCV in Huh7 cells and found that it markedly inhibited HCV infection in a dose-dependent manner (Fig. 3H), with EC 50 value of 5.13 nM (Fig. 3I). These results suggest that C156-P1 exhibits broad-spectrum antiviral activities against multiple viruses within the Flaviviri dae family.
## C156-P1 inhibited the infection of enveloped viruses via endocytosis entry
To determine whether C156-P1 could inhibit other viruses beyond the Flaviviridae family, we tested its efficacy against viruses from the Herpesviridae, Rhabdoviridae, Adenoviridae, and Paramyxoviridae families. First, we evaluated the inhibitory efficacies of C156-P1 against HSV-1-eGFP and VSV-eGFP in A549 cells and found that the virus titers released to the supernatant decreased in a dose-dependent manner with C156-P1 treatment (Fig. 4A, left panel), with EC 50 values of 6.21 nM for HSV-1-eGFP virus and 9.27 nM for VSV-eGFP virus determined by plaque assays (Fig. 4A, right panel). The reduction in virus titers was further confirmed by the levels of eGFP (Fig. 4B) and the number of eGFP-positive cells (Fig. 4C). Next, we tested whether C156-P1 showed inhibitory effect against adenovirus (AdV). We infected A549 cells with Ad5-eGFP virus and found no significant difference in the number of Ad5-eGFP-positive cells between DMSO-treated and C156-P1-treated cultures (Fig. 4D, left panel). Surprisingly, the mean fluorescence intensity (MFI) within the cells slightly increased after treatment with C156-P1 (Fig. 4D, right panel), and the expression levels of eGFP (Fig. 4E) and viral hexon mRNA (Fig. 4F) were also increased. Finally, we tested whether C156-P1 inhibited SeV infection by determining the mRNA expression level of the viral nucleocapsid protein (NP) gene (the SeV without reporter gene). Similar to Ad5-eGFP virus, C156-P1 did not inhibit SeV infection, and the viral NP mRNA levels were slightly increased in a C156-P1 concentra tion-dependent manner (Fig. 4G). Taken together, these results indicate that C156-P1 has a broad-spectrum antiviral activity against several Flaviviridae viruses (DENV, ZIKV, JEV, YFV, and HCV) as well as HSV-1 and VSV, but does not inhibit the infection of AdV and SeV.
We hypothesized that this may be related to the different pathways through which different viruses enter host cells. Both HSV-1 and VSV are enveloped viruses that utilize a cell-entry pathway similar to that of Flaviviridae viruses, involving endocytosis and pH-dependent fusion (30,31). In contrast, AdV is a non-enveloped virus that enters host cells through clathrin-mediated endocytosis. Unlike Flaviviridae viruses, the AdV partial and incubated with C156-P1 at the indicated concentrations for 24 h. The levels of SeV RNA were detected by RT-qPCR. In panels D-G, data are presented as means ± SEM of three independent experiments (Student's t-test; ns, no significant; *, P < 0.05; **, P < 0.01).
capsid uncoats in the endosomes to release protein VI, which ruptures the endosomal membrane, allowing partially disassembled virions to enter the cytoplasm (32). SeV, an enveloped virus, enters the host cell by direct fusion with the cell membrane (33). Our previous studies demonstrated that DP can inhibit endosomal acidification, preventing the fusion of the viral membrane with the endosomal membrane and subsequent infection (28). Taken together, we speculate that the nitrogen-containing DP derivative C156-P1 may inhibit the infections of viruses that enter host cells through endocytosis and fusion in a pH-dependent manner.
## C156-P1 inhibited flavivirus infection at the early stages after cell entry
To determine which step of the viral life cycle might be targeted by C156-P1, we performed a time-of-drug-addition assay. C156-P1 was administered to A549 cells either at the virus entry (before and co-incubation) or post-entry (after) stage during flavivirus infection. A control group treated with the solvent DMSO was included in parallel (Fig. 5A). The cells and supernatants of each group were harvested at 48 hpi for subsequent analysis. In the DENV-2 infection experiment, the before-, during-, and after-incubation groups all demonstrated significantly reduced levels of intracellular viral RNA, viral NS3 proteins, and extracellular infectious virus titers compared to the DMSO control group (Fig. 5B). Among the three treatment groups, the co-incubation group showed the lowest levels of both viral RNA and infectious titers, followed by the before-incubation group and then the after-incubation group (Fig. 5B). Similarly, in ZIKV (Fig. 5C) or JEV (Fig. 5D) infections, the co-incubation group exhibited the lowest levels of infection. These results suggest that C156-P1 inhibits flavivirus infection at an early stage.
Next, we tested whether C156-P1 inhibited the early stages of flavivirus entry into host cells, including binding and internalization steps (Fig. 5E). The results demonstrated that C156-P1 did not exhibit inhibitory effects on either the internalization or binding steps of DENV-2 (Fig. 5F), ZIKV (Fig. 5G), and JEV (Fig. 5H) infections. Taken together, these results indicate that C156-P1 inhibits flavivirus infection at an early stage after the viruses have entered the cells.
## C156-P1 restricted the release of the DENV-2 genome into the cytoplasm from endosomes
DENV infection triggers the type I interferon (IFN-I) response via host recognition of the released viral RNA in the cytoplasm (34,35). DENV enters host cells through endocytosis, where low pH-induced conformational changes in the E protein facilitate virus fusion with the endosomal membrane, thereby releasing the viral genome into the cytoplasm (10). If DENV is inhibited at the pre-fusion stage by C156-P1, it is possible that viral RNA may not be released into the cytosol, or only a limited amount of viral RNA would be released. Thus, the IFN-I response may not be activated or may be attenuated in DENV-infected cells in the presence of C156-P1. To this end, we investigated the IFN-I response in A549 cells following DENV-2 infection with and without C156-P1 treatment. The results showed that DENV-2 infection significantly increased the IFN-I responses at 12, 24, 36, and 48 hpi, as determined by elevated mRNA levels of IFNB1 and inter feron-stimulated genes (ISGs) such as IFIT1, ISG15, and OAS1 (Fig. 6A). Notably, the levels of IFNB1 and these ISGs were apparently suppressed by C156-P1 treatment (Fig. 6A). Concurrently, the phosphorylation level of TANK-binding kinase 1 (TBK1) and IFN regulatory factor 3 (IRF3) was increased by DENV-2 infection, and this increase was also suppressed by C156-P1 treatment (Fig. 6B). As a control, C156-P1 treatment alone, without virus infection, did not change the levels of phosphorylated TBK1 and IRF3 (Fig. 6B). These results suggest that C156-P1 inhibits IFN-I responses induced by DENV-2 infection, implying that less viral RNA may be released into the cytoplasm.
Toll-like receptor 3 (TLR3) is a pattern recognition receptor (PRR) predominantly localized to the endosomal membrane, where it is primarily activated by viral dou ble-stranded RNA (dsRNA) (36)(37)(38). Thus, the degree of TLR3 activation implies the sensing of viral dsRNA on the endosomal membrane. Here, we investigated the levels of TLR3 expression in A549 cells following DENV-2 infection and whether C156-P1 treatment modulated DENV-related TLR3 expression. The results showed that TLR3 mRNA levels increased in a time-dependent manner at 12, 24, 36, and 48 hpi, which correlated with DENV-2 genome replication (Fig. 2C). As expected, C156-P1 treatment significantly reduced the expression of TLR3 mRNA and abolished its time-depend ent pattern (Fig. 6C). These results indicate that C156-P1 suppressed TLR3 activation, implying that dsRNAs were restricted from exposure to the endosomal membrane. Taken together, C156-P1 inhibited IFN-I responses and endosomal TLR3 activation induced by DENV-2 infection, suggesting that C156-P1 treatment largely restricts DENV release from endosomes into the cytoplasm.
## C156-P1 mediated endosomal acidification by targeting ATP6V0A2 subunit of ATPase to restrict DENV residing in endosomes
We have demonstrated that C156-P1 inhibited DENV infection during the early stages post-cell entry and reduced the activation levels of IFN-I and TLR3. This leads to the hypothesis that C156-P1 attenuates or suppresses the release of the viral RNA into the cytoplasm. It is known that following DENV entry into endosomes, endosomal acidification is a prerequisite for the conformational rearrangement of the DENV E protein, which prepares it for membrane fusion and viral RNA release. It is established that, following entry into the cell via endocytosis, DENV undergoes a conformational rearrangement of its E protein in response to endosomal acidification. This process facilitates the fusion of the viral and cell endosomal membranes, ultimately resulting in the release of the viral genome into the cytoplasm (39,40). Therefore, we hypothesize that C156-P1 attenuates or suppresses the release of the viral genome into the cytoplasm, thereby reducing DENV infection.
Thus, we investigated whether C156-P1 inhibited DENV-2 infection by preventing endosomal acidification, as well as whether viral particles resided within the endosomes. First, we assessed whether C156-P1 affected the pH values of endosomes and lysosomes using an acid-sensitive indicator, DND-189 (41,42). A V-ATPase inhibitor, bafilomycin A1 (Baf-A1), was used as a positive control to prevent the acidification of endosomes. The results demonstrated that C156-P1 (blue) and Baf-A1 (red) treatments shifted the peak of FITC leftward compared to the DMSO mock treatment (purple) (Fig. 6D, left panel). Besides, C156-P1 significantly decreased the MFI of FITC in a dose-dependent manner (Fig. 6D, right panel). These results indicate that C156-P1 exhibits a mechanism similar to Baf-A1 in preventing endosomal acidification.
After internalization, DENV particles are delivered to Rab5-positive early endo somes, which subsequently mature into Rab7-positive late endosomes, where virus-cell membrane fusion occurs in a low pH-dependent manner (43). To further determine whether C156-P1 restricts viral particles residing within the endosome, A549 cells were transfected with mNeonGreen-Rab7A, then infected with DENV-2 in the presence of C156-P1 (50 nM) for 12 h (DENV E proteins were not detectable at 2, 4, and eight hpi) (Fig. S2). Our findings revealed that C156-P1 significantly enhanced the colocalization between DENV E protein and Rab7A (Fig. 6E). These results indicate that C156-P1 inhibits endosomal acidification and restricts DENV particles within the endosomes.
Next, we sought to determine the mechanism underlying the endosomal retention of DENV induced by C156-P1. Previous studies have demonstrated that DP modulates the expression of V-ATPase, thereby inhibiting the acidification of endosomes (16,28). Since C156-P1 induced endosomal acidification and restricted viral particles inside the endosomes, we proceeded to investigate whether V-ATPase played a role in this process. The V-ATPase complex comprises 32 subunits, and the ATP6V0A2 subunit plays a crucial role in endosomal proton transport, which is important for the function of V-ATPase (44,45). Thus, we investigated whether ATP6V0A2 is involved in DENV infection and its correlation with C156-P1 treatment. We observed a decrease in ATP6V0A2 mRNA expression following C156-P1 treatment of A549 cells, similar to the effect observed with PHY34 treatment, an inhibitor of ATP6V0A2 (Fig. 6F) (46). PHY34 treatment also inhibited endosomal acidification and DENV-2 infection in a dose-dependent manner, similar to C156-P1 treatment (Fig. 6G andH). We further examined whether C156-P1 and PHY34 affected V-ATPase activity. The results showed that C156-P1 and PHY34, similar (I) A549 cells were treated with DMSO, C156-P1 (100 nM or 1000 nM), PHY34 (100 nM), or Baf-A1 (100 nM) for 24 h, followed by the ATPase activity assay using 2 µg of total cell protein. (J) A549 cells were transfected with ATP6V0A2-specific or scramble siRNAs (NC) for 24 h, followed by infection with DENV-2 at an MOI of 2. The levels of ATP6V0A2 mRNA, intracellular viral RNA, DENV-2 NS3 protein, and extracellular virus titer were assessed at 48 hpi. (K) A549 cells were transfected with ATP6V0A2-specific or scramble siRNAs (NC) for 48 h, followed by the ATPase activity assay using 2 µg of total cell protein. In panels (A-G) and (I-K), data are presented as means ± SEM of three independent experiments (Student's t-test; ns, no significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001). to Baf-A1, significantly reduced V-ATPase activity. Additionally, the reduction of V-ATPase activity by C156-P1 was dose-dependent (Fig. 6I). Furthermore, siRNA knockdown of ATP6V0A2 markedly decreased DENV-2 RNA levels, NS3 protein expression, and infectious virus titers (Fig. 6J). Knockdown of the ATP6V0A2 subunit also resulted in a reduction in intracellular V-ATPase activity (Fig. 6K). Taken together, these results suggest that the ATP6V0A2 subunit was a target of C156-P1 and is primarily responsible for preventing endosomal acidification, thereby inhibiting flavivirus infection.
## C156-P1 inhibited DENV-2 infection in mice
To assess the therapeutic efficacy of C156-P1 on DENV infection, we infected AG129 mice, which are deficient in both IFN-α/β and IFN-γ receptors, with the DENV-2 strain 16681 and treated with C156-P1 (Fig. 7A). A dose of 0.2 mg/kg C156-P1 was selected for the mouse experiments based on dosage data from our previous animal studies with PTA in mice (28). We chose this dose because C156-P1 exhibited a lower EC 50 value than PTA, and both C156-P1 and PTA are DP-derived derivatives. In the group not receiving C156-P1 treatment, a mortality rate of 100% was observed on day 6 post-virus challenge.
In contrast, treatment with C156-P1 (0.2 mg/kg) protected the mice from this lethal dose challenge, resulting in a survival rate of 60% until the experiment was terminated on day 15 (Fig. 7B). In comparison to the DENV-2 group, treatment with C156-P1 mitigated the decline in body weight and enhanced behavioral signs in mice infected with DENV-2; the surviving mice began to recover from clinical symptoms from day 8 post-inoculation, and their body weight gradually recovered (Fig. 7C andD). None of the control group mice inoculated with DMEM exhibited any clinical symptoms or experienced weight loss. To further confirm the virus infection after injection, we determined the viral RNA loads in the blood samples (including blood cells) and found that C156-P1 markedly decreased DENV-2 in the blood of mice (Fig. 7E). These results indicate that C156-P1 exhibits therapeutic potential against DENV infection in vivo.
## DISCUSSION
There is currently no specific antiviral drug available for DENV infection, highlighting the urgent need to develop effective therapeutic agents for dengue. In this study, we discovered that C156-P1, a nitrogen-containing DP derivative, exhibited potent inhibitory effects against DENV-2 infection in vitro and in vivo. Furthermore, C156-P1 demonstrated broad-spectrum antiviral activities by inhibiting other serotypes of DENVs (DENV-1, DENV-3, and DENV-4) as well as other Flaviviridae viruses like ZIKV, JEV, YFV, and HCV. In addition, C156-P1 also inhibited the infections of enveloped viruses from other families, including HSV-1 and VSV. Mechanistic studies revealed that C156-P1 targeted the ATP6V0A2 subunit, affecting V-ATPase activity and suppressing endosomal acidification. This inhibitory mechanism restricted viral particles within the endosome, effectively blocking DENV infection both prior to and during the membrane fusion process. Collectively, these findings highlight the potentials of C156-P1 as a broad-spec trum antiviral agent, particularly against flaviviruses. The natural compound DP is an ANL isolated from the J. gendarussa plant. DP functions as a V-ATPase inhibitor and is regarded as a promising broad-spectrum antiviral candidate. The compound was found to be active against SARS-CoV-2 (18), influenza virus (20), tick-borne encephalitis virus, WNV, ZIKV, RVFV, rabies virus, HSV-1 (21), Chikungunya virus (47), and VSV (48) at low micromolar concentrations. Although DP is recognized as a broad-spectrum antiviral agent, its limited aqueous solubility and insufficient antiviral activity considerably hinder its clinical applicability. To improve the antiviral activity and bioavailability of DP, various DP nanoparticles and structurally modified derivatives have been reported. DP nanoparticles exhibited reduced cytotoxic ity and enhanced antiviral activity, as confirmed in influenza virus and feline infectious peritonitis virus (FIPV) (49,50). Glycosylated DP derivatives, such as Justiprocumin B and PTA, have been shown to enhance antiviral activity, inhibiting HIV and ZIKV at nanomolar concentrations, respectively (28,51). A recent study has found that various nitrogen-containing DP derivatives block Ebola virus entry at low nanomolar concentra tions (26). From a series of 8 N-ANLs out of 12 tested ANLs, we identified C156-P1, which effectively inhibited DENV-2 infection at low nanomolar concentrations (EC 50 < 5 nM), demonstrating antiviral efficacies in vitro across multiple monkey and human cell lines. The antiviral activity of C156-P1 against DENV-2 in A549 cells was nearly 300-fold greater compared to the parental compound DP. Notably, C156-P1 is structurally distinct from the other DP derivatives, as it is the only one containing a cyanide group. This nitrogen-and triple bond-containing cyanide group is positioned in close proximity to the DP skeleton. Although C70-P1 also has a cyanide group (Fig. S1), its cyanide group is located farther from the DP structure core. Compound 176-P1 has a structure very similar to C156-P1, but it contains only a carbon-carbon triple bond, rather than a cyanide group. The remaining compounds differ more significantly in structure from C156-P1. Furthermore, C156-P1 demonstrated potent antiviral effects against DENV-1, DENV-3, and DENV-4 infections at low nanomolar concentrations (EC 50 < 10 nM), highlighting its strong anti-DENV activity.
Numerous anti-flavivirus compounds have been identified in recent decades; however, only a limited number have advanced to preclinical trials due to concerns regarding efficacy, specificity, toxicity, and stability (52). A recent study identified a potent DENV inhibitor, JNJ-1802, which blocks the NS3-NS4B interaction and generates effective anti-DENV infection effects in mice and rhesus macaques; a phase I first-inhuman clinical trial has been completed for JNJ-1802 (53). However, JNJ-1802 lacks inhibitory effects against other Flaviviridae viruses. In contrast, C156-P1 is a broad-spec trum antiviral agent that not only potently inhibited DENV infection but also suppressed the infection of other Flaviviridae viruses. C156-P1 was demonstrated to effectively inhibit ZIKV, JEV, YFV, and HCV in vitro at low nanomolar concentrations (EC 50 < 6 nM), significantly below the cytotoxic concentration. Chloroquine, an FDA-approved drug for treating malaria, can be used to prevent ZIKV infections by inhibiting endosomal acidification; however, the inhibitory concentrations necessary for ZIKV infection in cell culture are in the micromolar range (54,55). The antiparasitic agent niclosamide can also inhibit DENV-2 infection by suppressing endosomal acidification, with an EC 50 value of approximately 10 µM (56). In addition to inhibiting Flaviviridae viruses, C156-P1 could also inhibit viruses that enter host cells through either endocytosis or low pH-dependent fusion, as demonstrated for HSV-1 and VSV. In contrast, C156-P1 did not inhibit SeV, which enters cells not through the endocytosis pathway but by directly fusing with the host cell membrane. Similarly, C156-P1 did not inhibit AdV, which enters host cells by clathrin-mediated endocytosis without the requirement for fusion with the endosomal membrane. These findings further suggest that C156-P1 selectively inhibits viruses that enter host cells through an endocytosis pathway and undergo low pH-dependent fusion.
In the mechanistic study, we found that C156-P1 suppressed flavivirus infection in the early stages after viral entry into host cells, without affecting virus binding and internalization. Furthermore, C156-P1 reduced the activation of the IFN-I response and endosomal TLR3 activation induced by DENV infection. Therefore, these results indicated that the inhibitory effect of C156-P1 on DENV infection involves pathways other than IFN-I activation. An attenuated IFN-I response indicates that the IFN-I activator of DENV RNA was restrictedly exposed to the RNA sensor in the cytoplasm. Finally, we experi mentally demonstrated that C156-P1 restricted DENV within the endosome, showing colocalization with Rab7A. Given that the early stages of the DENV life cycle after cell entry were inhibited, we found that C156-P1 targeted the pre-fusion stage or during the fusion of viral envelopes with endosomal membranes.
DP and PTA have been shown to inhibit the acidification of endosomes and lysosomes in a dose-dependent manner (28). C156-P1 is a nitrogen-containing derivative of DP, also demonstrated its ability to inhibit endosomal and lysosomal acidification in our study. DENV-2 enters host cells through clathrin-mediated endocytosis, deliv ering viral particles to Rab7-positive late endosomes for release into the cytoplasm via membrane fusion (43). Using confocal microscopy, we demonstrated that C156-P1 restricted DENV-2 to the late endosome. V-ATPase transports protons into the late endosome to maintain its acidic environment. HTP-013, a nitrogen-containing DP derivative, has been identified as a V-ATPase inhibitor that prevents lysosomal acidification by directly targeting ATP6V0A2 (17). Recent studies have shown that the loss of ATP6V0A2 expression significantly enhances the efficacy of HTP-013 in blocking EBOV infection (26). PHY34, structurally similar to DP, targets ATP6V0A2 and exhibits potent anticancer activity (46). We observed that both C156-P1 and PHY34 inhibited V-ATPase activity and downregulated ATP6V0A2 mRNA expression, thereby suppressing DENV-2 infection. Moreover, C156-P1 inhibits V-ATPase activity in a dose-dependent manner. Therefore, the inhibition of V-ATPase activity through targeting the ATP6V0A2 subunit is the primary mechanism by which C156-P1 suppressed DENV-2 infection. Nevertheless, the mode of action of C156-P1 on ATP6V0A2 requires further investigation, such as through structural biology, chemical biology, and other complementary approaches.
Altogether, our results indicate that C156-P1 inhibits V-ATPase activity by ATP6V0A2, thereby suppressing endosomal acidification and restricting DENV infection. Flaviviridae viruses (DENV, ZIKV, JEV, YFV, and HCV), Herpesviridae virus (HSV-1), and Rhabdoviri dae virus (VSV) all enter cells via endocytosis and are transported to the acidic late endosome where they undergo membrane fusion to release the viral genome. Thus, C156-P1 exhibits its broad-spectrum antiviral activity through a common mechanism by inhibiting V-ATPase activity and thereby endosomal acidification, highlighting the potential of C156-P1 to inhibit other viruses that use the same entry strategy as Flaviviridae viruses.
Given the insensitivity of DENV-2 to wild-type mice, AG129 mice are commonly utilized to establish DENV-2 infection models (57). C156-P1 prevented DENV-2-induced mortality in AG129 mice when administered concurrently with the virus, suggesting its potential to inhibit DENV-2 infection in vivo. Chloroquine and hydroxychloroquine, the FDA-approved medications for the treatment of malaria, also inhibit ZIKV infection in vivo and prevent endosomal acidification. However, in vivo experiments indicate that chloroquine and hydroxychloroquine require inhibitory concentrations of 100 and 40 mg/kg in mice, respectively (55,58). While further in vivo testing of C156-P1 is necessary, including dose optimization and drug formulation, our results suggested that C156-P1 effectively inhibited DENV-2 infection at a dosage of 0.2 mg/kg in mice.
In summary, we discovered that C156-P1, a novel nitrogen-containing DP derivative, exhibited broad-spectrum antiviral activities against five Flaviviridae viruses (DENV, ZIKV, JEV, YFV, and HCV) as well as two enveloped viruses (HSV-1 and VSV). C156-P1 targe ted the ATP6V0A2 subunit to inhibit V-ATPase activity, thereby preventing endosomal acidification and restricting viral activity within the late endosome. Targeting a host pathway essential for a number of viruses allows C156-P1 with significant potential for future development as an agent against diverse viral infections.
## Plasmid, antibodies, and reagents
Full-length human Rab7A cDNA was amplified by PCR. Rab7A cDNA was cloned into the pCDH-mNeonGreen expression plasmid to generate pCDH-mNeonGreen-Rab7A.
The primary antibodies used in this study were as follows: anti-DENV NS3 rabbit antibody (GeneTex, GTX124252), anti-DENV E rabbit antibody (GeneTex, GTX127277), anti-ZIKV NS3 rabbit antibody (GeneTex, GTX133309), anti-JEV E rabbit antibody (GeneTex, GTX125867), anti-YFV E rabbit antibody (GeneTex, GTX134024), anti-HCV Core C7-50 mouse antibody (Santa Cruz Biotechnology, sc-57800), anti-GFP tag mouse antibody (Proteintech, 66002-1-Ig), anti-horseradish peroxidase (HRP)-conjugated GAPDH antibody (Proteintech, HRP-60004), anti-HRP-conjugated Alpha Tubulin antibody (Proteintech, HRP-66031), anti-TBK1/NAK rabbit antibody (Abcam, ab40676), anti-phos pho-TBK1/NAK rabbit antibody (Cell Signaling Technology, Ser172, D52C2), anti-IRF3 rabbit antibody (Cell Signaling Technology, D6I4C), anti-phospho-IRF3 rabbit antibody (Cell Signaling Technology, Ser396, 4D4G), and anti-ATP6V0A2 rabbit antibody (Abcam, ab96803). The secondary antibodies used for western blotting analysis were goat anti-mouse IgG (H + L)-HRP (Ray Antibody Biotech, RM3001) and goat anti-rabbit IgG (H + L)-HRP (Ray Antibody Biotech, RM3002). The secondary antibodies used for immuno fluorescence were goat anti-rabbit IgG (H + L) highly cross-adsorbed secondary (Alexa Fluor Plus 488) (Invitrogen, A32731) and goat anti-rabbit IgG (H + L) highly cross-adsor bed secondary (Alexa Fluor Plus 555) (Invitrogen, A32732).
An ANL compound library was provided by Gihon Biotech Limited (Hong Kong, China). The chemical reagents bafilomycin A1 (Baf-A1, Selleck Chemicals, S1413), PHY34 (Targetmol, T37376), and LysoSensor Green DND-189 (Invitrogen, L7535) were pur chased.
## Cell viability assay
The cells were seeded in 96-well plates at a density of 8 × 10 3 cells/well and cultured at 37°C for 24 h. Serial dilutions of each compound, dissolved in 100 µL in DMSO/DMEM (2% FBS), were added to the cells. DMSO was used as a negative control. After incubation for 48 h, 10 µL of Cell Counting Kit-8 (CCK-8, Dojindo, CK04) was added to each well with gentle shaking, and the plates were incubated at 37°C for 30 min. The absorbance at 450 nm (OD 450 ) was determined using a microplate reader (BioTek, ELX800). The 50% cytotoxic concentration (CC 50 ) of a compound was calculated using GraphPad Prism 8.0 software (California, USA).
## Immunofluorescence and focus-forming units (FFUs) assay
The cells infected with DENV and HCV were analyzed by immunostaining. Briefly, virus-infected cells were fixed using methanol (-20°C) and incubated with primary antibody (anti-DENV NS3 or anti-HCV Core) at 4°C overnight, followed by incubation with secondary antibodies (conjugated with Alexa Fluor Plus 488) and Hoechst 33258 (Invitrogen, H1398) for 1 h at room temperature. Images were acquired using a fluorescence microscope (Leica DMI8, Germany), and the percentage of virus-positive cells was calculated by ImageJ software (National Institutes of Health, USA).
DENV infectivity titers were determined using a focus-forming assay as previously described (27). Briefly, Vero cells (8 × 10 3 cells/well) were seeded into a 96-well plate and incubated for 24 h. The cells were inoculated with 10-fold serial dilutions of DENV in DMEM for 2 h. Subsequently, the virus inoculum was removed, and the cells were cultured with DMEM (2% FBS) for 48 h. The immunofluorescence assay was performed as described above. The number of FFUs was enumerated manually under a fluorescence microscope (Leica DMI8, Germany).
## Plaque assay
JEV infectivity titers were determined by plaque assay using BHK-21 cells, while ZIKV, YFV, HSV-1-eGFP, and VSV-eGFP infectivity titers were assessed using Vero cells. Briefly, Vero or BHK-21 cells were seeded in a 12-well plate for 24 h and inoculated with 10-fold serial dilutions of viruses in DMEM. After 2 h, the virus inoculum was removed, and fresh DMEM with 2% FBS and 1% methyl cellulose (Sigma-Aldrich, M0512) was added. After 4-5 days, the cells were fixed using 4% paraformaldehyde and stained with 1% crystal violet. The visible plaques were counted after washing the plates with tap water, and the virus infectivity titers were calculated and presented as plaque-forming units per milliliter (PFU/mL) of culture supernatant.
## Reverse transcription-quantitative PCR (RT-qPCR)
Total RNA was extracted from cells by use of MagZol reagent (Magen, R4801-02) according to the manufacturer's instructions. For RT-qPCR, cDNA was synthesized using HiScript III RT SuperMix for qPCR Kit (Vazyme, R323-01) and subsequently amplified by qPCR using Magic SYBR Mixture reagent (CWBIO, CW3008M) with specific qPCR primers (Table 1). Relative mRNA expression of the interest genes was assessed using the 2 -ΔΔCt method and normalized to the GAPDH gene.
## Western blotting
The cells were harvested and lysed on ice for 30 min using lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1 mM DTT, and 0.5% TX-100, pH = 7.5), supplemented with a protease inhibitor cocktail (Sigma-Aldrich, P8849). The lysates were centrifuged at 4°C for 10 min at 12,000 rpm, and the supernatants were mixed with 1 × sodium dodecyl sulfate (SDS) loading buffer (Invitrogen, NP0008) and boiled for 10 min. Total protein was separated by 10% or 12% SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and then transferred onto 0.2 µm polyvinylidene fluoride membranes (Bio-Rad, 1620177). The membranes were blocked with 5% skimmed milk for 2 h at room temperature and then incubated with the appropriate primary antibodies at 4°C overnight. After three washes with Tris-buffered saline supplemented with 0.1% Tween-20 (TBST) (10 min per wash), the membranes were further incubated with HRP-conjugated anti-mouse or anti-rabbit secondary antibodies for 1 h at room temperature. After three additional washes with TBST (10 min per wash), the protein bands were detected using enhanced chemiluminescence reagents (Proteintech, PK10001) and visualized with Fuji medical X-ray film (Fujifilm, RX-N-C). Photoshop software. Pearson's correlation coefficient was calculated using the JACoP plugin for ImageJ software (National Institutes of Health) (71).
## ATPase activity assay
ATPase activity assay was performed as described previously with minor modifications (72). Total cellular protein was extracted using a lysis buffer (Beyotime, P0013J) supplemented with a protease inhibitor cocktail (Selleck Chemicals, B14001). Proteins were quantified using a Bicinchoninic Acid Protein Assay Kit (Beyotime, P0010) according to the manufacturer's instructions. ATPase activity was determined using 2 µg of cell lysate with the ATPase/GTPase Activity Assay Kit (Sigma-Aldrich, MAK113) according to the manufacturer's instructions. Following a 30-min incubation and a 30-min termination at room temperature, the samples were measured with a microplate reader (BioTek, ELX800) at 630 nm (OD 630 ).
## RNA interference
Small interference RNAs (siRNAs) targeting human ATP6V0A2 gene and scrambled siRNA controls were designed and synthesized by GenePharma (Suzhou, China) (Table 2). The siRNAs were transfected into A549 cells for 6 h using GP-transfect-Mate reagent (GenePharma, G04008) according to the manufacturer's instructions. At 48 h post-trans fection, the A549 cells were harvested for further analysis.
## Animal experiments
All animal experiments were approved by the Institutional Animal Care and Use Committee of Sun Yat-sen University Guangzhou, China (approval number: SYSU-IACUC-2024-B1498). AG129 mice (129/Sv mice deficient in interferon-α/β and interferonγ receptors) were from Prof. Jianping Zuo (Shanghai Institute of Materia Medica, Chinese Academy of Sciences, China) and maintained under specific-pathogen-free conditions at Sun Yat-sen University. Fifteen AG129 mice (6-8 weeks) were randomly divided into three groups. In the DENV-2 challenge group, five mice were inoculated with DENV-2 strain 16681 (1 × 10 6 PFU) without any additional treatment. In the C156-P1 treatment group, five mice were inoculated with DENV-2 strain 16681 (1 × 10 6 PFU) containing C156-P1 (0.2 mg/kg). C156-P1 was administered for three consecutive days following DENV-2 inoculation. In the mock infection group, five mice were injected with DMEM.
Each injection was performed intraperitoneally with 200 µL per dose for each mouse. Mortality, symptoms, and body weight of all mice were monitored for 15 days. Clinical severity was scored as previously described (73). Briefly, the scoring system includes five grades, namely grade 0 (healthy), grade 1 (ruffled hair or hunchbacked), grade 2 (reduced mobility), grade 3 (limb weakness), grade 4 (limb paralysis), and grade 5 (moribund or death). Blood samples (including blood cells) were collected from mice via tail cutting to determine virus copies in blood at 3, 5, and 7 days post-inoculation (dpi).
## Statistical analysis
All experimental data were presented as mean and standard error of mean (mean ± SEM) from at least three independent experiments. Statistical analysis was performed using Student's t-test with GraphPad Prism 8 software (California, USA), and P < 0.05 was considered significant.
## References
1. Ceconi, Ariën, Delputte (2024) "Diagnosing arthropod-borne flaviviruses: non-structural protein 1 (NS1) as a biomarker" *Trends Microbiol*
2. Diamond, Pierson (2015) "Molecular insight into dengue virus pathogenesis and its implications for disease control" *Cell*
3. Bhatt, Gething, Brady et al. (2013) "The global distribution and burden of dengue" *Nature*
4. Stanaway, Shepard, Undurraga et al. (2013) "The global burden of dengue: an analysis from the Global Burden of Disease Study" *Lancet Infect Dis*
5. Allan, Budge, Sauskojus (2023) "What sounds like Aedes, acts like Aedes, but is not Aedes? Lessons from dengue virus control for the management of invasive Anopheles" *Lancet Glob Health*
6. (2025) *Full-Length Text Antimicrobial Agents and Chemotherapy*
7. Pierson, Diamond (2020) "The continued threat of emerging flaviviruses" *Nat Microbiol*
8. Villordo, Carballeda, Filomatori et al. (2016) "RNA structure duplications and flavivirus host adaptation" *Trends Microbiol*
9. Screaton, Mongkolsapaya, Yacoub et al. (2015) "New insights into the immunopathology and control of dengue virus infection" *Nat Rev Immunol*
10. Wilder-Smith, Ooi, Horstick et al. (2019) *Dengue. Lancet*
11. Cruz-Oliveira, Freire, Conceição et al. (2015) "Receptors and routes of dengue virus entry into the host cells" *FEMS Microbiol Rev*
12. Welsch, Miller, Romero-Brey et al. (2009) "Composition and three-dimensional architecture of the dengue virus replication and assembly sites" *Cell Host Microbe*
13. Guzman, Harris (2015) *Dengue. Lancet*
14. Newman, Cragg (1981) "Natural products as sources of new drugs over the nearly four decades from 01" *J Nat Prod*
15. Zhang, Rumschlag-Booms, Guan et al. (2017) "Potent inhibitor of drug-resistant HIV-1 strains identified from the medicinal plant Justicia gendarussa" *J Nat Prod*
16. Li, Lu, Ma et al. (2022) "Synthesis and anti-tumor activity of nitrogencontaining derivatives of the natural product diphyllin" *Eur J Med Chem*
17. Shen, Zou, Chen et al. (2011) "Effects of diphyllin as a novel V-ATPase inhibitor on gastric adenocarci noma" *Eur J Pharmacol*
18. Wang, Pham, Lipps et al. (2019) "Previously uncharacterized vacuolar-type ATPase binding site discovered from structurally similar compounds with distinct mecha nisms of action" *ACS Chem Biol*
19. Stefanik, Strakova, Haviernik et al. (2021) "Antiviral activity of vacuolar ATPase blocker diphyllin against SARS-CoV-2"
20. Xu, Wang, Ku et al. (2019) "Anti-HIV lignans from Justicia procumbens" *Chin J Nat Med*
21. Chen, Cheng, Liu et al. (2013) "Inhibitory and combinatorial effect of diphyllin, a v-ATPase blocker, on influenza viruses" *Antiviral Res*
22. Štefánik, Bhosale, Haviernik et al. (2022) "Diphyllin shows a broad-spectrum antiviral activity against multiple medically important enveloped RNA and DNA viruses" *Viruses*
23. Xu, Shi, Zhu et al. (2024) "Novel SERCA2 inhibitor diphyllin displays anti-tumor effect in non-small cell lung cancer by promoting endoplasmic reticulum stress and mitochondrial dysfunction" *Cancer Lett*
24. Rao, Fang, Tzeng (2006) "Anti-inflammatory activities of constituents isolated from Phyllanthus polyphyllus" *J Ethnopharmacol*
25. Rao, Geethangili, Fang et al. (2007) "Antioxidant and cytotoxic activities of naturally occurring phenolic and related compounds: a comparative study" *Food Chem Toxicol*
26. Thamburaj, Ramaraj, Sethupathy et al. (2020) "Antibacterial and antibiofilm activities of diphyllin against fish pathogens" *Microb Pathog*
27. Sanford, Keiser, Fujii et al. (2024) "Evaluation of potency and metabolic stability of diphyllin-derived vacuolar-ATPase inhibitors" *Eur J Med Chem*
28. Hu, Li, Wu et al. (2023) "Identification of an arylnaphthalene lignan derivative as an inhibitor against dengue virus serotypes 1 to 4 (DENV-1 to -4) using a newly developed DENV-3 infectious clone and replicon" *Microbiol Spectr*
29. Martinez-Lopez, Persaud, Chavez et al. (2019) "Glycosylated diphyllin as a broad-spectrum antiviral agent against Zika virus" *EBioMedicine*
30. Vojnikova, Sukupova, Stefanik et al. (2024) "Nanoformu lation of the broad-spectrum hydrophobic antiviral vacuolar ATPase inhibitor diphyllin in human recombinant H-ferritin" *Int J Nanomedicine*
31. Nicola (2016) "Herpesvirus entry into host cells mediated by endosomal low pH" *Traffic*
32. Kim, Jenni, Stanifer et al. (2017) "Mechanism of membrane fusion induced by vesicular stomatitis virus G protein" *Proc Natl Acad Sci*
33. Wiethoff, Wodrich, Gerace et al. (2005) "Adenovirus protein VI mediates membrane disruption following capsid disassembly" *J Virol*
34. Fan, Sefton (1978) "The entry into host cells of Sindbis virus, vesicular stomatitis virus and Sendai virus" *Cell*
35. Ngono, Shresta (2018) "Immune response to dengue and Zika" *Annu Rev Immunol*
36. Dalrymple, Cimica, Mackow (2015) "Dengue virus NS proteins inhibit RIG-I/MAVS signaling by blocking TBK1/IRF3 phosphorylation: dengue virus serotype 1 NS4A is a unique interferon-regulating virulence determinant" *mBio*
37. Tominari, Akita, Matsumoto et al. (2022) "Endosomal TLR3 signaling in stromal osteoblasts induces prostaglandin E 2 -mediated inflammatory periodontal bone resorption" *J Biol Chem*
38. Alexopoulou, Holt, Medzhitov et al. (2001) "Recognition of double-stranded RNA and activation of NF-κB by Toll-like receptor 3" *Nature*
39. Garcia-Cattaneo, Gobert, Müller et al. (2012) "Cleavage of Toll-like receptor 3 by cathepsins B and H is essential for signaling" *Proc Natl Acad Sci*
40. Modis, Ogata, Clements et al. (2004) "Structure of the dengue virus envelope protein after membrane fusion" *Nature*
41. Sánchez-San Martín, Liu, Kielian (2009) "Dealing with low pH: entry and exit of alphaviruses and flaviviruses" *Trends Microbiol*
42. Bagh, Peng, Chandra et al. (2017) "Misrouting of v-ATPase subunit V0a1 dysregulates lysosomal acidification in a neurodegenerative lysosomal storage disease model" *Nat Commun*
43. Li, Lang, Ji et al. (2019) "A scorpion venom peptide Ev37 restricts viral late entry by alkalizing acidic organelles" *J Biol Chem*
44. Van Der Schaar, Rust, Chen et al. (2008) "Dissecting the cell entry pathway of dengue Full-Length Text Antimicrobial Agents and Chemotherapy November"
45. "virus by single-particle tracking in living cells" *PLoS Pathog*
46. Wang, Wu, Robinson et al. (2020) "Structures of a complete human V-ATPase reveal mechanisms of its assembly" *Mol Cell*
47. Esmail, Kartner, Yao et al. (2018) "Molecular mechanisms of cutis laxa-and distal renal tubular acidosiscausing mutations in V-ATPase a subunits, ATP6V0A2 and ATP6V0A4" *J Biol Chem*
48. Salvi, Young, Huntsman et al. (2022) "PHY34 inhibits autophagy through V-ATPase V0A2 subunit inhibition and CAS/CSE1L nuclear cargo trafficking in high grade serous ovarian cancer" *Cell Death Dis*
49. Castro, Álvarez (2024) "New highly selective antivirals for chikungunya virus identified from the screening of a drug-like compound library" *Curr Microbiol*
50. Asano, Chiba, Tada et al. (1996) "Antiviral activity of lignans and their glycosides from Justicia procumbens" *Phytochemistry*
51. Hu, Chen, Fang et al. (2018) "Antiviral efficacy of nanoparticulate vacuolar ATPase inhibitors against influenza virus infection" *Int J Nanomedicine*
52. Hu, Chang, Fang et al. (2017) "Nanoparticulate vacuolar ATPase blocker exhibits potent host-targeted antiviral activity against feline coronavirus" *Sci Rep*
53. Zhang, Rumschlag-Booms, Guan et al. (2017) "Anti-HIV diphyllin glycosides from Justicia gendarussa" *Phytochemistry*
54. Troost, Smit (2020) "Recent advances in antiviral drug development towards dengue virus" *Curr Opin Virol*
55. Goethals, Kaptein, Kesteleyn et al. (2023) "Blocking NS3-NS4B interaction inhibits dengue virus in nonhuman primates" *Nature*
56. Delvecchio, Higa, Pezzuto et al. (2016) "Chloroquine, an endocytosis blocking agent, inhibits Zika virus infection in different cell models" *Viruses*
57. Li, Zhu, Ji et al. (2017) "Chloroquine, a FDA-approved drug, prevents Zika virus infection and its associated congenital microcephaly in mice" *EBioMedicine*
58. Kao, Huangfu, Tsai et al. (2018) "The antiparasitic drug niclosamide inhibits dengue virus infection by interfering with endosomal acidification independent of mTOR" *PLoS Negl Trop Dis*
59. Kaptein, Goethals, Kiemel et al. (2021) "A pan-serotype dengue virus inhibitor targeting the NS3-NS4B interaction" *Nature*
60. Cao, Parnell, Diamond et al. (2017) "Inhibition of autophagy limits vertical transmission of Zika virus in pregnant mice" *J Exp Med*
61. Hao, Li, Zhang et al. (2023) "NLRC5 restricts dengue virus infection by promoting the autophagic degradation of viral NS3 through E3 ligase CUL2 (cullin 2)" *Autophagy*
62. Pu, Wu, Tsai et al. (2014) "A novel approach to propagate flavivirus infectious cDNA clones in bacteria by introducing tandem repeat sequences upstream of virus genome" *J Gen Virol*
63. Deng, Zhao, Li et al. (2016) "Isolation, identification and genomic characterization of the Asian lineage Zika virus imported to China" *Sci China Life Sci*
64. Li, Dong, Wang et al. (2018) "Development of a chimeric Zika vaccine using a licensed liveattenuated flavivirus vaccine as backbone" *Nat Commun*
65. Qian, Wu, Tang et al. (2022) "Rifapentine is an entry and replication inhibitor against yellow fever virus both in vitro and in vivo" *Emerg Microbes Infect*
66. Wakita, Pietschmann, Kato et al. (2005) "Production of infectious hepatitis C virus in tissue culture from a cloned viral genome" *Nat Med*
67. Zhong, Gastaminza, Cheng et al. (2005) "Robust hepatitis C virus infection in vitro" *Proc Natl Acad Sci*
68. Xie, Tian, Yang et al. (2022) "OTUD7B deubiquitinates SQSTM1/p62 and promotes IRF3 degradation to regulate antiviral immunity" *Autophagy*
69. Chen, Meng, Qin et al. (2016) "TRIM14 inhibits cGAS degradation mediated by selective autophagy receptor p62 to promote innate immune responses" *Mol Cell*
70. Sun, Feng, Zhang et al. (2012) "Circumventing antivector immunity by using adenovirus-infected blood cells for repeated application of adenovirus-vectored vaccines: proof of concept in rhesus macaques" *J Virol*
71. Dejarnac, Hafirassou, Chazal et al. (2018) "TIM-1 ubiquitination mediates dengue virus entry" *Cell Rep*
72. Mauvezin, Nagy, Juhász et al. (2015) "Autophagosomelysosome fusion is independent of V-ATPase-mediated acidification" *Nat Commun*
73. Kim, Kim, Sung et al. (2016) "Interferoninducible protein SCOTIN interferes with HCV replication through the autolysosomal degradation of NS5A" *Nat Commun*
74. Carapito, Konantz, Paillard et al. (2017) "Mutations in signal recognition particle SRP54 cause syndromic neutropenia with Shwachman-Diamond-like features" *J Clin Invest*
75. Chen, Yang, Cao et al. (2023) "Liang-Ge-San inhibits dengue virus serotype 2 infection by reducing caveolin1-induced cytoplasmic heat shock protein 70 translocation into the plasma membrane" *Phytomedicine* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12751207&blobtype=pdf | # Metagenomic identification, isolation, and complete genome characterization of two novel picornaviruses in wild duck from Northeastern Siberia
Nikita Dubovitskiy, Olga Kurskaya, Mariya Solomatina, Arina Loginova, Anastasiya Derko, Anna Khozyainova, Evgeny Denisov, Evgeniy Shemyakin, Alexander Shestopalov, Kirill Sharshov
## Abstract
Duck hepatitis A virus (DHAV, Avihepatovirus ahepati species) is a well-known pathogen of domestic ducks, causing fatal disease in ducklings and posing a significant burden to duck poultry farms. Avihepatovirus ahepati is the sole species within the genus Avihepatovirus and is classified into three distinct genotypes: DHAV-1, DHAV-2, and DHAV-3. In our study, we detected and isolated two strains belonging to the Picornaviridae family from Anas crecca in Northeastern Asia. One of the isolates is related to the DHAV-1 group; however, its polyprotein gene sequence shares only 77.83% nucleotide identity and 89.68% amino acid identity with the most closely related DHAV-1 sequence available, suggesting it represents a highly divergent lineage. The second isolated duck picornavirus shows 60.16% nucleotide identity to the polyprotein gene sequence of the previously described duck picornavirus strain detected during an outbreak in domestic ducks in China. The identification of these two novel picornaviruses in wild ducks, along with their efficient replication in duck embryos and primary cell cultures, emphasizes the need for comprehensive studies of their prevalence in wild ducks and their biological traits to assess potential risks for wildlife and poultry farming. The obtained complete genome sequences and viral isolates enhance our understanding of the diversity, evolution, and ecology of avian picornaviruses.
## Introduction
Viruses can impose a substantial burden on poultry farming, causing diseases with high morbidity and mortality among birds, and resulting in significant economic losses [1]. A wide variety of avian viruses, including those from the Astroviridae, Orthomyxoviridae, Paramyxoviridae, Parvoviridae, and Sedoreoviridae families, circulate among wild and domestic bird populations. While some avian viruses have been extensively characterized due to their agricultural and public health importance, many other remain poorly understood, and their diversity, host range, and ecological roles are still being uncovered.
Picornaviridae is a family of small non-enveloped viruses with positive-sense single-stranded RNA genome. Following International Committee on Taxonomy of Viruses (ICTV), the Picornaviridae family is divided into 159 species belonging to 68 genera based on genomic data. The genome length of Picornaviridae species varies from 6.7 to 10.1 kb [2]. The open reading frame (ORF), flanked by 5' and 3' untranslated regions (UTRs), encodes a polyprotein that is subsequently processed in structural and non-structural proteins. VPg protein covalently linked to 5'-terminus of genomic RNA.
Viruses of the Picornaviridae family are capable of infecting vertebrates, causing diseases in a wide range of animal species as well as in humans [2]. A comprehensive study from China performed an investigation of wild bird cloaca viromes of 3182 birds and found 32 genomes of picornaviruses [3]. At present, four designated picornavirus species have been identified in association with domestic ducks: duck hepatitis A virus (DHAV/Avihepatovirus ahepati), duck egg-reducing syndrome virus (DERSV), duck megrivirus (Megrivirus aturhepa), duck aalivirus (Aalivirus apekidu), duck anativirus (Anativirus aductai) [4][5][6][7]. The most well-known picornavirus-associated disease in domestic ducks is caused by duck hepatitis A virus. Duck hepatitis A virus is the pathogen that causes Duck (viral) hepatitis type I, while type II and type III are caused by Astroviridae family viruses.
The genome length of duck hepatitis A virus is approximately 7.8 kb. The 5'-UTR of DHAV genome contains a type IV internal ribosome entry site (IRES) [8]. The single ORF's polyprotein product after translation cleaved by viral 2 A protease into 2 subunits: P1 and P2-P3. Next, protease 3 C cleaves: P1 into structural proteins VP0, VP1, and VP3; P2 into non-structural 2A1, 2A2, 2A3, 2B, and 2 C proteins; P3 into non-structural 3 A, 3B, 3 C, and 3D proteins [9]. Genetically, the duck hepatitis A virus is divided into 3 different genotypes DHAV-1, DHAV-2, and DHAV-3. Pathogenetically, DHAV-1 primarily infects 1-2 weeks ducklings causing liver enlargement and hemorrhages. However, another detected pathotype of DHAV-1, designated as pancreatitis-type, induces
## Graphical abstract
pancreatitis without pathological changes in liver [10]. Additionally, it has been shown that the DHAV-1 isolate FC16115 caused egg drop and reduced feed consumption in experimentally infected Cherry Valley laying ducks [11]. DHAV-3 was also detected in the spleen and kidney of ducklings, where it exhibited a higher replication level [9]. Notably, a recent study demonstrated the potential vertical transmission of DHAV-1 from infected ducks to their embryos [12].
DHAV-3 has been reported in Korea, Vietnam, and China, whereas DHAV-2 was first identified in Taiwan and subsequently detected in India [9,13]. DHAV-1 is the most globally widespread genotype detected in Eurasia, Africa, and North America [9].
Recently, with the adoption of a metagenomic approach, several studies have reported the detection of picornaviruses in domestic ducks exhibiting disease symptoms. Picornavirus with a relatively low identity to DHAV was isolated from ducks with short beak and dwarfism syndrome (SBDS) in 2015 in China proposing a new species of Avihepatovirus genus [14]. Moreover, another picornavirus was isolated in China in 2017 from domestic ducklings with paralysis and neck twist [15]. The virus was related to the Avihepatovirus genus; hence, the amino acid identity was relatively low (polyprotein 41.45%). An Avihepatovirus-like group of viruses was detected in wild ducks of the Anas species in Australia, forming a phylogenetically distinct sister clade to known Avihepatovirus variants [16].
The territory of Russia is crossed by plural intra-and intercontinental migratory routes of wild waterfowl and serves as breeding areas for birds in the summer season, providing the possibility to transmit and replicate a variety of avian viruses among different bird species. In the Asian territories of Russia, the circulation of avian influenza viruses, avian coronaviruses, avian paramyxoviruses, and avian astroviruses has been documented [17][18][19]. To date, no genomic sequences of DHAVs circulating in Russia are available in Genbank; however, serological data confirm the presence of DHAV-specific antibodies in duck farms [20].
Therefore, the aim of our study was to perform virological and genetic analysis of newly identified and isolated picornaviruses, including the duck hepatitis A virus, initially detected through metagenomic sequencing of the common teal (Anas crecca) feces virome. This study reports the first identification of these viruses in the northern habitats of wild birds (Yakutia region, Russia) and demonstrates their ability to cause mortality in duck embryos. These viruses also exhibit substantial genetic divergence from picornaviruses previously detected in domestic ducks, which are of veterinary significance.
## Materials and methods
## Sample collection and virus-like particles (VLP) enrichment
Feces of 3 common teals (Anas crecca) were collected in 2 mL individual tubes in the Yakutia region of Northeast Siberia (Russia) and stored in liquid nitrogen before delivery to the laboratory. Detailed information on the collected samples is provided in Supplementary Material, Table S1. Further preparation was performed following the NetoVir protocol with modifications [21]. A 20% suspension was prepared from individual feces samples. The suspension was centrifuged for 3 min at 17,000 g and 200 µL of supernatant was transferred to a 0.8 μm PES centrifugal filter (Sartorius, Germany). Sample was filtered for 1 min at 17,000 g. Filtrates of 3 samples were pooled in equal volumes and 130 µL of the filtrate was added to a new 1.5 mL tube with 7 µL of 20X nuclease buffer added. Then, 2 µL of bensonase and 1 µL of micrococcal nuclease was added and the sample was incubated for 2 h at 37 °C following termination of treatment with 7 µL of 0.2 M EDTA.
## Detection of viruses via metagenomic approach
Nucleic acid extraction from a VLP-enriched sample was performed with a column-based RNA extraction kit (Biolabmix, Russia) following manufacture protocol. For whole transcriptome amplification, WTA2 kit (Sigma Aldrich, Germany) was used. 2.82 µL of extracted nucleic acids was mixed by pipetting with 0.5 µL of Library Synthesis Solution and incubated at 95 °C for 2 min following cooling to 18 °C. After that 1.68 µL of premixed solution prepared on ice (0.5 µL of Library Synthesis Buffer, 0.78 µL of dH 2 O, and 0.4 µL of Library Solution Enzyme) was added and mixed by pipetting. Sample was incubated at 18 °C for 10 min, 25 °C for 10 min, 37 °C for 30 min, 42 °C for 10 min, 70 °C for 20 min, and 4 °C. Then 7.5 µL of Amplification mix, 60.2 µL of dH 2 O, 1.58 µL of dNTP mix, 0.75 µL of Amplification Enzyme, and 5 µL of DNA library was mixed by pipetting on ice. Samples were incubated in a thermocycler with the following protocol: 94 °C for 2 min, followed by 17 cycles of 94 °C for 30 s, and 70 °C for 5 min, and cooling down to 4 °C. Amplified libraries were purified with a column-based purification kit (Biolabmix, Russia) following manufacture protocol. Further library preparation was performed with a Syntera library preparation kit (Syntol, Russia). Sequencing was performed in Tomsk National Research Medical Center on the Genolab M platform (Genemind, China) and 2 × 150 bp reads were obtained. Paired-end reads were used to obtain metagenomic assembly using ViPER pipeline v.2.3.1 [22] with the following arguments -m 200 -tripleassembly; reads aligned to host genome (Anas crecca genome, GCA_036873605) were removed from assembly.
## RT-PCR confirmation of viral RNA presence
We confirmed the RNA presence of viruses detected metagenomically with RT-PCR. For that, we designed oligonucleotides, flanking the region 1093-1455
of detected DPiV genome. RNA was isolated with a column-based RNA extraction kit (Biolabmix, Russia) following manufacturer protocol.
BioMaster RT-PCR -Premium kit (Biolabmix, Russia) was used to perform RT-PCR. We added 6.25 µL of RT-PCR-Premium to 0.5 µL of BioMaster-Premium-mix following 5 pmol of each primer, 2.5 µL of RNA, and dH 2 O to a final volume of 12.5 µL. DHAV and DPiV RT-PCRs were performed separately in the temperature profile of reverse transcription at 45 °C for 30 min, preliminary denaturation at 93 °C for 5 min, following 35 cycles: 93 °C for 15 s, 58 °C for 20 s, and 68 °C for 30 s; and final elongation at 68 °C for 7 min.
PCR products were visualized in 1.5% agarose gel electrophoresis.
$$(DHAVY14F-A G T T T G G T T C C C C T A C C C A T T C A A; DHAVY14R-A C A G G C T T A T G G A T T G G T C T C C T C) of detected DHAV genome and the region 2715-3203 (DPIVY14F-G A C C A G G G T T T T G A T G A G G T G G A T T, DPiVY14R-G A C A T C C C C T T G T G T T A A G C C A G A A)$$
## Isolation of viruses with duck embryo and duck embryo fibroblast (DEF) primary cell culture
Twelve-day duck embryos were used for virus isolation. A 100 µL volume of the suspension supernatant was inoculated into the allantoic cavity of the embryo. Inoculated embryos were incubated for 5 days at 37 °C. The condition of the embryos was monitored daily by egg candling throughout the incubation period, and allantoic fluid was collected from embryos that died on the day of detection. Collected samples were used to confirm successful isolation and for further assays and passages.
Duck embryo fibroblast (DEF) primary cells from 12-day-old Pekin duck embryos were prepared following protocol for chicken embryo fibroblast isolation as described previously [23] with modifications. Briefly, muscle tissues from duck embryo were cut into 1 mm pieces, washed, and then digested in 0.1% trypsin at room temperature for 10 min, with gentle mixing of the tube every 3 min. Trypsin was removed after centrifugation for 3 min at 5000 g and cells were washed with PBSS 3 times. Prepared DEF cells were seeded into the 96-well plates with 3 × 10 4 cells per well in a growth medium Dulbecco's Modified Eagle Medium (DMEM, with 10% FBS (Capricorn Scientific, Germany) and 50 µg/ mL of gentamicin sulfate (BioloT, Russia). In 24 h of the incubation cell monolayer was pre-washed with a Hanks solution; then the 100 µL of four-fold dilutions of allantoic fluid containing the virus were inoculated into cells, and plates were incubated at 37 °C and 5% CO 2 for 1 h for virus adsorption. Upon removing the supernatant, a maintenance medium consisting of DMEM (Capricorn Scientific, Germany) with 2% of FBS (Capricorn Scientific, Germany) and 100 µg/mL of gentamicin sulfate (BioloT, Russia) was added into the wells. Cytopathic effect of viruses on DEFs was assigned visually using the microscope (Micromed I, Russia). At the 4th and 6th d.p.i. aliquots of supernatant were collected for further PCR-detection of the virus.
## ELD50 Estimation for DHAV
We used the method of Reed and Mench to estimate 50% Embryo Lethal Dose (ELD 50 ) of detected DHAV. Two-fold serial dilutions of the virus isolate were inoculated into duck embryos, with 100 µL of diluted sample administrated per embryo. The embryos were incubated then for 7 days. ELD 50 was calculated with the following formula:
where A -percent of mortality at the dilution immediately below the target 50% dose.
B -percent of mortality in dilution immediately above the target 50% dose.
$$X = (A -50)/(A -B)$$
## 3' RACE of DHAV
We used the Mint RACE cDNA amplification set (Evrogen, Russia) to confirm the 3'-end of the DHAV genome. Extracted RNA was used for cDNA synthesis and amplification with Mint RT and Encyclo PCR kits (Evrogen, Russia) following manufacturer protocol. For the
## 3'-RACE first round, we used DHAVY14SO3-T T A A G T C C C G A T G C C C T G T C C oligonucleotide and for the second round, DHAVY14SO4-G A A A T T C
For the first round of RACE, we diluted 2 µL of amplified cDNA with 38 µL of dH 2 O. 2 µL of diluted product was added to reaction mix (40 µL of dH 2 O, 5 µL of 10x Encyclo buffer, 1 µL of dNTP mix, 1 µL of 10 pmol/µL DHAVY14SO3, and 1 µL of Encyclo polymerase). We divided the reaction volume into two 200 µL tubes 25 µL each and added 1 µL of 25x Step-out primer mix-1 to one of them. The second tube was used as a control. RACE was performed in the following conditions: 95 °C for 1 min following 29 cycles: 95 °C for 15 s, 58 °C for 20 s, and 72 °C for 3 min. The second step of amplification was performed with a 1:20 diluted amplification product from the first step with DHAVY14SO4 oligonucleotide in the same conditions. The product of amplification was observed with 1.5% agarose gel electrophoresis and confirmed via Sanger sequencing.
$$A A C C C T G G G G C G T C was used.$$
## Genome annotation and phylogenetic analysis
The genome sequence of DHAV was annotated with VADR v.1.6.4 and reference DHAV genome NC_008250 [24]. The genome sequence of unclassified DPiV was annotated manually due to low identity using reference DPiV sequence from GenBank (MT681985). Annotated genome sequences were visualized with the DNA Features Viewer tool [25]. Alignment of amino acid sequences of polyprotein was created with MAFFT v.7.505 [26]. The phylogenetic tree was constructed with IQTREE2 v.2.3.6 using a maximum likelihood algorithm and best-fit substitution model (LG + F + I + R4), with node support evaluated through 1,000 bootstrap replicates [27]. The resulting tree was visualized with the ggtree R package [28]. Simplot of complete genome sequences of Avihepatovirus was constructed with seqcombo R package. For IRES RNA sequence secondary structure prediction RNAfold of ViennaRNA Package 2.0 [29] was used with default parameters with further manual visualization using RiboSketch [30].
## Results
## Metagenomic detection of picornaviruses presence
Pooled metagenomic sequencing of fecal samples from three common teal individuals generated 19,577,424 paired-end reads. After metagenomic assembly, 764,678 reads were re-aligned to viral contigs, with 138,826 reads assigned to the Picornaviridae family. The most abundant contig (Y14_DHAV) belonging to the Picornaviridae family was assigned to species Avihepatovirus ahepati. The next most abundant contig (Y14_DPiV), supported by 63,006 reads, was identified as Duck picornavirus (DPiV) of an unassigned species. However, the top hit in the BLASTx search corresponded to a Duck picornavirus associated with short beak and dwarfism syndrome in domestic ducklings in China [14]. The length of trimmed contigs was 7731 and 7431 with average coverage of 268× and 235× for Y14_DHAV and Y14_DPiV, respectively. To confirm the presence of viral RNA in the sample and to identify its source within the metagenomic pool, RNA was extracted from individual feces samples comprising the pool. RT-PCR using two pairs of oligonucleotides (DHAVY14, DPiVY14) confirmed the presence of viral RNA corresponding to two detected viruses in a single sample (18Yak).
## Isolation of DHAV and DPiV
Then, we inoculated 12-days duck embryos with feces suspension and after incubation, we confirmed the presence of two viruses in allantoic fluid with RT-PCR. Subsequently, a primary duck embryo fibroblast (DEF) cell culture was used in an attempt to obtain viral monoisolates. A serial 1:4 dilutions of the positive allantoic fluid were inoculated onto DEF cells, but no cytopathic effect (CPE) was observed seven days post-inoculation. However, we performed RT-PCR of individual wells and detected wells with monoisolates of DHAV and DPiV in 1:4, 1:16, and 1:64 dilutions which we named as isolates DHAV/18Yak and DPiV/18Yak, respectively. Monoisolates were used for cultivation in the second passage of DEF. DHAV was detected in the second passage with no CPE, and RT-PCR of DPiV monoisolates was negative in the second passage.
Monoisolates from the first passage of DEF cultivation were inoculated into the allantoic cavity of 12-day duck embryos and after cultivation, both viruses were confirmed to replicate with RT-PCR. Signs of pathogenicity of viruses in duck embryos were observed. The pathogenicity of the DHAV isolate in duck embryos was confirmed by an embryonic lethal dose 50% (ELD 50 ) assay, which was determined to be 10 1.55 ELD 50 /mL. Hence, future characterization of the isolates should include deep sequencing and electron microscopy (EM) to exclude the possibility of co-cultivation with other viruses or bacteria that could influence the observed results.
## Genetic characterization
We confirmed the completeness of the 3'-end of the genome of DHAV with 3'-RACE; however, the 5'-end could not be verified with RACE. The length of contigs and coverage points to the significant reliability of observed genome assembly. DHAV has 5' UTR of 646 nucleotides and 3' UTR of 312 nucleotides (polyA excluded). The polyprotein of the detected virus has a length of 6750 nucleotides and 2250 amino acids, respectively, while reference Duck hepatitis A virus type 1 has 2249 amino acids (6747 nucl.). The polyprotein gene identity level is only 77.43% against the reference DHAV-1 sequence (NC_008250), with the maximum value (77.84%) for strain DRL-62 (DQ219396). The polyprotein structure is typical for DHAV: a large polyprotein cleaved onto three subunits P1, P2, P3 -which further cleaved into three structural (VP0, VP3, VP1) and nine non-structural proteins (2A1, 2A2, 2A3, 2B, 2C, 3A, 3B, 3C, and 3D) (Fig. 1).
Interestingly, the P1 region of polyprotein gene encoding structural proteins has the lowest identity values against DHAV-1 reference, while sequences of P1 have higher values of identity in comparison with DHAV-2 (Fig. 2, Figure S2). In particular, the VP1 amino acid sequence of strain from our study has 88.66% of identity with DHAV-2 isolate from India (OQ862826) and only 70.04% of identity with DHAV-1 strain (DQ219396). To investigate potential recombination, six statistical recombination detection methods were applied using RDP5 [31]. Five detection tests supported a recombination event when GENECONV test showed a negative result. Detailed gene and protein characteristics are provided in Table 1. VP1 protein has the lowest amino acid identity value (70.04%), while 2A3 and 2B are highly conservative with a 100% level of amino acid identity. Previously described conservative region of neutralizing linear B-cell epitope 75 GEIILT 80 in DHAV-1 and 75 GEVILT 80 in DHAV-3 VP1 differs in detected DHAV -77 GELVVT 82 , which has more similarity to DHAV-2 region 77 GEL-VIT 82 [32]. VP3 genotype-1-specific epitope 205 PSNI 208 [33] has less similarity with detected strain sequence protein possesses previously described in DHAV-1 and DHAV-2 strains ribosomal skipping site 14 GVEPNPGP 21 [13,34].
Phylogenetic analysis of DHAV polyprotein amino acid sequences indicates that the newly identified strain shares a most recent common ancestor with the DHAV-1 clade, yet forms a distinct sister branch (Fig. 3). A relatively high distance between the DHAV-1 clade and the detected strain can support the hypothesis of antigenic difference between them, however, the complex antigenic study is needed.
Similar to previously characterized DHAV strains, the detected DHAV possesses a similar to type IV internal ribosome entry site (IRES) secondary structure in region 361-633 of the genome (Fig. 4). The length of IRES sequence is 266 nucleotides when similar region of reference strain (NC_008250) has 261 nucleotides. Sequences have 83% of nucleotide identity.
The genetic structure of strain DPiV/18Yak is similar to the described Duck/FC22/China/2017 strain: 5'UTR-
The 5′UTR is 578 nucleotides in length, although its completeness has not been confirmed. The 3′UTR, excluding the poly(A) tail, is 425 nucleotides long. The complete polyprotein of the isolate is 6,405 nucleotides long, encoding 2,135 amino acids, whereas the reference strain contains 6,420 nucleotides and encodes 2,140 amino acids. The strain has a low amino acid identity of polyprotein against the described picornavirus (61.89%) (Table 2). 3 A protein amino acid sequence has the lowest identity level (44.64%) and 2A1 has the highest value (81.25%).
$$L-VP0-VP3-VP1-2A1-2A2-2B-2-3 A-3B-3 C-3D-3'UTR.$$
## Discussion
Identifying virus reservoirs of veterinary importance in the wild poses a significant challenge. For many viruses, virus-host interactions, the role of wild animals in viral evolution, and the diversity of the genetic pool remain poorly understood. Wild migratory birds, particularly those from remote and poorly studied breeding areas in northern latitudes, such as northeast Siberia -serve as important reservoirs. Monitoring these birds is essential due to their potential role in maintaining and transmitting viruses across different ecosystems.
In our study, we detected two avian picornaviruses in common teal feces collected in 2022 in the Northeast region of Russia -Yakutia region. Both viruses were passaged in domestic duck embryos and DEF primary cell culture to obtain two separate monoisolates. The first viral genome was related to the DHAV-1 genotype; however, it exhibited a relatively high pairwise genetic distance from the known DHAV-1 cluster. The coding Fig. 3 Maximum-likelihood phylogenetic tree of polyprotein amino acid sequences of Picornaviridae genera Avihepatovirus, Aalivirus, Grusopivirus and unclassified Picornviridae. Viruses, described in the study marked with red and blue rectangles. List of sequences used for phylogenetic analysis provided in Supplementary Material, Table S4 region of structural proteins (P1 subunit) has 69.60% and 75.68% of nucleotide sequence identity with reference DHAV-1 and DHAV-2 genome, respectively, highlighting the possibility of antigenic differences with known DHAV genotypes, as it codes 3 capsid proteins, where VP1 possesses multiple antigenic epitopes [36]. Particularly, VP1 amino acid sequence similarity of DHAV-1 (DQ219396) in comparison with DHAV-2 (OQ862826) is 70.04%, DHAV-1 with DHAV/18Yak -70.04%, DHAV-1 with DHAV-3 (DQ812093) -76.47%. DHAV/18Yak VP1 amino acid sequence shares 88.66% of similarity with DHAV-2. The DHAV/18Yak epitopes in VP1 and VP3 differ from previously described DHAV epitopes, which may additionally indicate potential antigenic divergence. 2A3 and 2B proteins have 100% amino acid identity with the DHAV-1 genome, assuming the conservative role in virus replication. Previous study have reported multiple inter-and intragenic recombination events in DHAVs, with a particularly high frequency detected in the 5' region and the upstream of the capsid proteincoding region [4]. Simplot, pairwise similarity comparisons, and individual protein phylogenetic trees reveal a relatively close relationship between the P1 region of the isolated DHAV/18Yak and the DHAV-2 genotype (Fig. 2, Supplementary Material: Figure S2, Figure S4), whereas the remainder of the genome shows greater similarity to DHAV-1. Together with the RDP5 recombination tests, this suggests a potential recombination event, however, additional genomic data from the newly identified wild duck-type DHAV cluster is needed to determine whether the observed signal results from recombination or can be attributed to convergent evolution. Understanding the antigenic properties and potential recombination events is crucial for the early detection of novel viral variants that may have significant implications for wildlife diseases and veterinary medicine, particularly as a threat to the health of diverse bird species.
Disease reported for Avihepatovirus in China had high mortality and morbidity among 20-25-day-old ducklings with hemorrhage in the liver and pancreas [9]. The virus detected in China was isolated with SPF chicken embryos with 100% mortality in embryos after 4 passages. In our study, we have not observed CPE in DEF inoculated with isolates, however, both isolates showed limited pathogenicity in duck embryos. Despite detected viruses were able to replicate in DEF cells (limited replication for DPiV), we assume that the virus titer was low which corresponds to the data that DHAV can replicate in duck and chicken EF with low efficacy [9]. Low nucleotide identity of viral capsid protein genes against known variants in composition with observed signs of embryonic pathogenicity highlights the possibility of domestic duck infection with wild duck virus variants. However, experimental confirmation using animal model infection is needed.
The second virus genome (DPiV/18Yak strain) was basal to a genomically diverse clade of unclassified picornaviruses of wild birds. The virus with the highest identity level was previously reported to be detected in domestic ducks with short beak and dwarfism syndrome outbreak in China, and based on similarity level, authors propose a new species of Avihepatovirus genus [14].
According to the ICTV criteria for the demarcation of picornavirus genera, our DPiV strain qualifies as a new genus based on the protein similarity in P1 being less than 66%, as well as significant differences observed in proteins L, 2B, 3A, and 3B. However, it does not meet the criterion related to similarity levels in proteins 2C, 3C, and 3D, which must be less than 64%.
Considering all the above, we conclude that the DPiV strain belongs to an unclassified genus and represents a distinct species within this genus. This conclusion is supported by: (1) its formation of a sister clade to the unclassified group; and (2) significant protein differences compared to the closest members of the unclassified group.
In summary, based on genome organization and phylogenetic analysis, we propose that one of the isolated viruses belongs to the species Avihepatovirus ahepati, sharing its most recent common ancestor with the DHAV-1 genotype. However, it exhibits substantial genetic divergence from known isolates, particularly within the structural protein-coding region, which may reflect antigenic differences. Notably, the virus was isolated from a wild duck, whereas previously characterized strains have been associated with domestic duck populations. Taken together, these findings support the designation of the isolated virus as a novel genotype. Further characterization of the isolate with electron microscopy, metagenomics, and metatranscriptomics deep sequencing to confirm its homogeneity is needed. Additional immunological comparative analysis of the isolate and comprehensive surveillance of DHAV diversity in wild ducks in regions crossed by migratory flyways using new genetic data would provide deeper insights into viral evolution and potential risks. DPiV/18Yak isolate genome has lower identity levels with known viruses and basal even to Unclassified bird picornaviruses clade. Genetic distances partially satisfy ICTV criteria for new genus demarcation in the Picornaviridae family. Particularly, P1 protein amino acid sequence identity is less than 64% (51.33%). However, other identity levels are lower than the threshold for genus demarcation. Therefore, although the taxonomy remains uncertain, it is essential to investigate the biological properties of genetically distanced viruses detected in wild ducks in the remote far northeastern region, in order to better assess their potential impact on domestic duck populations.
## References
1. Djunaidi, Djunaidi "The Economic Impacts of Avian Influenza on World Poultry Trade and the U.S. Poultry Industry: A Spatial Equilibrium Analysis" *Journal of Agricultural and Applied Economics. Cambridge Core*
2. Zell (2017) "Picornaviridae-the ever-growing virus family" *Arch Virol*
3. Shan, Yang, Wang et al. (2022) "Virome in the cloaca of wild and breeding birds revealed a diversity of significant viruses" *Microbiome*
4. Yang, Shah, Bahoussi et al. (1986) "Duck hepatitis a virus: Full-length genome-based phylogenetic and phylogeographic view during"
5. (2023) *Virus Res*
6. Su, Shuo, Luo et al. (2022) "An emerging Duck Egg-Reducing syndrome caused by a novel picornavirus containing seven putative 2A peptides" *Viruses*
7. Liao, Zheng, Yuan et al. (2014) "Genomic characterization of a novel picornavirus in Pekin ducks" *Vet Microbiol*
8. Wang, Liu, Wang et al. (2014) "Genetic characterization of a novel duck-origin picornavirus with six 2A proteins"
10. Pan, Yang, Zhou et al. (2012) "Duck hepatitis A virus possesses a distinct type IV internal ribosome entry site element of picornavirus" *J Virol*
11. Zhao, Wu, Wang et al. (2024) "Advances in the Duck hepatitis A virus and lessons learned from those in recent years" *Microb Pathog*
12. Zhang, Wu, Liu et al. (1080) "Current status and future direction of Duck hepatitis A virus vaccines" *Avian Pathol*
13. Zhang, Chen, Zhang et al. (2018) "Novel Duck hepatitis A virus type 1 isolates from adult Ducks showing egg drop syndrome" *Vet Microbiol*
14. Zhang, Yang, Lan et al. (2021) "Evidence of possible vertical transmission of Duck hepatitis A virus type 1 in Ducks" *Transbounding Emerg Dis*
15. Rajendran, Sundaram, Rao et al. (2025) "Molecular and phylogenetic analyses of the first complete genome sequence of Duck hepatitis A virus genotype 2 from India" *Gene*
16. Li, Shan, Chen et al. (2020) "Isolation and complete genome analysis of a novel Duck picornavirus in China" *Vet Microbiol*
17. Li, Wang, Yu et al. (2020) "Identification and genome characterization of a novel picornavirus from ducks in China" *Arch Virol*
19. Wille, Eden, Shi et al. (2018) "Virus-virus interactions and host ecology are associated with RNA Virome structure in wild birds" *Mol Ecol*
20. Sharshov, Dubovitskiy, Derko et al. (2023) "Does avian coronavirus Co-Circulate with avian paramyxovirus and avian influenza virus in wild ducks in siberia?" *Viruses*
21. Marchenko, Danilenko, Kolosova et al. (1038) "Diversity of gammacoronaviruses and deltacoronaviruses in wild birds and poultry in Russia" *Sci Rep*
22. Zhirov, Dubovitskiy, Derko et al. (0213) "First detection and diversity of astroviruses in wild migratory birds of Sakhalin Island, North Pacific" *Virus Genes*
23. Nikitina, Yavdoshak, Leonov et al. (1968) "Serological monitoring of Duck viral hepatitis type I in Russian federation" *Ptitsevodstvo*
24. Conceição-Neto, Yinda, Van Ranst et al. (2018) "Modular approach to customize sample Preparation procedures for viral metagenomics"
25. De Coninck, Basler, Jansen et al. (2023)
26. Tian, Cai, He et al. "Frontiers | isolation and selection of Duck primary cells as pathogenic and innate Immunologic cell models for Duck Plague Virus"
27. Schäffer, Hatcher, Yankie et al. (1186) "VADR: validation and annotation of virus sequence submissions to GenBank" *BMC Bioinformatics*
28. Zulkower, Rosser (2020) "DNA features viewer: a sequence annotation formatting and plotting library for python" *Bioinformatics*
29. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
31. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol*
32. Yu, Smith, Zhu et al. (0210) "8. ggtree: an r package for visualization and annotation of phylogenetic trees with their covariates and other associated data" *Methods Ecol Evol*
33. Lorenz, Bernhart, Höner Zu Siederdissen et al. (2011) "ViennaRNA Package 2.0. Algorithms for Molecular Biology"
34. Lu, Bindewald, Kasprzak et al. (2018) "RiboSketch: versatile visualization of multi-stranded RNA and DNA secondary structure" *Bioinformatics*
35. Martin, Varsani, Roumagnac et al. (2021) "RDP5: a computer program for analyzing recombination in, and removing signals of recombination from, nucleotide sequence datasets" *Virus Evol*
36. Zhang, Zhou, Xin et al. (2015) "Identification of a conserved neutralizing linear B-cell epitope in the VP1 proteins of Duck hepatitis A virus type 1 and 3" *Vet Microbiol*
37. Wu, Zhang, Meng et al. (1038) "Mapping a type-specific epitope by monoclonal antibody against VP3 protein of Duck hepatitis A type 1 virus" *Sci Rep*
38. Yang, Zeng, Wang et al. (2018) "Frontiers | DHAV-1 2A1 Peptide -A newly discovered Co-expression tool that mediates the ribosomal skipping function"
39. Hellen, De Breyne (2007) "A distinct group of Hepacivirus/Pestivirus-Like internal ribosomal entry sites in members of diverse picornavirus genera: evidence for modular exchange of functional noncoding RNA elements by recombination" *J Virol*
40. Ma, Sheng, Huang et al. (2015) "Molecular evolution and genetic analysis of the major capsid protein VP1 of Duck hepatitis A viruses: implications for antigenic stability" *PLoS ONE* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12833367&blobtype=pdf | # AI-powered analysis of viral metagenomic sequencing data for rapid outbreak investigation and novel pathogen discovery
David Chisompola, Emmanuel Luwaya, John Nzobokela, Phinnoty Mwansa, Martin Chakulya, Deepak Patil, Srivastava Vartika, Cleveland Clinic, United States, Saeed Soleiman-Meigooni
## Abstract
Emerging viral outbreaks continue to pose a persistent global health threat, underscoring the urgent need for a shift from reactive to proactive health security strategies. Viral metagenomic next-generation sequencing (mNGS) offers an unbiased, powerful approach to pathogen detection and discovery, yet its utility has been constrained by the computational complexity and slow turnaround time of data analysis during outbreak crises. The integration of artificial intelligence (AI) and mNGS is dismantling these barriers, enabling faster, more scalable outbreak response. This review synthesizes how AI-driven analytics are transforming mNGS applications, from genome assembly to sequence classification, using advanced architectures such as convolutional neural networks, recurrent neural networks, and transformers. Beyond accelerating workflows, AI's capacity for pattern recognition outperforms traditional homology-based methods, facilitating the discovery of novel viral families and tracing hidden transmission chains through anomaly detection. Nonetheless, critical challenges remain, including limited training data, the interpretability of AI models, and resource-intensive computational demands that risk widening an "AI divide" in global health. We evaluate these obstacles and highlight forward-looking strategies, including federated learning for privacy-preserving data sharing and explainable AI for improving trust and biological insight. Looking ahead, we envision an "AI-first" paradigm for outbreak preparedness, anchored in integrated "Digital Immune Systems" for continuous, global-scale surveillance. By framing the synergy between mNGS and AI as a transformative leap, this review underscores its potential to strengthen resilience against future pandemics.
## 1 Introduction
Emerging and re-emerging viral pathogens continue to threaten global health security, as evidenced by outbreaks of Ebola virus, Zika virus, SARS-CoV-2, and, most recently, Mpox (Han et al., 2023). These events highlight how infectious diseases can trigger epidemics and pandemics that overwhelm healthcare systems and cause widespread societal and economic disruption (Peters et al., 2020;Bavinger et al., 2020). The frequency of such outbreaks, particularly those of zoonotic origin, is increasing, driven by factors including climate change, ecological disruption, and intensified global connectivity (Wilder-Smith, 2021). A critical determinant in mitigating the impact of these events is the speed of the public health response. The rapid and accurate identification of the causative pathogen is the essential first step for presence of viral proteins, providing results in 15-30 min at the point of care. Their primary role is rapid case identification and triage, enabling immediate isolation and the prompt initiation of contact tracing to break chains of transmission in an outbreak's early stages (Yimer et al., 2024). One of their key strengths is that RATs can be deployed without the need for specialized infrastructure, requiring only minimal tools, which has made them particularly valuable in resource-limited settings. However, their analytical performance, particularly sensitivity and specificity, has shown considerable variability, necessitating confirmation with PCR-based methods (Yimer et al., 2024;Hirabayashi et al., 2024).
PCR based techniques remain the gold standard for detecting viral infections, including SARS-CoV-2 (Cantón Cruz et al., 2025), and mpox in recent outbreaks (da Silva et al., 2023;Li et al., 2010). Despite their high analytical performance, particularly in terms of sensitivity, specificity, and throughput capacity for testing large volumes of suspected cases within a limited time, PCR methods are constrained by high costs, long turnaround times, reliance on skilled personnel, and potential exposure risks at testing sites. Nonetheless, PCR continues to be the preferred method at many facilities, while RATs are increasingly adopted as complementary tools in outbreak settings (Cantón Cruz et al., 2025).
During outbreaks, neither RATs nor PCR techniques have demonstrated the capacity to identify novel viral pathogens, as both depend on prior knowledge of existing viruses for their design and clinical utility. This limitation underscores the need for advanced sequencing technologies to enable novel pathogen discovery. Several sequencing platforms are available, including first-generation methods such as Sanger sequencing, second-generation platforms like Illumina, and third-generation technologies such as PacBio (Heather and Chain, 2016).
Sanger sequencing, also known as the chain-termination method, was developed by Frederick Sanger in 1977 and is considered the firstgeneration DNA sequencing technology (Eren et al., 2022). It relies on the selective incorporation of chain-terminating dideoxynucleotides (ddNTPs) during DNA synthesis, producing fragments of varying lengths that can be resolved by capillary electrophoresis to determine the nucleotide sequence (Heather and Chain, 2016;Eren et al., 2022). Although limited by its application in large-scale outbreak settings, relatively low throughput, short read lengths, and higher costs compared to next-generation sequencing (NGS) methods, Sanger sequencing remains widely used due to its high accuracy, reliability, and suitability for small-scale projects such as gene validation, clinical diagnostics, and confirmatory sequencing. Its precision in detecting single nucleotide variants continues to make it a valuable reference standard, even in the era of high-throughput sequencing technologies.
Illumina sequencing delivers high-fidelity data for definitive analysis. Typically deployed on PCR or RAT-positive samples, Illumina's high accuracy and throughput are the cornerstone of genomic epidemiology (Huang et al., 2019). It enables precise reconstruction of transmission chains through whole-genome sequencing, distinguishes between multiple introductions of a virus, and powers large-scale surveillance to monitor for variants of concern. These attributes make Illumina particularly well-suited for outbreak investigations and the detection of novel viral pathogens.
Oxford Nanopore Technologies (ONT) sequencing provides realtime genomic intelligence. The portability of devices like the MinION allows for sequencing to be deployed directly in the field or in regional laboratories (Lu et al., 2016). This facilitates rapid initial characterization of an outbreak, enabling immediate detection of genetic drift or the emergence of a novel variant. Most significantly, its capacity for long-read, unbiased metagenomic sequencing makes it a powerful tool for de novo viral discovery when targeted tests are negative.
Sequencing platforms generate massive datasets that necessitate robust bioinformatic analysis. The sheer volume of data makes timely sequence interpretation challenging, and bioinformatic tools are essential for identifying novel viral pathogens. However, traditional bioinformatics is often constrained by the computational resources, costs, and specialized expertise required, limitations that are especially pronounced in resource-limited settings. Since the emergence of artificial intelligence (AI), there has been a growing effort to develop and validate automated AI-driven tools capable of analyzing sequencing data rapidly, enabling near real-time insights to support public health responses during outbreaks.
The full potential of this approach is realized through integration (Figure 2). In this framework, samples collected from humans, animals, or environmental sources are routed for sequencing. ONT provides rapid, near-source intelligence for initial outbreak characterization, while Illumina supplies high-fidelity data for definitive reconstruction and long-term surveillance. This combination of speed, portability, and accuracy forms a robust system for mitigating the impact of viral outbreaks and accelerating the discovery of emerging pathogens.
## 4 Traditional outbreak screening methods and their limitations
In the recent past, outbreak screening has facilitated the development and refinement of a wide range of diagnostic methods for the early detection of infectious diseases (Watkins et al., 2006). These include culture-based techniques, direct microscopy, immunoassays (antigen and antibody detection), and targeted NAATs such as PCR (Ieven, 2007). Traditional methods have been instrumental not only in guiding clinical diagnosis but also in paving the way for more advanced techniques. However, evidence suggests that there is considerable variability in their application during outbreak investigations, with no single method consistently preferred across different pathogens or public health settings. This inconsistency stems largely from differences in analytical performance, the availability of technical expertise, infrastructure demands, turnaround times, and associated costs.
Traditional outbreak screening methodologies have primarily relied on a combination of clinical suspicion and targeted laboratory testing (Abat et al., 2016). Symptoms alone are rarely sufficient for accurate identification, especially given that many pathogens produce overlapping clinical syndromes. As a result, clinicians, including physicians, veterinarians, nurse practitioners, and pathologists, play a central role in identifying suspected cases by screening patients presenting with compatible symptoms, collecting appropriate specimens, and initiating laboratory confirmation (Wagner et al., 2006). In this way, clinical suspicion provides a gateway for standardized case definitions and subsequent confirmatory laboratory testing. Conventional diagnostic tools such as real-time RT-PCR and viral culture have thus been central to outbreak detection (Reintjes and Zanuzdana, 2009). Nonetheless, despite their long history of use, these traditional approaches face significant challenges related to speed, accuracy, cost, and scalability in the context of modern outbreaks.
## 4.1 Direct microscopy
Direct microscopy, including electron microscopy (EM), remains a rapid tool for the presumptive identification of pathogens, allowing direct visualization of viral particles without prior knowledge (Apollon et al., 2022). Historically, EM played a decisive role in virus discovery, such as the 1948 differentiation of smallpox from chickenpox (Goldsmith and Miller, 2009). advances like transmission electron microscopy (TEM) and cryogenic electron microscopy (cryo-EM) (Richert-Pöggeler et al., 2019), sensitivity is generally lower than culture or PCR, and diagnostic error is a risk, as seen during coronavirus outbreaks (Bullock et al., 2022;Curry, 2003). Furthermore, in the case of SARS-CoV-2, EM confirmation of viral presence has been limited to select tissues (lung, heart, olfactory mucosa, and placenta), with inconsistent findings elsewhere (Birkhead et al., 2021). Such limitations highlight that, while EM remains invaluable for the discovery and confirmation of novel or unusual pathogens, it is not suitable for routine or large-scale outbreak diagnostics.
## 4.2 Polymerase chain reaction methods
The advent of nucleic acid amplification techniques, particularly PCR, transformed diagnostic virology beginning in the 1970s (Leland and Ginocchio, 2007). PCR enables sensitive and highly specific detection of viral nucleic acids without the need for viral propagation in culture, making it a faster and more versatile tool than traditional methods. Modern real-time PCR and multiplex NAAT platforms now allow simultaneous detection of up to 15 viruses and 4 bacteria in a single assay, representing a major advancement in outbreak screening, particularly for respiratory infections (Das et al., 2015).
PCR methods typically achieve sensitivity near 95% and specificity approaching 100%, which has established them as the gold standard in many diagnostic contexts. Their ability to detect low viral loads early in infection is particularly advantageous during outbreak investigations, where rapid case identification is critical. However, PCR is not without limitations. False positives may occur due to sample contamination, while false negatives may result from poor sample quality, inhibitors, or genetic mutations in the target region (Ieven, 2007;Iwata, 2020). Furthermore, high costs of reagents, consumables, and equipment, as well as the requirement for reliable electricity and trained staff, limit widespread accessibility in many resource-constrained settings. Despite these barriers, PCR remains the most widely adopted tool for outbreak detection, bridging the gap between clinical suspicion and definitive laboratory confirmation. Integration of viral metagenomics with AI intelligence, machine learning and deep learning.
## 4.3 Culture-based techniques
For much of the 20th century, culture-based methods were the cornerstone of viral diagnostics and were long regarded as the gold standard for pathogen identification (Leland and Ginocchio, 2007). Viruses such as vaccinia, smallpox, and yellow fever were among the earliest to be propagated in culture between 1913 and the 1950s, with subsequent breakthroughs following the discovery that poliovirus could grow in non-neural cell lines. Viral culture remains unmatched in its ability to generate live isolates for further characterization, including drug susceptibility testing, antigenic typing, and vaccine development.
However, culture-based techniques are inherently slow, often requiring days to weeks to yield results, an unacceptable delay in the context of rapidly evolving outbreaks. They are also technically demanding, requiring specialized laboratory facilities, strict biosafety protocols, and highly trained personnel. Contamination risks further complicated interpretation, sometimes necessitating repeat cultures and extending diagnostic timelines. In many low-and middle-income countries, inadequate infrastructure and resource constraints have restricted the use of viral culture in routine outbreak surveillance.
While culture retains value for research, reference laboratories, and vaccine development, its role in frontline outbreak detection has largely been superseded by faster and more sensitive molecular methods.
## 4.4 Immunoassays (antigen and antibody detection)
Immunoassays remain a cornerstone in viral discovery due to their specificity, sensitivity, and relative ease of implementation (Wang et al., 2023). These assays leverage highly selective interactions between viral antigens and host antibodies or between antibodies and viral antigens, enabling both direct and indirect detection of viral pathogens. The two primary categories-antigen detection and antibody detection-serve complementary roles in uncovering known and novel viruses (Pavia and Plummer, 2021).
Antigen-based immunoassays detect viral proteins directly in clinical or environmental samples, providing evidence of active infection (Louten, 2016). Techniques such as enzyme-linked immunosorbent assays (ELISAs), lateral flow assays, and chemiluminescent immunoassays utilize monoclonal or polyclonal antibodies to capture and quantify viral antigens. In the context of viral discovery, these assays can rapidly screen large sample sets, flagging potential cases for more detailed molecular characterization. For example, during early outbreak investigations, antigen assays have been critical in identifying emerging influenza strains or novel coronaviruses (Cantón Cruz et al., 2025;Hirabayashi et al., 2024;Pavia and Plummer, 2021), often preceding nucleic acid-based confirmation.
Serological immunoassays detect host antibodies generated in response to viral infection, offering insights into exposure history and immune response dynamics. ELISA, Western blotting, and multiplex immunoassays allow high-throughput screening for IgM, IgG, or IgA antibodies against viral antigens. In viral discovery, antibody detection is particularly valuable for identifying past or subclinical infections, uncovering viruses that may evade direct detection (Louten, 2016). Serology can also guide epidemiological investigations, revealing the prevalence and distribution of previously unrecognized viral pathogens in populations.
Immunoassays are often used in tandem with molecular techniques to increase detection sensitivity and validate findings (Wang et al., 2023). For emerging viruses with limited genomic information, immunoassays can provide the first clues of viral presence by recognizing conserved structural proteins or crossreactive epitopes. Furthermore, advances in recombinant antigen production, high-affinity antibody engineering, and multiplexed assay platforms have expanded the ability to detect multiple viral targets simultaneously, accelerating the pace of discovery (Matsunaga and Tsumoto, 2025).
Despite their utility, immunoassays face several limitations. Crossreactivity with related viruses may produce false positives, particularly in antibody-based assays (Luvira et al., 2022). Antigen assays may have reduced sensitivity in low-viral-load samples, while serology is limited by the window period between infection and detectable antibody production. Nevertheless, when carefully designed and interpreted in conjunction with complementary techniques such as metagenomic sequencing, immunoassays provide a rapid, cost-effective, and scalable approach for identifying novel viruses and monitoring emerging infectious threats.
## 4.5 DNA microarray
DNA microarray technology emerged as an advanced diagnostic tool for infectious diseases, designed to enable the simultaneous and specific detection of a wide range of pathogens (Asmare and Erkihun, 2023). The principle of detection relies on solid-phase hybridization, where pathogen-specific oligonucleotide probes are immobilized on a solid surface and hybridize with complementary sequences from a mixture of fluorescently labeled nucleic acids (Martínez et al., 2014). Over time, diverse microarray platforms were developed to target pathogens associated with respiratory, hemorrhagic, blood-borne, and central nervous system syndromes (Martínez et al., 2014), while broader-spectrum microarrays were designed for virus discovery and surveillance (Wang et al., 2002).
This technology represented a pivotal step forward in molecular diagnostics, as it enabled the parallel screening of thousands of predefined viral sequences on a single chip through probe-target hybridization (Wang et al., 2003). Compared with single-plex PCR, microarrays offered a much wider scope of detection. However, they remained fundamentally targeted: probe design required prior knowledge of pathogen genomic sequences, restricting their ability to identify novel or highly divergent agents. Additional challenges included cross-hybridization artifacts, which could compromise specificity, and a generally lower sensitivity compared to amplificationbased methods.
As a result, while DNA microarrays played an important transitional role in broad-spectrum pathogen detection, they were eventually surpassed in pathogen discovery by metagenomic nextgeneration sequencing. Unlike microarrays, metagenomic nextgeneration sequencing provides a truly unbiased approach, capable of identifying both known and previously uncharacterized pathogens, making it the focus of current and future innovations in infectious disease diagnostics. Viral metagenomics next-generation sequencing provides a fast, sensitive, and robust approach for detecting viruses, including those that remain undetectable by traditional culture techniques and sequence-dependent assays (Mokili et al., 2012). This unique capability has firmly established mNGS as a leading tool in the discovery of novel viruses. Unlike conventional diagnostic assays, which rely on prior knowledge of target sequences, mNGS employs an unbiased strategy that enables the simultaneous detection of both known and unknown viral pathogens. This makes it invaluable in situations where the causative agent of an outbreak is unknown, representing a critical first step in mounting an effective and timely outbreak response (Greninger, 2018). When both culture-based methods and advanced molecular assays fail to detect a pathogen, mNGS has often served as the ultimate diagnostic approach, leading to the identification of novel viruses (Figure 2; Roux et al., 2021).
Since the first viral genome was sequenced using metagenomic methods in 2002, the pace of virus discovery has accelerated dramatically (Mokili et al., 2012;Dutilh et al., 2017). A landmark example is the identification of SARS-CoV-2 in 2019, where mNGS enabled rapid characterization of the novel coronavirus and provided genomic data that informed early diagnostic test development, epidemiological modeling, and vaccine design. Two decades later, mNGS remains at the forefront of virology, now enhanced by the integration of artificial intelligence (AI), machine learning (ML), and deep learning tools, which increase the speed, accuracy, and interpretability of vast sequencing datasets.
The strength of mNGS lies in its flexibility and broad applicability. Unlike targeted assays such as PCR, which require specific primers, mNGS can be applied to a wide variety of sample types, including blood, respiratory swabs, stool, plant tissues, and environmental reservoirs such as wastewater, while still generating high-quality sequence data (Roux et al., 2021). This versatility is particularly significant in today's context of frequent human-animal-environment interactions, where zoonotic spillover events have led to the emergence of high-impact pathogens such as Ebola virus, mpox virus, and coronaviruses. Beyond pathogen discovery, mNGS has proven valuable in detecting co-infections, characterizing viral diversity within hosts, and monitoring viral evolution. For example, it has successfully identified viral-bacterial co-infections such as varicella zoster virus with herpes simplex virus-2 (Schuele et al., 2025). Similarly, Slavov, reported that clinically important viruses, including measles virus, SARS-CoV-2, hepatitis B virus, parvovirus B19, adenovirus, and human herpesviruses, were detected alongside commensal members of the blood virome such as anelloviruses (Slavov, 2025). These findings highlight mNGS's ability not only to detect pathogens but also to provide insights into the broader viral ecosystem associated with human health and disease.
The workflow of mNGS begins with specimen collection, which can include clinical samples (e.g., blood, cerebrospinal fluid, nasopharyngeal swabs), animal reservoirs, or environmental sources (e.g., soil and water). Viral nucleic acids (DNA and/or RNA) are then extracted, followed by random or targeted amplification and library preparation (Morgan et al., 2010). Sequencing is performed using high-throughput platforms such as Illumina, Oxford Nanopore Technologies, or Pacific Biosciences (PacBio), each offering distinct advantages in terms of read length, throughput, and error profile. The raw sequence data undergoes comprehensive bioinformatics processing, including quality filtering, host read subtraction, de novo assembly, and taxonomic classification, to achieve viral identification and characterization (Slavov, 2025;Alcolea-Medina et al., 2024;Chiu and Miller, 2019).
Despite its transformative power, the application of mNGS requires specialized expertise, sophisticated equipment, significant computational resources, and robust laboratory and bioinformatics infrastructure. These requirements currently limit their widespread use in low-and middle-income countries where outbreak risks are often highest. Nevertheless, the unbiased nature of mNGS makes it uniquely suited for outbreak investigations and pathogen surveillance. Its real-world impact has been repeatedly demonstrated, such as in the rapid identification of SARS-CoV-2 in Wuhan in 2019, the genomic characterization of mpox outbreaks, and the elucidation of viral genomes during Zika and Ebola epidemics. Moving forward, continued improvements in sequencing technology, data analysis pipelines, and cost reduction, alongside integration with AI-driven analytics, will further strengthen the role of mNGS in global health security and pandemic preparedness.
## 5.1 Metagenomic next generation sequencing technologies
Metagenomic next-generation sequencing platforms are now widely employed not only for targeted sequencing of specific genes or genomic regions but also for comprehensive, sequence-based association analyses that drive pathogen discovery and characterization (Harismendy et al., 2009). Their growing adoption is largely fueled by the urgent need for faster, more accurate, and versatile diagnostic tools in infectious disease management. Beyond diagnostics, mNGS has opened entirely new avenues for research by allowing scientists to interrogate genetic information at an unprecedented scale and resolution, thereby advancing our understanding of microbial diversity, host-pathogen interactions, and evolutionary dynamics (Goodwin et al., 2016).
Technology's integration of high-throughput performance with steadily improving affordability has solidified its role as a cornerstone in fields spanning from fundamental biology and epidemiology to precision medicine and clinical diagnostics. While debates persist regarding their overall cost-effectiveness, particularly in low-resource settings, the practical utility of mNGS in accelerating discovery and improving diagnostic sensitivity has proven invaluable. Importantly, one of its defining advantages lies in its ability to generate vast volumes of sequence data, typically ranging from 300 to 500 cycles per run, enabling deep coverage and comprehensive genomic profiling (Kozich et al., 2013).
Nevertheless, sequencing performance varies considerably across platforms, with notable differences in read length, accuracy, throughput, and error profiles that directly impact downstream analyses and clinical applications. Some technologies are optimized for generating short, highly accurate reads, whereas others prioritize long-read sequencing, which is advantageous for genome assembly and structural variant detection. Historically, Illumina's short-read (second-generation) sequencing has dominated mNGS because of its high accuracy and throughput. In contrast, the advent of third-generation platforms such as Oxford Nanopore Technologies (ONT) and Pacific Biosciences (PacBio) has revolutionized the field by enabling long-read sequencing, often spanning thousands of bases, with steadily improving accuracy (Han et al., 2024). While short reads (<300 bp) can lead to fragmented assemblies and may overlook structural variants, long reads provide the ability to resolve repetitive regions and capture full-length genes, thereby offering significant advantages in metagenomic applications (Han et al., 2024). Below we compare Illumina, ONT, and PacBio across key metrics for metagenomics (Table 1).
## 5.1.1 Illumina sequencing platforms
Illumina-based sequencing technology is one of the most widely adopted platforms in metagenomics, providing high-throughput and cost-effective sequencing of DNA and RNA (Xia et al., 2023;Schirmer et al., 2016). It has become a cornerstone for microbial community analysis, particularly for taxonomic profiling, functional annotation, and pathogen detection (Elbehiry and Abalkhail, 2025). The platform's short-read approach delivers massive sequencing depth at relatively low cost, making it ideal for large-scale studies across clinical, environmental, and engineered systems (Xia et al., 2023). Illumina short reads enable sensitive and robust detection of microbial community composition, antimicrobial resistance genes, and metabolic pathways, provided sufficient coverage is achieved. Furthermore, metagenomic Illumina tags ( mi Tags) can mitigate PCR amplification biases, offering more accurate estimates of microbial richness and evenness compared with traditional amplicon-based methods (Logares et al., 2014). Nonetheless, with modern instruments such as the NovaSeq 6000 and NovaSeq X series, Illumina can yield terabase-scale data in a single run, supporting large-scale environmental surveys and deep sequencing for rare pathogen detection.
Despite these strengths, Illumina sequencing introduces several challenges and biases. Position-and motif-specific errors, most commonly substitution errors linked to specific nucleotide motifs, may persist even after quality filtering and affect downstream analyses (Schirmer et al., 2016). The reliance on short reads (<300 bp) can lead to fragmented assemblies, particularly in complex communities or genomes with high sequence similarity, thereby limiting recovery of complete genomes and structural variants (González et al., 2025). • Nanopore-based electronic sensing of DNA/RNA strands (Ni et al., 2019).
• Single Molecule, Real-Time (SMRT)
sequencing with circular consensus sequencing (CCS) (Zhang et al., 2025).
Typical read length
• 50-300 bp (short) (Kozich et al., 2013).
• 10 kb-100+ kb (ultra-long) (Moustakli et al., 2025;Shafin et al., 2020).
• Long & Accurate (15-25 kb) (HiFi reads) (Wen and Tang, 2025;Travers et al., 2010;Wenger et al., 2019).
## Throughput and speed
• Very High throughput (Terabases/ run). Run time: hours to days (Liu et al., 2021).
• Scalable & Real-Time. From portable (MinION) to high throughput (PromethION). Data is available immediately. Scalable (MinION: ~30 Gb; PromethION:
~13 Tb) (Chen and Xu, 2023).
• High (Revio: ~100-120 Gb per SMRT Cell) (Zhang et al., 2025).
## Accuracy profile
• Very High (>99.9%); substitution errors (Schirmer et al., 2016).
• Moderate-High (Raw: ~99%; with Q20+: >99.9%); indel errors in homopolymers (Sereika et al., 2022;Chen et al., 2024).
• Extremely High (>99.9% with HiFi mode) (Wen and Tang, 2025;Travers et al., 2010;Wenger et al., 2019).
## Primary strengths
• Gold standard for cost-effective, high-depth sequencing.
• Excellent for SNP/variant calling and quantitative abundance (Xia et al., 2023;Logares et al., 2014).
• Real-time analysis for immediate insights.
• Extreme portability for field use.
• Best for detecting structural variants and de novo assembly.
• Direct RNA sequencing (De Coster et al., 2019).
• Unparalleled combination of long reads and high accuracy.
• Ideal for resolving complex regions, haplotype phasing, and high-quality de novo assembly (Slavov, 2025;Alcolea-Medina et al., 2024;Chiu and Miller, 2019).
## Primary limitations
• Short reads fail to resolve repeats, leading to fragmented assemblies.
• Cannot phase haplotypes or detect large Structural Variants s easily (González et al., 2025).
• Historically higher error rate, though improving rapidly.
• Requires substantial bioinformatics for basecalling and error correction (Sereika et al., 2022;Chen et al., 2024).
• Highest per-sample cost.
• Requires high-quality, high molecular weight DNA input.
• Less suitable for rapid, real-time applications (Han et al., 2024).
Best suited for mNGS application
• Large-scale surveillance and detection: Sensitive pathogen identification in complex samples (e.g., microbiome, plasma).
• High-resolution SNP analysis for outbreak tracing. (Xia et al., 2023).
• Rapid Real-time pathogen detection and outbreak response in the field.
• De novo assembly of unknown pathogens.
• Epigenetic modification detection (Marić et al., 2024).
• Gold-standard de novo genome assembly of novel viruses/bacteria.
• Resolving complex viral communities and strain variants.
• Full-length 16S/18S sequencing without amplification bias (Slavov, 2025;Alcolea-Medina et al., 2024;Chiu and Miller, 2019).
Frontiers in Microbiology 09 frontiersin.org Low-abundance organisms or genes may also be difficult to detect without very deep sequencing (e.g., ~30 million reads for 1% abundance) (Rooney et al., 2022).
To overcome these limitations, hybrid strategies that combine Illumina with long-read platforms such as Oxford Nanopore Technologies (ONT) or Pacific Biosciences (PacBio) are increasingly employed (Xia et al., 2023;Sevim et al., 2019). These approaches improve assembly contiguity, genome completeness, and the resolution of structural variants, while retaining Illumina's advantage in low error rates and reliable genome recovery. Ultimately, Illumina remains foundational in metagenomics for its accuracy, throughput, and affordability, but optimal study design requires careful consideration of its error profiles and assembly constraints. In many cases, hybrid or long-read approaches provide a more comprehensive view of microbial diversity and genome structure.
## 5.1.2 Oxford nanopore technologies
Oxford Nanopore Technologies (ONT) has transformed metagenomics by enabling real-time, portable, and long-read sequencing of complex microbial communities (Ni et al., 2019). ONT devices employ protein nanopores embedded in membranes to directly read native DNA or RNA molecules as they translocate through the pore, generating reads of virtually unlimited length, routinely exceeding 100 kilobases (kb) and occasionally surpassing 1 megabase (Mb) (Moustakli et al., 2025;Shafin et al., 2020). This ability to generate ultra-long reads is ONT's defining feature, enabling more contiguous genome assemblies, improved detection of structural variations, and strain-level resolution in metagenomic samples. Compared with short-read platforms, ONT excels at resolving repetitive regions and reconstructing complex genomes.
ONT platforms, including the portable MinION and highthroughput PromethION, are increasingly applied across clinical diagnostics, outbreak investigations, environmental monitoring, and food safety (De Coster et al., 2019). Technology is particularly valuable for rapid pathogen detection, often providing actionable results within hours. For example, ONT sequencing has been deployed in real-time outbreak response and field-based surveillance due to its portability and minimal infrastructure requirements (Oehler et al., 2023). Output scales from ∼30-50 Gb on a MinION flow cell to up to ∼13 Tb on a PromethION run (48 flow cells, 72 h), making it suitable for both targeted and large-scale applications (Chen and Xu, 2023).
Despite these advantages, ONT historically faced limitations due to higher raw error rates (5-10%), predominantly indels in homopolymeric regions. Although consensus polishing with short reads was often required, recent advances, including R10.4.1 nanopores and Q20 + chemistry-have markedly improved basecalling performance, achieving raw-read accuracies exceeding 99% (Sereika et al., 2022;Chen et al., 2024). Nevertheless, homopolymer-associated errors remain a challenge, and robust bioinformatics workflows are required for error correction, host DNA contamination filtering, and metagenome-assembled genome (MAG) reconstruction. Hybrid approaches that integrate ONT with Illumina sequencing remain the gold standard for producing highly complete and accurate assemblies.
Specialized tools (e.g., Pike for OTU-level analysis) and the development of field-adapted extraction protocols continue to expand ONT's utility, offering flexible and cost-effective solutions for microbial surveillance and biodiversity studies (Krivonos et al., 2025).
Overall, ONT's key strength lies in its real-time sequencing capability, enabling rapid clinical diagnostics, novel pathogen surveillance, and metagenomic assemblies requiring ultra-long reads. However, it's per-base cost remains higher than Illumina for very large-scale projects, and careful consideration of study goals is necessary when selecting ONT as a primary sequencing platform.
## 5.1.3 Pacific biosciences (PacBio)
Pacific Biosciences has established itself as the leader in highfidelity long-read sequencing through its circular consensus sequencing (CCS) strategy, yielding HiFi reads that combine longread lengths (typically 15-25 kb) with accuracies exceeding 99.9% (Q30 or higher) (Wen and Tang, 2025;Travers et al., 2010;Wenger et al., 2019). The latest Revio platform generates ~100-120 Gb per SMRT Cell, with up to four cells running in parallel, enabling hundreds of gigabases of highly accurate long-read data per day (Zhang et al., 2025). PacBio's requirement for high molecular weight DNA and relatively complex library preparation workflows pose technical challenges, but the resulting data are uniquely powerful. HiFi reads preserve single-nucleotide accuracy while spanning long genomic regions, enabling recovery of complete metagenome-assembled genomes (MAGs), structural variant resolution, and improved taxonomic resolution for rare or novel taxa. Comparative studies have shown that PacBio recovers more low-abundance lineages than Illumina or ONT, due to its combination of read length and accuracy. The high capital cost of PacBio systems and consumables may restrict adoption in resource-limited settings, but for high-resolution metagenomics, especially in research on complex microbial communities, PacBio HiFi data are considered the gold standard.
## 6 Current approaches to data analysis
Traditional alignment-assembly-annotation pipelines remain the backbone of viral metagenomic sequencing for outbreak investigation (Yang et al., 2024). They provide interpretable and clinically actionable results: read classification enables rapid confirmation of suspected pathogens, genome assemblies allow high-resolution phylogenetic analyses, and annotation facilitates detection of resistance or virulence markers (Song et al., 2021;Rosenboom et al., 2022). Benchmarking studies of clinical metagenomic pipelines confirm that these workflows are highly specific for known pathogens and reproducible across laboratories, supporting their integration into surveillance and public health responses (de Vries et al., 2021). The assembled genomes were central in tracking SARS-CoV-2 lineage dynamics during the COVID-19 pandemic, enabling timely insights into transmission, mutation hotspots, and global spread (Saravanan et al., 2022). The maturity of these pipelines, coupled with standardized platforms such as Nextflow and Snakemake, ensures reproducibility and traceability, critical strengths during time-sensitive outbreak responses (Langer et al., 2025).
On the other hand, recent studies confirm important limitations when applying traditional pipelines for outbreak investigation and novel pathogen discovery. Wet-lab protocol comparisons have demonstrated that enrichment via capture panels dramatically increases sensitivity in low viral load samples: for respiratory pathogens such as SARS-CoV-2 or influenza A, target capture sequencing yielded 180-2,000-fold higher viral read counts compared to untargeted metagenomics in some clinical specimens (Takemae et al., 2024). Benchmarking of virus identification tools using real-world metagenomic datasets found that while some tools perform well under default settings, there is a trade-off between sensitivity and specificity and many tools fail to recover virus contigs when genomes are fragmented or diverged (Takemae et al., 2024). Therefore, reliance solely on traditional pipelines can delay detection of novel or low-abundance pathogens in outbreak settings, unless supplemented with optimized sample preparation, high sequencing depth, and tools designed to handle divergent sequences (Wu et al., 2024).
## 7 AI and machine learning applications
The rapid expansion of metagenomic sequencing has generated unprecedented volumes of complex and heterogeneous data, necessitating advanced analytical frameworks beyond conventional bioinformatics (Pita-Galeana et al., 2025). AI, particularly ML and DL, has emerged as a transformative tool for extracting biologically meaningful insights from metagenomic datasets. Its applications span the entire analytical pipeline, from raw data preprocessing to functional inference and clinical translation.
One of the central challenges in viral metagenomics is accurate classification of sequences, especially when viral genomes exhibit high mutation rates or when reference databases are incomplete. Traditional bioinformatics pipelines rely on alignment-based methods (e.g., BLAST, Bowtie) or k-mer frequency approaches (Pita-Galeana et al., 2025;Wu et al., 2021). These approaches are limited when viral sequences diverge significantly from known references.
Deep learning algorithms, particularly convolutional neural networks (CNNs) (de Souza et al., 2023), and recurrent neural networks (RNNs) (Deif et al., 2021), have been developed to overcome these limitations. By learning hierarchical sequence features directly from raw data, deep learning models can classify viral sequences with higher accuracy and generalize better to novel or divergent genomes. For instance, CNN-based models have been applied to detect viral families from short sequencing reads without requiring genome assembly (Tampuu et al., 2019).
Autoencoders and attention-based models (e.g., transformers) further extend classification performance by capturing long-range dependencies and sequence motifs relevant to viral taxonomy (Mswahili and Jeong, 2024). Importantly, these models can recognize "novelty signatures, " allowing for classification at higher taxonomic ranks when species-level resolution is not possible (Table 2). This feature is critical in outbreak investigations where the pathogen may belong to an underrepresented or previously unknown viral group.
## 7.1 Key architectures and applications
## 7.1.1 Convolutional neural networks (CNNs)
Convolutional neural networks, initially developed for computer vision, have proven highly effective in biological sequence analysis due to their ability to detect local, position-invariant patterns, a defining property of genomic data (Rives et al., 2021;Rives et al., 2021). By applying learnable filters across nucleotide or amino acid sequences, CNNs act as motif discovery engines, identifying conserved short patterns such as transcription factor binding sites or protease cleavage motifs (Alley et al., 2019;Jumper et al., 2021;Senior et al., 2020). Sequences are typically represented numerically through one-hot encoding, where nucleotides or amino acids are mapped into binary vector space, forming input matrices analogous to images (Jumper et al., 2021). Convolutional and pooling layers then generate feature maps that summarize motif occurrence and position, conferring robustness to sequence variability (Unsal et al., 2022;Bileschi et al., 2022).
This hierarchical feature extraction enables CNNs to detect both simple motifs and complex higher-order structures, such as protein domains or viral polymerases (Hie et al., 2021;Ming et al., 2023;Lee, 2023). Consequently, CNNs can classify viral from non-viral sequences, annotate regulatory regions, and identify taxonomic signatures, even in cases of low sequence similarity to reference genomes (104). This capability to detect motifs without high-sequence homology is particularly advantageous over alignment-based methods like BLAST during the investigation of a novel outbreak. For instance, a CNN can enable family-level classification of an unknown virus based on conserved polymerase motifs, providing a crucial first clue for public health responders within hours, even when the virus shares less than 50% sequence similarity to any known reference. Tools like DeepVirFinder leverage CNN architectures to uncover viral sequences in metagenomic assemblies by integrating k-mer compositions with contextual genomic signals, outperforming alignment-based methods in novel virus discovery (Ren et al., 2020;Ren et al., 2020). As sensitive pattern detectors, CNNs thus provide a scalable solution for viral genome annotation and pathogen discovery in metagenomics.
## 7.1.2 Recurrent neural networks (RNNs)
Recurrent Neural Networks are designed specifically for sequential data and thus provide a natural framework for nucleotide and amino acid sequence analysis (Chandra et al., 2023). Unlike CNNs, which specialize in local motif detection, RNNs capture dependencies across positions by processing sequences element-byelement while maintaining a hidden state that reflects prior context (Graves, 2012). This allows the model to incorporate the meaning of a nucleotide or codon in relation to its surrounding sequence, which is critical for recognizing reading frames, splice sites, and regulatory elements spanning long genomic distances (Rumelhart et al., 1986;Auslander et al., 2021).
However, conventional RNNs are hindered by the vanishing gradient problem, limiting their ability to learn long-range dependencies (Bengio et al., 1994). Despite this, they remain valuable for tasks requiring short-to medium-range contextual modeling, including identifying conserved sequence patterns, functional annotation of genes, and detecting short-range evolutionary constraints.
## 7.1.3 Long short-term memory networks (LSTMs)
Long short-term memory networks extend RNNs by incorporating gating mechanisms, input, forget, and output gates, that regulate the retention and flow of information (Gers et al., 2000). This architecture mitigates vanishing gradients, enabling learning across long genomic regions where distant interactions carry biological significance.
In viral metagenomics, LSTMs are particularly effective in modeling genome-wide signals such as codon usage bias, oligonucleotide frequencies, and co-evolutionary patterns between Frontiers in Microbiology 11 frontiersin.org viruses and hosts (Nayfach et al., 2021). They also excel at identifying start/stop codons, splice sites, and functional domains separated by introns or long intergenic regions (Pasolli et al., 2019). This is critical for modeling viral evolution within a host during a prolonged outbreak, such as an Ebola or SARS-CoV-2 epidemic, allowing researchers to track the emergence of quasi-species that may impact transmission or treatment efficacy. For phylogenetic classification, LSTMs capture patterns of mutation and conservation across entire genomes, producing robust evolutionary inferences beyond local homology (Singh et al., 2016). Although newer architectures such as Transformers (Tampuu et al., 2019), offer advantages in handling very long sequences with parallelization, LSTMs remain widely used due to their strong performance in tasks where sequential order and contextual dependencies are biologically essential (Jurtz et al., 2017).
## 7.2 Transformers and attention mechanisms
Transformers mark a major advancement in sequence analysis, diverging from convolutional and recurrent architectures through their core innovation: the self-attention mechanism (Choi et al., 2022). Unlike CNNs, which emphasize local motifs, or RNNs, which process sequences sequentially, transformers assign weights to all positions
## Architecture
## Core mechanism Key strengths in viral metagenomics
## Primary applications and examples
Convolutional neural networks (CNNs)
Applies learnable filters to detect local, position-invariant patterns (Unsal et al., 2022;Bileschi et al., 2022).
Excellent motif discovery; robust to sequence variability; does not require high sequence homology (Ren et al., 2020;Ren et al., 2020).
• Viral sequence classification:
Distinguishing viral from host sequences (e.g., DeepVirFinder) (Ren et al., 2020;Ren et al., 2020).
• Family-level identification: Classifying unknown viruses based on conserved protein motifs (e.g., polymerase) (Mswahili and Jeong, 2024).
## Recurrent neural networks (RNNs)/
Long short-term memory (LSTMs)
Processes sequences step-by-step, maintaining a "memory" of previous context via hidden states (Nayfach et al., 2021).
Models short-to-long-range dependencies;
captures sequential context and temporal dynamics (Bengio et al., 1994).
• Modeling viral evolution: Tracking intra-host evolution and quasi-species dynamics during prolonged outbreaks (Singh et al., 2016).
• Functional annotation: Identifying gene boundaries (start/stop codons) and regulatory elements (Pasolli et al., 2019).
## Transformers and attention mechanisms
Uses self-attention to weigh the importance of all positions in a sequence simultaneously (Bigness et al., 2022;Choi and Lee, 2023).
Captures long-range, global dependencies efficiently; enables parallel processing;
highly adaptable via pre-training (Bigness et al., 2022;Choi and Lee, 2023).
• De novo genome assembly: Improved assembly of novel viral genomes from complex metagenomic data (Slavov, 2025;Alcolea-Medina et al., 2024;Chiu and Miller, 2019).
• Host prediction and pathogenicity:
Learning generalizable representations for tasks like host prediction (e.g., ViralBERT, PathogenTransformer) (Abràmoff et al., 2023;Vashisht et al., 2023;Choi and Lee, 2023).
## Graph neural networks (GNNs)
Operates on graph structures where nodes (e.g., individuals, variants) are connected by edges (e.g., transmissions) (Bileschi et al., 2022;Liu et al., 2020).
## Models relational and network data;
integrates heterogeneous data types (genomic, mobility, contact) (Nasir et al., 2023).
• Transmission dynamics: Reconstructing transmission chains and identifying superspreader events (Choi et al., 2022).
• Variant spread tracking: Integrating genomic and mobility data to project geographic spread of variants (Choi et al., 2022).
## Autoencoders (for anomaly detection)
Compresses input data into a latent space and reconstructs it; high reconstruction error indicates anomalies (Marić et al., 2024;Zhao et al., 2023).
Unsupervised learning; does not require labeled data for novel pathogens; identifies deviations from known sequences (Barredo Arrieta et al., 2020).
• Novel pathogen discovery: Flagging unclassified or highly divergent genomic fragments as potential novel viruses (Barredo Arrieta et al., 2020).
• Outbreak early warning: Detecting anomalous sequence clusters in surveillance data (Kuo and Ying, 2023).
simultaneously, enabling direct modeling of long-range dependencies (Bigness et al., 2022;Choi and Lee, 2023). This "global receptive field" allows the model to capture distant but functionally linked genomic features, such as promoter-coding region interactions or co-evolutionary signals across protein domains (Marić et al., 2024;Zhao et al., 2023).
The multi-head attention framework further enhances representational capacity by attending to different subspaces in parallel, capturing both syntactic (e.g., reading frames) and semantic (e.g., protein domain function) dimensions of genomic sequences (Marić et al., 2024). By modeling entire genomes holistically, transformers facilitate more accurate de novo assembly of novel viral genomes from complex metagenomic samples, directly addressing the challenge of fragmented assemblies that can delay the development of confirmatory PCR tests. Pre-trained models such as ViralBERT and PathogenTransformer leverage massive viral sequence corpora to learn generalizable representations that can be fine-tuned for specific tasks, including host prediction, gene annotation, and pathogenicity classification (Abràmoff et al., 2023;Vashisht et al., 2023;Choi and Lee, 2023). More recent architectures, such as MetaViT, extend transformer applications to metagenomics, effectively identifying novel viral sequences by recognizing global genomic signatures beyond local homology (Ji et al., 2021). By modeling genomes holistically, transformers advance viral genomics toward contextaware interpretation and accelerate novel virus discovery.
## 7.3 AI in anomaly and outlier detection
Early identification of novel pathogens is a critical application of AI in outbreak surveillance (Jurtz et al., 2017). Traditional approaches, which rely on known genetic signatures or symptom patterns, are limited in detecting truly novel threats (Tisza and Buck, 2021). In contrast, unsupervised and semi-supervised learning models excel at anomaly detection by learning baseline distributions of genomic or clinical data and flagging deviations (Barredo Arrieta et al., 2020). Within metagenomic sequencing datasets, algorithms such as isolation forests and autoencoders can detect unclassified genomic fragments as potential novel viruses (Kuo and Ying, 2023). For example, high reconstruction error in autoencoders indicates sequences that diverge from known distributions, serving as a quantifiable anomaly score (Marić et al., 2024;Zhao et al., 2023).
Beyond genomics, AI pipelines integrate clinical and epidemiological metadata, such as geographic location, travel history, and symptom onset, with sequence anomalies to detect clusters of unexplained infections (Edwards et al., 2016). This proactive detection framework can generate early outbreak warnings well before traditional confirmation methods, potentially reducing response delays by weeks (Willmington et al., 2022;Rudin, 2019).
## 7.4 Predictive models for transmission dynamics
Following pathogen identification, AI-driven models enhance prediction of transmission dynamics, informing timely public health interventions (Lee et al., 2023). Traditional SEIR (Susceptible-Exposed-Infectious-Recovered) frameworks provide a foundation but rely on static parameters (Auslander et al., 2021;Cheohen et al., 2025). AI augments these models by integrating real-time data and adapting parameters dynamically. Time-series methods, particularly LSTMs, can incorporate case counts, human mobility, climate data, and social media signals to forecast short-term epidemic trends with greater accuracy (Choi and Lee, 2023;Kuo and Ying, 2023).
Graph neural networks (GNNs) extend predictive power by modeling transmission chains, representing individuals or communities as nodes and their interactions as edges (Bileschi et al., 2022;Liu et al., 2020). Such models can identify superspreader events, transmission hubs, and potential intervention points. Moreover, by incorporating genomic data, GNNs can track pathogen evolution alongside mobility-driven spread, enabling projections of both geographic expansion and variant dominance (Nasir et al., 2023). These integrative models support resource prioritization and targeted containment strategies, bridging epidemiological forecasting with genomic surveillance.
## 7.5 Case studies
## 7.5.1 SARS-CoV-2
The COVID-19 pandemic served as a large-scale proving ground for AI in virology (Rives et al., 2021;Shkoporov et al., 2022). Deep learning models, most notably AlphaFold2, accurately predicted the 3D structure of the SARS-CoV-2 spike protein, which significantly accelerated rational vaccine design and therapeutic development (Jumper et al., 2021). AI-driven genomic surveillance systems played a crucial role in monitoring viral evolution by classifying variants of concern (e.g., Alpha, Delta, Omicron) through detection of mutational signatures associated with increased transmissibility, pathogenicity, and immune escape In parallel, natural language processing (NLP) tools enhanced real-time situational awareness by rapidly scanning global research articles, and news outlets to identify and synthesize emerging scientific insights (Jumper et al., 2021;Ren et al., 2020;Capponi et al., 2021). Collectively, these advances highlighted the transformative role of AI in outbreak response and set the stage for its broader integration into future pandemic preparedness strategies.
## 7.5.2 Ebola virus
During the 2018-2020 Kivu outbreak in the Democratic Republic of the Congo, AI facilitated predictive risk mapping by integrating satellite imagery, climate data, and animal habitat distributions to identify spillover hotspots (Willmington et al., 2022;Pigott et al., 2014). Machine learning models were employed to differentiate between local transmission chains and novel viral introductions, thereby informing containment strategies and resource allocation. Furthermore, AI-driven phylodynamic frameworks provided critical insights into the evolutionary dynamics and geographic spread of the virus. Complementarily, network-based analyses of contact-tracing data identified key transmission pathways, which guided targeted vaccination campaigns in resource-constrained and conflict-affected settings.
## 7.5.3 Mpox
The 2022 global Mpox outbreak highlighted AI's potential in detecting atypical transmission. Machine learning analyses of genomic data confirmed sustained human-to-human spread and revealed 10. 3389/fmicb.2025.1717859 Frontiers in Microbiology 13 frontiersin.org hidden transmission chains beyond endemic regions (Gigante et al., 2022). Models integrating air travel and case data further predicted high-risk cities for importations, supporting proactive surveillance and public health messaging. In clinical diagnostics, CNN-based models distinguished Mpox from other skin lesions with accuracies ranging from 78 to 98.8% across multiple datasets and architectures (Chadaga et al., 2023), underscoring AI's potential for rapid and reliable Mpox detection.
## 7.5.4 Influenza
AI applications in influenza span routine forecasting and pandemic preparedness. In seasonal surveillance, U.S. CDC forecasts are augmented with models incorporating viral genomics, search engine data, and historical trends to predict epidemic timing and intensity (Reich et al., 2019). For pandemic risk assessment, AI evaluates avian influenza strains (e.g., H5N1, H7N9), predicting traits such as receptor binding specificity and antigenic drift to inform pre-pandemic vaccine libraries (Lou et al., 2024).
## 7.6 Advantages over traditional methods
The integration of AI into virology and epidemiology provides substantial advantages over conventional approaches, particularly in speed, scalability, and predictive power.
## 7.6.1 Speed and automation
Traditional sequence analyses, such as BLAST searches and phylogenetic reconstructions, are computationally intensive and require manual curation. In contrast, trained AI models can process millions of sequences in hours, enabling real-time surveillance. Automated pipelines convert raw sequencing reads into variant calls and lineage assignments with minimal human intervention, allowing experts to focus on interpretation rather than data processing (Lee, 2023).
## 7.6.2 Handling high-dimensional data
AI excels at integrating diverse datasets, including genomic sequences, protein structures, clinical outcomes, mobility patterns, and environmental variables, revealing complex, non-linear relationships that traditional methods cannot capture. Whereas logistic regression may detect a few predictors, machine learning models such as random forests or neural networks can uncover intricate interactions to predict patient severity or outbreak hotspots (Soenksen et al., 2022).
## 7.6.3 Discovery of novel patterns
Unlike hypothesis-driven methods, AI can identify previously unrecognized patterns via unsupervised learning. This has enabled the discovery of novel CRISPR systems and microbial defense mechanisms (Doron et al., 2018). In virology, such approaches facilitate the detection of novel viral families and unconventional pathogenic mechanisms that might be overlooked by conventional analyses.
## 7.6.4 Predictive accuracy and adaptability
Classical compartmental models, such as SEIR, rely on fixed parameters. AI-enhanced models continuously assimilate new data, adapting forecasts as outbreaks evolve. This adaptability improves short-term predictive accuracy, as demonstrated by ensemble models that consistently outperformed traditional methods during the COVID-19 pandemic (Cramer et al., 2022).
## 8 Challenges and proposed solutions
The integration of AI into viral metagenomics offers transformative potential for outbreak response, yet its journey toward widespread, reliable, and equitable implementation is constrained by significant challenges Critically examining these limitations is not intended to diminish the technology's promise, but rather to provide a roadmap for guiding its continued evolution. This section highlights the central barriers, ranging from data availability and quality to model interpretability, to infrastructural and resource constraintsand aligns them with emerging research directions and technological innovations that seek to address these gaps.
## 8.1 Data scarcity and labeling bottlenecks
A foundational challenge is the "data requirements" of deep learning models, which require vast quantities of high-quality, accurately labeled sequences for training (Ren et al., 2020;Greener et al., 2022). The performance and generalizability of models are directly correlated with the volume and quality of their training data. However, the ground truth in virology is often elusive; labeling sequences as "viral" or assigning taxonomy requires slow, manual experimental validation or high-confidence homology, creating a fundamental data bottleneck (Edwards et al., 2016;Roux et al., 2019). As a result, viral sequence datasets remain orders of magnitude smaller than those used to train foundational models in other domains. Researchers often resort to data augmentation, semi-supervised learning, or transfer learning to partially overcome these limitations (Li et al., 2021;Santiago-Rodriguez and Hollister, 2022). While useful, these approaches are ultimately stopgap solutions; they cannot fully replace large-scale, high-fidelity, experimentally validated data.
More pernicious than sheer quantity is the profound taxonomic bias embedded within existing genomic databases. Public repositories like GenBank and RefSeq are overwhelmingly skewed toward viruses of established clinical and agricultural importance (e.g., influenza, HIV, SARS-CoV-2) (Nayfach et al., 2021;Tisza and Buck, 2021;Shkoporov et al., 2022;Schulz et al., 2020). In contrast, viruses from environmental niches, extreme ecosystems, and non-model organisms are severely underrepresented, creating a vast "viral dark matter" (Li et al., 2021;Santiago-Rodriguez and Hollister, 2022). This imbalance creates a "long-tail" distribution problem where DL models become highly accurate at recognizing common human pathogens but fail to identify novel or underrepresented viral families from under-sampled ecosystems, potentially delaying the response to a novel zoonotic spillover event (Nazer et al., 2023;Rampelli et al., 2020).
## 8.1.1 Emerging solutions and research directions
A multi-faceted approach is being developed to combat data limitations:
1 Transfer learning and pre-trained models: Researchers are increasingly leveraging models pre-trained on massive, 10.3389/fmicb.2025.1717859
Frontiers in Microbiology 14 frontiersin.org general-purpose protein or nucleotide sequence databases (e.g., models inspired by AlphaFold, DNABERT) (Jumper et al., 2021;Capponi et al., 2021). These models learn fundamental biological "grammar" and can be fine-tuned for specific viral classification tasks with much smaller, viral-specific datasets, thereby reducing the burden of data scarcity (Camargo et al., 2023). 2 Data augmentation and few-shot learning: Advanced techniques are being employed to artificially expand training datasets by generating realistic synthetic viral sequences (Ji et al., 2021). Furthermore, "few-shot learning" algorithms are being designed to learn effectively from a very small number of examples, which is critical for rare or novel viral families. 3 Global sequencing initiatives: Concerted efforts to systematically sequence diverse environments (e.g., the Global Virome Project, Earth Virome) are crucial for populating databases with novel viral sequences, thereby gradually correcting taxonomic biases and providing a more representative ground truth for model training (Rampelli et al., 2020).
## 8.2 The black box problem: interpretability and explainable AI
The predictive power of deep learning models is often tempered by their lack of interpretability, rendering them as inscrutable "black boxes" (Rudin, 2019;Greener et al., 2022). Although models may achieve high accuracy in distinguishing viral from host sequences, the underlying basis of their predictions often remains opaque. Identifying which specific nucleotides, motifs, or genomic structures drive a given decision is a persistent challenge. This lack of interpretability poses a critical barrier for virologists and public health officials, who need not only accurate classifications but also biologically meaningful and actionable insights to guide experimental validation and inform public health interventions (Barredo Arrieta et al., 2020;Samek et al., 2017).
## 8.2.1 Emerging solutions and research directions
The field of Explainable AI (XAI) is becoming indispensable for building trust and transforming predictions into scientific discovery.
1 Saliency maps and gradient-based techniques: Methods like Grad-CAM can highlight the nucleotides in an input sequence that most strongly influence the model's output, creating a "heatmap" of importance across the genome (Singh et al., 2016). For instance, when a model classifies a sequence as a coronavirus, a saliency map might pinpoint the receptorbinding domain, providing immediate, biologically plausible validation. 2 Feature attribution methods: Frameworks like SHAP (SHapley Additive exPlanations) quantify the contribution of each input feature to the final prediction (Lundberg and Lee, 2017). In viral host prediction, SHAP can reveal if a model is relying on codon usage bias or specific promoter sequences, thereby uncovering genomic signatures of co-evolution and generating testable hypotheses (Auslander et al., 2021). 3 XAI for discovery: Crucially, XAI extends beyond model debugging to enable biological insight. For instance, if an XAI model consistently highlights a non-structural protein gene in novel viruses associated with severe disease, it could point toward a previously uncharacterized virulence factor, guiding subsequent experimental research (Li et al., 2021;Choi et al., 2022). The development of domain-specific XAI tools is a critical research frontier for making AI a collaborative partner in virology.
## 8.3 Generalization and computational barriers
The development and training of state-of-the-art AI models require substantial GPU power and memory, creating a high financial and infrastructural barrier to entry for many academic and public health laboratories (Slavov, 2025). This is especially problematic in resource-limited settings where outbreak risks are often highest, threatening to create a new "AI divide" in global health security.
Furthermore, models trained on data from specific environments (e.g., human respiratory samples) often suffer from poor generalizability when applied to new contexts (e.g., seawater or animal vectors), a phenomenon known as overfitting (Li et al., 2023). A model that excels at identifying respiratory viruses in Illumina data from a US hospital may perform poorly on Nanopore data from bat samples in Southeast Asia, limiting its utility for proactive surveillance at the human-animal interface.
## 8.3.1 Emerging solutions and research directions
Innovations in computational infrastructure and model design are beginning to mitigate these barriers:
1 Cloud-based platforms and pre-trained models: The growth of cloud computing allows researchers to access highperformance computing on demand (Le Piane et al., 2024). More importantly, the sharing of pre-trained models means that end-users can fine-tune existing powerful models for their specific tasks, bypassing the immense cost of training from scratch. 2 Federated learning: This distributed approach enables AI models to be trained collaboratively across multiple decentralized datasets (e.g., from hospitals in different countries) without the raw data ever leaving its local environment (Nazer et al., 2023); (Yurdem et al., 2024). This preserves data privacy and sovereignty while allowing for the creation of more robust and generalizable models from diverse data sources, directly addressing the generalizability challenge. 3 Model optimization and lightweight architectures: Active research into model compression, quantization, and the development of more efficient neural network architectures aims to create powerful yet lean models that can be deployed on less powerful hardware, including at the point-of-care with portable sequencers (Dantas et al., 2024).
## 8.4 Ethical considerations and governance
Beyond technical and computational barriers, the deployment of AI-powered metagenomics raises profound ethical questions that Johnson et al., 2025). As NGS becomes increasingly integrated into clinical practice, the development of comprehensive, standardized regulations will be essential to effectively address its associated ethical challenges.
A primary concern is data privacy and consent. Clinical mNGS often sequences all nucleic acids in a sample, including the human host genome (Elbehiry and Abalkhail, 2025). This raises critical questions about patient autonomy and informed consent, as it is impossible to predict all pathogens that might be found. Furthermore, the integration of genomic data with clinical and mobility information in AI models creates rich datasets that are potentially re-identifiable, posing significant privacy risks if breached or misused.
Secondly, algorithmic bias and equity are major concerns (Joseph, 2025). As discussed in Section 6.1, models trained on biased data will perpetuate and potentially amplify these biases in their predictions. This can lead to systemic blind spots where pathogens circulating in under-surveilled regions are not detected, or where diagnostic AI tools perform poorly for certain populations. This could exacerbate global health inequities, directing resources and attention away from the most vulnerable communities.
Finally, the rapid identification of a novel pathogen with pandemic potential triggers complex questions about data sharing and dual-use risk. While rapid, open data sharing is crucial for a coordinated global response, it also creates a tension with national security and the risk of "dual use" research, where the same genomic information used to develop vaccines and diagnostics could theoretically be misused (Flores-Coronado et al., 2025). Establishing norms for the responsible communication of high-consequence findings to public health authorities without causing undue panic or stigma is a critical challenge.
## 8.4.1 Emerging solutions and governance frameworks
Developing robust ethical and governance structures is as important as advancing the technology itself.
1 Strengthening consent frameworks: Moving toward dynamic or tiered consent models for metagenomic testing, alongside the development and use of robust data anonymization and secure, federated learning techniques, can help protect individual privacy (Elbehiry and Abalkhail, 2025). 2 Bias audits and equity-focused design: Implementing mandatory algorithmic bias audits and actively promoting the sequencing of diverse viromes are essential to build fair and representative models. The "fairness" of AI models must be a key performance metric alongside accuracy (Chen et al., 2023). 3 International governance and policy: The establishment of clear international guidelines and agreements on the timely sharing of pathogen genomic data, coupled with frameworks to manage dual-use concerns, is urgently needed. Organizations like the WHO are pivotal in facilitating this dialog to ensure that these powerful tools serve global public health interests equitably and responsibly (Johnson et al., 2025).
Despite these challenges, the integration of deep learning, especially transformer-based models, into viral metagenomics is reshaping the field. By moving from reliance on known references toward data-driven discovery, these approaches are essential for both rapid outbreak characterization and systematic exploration of the global virosphere. Overcoming data scarcity, reducing bias, improving interpretability, ensuring generalizability and ethical consideration will be critical for unlocking their full potential.
By confronting these challenges with the outlined strategies, the field is moving steadily toward the development of robust, interpretable, and globally accessible AI-powered metagenomic tools. The goal is not to create perfect models, but to build resilient systems where the combined strengths of mNGS and AI can be reliably leveraged in the high-stakes, time-sensitive environment of an emerging outbreak.
## 9 Future perspectives
The trajectory of AI and viral metagenomics points toward a fundamental shift in how we monitor, detect, and respond to infectious disease threats. The future lies not merely in refining individual technologies, but in their deep integration into proactive, intelligent, and equitable global health systems. This section outlines several concrete paradigms and key milestones that will define the next decade of outbreak response.
The concept of an "AI-first outbreak response" heralds a paradigm shift wherein AI technologies transition from being supportive tools to leading players in epidemiological investigations and response strategies (Kaur and Butt, 2025). Traditionally, outbreak management has been largely human-driven, relying heavily on expert input for hypothesis generation, prioritizing laboratory testing, and coordinating contact tracing (Kaur and Butt, 2025). In contrast, AI-based systems can now autonomously generate hypotheses about pathogen origins and transmission pathways by integrating sequencing data with epidemiological and mobility information (Ye et al., 2025). They can also accelerate pathogen detection through rapid sequence classification and predict transmission hotspots for targeted interventions (Srivastava et al., 2025). By adopting an AI-first approach, public health systems can become significantly more proactive, adaptive, and scalable, often anticipating outbreaks before they fully manifest (CR et al., 2023). This strategy stands to drastically reduce the delay between pathogen identification and public health action, mitigating outbreak impacts and saving lives on a global scale (Villanueva-Miranda et al., 2025). As AI capabilities continue to advance, this proactive framework is poised to become a cornerstone of modern infectious disease control (Kaur and Butt, 2025).
The integration of AI with mNGS is poised to transform outbreak response from a largely reactive process into a proactive, predictive, and globally coordinated system. Several developments will shape this future trajectory. First, the concept of a "Digital Immune System" envisions an AI-driven global surveillance network capable of continuously analyzing metagenomic, clinical, environmental, and mobility data streams (Afshinnekoo et al., 2015). Such systems would detect anomalies and novel genomic signatures with sufficient accuracy to trigger automated early warnings, potentially identifying outbreaks weeks before traditional reporting. Future developments should focus on creating AI systems that seamlessly mesh with existing public health surveillance infrastructures such as hospital reporting networks and environmental monitoring stations allowing for continuous, automated data ingestion and pathogen detection (Alwakeel, 2025;Kaur and Butt, 2025). This synergy will support the development of an early warning ecosystem capable of flagging viral emergence or mutations in real time, thereby allowing proactive measures to prevent large-scale spread (Villanueva-Miranda et al., 2025). Second, advances in point-of-care metagenomics will enable rapid, field-deployable sequencing workflows, achieving "sample-toanswer" diagnostics in under 2 h (de Olazarra and Wang, 2023). Breakthroughs in portable hardware, lightweight AI algorithms, and curated "reference-on-a-chip" databases will make real-time sequencing feasible even in remote or resource-limited settings.
Third, federated learning frameworks will address data sovereignty and privacy concerns by enabling collaborative training of AI models without transferring raw genomic data across borders (Yurdem et al., 2024). This will foster equity, reduce taxonomic bias, and ensure that models remain generalizable across diverse populations and geographic regions. Cloud-based platforms combined with federated learning frameworks offer an innovative solution to the long-standing challenge of data sharing in pathogen surveillance (Aswini et al., 2025). Conventional centralized repositories often raise issues of privacy, sovereignty, and security, which discourage laboratories and countries from freely exchanging sensitive genomic data (Chourasia et al., 2024). Federated learning circumvents this by enabling AI models to be trained collaboratively across multiple decentralized datasets held by different organizations or countries, without the raw data ever leaving their local environments (Zwiers et al., 2024). This preserves patient confidentiality and national data ownership while harnessing the breadth of diverse, globally distributed datasets to produce more robust and generalizable AI models. Importantly, fostering such international collaboration is vital for pandemic preparedness and the detection of novel pathogens, as it enhances transparency, broadens surveillance capacity, and accelerates coordinated global responses (Zwiers et al., 2024;Calvino et al., 2024).
Fourth, the evolution of XAI will move AI beyond black-box predictions toward interpretable outputs that highlight key genomic features driving classification (Ali et al., 2023). This capability will enhance trust, accelerate biological discovery, and support evidence-based decision-making by public health authorities. The adoption of explainable AI will be critical for ensuring broad trust in AI-powered outbreak analytics among public health officials, researchers, and policymakers. Although AI can uncover intricate patterns and generate predictions from highdimensional metagenomic data, its "black-box" nature risks undermining confidence if the reasoning behind outputs remains unclear (Giuste et al., 2023). To bridge this gap, explainability frameworks tailored for viral genomics and metagenomics will need to be developed, capable of providing domain-relevant insights into how models prioritize mutations, classify sequences, and generate risk assessments (Yagin et al., 2023). Such transparency not only enables experts to validate AI findings but also supports effective communication with non-specialist stakeholders, thereby strengthening decision-making and public understanding (Msomi et al., 2025;Abe et al., 2023). As these explanation tools evolve, they will empower epidemiologists to better interpret AI-driven alerts, identify false positives or novel biology, and ultimately improve the overall accuracy and credibility of outbreak investigations.
Together, these innovations point toward an AI-first outbreak response paradigm, where intelligent systems autonomously analyze data, generate hypotheses, predict transmission dynamics, and guide interventions, while human experts provide oversight and strategy. By making surveillance faster, more interpretable, and globally inclusive, AI-powered mNGS could become the cornerstone of a resilient defense against future pandemics.
## 10 Conclusion
The integration of artificial intelligence with viral metagenomics marks a paradigm shift in outbreak response, moving us from reactive diagnostics to proactive pandemic preparedness. AI directly addresses the core bottleneck of mNGS-data complexityby enabling rapid pathogen identification, novel virus discovery beyond traditional methods, and predictive modeling of outbreaks. While challenges of data scarcity, model interpretability, and equitable access remain, emerging solutions like explainable AI and federated learning provide a clear path forward. This powerful synergy is forging a new "AI-first" frontier in global health, paving the way for intelligent surveillance systems capable of defending against future viral threats. and human generated. AI-based tools were used solely for language proofreading and rephrasing to enhance clarity and readability.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
## References
1. Abat, Chaudet, Rolain et al. (2016) "Traditional and syndromic surveillance of infectious diseases and pathogens" *Int. J. Infect. Dis*
2. Abe, Tago, Yokoyama et al. (2023)
3. "Explainable AI for estimating pathogenicity of genetic variants using large-scale knowledge graphs" *Cancers*
4. Abràmoff, Tarver, Loyo-Berrios et al. (2023) "Considerations for addressing bias in artificial intelligence for health equity" *NPJ Digit. Med*
5. Afshinnekoo, Meydan, Chowdhury et al. (2015) "Geospatial resolution of human and bacterial diversity with City-scale metagenomics" *Cell Systems*
6. Alcolea-Medina, Alder, Snell et al. (2024) "Unified metagenomic method for rapid detection of microorganisms in clinical samples" *Commun. Med*
7. Ali, Abuhmed, El-Sappagh et al. (2023) "Explainable artificial intelligence (XAI): what we know and what is left to attain trustworthy artificial intelligence" *Inf. Fusion*
8. Alley, Khimulya, Biswas et al. (2019) "Unified rational protein engineering with sequence-based deep representation learning" *Nat. Methods*
9. Alwakeel (2025) "AI-assisted real-time monitoring of infectious diseases in urban areas" *Mathematics*
10. Apollon, Kamaraj, Vidales-Contreras et al. (2022) "A beginner's guide to different types of microscopes"
11. Kamaraj, Thirumurugan, Dhanabalan et al.
12. Asmare, Erkihun (2023) "Recent application of DNA microarray techniques to diagnose infectious disease" *PLMI*
13. Aswini, Saranya, Gayathri et al. (2025) "Revolutionizing infectious disease surveillance: multi-omics technologies and AI-driven integration" *Decoding Infect Trans*
14. Auslander, Gussow, Koonin (2021) "Incorporating machine learning into established bioinformatics frameworks" *Int. J. Mol. Sci*
15. Barredo Arrieta, Díaz-Rodríguez, Del Ser et al. (2020) "Explainable artificial intelligence (XAI): concepts, taxonomies, opportunities and challenges toward responsible AI" *Inf. Fusion*
16. Bavinger, Shantha, Yeh (2020) "Ebola, COVID-19 and emerging infectious disease: lessons learned and future preparedness" *Curr. Opin. Ophthalmol*
17. Bengio, Simard, Frasconi (1994) "Learning long-term dependencies with gradient descent is difficult" *IEEE Trans. Neural Netw*
18. Bigness, Loinaz, Patel et al. (2022) "Integrating longrange regulatory interactions to predict gene expression using graph convolutional networks" *J. Comput. Biol*
19. Bileschi, Belanger, Bryant et al. (2021) "Ultrastructural evidence for vertical transmission of SARS-CoV-2" *Int. J. Infect. Dis*
20. Bullock, Goldsmith, Miller (2022) "Detection and identification of coronaviruses in human tissues using electron microscopy" *Microsc. Res. Tech*
21. Calvino, Peconi, Strafella et al. (2024) "Federated learning: breaking down barriers in global genomic research" *Genes*
22. Camargo, Nayfach, Chen et al.
23. Img "VR v4: an expanded database of uncultivated virus genomes within a framework of extensive functional, taxonomic, and ecological metadata" *Nucleic Acids Res*
24. Cruz, Durán Barrón, Morales Lozada et al.
25. Sánchez, Deloya Brito, García Colín (2025) "Detection of SARS-CoV-2 using the Abbott™ PANBIO™ COVID-19 SELF-TEST rapid test in patients seen at INER" *Biomedicine*
26. Capponi, Wang, Navarro et al. (2021) "AI-driven prediction of SARS-CoV-2 variant binding trends from atomistic simulations" *Eur. Phys. J. E Soft Matter*
27. Chadaga, Prabhu, Sampathila et al. (2023) "Application of artificial intelligence techniques for monkeypox: a systematic review" *Diagnostics (Basel)*
28. Chandra, Tünnermann, Löfstedt et al. (2023) "Transformer-based deep learning for predicting protein properties in the life sciences"
29. Chen, Clayton, Novak et al. (2023) "Human-Centered design to address biases in artificial intelligence" *J. Med. Internet Res*
30. Chen, Dong, Ge et al. (2013) "The history and advances of reversible terminators used in new generations of sequencing technology" *Genomics Proteomics Bioinformatics*
31. Chen, Grim, Ramachandran et al. (2024) "Advancing metagenomeassembled genome-based pathogen identification: unraveling the power of long-read assembly algorithms in Oxford nanopore sequencing" *Microbiol. Spect*
32. Chen, Xu (2023) "Application of nanopore sequencing in the diagnosis and treatment of pulmonary infections" *Mol. Diagn. Ther*
33. Cheohen, Gomes, Machado Schelk et al. (2025) "CNN-LSTM hybrid model for AI-driven prediction of COVID-19 severity from spike sequences and clinical data" *arXiv*
34. Chiu, Miller (2019) "Clinical metagenomics" *Nat. Rev. Genet*
35. Choi, Lee (2023) "Transformer architecture and attention mechanisms in genome data analysis: a comprehensive review" *Biology*
36. Choi, Pyenson, Liebig et al. (2022) "Integrating artificial intelligence in healthcare: applications, challenges, and future directions" *Artif. Life Robot*
37. Chourasia, Lonkar, Ali et al. (2023) "Artificial intelligence in public health: the potential of epidemic early warning systems" *J. Int. Med. Res*
38. Cramer, Ray, Lopez et al. (2022) "Evaluation of individual and ensemble probabilistic forecasts of COVID-19 mortality in the United States" *Proc. Natl. Acad. Sci. USA*
39. Curry (2003) "Electron microscopy and the investigation of new infectious diseases" *Int. J. Infect. Dis*
40. Silva, Da, Kohl et al. (2023) "Clinical and laboratory diagnosis of monkeypox (mpox): current status and future directions iScience"
41. Dantas, Da Sabino Silva, Cordeiro et al. (2024) "A comprehensive review of model compression techniques in machine learning" *Appl. Intell*
42. Das, Floch, Houhou et al. (2015) "Viruses detected by systematic multiplex polymerase chain reaction in adults with suspected community-acquired pneumonia attending emergency departments in France" *Clin. Microbiol. Infect*
43. De Coster, De Rijk, De Roeck et al. (2019) "Structural variants identified by Oxford nanopore PromethION sequencing of the human genome" *Genome Res*
44. De Olazarra, Wang (2023) "Advances in point-of-care genetic testing for personalized medicine applications" *Biomicrofluidics*
45. De Souza, Azevedo, De Souza et al. (2023) "New proposal of viral genome representation applied in the classification of SARS-CoV-2 with deep learning" *BMC Bioinformatics*
46. De Vries, Brown, Fischer et al. (2021) "Benchmark of thirteen bioinformatic pipelines for metagenomic virus diagnostics using datasets from clinical samples" *J. Clin. Virol*
47. Deif, Solyman, Kamarposhti et al. (2021) "A deep bidirectional recurrent neural network for identification of SARS-CoV-2 from viral genome sequences" *MBE*
48. Dhaarani, Reddy (2025) "Progressing microbial genomics: artificial intelligence and deep learning driven advances in genome analysis and therapeutics" *Intell. Based Med*
49. Doron, Melamed, Ofir et al. (2018) "Systematic discovery of antiphage defense systems in the microbial pangenome" *Science*
50. Dutilh, Reyes, Hall et al. (2017) "Editorial: virus discovery by metagenomics: the (Im)possibilities" *Front. Microbiol*
51. Edwards, Mcnair, Faust et al. (2016) "Computational approaches to predict bacteriophage-host relationships" *FEMS Microbiol. Rev*
52. Elbehiry, Abalkhail (1991) "Metagenomic next-generation sequencing in infectious diseases: clinical applications, translational challenges, and future directions" *Diagnostics*
53. Eren, Taktakoğlu, Pirim (2022) "DNA sequencing methods: from past to present" *Eurasian J Med*
54. Flores-Coronado, Alanis-Valdez, Herrera-Saldivar et al. (2025) "Awareness of the dual-use dilemma in scientific research: reflections and challenges to Latin America" *Front. Bioeng. Biotechnol*
55. Gers, Schmidhuber, Cummins (2000) "Learning to forget: continual prediction with LSTM" *Neural Comput*
56. Gigante, Korber, Seabolt et al. (2022) "Multiple lineages of monkeypox virus detected in the United States, 2021-2022" *Science*
57. Giuste, Shi, Zhu et al. (2023) "Explainable artificial intelligence methods in combating pandemics: a systematic review" *IEEE Rev. Biomed. Eng*
58. Goldsmith, Miller (2009) "Modern uses of Electron microscopy for detection of viruses" *Clin. Microbiol. Rev*
59. González, Fullaondo, Odriozola (2025) "Why are long-read sequencing methods revolutionizing microbiome analysis?" *Microorganisms*
60. Goodwin, Mcpherson, Mccombie (2016) "Coming of age: ten years of next-generation sequencing technologies" *Nat. Rev. Genet*
61. Graves (2012) "Supervised sequence labelling with recurrent neural networks"
62. Greener, Kandathil, Moffat et al. (2022) "A guide to machine learning for biologists" *Nat. Rev. Mol. Cell Biol*
63. Greninger (2018) "A decade of RNA virus metagenomics is (not) enough" *Virus Res*
64. Han, He, Li et al. (2024) "Unlocking the potential of metagenomics with the PacBio high-Fidelity sequencing technology" *Microorganisms*
65. Han, Song, Pierson et al. (2023) "Emerging infectious diseases are virulent viruses-are we prepared? An overview" *Microorganisms*
66. Hanna, Pantanowitz, Dash et al. (2025) "Future of artificial intelligence-machine learning trends in pathology and medicine" *Mod. Pathol*
67. Harismendy, Ng, Strausberg et al. (2009) "Evaluation of next generation sequencing platforms for population targeted sequencing studies" *Genome Biol*
68. Haug, Drazen (2023) "Artificial intelligence and machine learning in clinical medicine" *N. Engl. J. Med*
69. Heather, Chain (2016) "The sequence of sequencers: the history of sequencing DNA" *Genomics*
70. Hie, Zhong, Berger et al. (2021) "Learning the language of viral evolution and escape" *Science*
71. Hirabayashi, Mercado, Hull et al. (2024) "Comparison of diagnostic accuracy of rapid antigen tests for COVID-19 compared to the viral genetic test in adults: a systematic review and meta-analysis" *JBI Evid Synth*
72. Huang, Jennison, Whiley et al. (2007) "Currently used nucleic acid amplification tests for the detection of viruses and atypicals in acute respiratory infections" *J. Clin. Virol*
73. Iwata (2020) "PCR detection and new therapies for COVID-19" *J Periodontal Implant Sci*
74. Ji, Zhou, Liu et al. (2021) "DNABERT: pre-trained bidirectional encoder representations from transformers model for DNA-language in genome" *Bioinformatics*
75. Johnson, Jamrozik, Ramachandran et al. (2025) "Clinical metagenomics: ethical issues" *J. Med. Microbiol*
76. Joseph (2025) "Algorithmic bias in public health AI: a silent threat to equity in lowresource settings" *Front. Public Health*
77. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
78. Jurtz, Johansen, Nielsen et al. (2017) "An introduction to deep learning on biological sequence data: examples and solutions" *Bioinformatics*
79. Kaur, Butt (2025) "AI-driven epidemic intelligence: the future of outbreak detection and response" *Front. Artif. Intell*
80. Khan, Rathod, Gupta et al. (2024) "Evolution and impact of nucleic acid amplification test (NAAT) for diagnosis of coronavirus disease" *Anal. Chem*
81. Kozich, Westcott, Baxter et al. (2013) "Development of a dual-index sequencing strategy and curation pipeline for Analyzing amplicon sequence data on the MiSeq Illumina sequencing platform" *Appl. Environ. Microbiol*
82. Krivonos, Fedorov, Konanov et al. (2025) "Pike: OTU-level analysis for Oxford nanopore amplicon metagenomics" *Int. J. Mol. Sci*
83. Kuo, Ying (2023) "An unsupervised deep learning framework for anomaly detection" in Intelligent information and database systems"
84. Langer, Amaral, Baudement et al. (2025) "Empowering bioinformatics communities with Nextflow and NF-core"
85. *Genome Biol*
86. Le Piane, Vozza, Baldoni et al. (2024) "Integrating highperformance computing, machine learning, data management workflows, and infrastructures for multiscale simulations and nanomaterials technologies" *Beilstein J. Nanotechnol*
87. Lee (2023) "Deep learning techniques with genomic data in Cancer prognosis: a comprehensive review of the 2021-2023 literature" *Biology (Basel)*
88. Lee (2023) "The principles and applications of high-throughput sequencing technologies" *Dev Reprod*
89. Lee, Jansen, Sanderson et al. (2023) "Public health emergency preparedness for infectious disease emergencies: a scoping review of recent evidence" *BMC Public Health*
90. Leland, Ginocchio (2007) "Role of cell culture for virus detection in the age of technology" *Clin. Microbiol. Rev*
91. Li, Decost, Choudhary et al. (2023) "A critical examination of robustness and generalizability of machine learning prediction of materials properties" *NPJ Comput. Mater*
92. Li, Wang, Li et al. (2021) "A deep learning-based method for identification of bacteriophage-host interaction" *IEEE/ACM Trans. Comput. Biol. Bioinform*
93. Li, Zhao, Wilkins et al. (2010) "Real-time PCR assays for the specific detection of monkeypox virus west African and Congo Basin strain DNA" *J. Virol. Methods*
94. Linares, Pérez-Tanoira, Carrero et al. (2020) "Panbio antigen rapid test is reliable to diagnose SARS-CoV-2 infection in the first 7 days after the onset of symptoms" *J. Clin. Virol*
95. Liu (2025) "Bracing the artificial intelligence technology in viral infectious disease control" *Inf. Med*
96. Liu, Han, Zhou et al. (2021) "Comparative performance of the GenoLab M and NovaSeq 6000 sequencing platforms for transcriptome and LncRNA analysis" *BMC Genomics*
97. Liu, Li, Wang et al. (2020) "Application of deep learning in genomics" *Sci. China Life Sci*
98. Logares, Sunagawa, Salazar et al. (2014) "Metagenomic 16S rDNA Illumina tags are a powerful alternative to amplicon sequencing to explore diversity and structure of microbial communities" *Environ. Microbiol*
99. Lou, Liang, Cao et al. (2024) "Predictive evolutionary modelling for influenza virus by site-based dynamics of mutations" *Nat. Commun*
100. Louten (2016) "Detection and diagnosis of viral infections. Essential Human Virol"
101. Lu, Giordano, Ning (2016) "Oxford nanopore MinION sequencing and genome assembly" *Genomics Proteomics Bioinformatics*
102. Lundberg, Lee (2017) "A Unified Approach to Interpreting Model Predictions"
103. Luvira, Leaungwutiwong, Thippornchai et al. (2022) "False positivity of anti-SARS-CoV-2 antibodies in patients with acute tropical diseases in Thailand" *Trop Med Infect Dis*
104. Marić, Križanović, Riondet et al. (2024) "Comparative analysis of metagenomic classifiers for long-read sequencing datasets" *BMC Bioinformatics*
105. Martínez, Del Soto-Río, Gutiérrez et al. (2014) "DNA microarray for detection of gastrointestinal viruses" *J. Clin. Microbiol*
106. Martinez-Martin (2019) "Privacy and ethical challenges in nextgeneration sequencing" *Expert Rev. Precis. Med. Drug Dev*
107. Matsunaga, Tsumoto, Miller et al. (2013) "Accelerating antibody discovery and optimization with high-throughput experimentation and machine learning" *J. Biomed. Sci*
108. Ming, Chen, Wang et al. (2023) "HostNet: improved sequence representation in deep neural networks for virus-host prediction" *BMC Bioinformatics*
109. Mokili, Rohwer, Dutilh (2012) "Metagenomics and future perspectives in virus discovery" *Curr. Opin. Virol*
110. Morgan, Darling, Eisen (2010) "Metagenomic sequencing of an in vitro-simulated microbial community" *PLoS One*
111. Moustakli, Christopoulos, Potiris et al. (2025) "Long-read sequencing and structural variant detection: unlocking the hidden genome in rare genetic disorders" *Diagnostics (Basel)*
112. Msomi, Levy, Matteson et al. (2025)
113. "Wastewater-integrated pathogen surveillance dashboards enable real-time, transparent, and interpretable public health risk assessment and dissemination" *PLOS Glob Public Health*
114. Mswahili, Jeong (2024) "Transformer-based models for chemical SMILES representation: a comprehensive literature review" *Heliyon*
115. Nasir, Aamir, Kanji et al. (2023) "Tracking SARS-CoV-2 variants through pandemic waves using RT-PCR testing in low-resource settings" *PLOS Glob Public Health*
116. Nayfach, Páez-Espino, Call et al. (2021) "Metagenomic compendium of 189,680 DNA viruses from the human gut microbiome" *Nat. Microbiol*
117. Nazer, Zatarah, Waldrip et al. (2023) "Bias in artificial intelligence algorithms and recommendations for mitigation" *PLOS Digit Health*
118. Ni, Huang, Zhang et al. (2019) "DeepSignal: detecting DNA methylation state from nanopore sequencing reads using deep-learning" *Bioinformatics*
119. Oehler, Wright, Stark et al. (2023) "The application of long-read sequencing in clinical settings" *Hum. Genomics*
120. Pasolli, Asnicar, Manara et al. (2019) "Extensive unexplored human microbiome diversity revealed by over 150,000 genomes from metagenomes spanning age, geography, and lifestyle" *Cell*
121. Pavia, Plummer (2021) "The evolution of rapid antigen detection systems and their application for COVID-19 and other serious respiratory infectious diseases" *J. Microbiol. Immunol. Infect*
122. Peters, Vetter, Guitart et al. (2020) "Understanding the emerging coronavirus: what it means for health security and infection prevention" *J. Hosp. Infect*
123. Pigott, Golding, Mylne et al. (2014) "Mapping the zoonotic niche of Ebola virus disease in Africa"
124. Pita-Galeana, Ruhle, López-Vázquez et al. (2025) "Computational metagenomics: state of the art" *Int. J. Mol. Sci*
125. Rampelli, Soverini, D' Amico et al. (2020) "Shotgun metagenomics of gut microbiota in humans with up to extreme longevity and the increasing role of xenobiotic degradation"
126. Reich, Mcgowan, Yamana et al. (2019) "Accuracy of real-time multi-model ensemble forecasts for seasonal influenza in the U.S" *PLoS Comput. Biol*
127. Reintjes, Zanuzdana (2009) "Outbreak investigations. Mod"
128. Ren, Song, Deng et al. (2020) "Identifying viruses from metagenomic data using deep learning" *Quant Biol*
129. Richert-Pöggeler, Franzke, Hipp et al. (2019) "Electron microscopy methods for virus diagnosis and high resolution analysis of viruses" *Front. Microbiol*
130. Rives, Meier, Sercu et al. (2021) "Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences" *Proc. Natl. Acad. Sci. USA*
131. Rooney, Raphenya, Melano et al. (2022) "Performance characteristics of next-generation sequencing for the detection of antimicrobial resistance determinants in Escherichia coli genomes and metagenomes"
132. Rosenboom, Scheithauer, Friedrich et al. (2022) "Wochenende -modular and flexible alignment-based shotgun metagenome analysis" *BMC Genomics*
133. Roux, Adriaenssens, Dutilh et al. (2019) "Minimum information about an uncultivated virus genome (MIUViG)" *Nat. Biotechnol*
134. Roux, Matthijnssens, Dutilh (2021) "Metagenomics in virology. Encyclopedia Virol"
135. Rudin, Rumelhart, Hinton et al. (1986) "Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead" *Nat Mach Intell*
136. Samek, Wiegand, Müller (2017) "Explainable artificial intelligence: understanding, visualizing and interpreting deep learning models"
137. Santiago-Rodriguez, Hollister (2022) "Unraveling the viral dark matter through viral metagenomics" *Front. Immunol*
138. Saravanan, Panigrahi, Kumar et al. (2022) "Role of genomics in combating COVID-19 pandemic" *Gene*
139. Schirmer, D' Amore, Ijaz et al. (2016) "Metagenomic sequencing of mpox virus clade Ib lesions identifies possible bacterial and viral co-infections in hospitalized patients in eastern DRC" *BMC Bioinformatics*
140. Schulz, Roux, Paez-Espino et al. (2020) "Improved protein structure prediction using potentials from deep learning" *Nature*
141. Sereika, Kirkegaard, Karst et al. (2022) "Oxford nanopore R10.4 long-read sequencing enables the generation of near-finished bacterial genomes from pure cultures and metagenomes without short-read or reference polishing" *Nat. Methods*
142. Sevim, Lee, Egan et al. (2019) "Shotgun metagenome data of a defined mock community using Oxford nanopore, PacBio and Illumina technologies" *Sci Data*
143. "Nanopore sequencing and the Shasta toolkit enable efficient de novo assembly of eleven human genomes" *Nat. Biotechnol*
144. Shkoporov, Stockdale, Lavelle et al. (2016) "DeepChrome: deep-learning for predicting gene expression from histone modifications" *Nat. Microbiol*
145. Slavov (2025) "Routine detection of viruses through metagenomics: where do we stand?" *Am. J. Trop. Med. Hyg*
146. Soenksen, Ma, Zeng et al. (2022) "Integrated multimodal artificial intelligence framework for healthcare applications" *NPJ Digit Med*
147. Song, Ma, Xu et al. (2021) "Role of artificial intelligence in early diagnosis and treatment of infectious diseases" *BMC Med. Genet*
148. Takemae, Kuba, Oba et al. (2024) "Direct genome sequencing of respiratory viruses from low viral load clinical specimens using the target capture sequencing technology" *Microbiol. Spect*
149. Tampuu, Bzhalava, Dillner et al. (2019) "ViraMiner: deep learning on raw DNA sequences for identifying viral genomes in human samples" *PLoS One*
150. Thomas, Gilbert, Meyer (2012) "Metagenomics -a guide from sampling to data analysis" *Microb Inform Exp*
151. Tisza, Buck (2021) "A catalog of tens of thousands of viruses from human metagenomes reveals hidden associations with chronic diseases" *Proc. Natl. Acad. Sci. USA*
152. Travers, Chin, Rank et al. (2010) "A flexible and efficient template format for circular consensus sequencing and SNP detection" *Nucleic Acids Res*
153. Unsal, Atas, Albayrak et al. (2022) "Learning functional properties of proteins with language models" *Nat. Mach. Intell*
154. Vashisht, Vashisht, Mondal et al. (2023) "Genomics for emerging pathogen identification and monitoring: prospects and obstacles" *Biomed Informatics*
155. Villanueva-Miranda, Xiao, Xie (2025) "Artificial intelligence in early warning systems for infectious disease surveillance: a systematic review" *Front. Public Health*
156. Wagner, Gresham, Dato (2006) "Case detection, outbreak detection, and outbreak characterization" *Handb. Biosurveillance*
157. Wang, Chen, Xiang et al. (2023) "Recent advances in immunoassay technologies for the detection of human coronavirus infections" *Front. Cell. Infect. Microbiol*
158. Wang, Coscoy, Zylberberg et al. (2002) "Microarray-based detection and genotyping of viral pathogens" *Proc. Natl. Acad. Sci. USA*
159. Wang, Urisman, Liu et al. (2003) "Viral discovery and sequence recovery using DNA microarrays" *PLoS Biol*
160. Watkins, Eagleson, Hall et al. (2006) "Approaches to the evaluation of outbreak detection methods" *BMC Public Health*
161. Wen, Tang (2025) "Single-cell omics sequencing technologies: the long-read generation" *Trends Genet*
162. Wenger, Peluso, Rowell et al. (2019) "Accurate circular consensus long-read sequencing improves variant detection and assembly of a human genome" *Nat. Biotechnol*
163. Wilder-Smith (2021) "COVID-19 in comparison with other emerging viral diseases: risk of geographic spread via travel" *Trop. Dis. Travel Med. Vaccines*
164. Willmington, Belardi, Murante et al. (2022) "The contribution of benchmarking to quality improvement in healthcare. A systematic literature review" *BMC Health Serv. Res*
165. Wu, Fang, Tan et al. (2021) "DeePhage: distinguishing virulent and temperate phage-derived sequences in metavirome data with a deep learning approach" *Gigascience*
166. Wu, Wijesekara, Piedade et al. (2024) "Benchmarking bioinformatic virus identification tools using realworld metagenomic data across biomes" *Genome Biol*
167. Xia, Li, Wu et al. (2023) "Strategies and tools in illumina and nanopore-integrated metagenomic analysis of microbiome data"
168. Yagin, Cicek, Alkhateeb et al. (2023) "Explainable artificial intelligence model for identifying COVID-19 gene biomarkers" *Comput. Biol. Med*
169. Yang, Shan, Liu et al. (2024) "VirID: beyond virus discovery-an integrated platform for comprehensive RNA virus characterization" *Mol. Biol. Evol*
170. Ye, Pandey, Bawden et al. (2024) "Integrating artificial intelligence with mechanistic epidemiological modeling: a scoping review of opportunities and challenges" *BMJ Glob. Health*
171. Yurdem, Kuzlu, Gullu et al. (2024) "Federated learning: overview, strategies, applications, tools and future directions" *Heliyon*
172. Zhang, Liu, Li et al. (2025) "Adaptation of single molecule real time (SMRT) sequence technology for hepatitis C virus genome sequencing and identification of resistance-associated substitutions" *Virology*
173. Zhao, Li, Peng et al. (2023) "A structured multi-head attention prediction method based on heterogeneous financial data" *PeerJ Comput Sci*
174. Zwiers, Grobbee, Uijl et al. (2024) "Federated learning as a smart tool for research on infectious diseases" *BMC Infect. Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12172453&blobtype=pdf | # Correction for Li et al., "Mitochondria-mediated ferroptosis contributes to the inflammatory responses of bovine viral diarrhea virus (BVDV) in vitro"
Zhijun Li, Bao Zhao, Ying Zhang, Wenqi Fan, Qinghong Xue, Xiwen Chen, Jingyu Wang, Xuefeng Qi
## Abstract
Page 8, lines 46 and 48: "(Fig. 6C andD)" should read "(Fig. 5C andD). " Page 8, line 53: "(Fig. 6E andF)" should read "(Fig. 5E andF). " Page 10, line 4: "(Fig. 6H andI)" should read "(Fig. 5H andI). " Page 10, line 5: "(Fig. 6J)" should read "(Fig. 5J). " Page 10, line 14: "(Fig. 7A)" should read "(Fig. 6A), " "(Fig. 7B)" should read "(Fig. 6B), " and "(Fig. 7C)" should read "(Fig. 6C). "Page 10, lines 22, 23, and 25: "(Fig. 5A)" should read "(Fig. 7A). " Page 10, line 30: "(Fig. 5B)" should read "(Fig. 7B). " Page 10, lines 31, 35, and 37: "(Fig. 5C)" should read "(Fig. 7C). " Page 12, line 2: "(Fig. 5D)" should read "(Fig. 7D). " Page 12, lines 5 and 8: "(Fig. 5E)" should read "(Fig. 7E). " Page 15, lines 3 to 8: "Although knockdown of Parkin had no significant effects on the expression of IFN-β (Fig. 8F), IL-18 (Fig. 8G), and IL-1β (Fig. 8H) in CP BVDV-infected cells compared to control siRNA-transfected cells, an increased level of cytokines indicated was detected in NCP BVDV-infected cells pretransfected with siParkin compared to control siRNA-transfected cells (Fig. 8F through H). " should read "Although knockdown of Parkin significantly decreased the expression of IFN-β (Fig. 8G), IL-18 (Fig. 8H), and IL-1β (Fig. 8I) in CP BVDV-infected cells compared to control siRNA-transfected cells, an increased level of the indicated cytokines was detected in NCP BVDV-infected cells pretransfected with siParkin compared to control siRNA-transfected cells (Fig. 8G through I). "We apologize for these errors. This correction does not affect the conclusions of the study.
AUTHOR ORCIDs |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12747146&blobtype=pdf | # Communication and involvement of students and young researchers in animal science: an European perspective
Serban Morosan, Federica Riva, Giuliano Grignaschi, Sabine Chourbaji
## Introduction
The general public's opinion is controversial and divided on the use of animals in research. In part, this could be due to a lack of openness and transparency about why and how animals are used in research, and a lack of awareness of the stringent rules and regulations under which such experiments are performed. Universities (as well as other research players) could be more vocal about the benefits which animal use brings to both basic and applied research.
The Concordat on Openness, launched in May 2014 by Understanding Animal Research, represents an important initiative aimed at enhancing transparency and accountability in the use of animals for research purposes (Jarrett, 2016). Over the years, the openness movement has gained momentum, with many organizations and researchers adopting similar practices across various fields. This has led to greater awareness and constructive dialog surrounding the ethical issues related to animal use in research.
As LERU noted in 2020 in the publication on Good Practice in Communicating Animal Research at Universities, developing an open and transparent approach to animal research within a university can help increase the awareness of the public about animal research (Morosan and Dinner, 2020). Students and young scientists are often very willing to communicate their research, but can lack confidence and a nuanced approach in communicating this sensitive topic.
Effective engagement and communication with students and early-career researchers are essential to the long-term success and sustainability of the animal science, veterinary, and biomedical profession. As societal expectations around animal welfare, scientific transparency, and interprofessional collaboration evolve, so too must the educational strategies that shape the next generation of animal scientists and veterinarians. The challenge lies not only in transmitting technical knowledge, but also in fostering motivation, ethical awareness, and a sense of professional identity through meaningful involvement and communication. This is also true for interacting with so-called activist groups, which may enrich the impact of communication.
Studies in veterinary and animal science education highlight a recurring pattern: students often enter these fields with strong intrinsic motivations, particularly a desire to care for animals, yet they may lack the practical experience and broader contextual understanding required for success (Whitaker et al., 2020). An Irish study found that while veterinary nursing students are deeply committed to animal care, many begin their training with limited animal-handling experience and uneven exposure to veterinary practice. This gap between motivation and preparedness underscores the importance of early, supportive, and realistic engagement strategies to build both skill
## Implications
Students and young researchers in animal science, veterinary, and biomedical professions encounter motivation and communication problems; it is important to look for new strategies to communicate and involve students and young researchers in animal science, veterinary and biomedical professions; LERU workshops in 2022 and 2024 proposed new methods for student and young researcher involvement by fostering reflection, peer exchange, and engagement with stakeholders from industry, academia, and patient organizations. and confidence and avoid the abandonment of their profession (Whitaker et al., 2020).
Moreover the attitude of students to animals depends on different factors: demographic data of the student (age; gender; program; year; meat eater, vegetarian, or vegan; having a current animal; having an animal as a child; previous study; employed work outside university studies) and category of animals (pet, pest, or livestock) (Hazel et al., 2011). The Hazel and colleagues study demonstrated that educational interventions (specifically, courses on animal welfare and ethics) can positively shift students' attitudes toward animals, particularly those in less emotive categories such as livestock or pests (Hazel et al., 2011). Notably, the study found that female students and those with prior animal experience showed higher levels of empathy, suggesting that student background significantly influences how information is received and internalized (Hazel et al., 2011). These findings point to the need for educators to communicate not only facts but values, using inclusive approaches that resonate across a diverse student body (Hazel et al., 2011).
So it is very important to outline the ideal profile for the teachers in veterinary, animal science, and biomedical fields that could better involve students, transmit the knowledge, and form highly qualified professional figures. An American study analyzed the features of the top teachers in animal science, exploring what makes teaching in animal science truly impactful (Whitaker et al., 2020). The authors revealed that award-winning educators go beyond content delivery; they engage students by building rapport, setting high expectations, incorporating humor, and making material relevant and applicable (Dunne et al., 2018).
These teachers prioritize interaction, actively involve students in the learning process, and adjust their communication to meet learners' needs-all of which are critical for fostering deeper involvement and professional development (Dunne et al., 2018).
Together, these studies suggest that involvement and communication must be intentional, empathetic, and contextaware. For students and young researchers to thrive, they must be engaged as active participants in their own learning and development, supported by instructors who recognize the social, emotional, and experiential dimensions of education.
In late 2021, the LERU Protection of Animals for Scientific Purposes (ANIM) group proposed to hold a student workshop on effective communications, including the use of animals in scientific research. The University of Heidelberg's Communication and Marketing Department kindly agreed to organize the event, which took place in 2022, with the participation of 19 students coming from different EU Universities (10 LERU universities represented: 2 from Heidelberg, 17 from other LERU universities).
The aim of this 3-d event was to help early-stage researchers better explain animal science research to a nonspecialist audience. Experts from different sectors informed the participants about the political framework surrounding transparency and the benefits that transparency and communication can bring.
The students heard the experiences of both industry and academia, and considered the effectiveness of external communication tools and how they could better explain their own research. The students also learned about regulation, communication skills, ethics, and new methodological approaches.
When meeting in person, we found that despite differing opinions among student groups regarding animal research, they shared common ground on key issues related to animal research.
Effective communication of students regarding animal research involves addressing the audience's core questions and concerns. Responding to public concerns requires a solid understanding of scientific principles, research design, speciesspecific animal welfare and care practices, ethical considerations, and regulatory frameworks. Since individuals within a research institution may possess expertise in some, but not all, of these areas, a team-based approach is often the most effective way to communicate this information to the public. Media training for all team members, provided by communication experts, is a crucial element of success. Engaging respectfully with the public differs significantly from communicating with fellow scientists. Whether we are students, scientists, or veterinarians, we are much more likely to succeed in public interactions after receiving professional guidance on content, vocabulary, presentation techniques, and, in some cases, on developing patience and managing emotional responses (Figure 1).
Due to the great success of the Heidelberg workshop, LERU decided to give another opportunity to early-stage researchers to share good practice in the communication of animal research, and in September 2024, a second workshop was held at the University of Milano. Experts from patient and industry associations, policy-makers, journalists, and university teachers discussed with the 22 participants (coming from 15 different EU Universities) about the benefits that efficacious communication and increased transparency on research using animals can bring, and how this can be done in an effective manner.
At the end of the workshop, participants were asked to answer some questions. When they were asked, "What was the most valuable part of the workshop for you?", the most frequent answer was that the most valuable part was definitely the interaction with the other workshop participants. It was great to meet a vast variety of fellow young researchers encountering similar problems in communicating their research. Additionally, it was very valuable for the participant to reflect on my own morals regarding animal use in research and in general. This was triggered by the dialog with experts and the other participants alike. When you are certain in your own view that the use of animals in research is justified and reasonable, you can communicate and educate other people regarding animal experimentation much more effectively. The participants also recognized that the actual core purpose of the workshop, communicating animal research and the accompanying themes such as transparency and dialogue, was invaluable. This summer school experience offered a unique chance to focus on the very real challenges of communicating animal research.
The emphasis on transparency in communicating the research results was a take on a message that is so simple and applicable, yet often not widely known or implemented.
Another valuable aspect highlighted by the participants was the possibility to discuss the relevance of animal experimentation with the patients and their families, but also have the opportunity to approach the topic at different levels (e.g. industry's approach, politics, etc.).
Another question was to report any additional comments or suggestions. The suggestions for the improvement of such an event consisted of the insertion of a topic on how to communicate and transfer to participant PI, professors, and universities all the knowledge learn during the LERU Summer school.
After this 2 workshops on effective communication, the young participants revealed that they learned some very important lessons: (a) the principles of good science form the foundation for selecting appropriate models and methods, as well as understanding their limitations; (b) an open dialog is necessary, recognizing that different research fields require different methodologies; (c) the importance of the respect for all kind of public audience both in written and spoken communication; (d) an effective way to communicate about animal research is essential to address the audience's key questions and concerns:
-Why is the study important? What are the intended benefits, and are those benefits realistically achievable? -Why is the use of animals Why can't these questions be answered using alternative methods?
-What is the experience of the animals? What happens to them before, during, and after the study? -What ethical standards guide the research? How was the use of animals ethically justified? -Who approves the research? What individuals or committees are responsible for evaluating and authorizing the study?
The participants evaluated the LERU Summer School as a meaningful experience that allowed them to enrich personally and professionally.
We could conclude that such events represent valuable instruments to reach out to especially young scientists. I had never had the chance to interact with researchers from different institutions and actually discuss and compare our research and how we feel about it. It was deeply useful for my own work and professional development, as well as my own personal development in this area. I plan on disseminating this information to my own institution now, and even if they take on only 10% of what was taught, that makes this workshop so worth it.
## References
1. Dunne, Brereton, Duggan et al. (2018) "Motivation and prior animal experience of newly enrolled veterinary nursing students at two Irish third-level institutions" *J. Vet. Med. Educ*
2. Hazel, Signal, Taylor (2011) "Can teaching veterinary and animal-science students about animal welfare affect their attitude toward animals and human-related empathy?" *J. Vet. Med. Educ*
3. Jarrett (2016) "The concordant on openness and its benefits to animal research" *Lab Anim (NY)*
4. Morosan, Dinner (2020) "Good practice in communicating animal research at universities"
5. Whitaker, Lamberson, Smith et al. (2020) "What best animal science teachers do" *Transl Anim Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12526248&blobtype=pdf | # Population Genetic Analysis Reveals Recent Demographic Expansion and Local Differentiation of Areca Palm Velarivirus 1 in Hainan Island
Xiaoqing Niu, Zhongtian Xu, Zhaowei Lin, Qinghua Tang, Zhenguo Du, Fangluan Gao
## Abstract
Areca palm velarivirus 1 (APV1), the causal agent of yellow leaf disease (YLD), poses a serious threat to the economically important areca palm industry in the Hainan Province, China, yet its evolutionary dynamics remain poorly understood. Here, we performed a large-scale molecular survey by sequencing the coat protein (CP) gene from 364 APV1infected samples collected across major cultivation regions of Hainan. Population genetic analyses revealed extremely high haplotype diversity (H d = 0.997) but very low nucleotide diversity (π = 0.017). Neutrality tests (Tajima's D = -2.266; Fu's F S = -23.697) and a unimodal mismatch distribution supported a scenario of recent demographic expansion from a restricted ancestral pool. Evolutionary analyses indicated that the CP gene is subject to strong purifying selection, although eight codons exhibited episodic positive selection, suggesting ongoing viral adaptation. Furthermore, we identified three distinct genetic clusters with significant geographic structuring, indicating that viral dissemination is shaped by local factors. Together, these results reveal a recent explosive invasion of APV1 characterized by rapid island-wide expansion and emerging local differentiation. This work provides novel insights into the evolutionary trajectory of APV1 and establishes a genomic basis for improved surveillance and management of YLD.
## 1. Introduction
The areca palm (Areca catechu L.) is a cash crop of substantial economic importance, cultivated extensively across South and Southeast Asia, the Pacific Islands, and parts of Africa [1]. In China, the Hainan Province, which is located on an island in the South China Sea, dominates production and accounts for approximately 99% of the national output [2]. This industry plays a pivotal role in the regional economy, supporting the livelihoods of nearly five million residents and contributing over 30% of annual household income for local farmers. However, areca palm cultivation faces a severe threat from yellow leaf disease (YLD), a devastating condition first reported in India in 1914 [3]. YLD manifests as progressive leaf yellowing and reduced plant vigor, leading to significant losses in yield and quality [1]. Although its etiology has been debated for decades-with early studies implicating phytoplasmas [4]-mounting evidence now identifies areca palm velarivirus 1 (APV1) as the primary causal agent in Hainan [3,5,6].
APV1, a member of the genus Velarivirus (family Closteroviridae), possesses a positivesense, single-stranded RNA genome containing 11 open reading frames (ORFs) [7]. Its first open reading frame, ORF1a, encodes a large multifunctional polyprotein that includes papain-like protease (P-Pro), methyltransferase (Met), and helicase (Hel) domains. ORF1b, expressed via a -1 ribosomal frameshift from ORF1a, encodes the RNA-dependent RNA polymerase (RdRp). The downstream ORFs encode a small hydrophobic protein (ORF2), a heat shock protein 70 homolog (HSP70h, ORF3), a 64 kDa polypeptide (ORF4), and a protein of unknown function (ORF5). ORF6 and ORF7 encode the major coat protein (CP) and minor coat protein (CPm), respectively, while the functions of proteins from ORFs 8-10 remain uncharacterized [8].
The causal association between APV1 and YLD is substantiated by multiple lines of compelling evidence. APV1 was first identified in symptomatic areca palms from Hainan via small RNA sequencing [3]. Subsequent RT-PCR surveys confirmed its high prevalence in diseased plants across the province, with only minimal detection in asymptomatic counterparts [5,9]. Quantitative analyses further revealed a strong positive correlation between APV1 titers and symptom severity [10]. Moreover, transmission experiments demonstrated that APV1 is efficiently vectored to healthy seedlings by mealybugs (Pseudococcus cryptus and Ferrisia virgata), which subsequently develop typical YLD symptoms [6].
Despite this etiological clarity, significant gaps persist in our understanding of APV1 ′ s epidemiology and evolutionary dynamics. Key unanswered questions include whether the APV1 population in Hainan resulted from long-term local evolution or a recent introduction, and how its spatial spread and adaptive processes are governed. To address these questions, we conducted a comprehensive molecular survey of APV1 across the Hainan Province. By sequencing the CP gene from over 360 infected samples collected from major production areas, we aimed to delineate the virus's genetic diversity, population structure, and evolutionary patterns. This work provides a crucial foundation for understanding APV1 epidemiology and developing genome-informed management strategies for YLD.
## 2. Results
## 2.1. APV1 Exhibits High Haplotype Diversity but Low Nucleotide Diversity
A total of 364 APV1 isolates were randomly collected from areca palm (Areca catechu L.) across 76 towns spanning 18 counties/cities in the Hainan Province between 2020 and 2022 (Table S1). The sampling locations spanned latitudes from 18.34 • N to 19.89 • N and longitudes from 108.86 • E to 110.58 • E, representing the major cultivation areas across the island (Figure 1). Sequence analysis of the CP sequences from these isolates revealed 287 distinct haplotypes, corresponding to an exceptionally high haplotype diversity (H d = 0.997). This indicates a rich variety of sequence variants within the population. Despite this, nucleotide diversity was remarkably low (π = 0.017 ± 0.0003), with the minimum pairwise identity of 95.8% (Table 1). This combination-high haplotype diversity coupled with low nucleotide diversity-is a hallmark of populations that have recently expanded from a limited ancestral pool. To contextualize this finding, we calculated the nucleotide diversity for two other velariviruses, grapevine leafroll-associated virus 7 (GLV7) and little cherry virus 1 (LCV1). Both viruses exhibited significantly higher diversity (π = 0.067 ± 0.008 and 0.206 ± 0.007, respectively), indicating that the low nucleotide diversity observed in APV1 is an unusual feature for this genus and likely reflects its unique evolutionary history in Hainan (Table 1).
## 2.2. Neutrality Tests and Mismatch Distribution Support APV1 Population Expansion
Neutrality tests yielded significantly negative values for both Tajima's D (-2.266, p < 0.01) and Fu's F S (-23.697, p < 0.01) (Table 1). These results are consistent with an excess of low-frequency mutations, a pattern indicative of either population growth or a recent selective sweep. To distinguish between these possibilities, we performed a mismatch distribution analysis. The observed distribution of pairwise nucleotide differences was distinctly unimodal and smooth, fitting well with the expectations of a spatial expansion model (SSD = 0.002, p > 0.05; H rag = 0.005, p > 0.05) (Figure 2). Collectively, while negative neutrality statistics alone can be ambiguous, the unimodal mismatch distribution strongly suggests that demographic expansion is the primary driver shaping the genetic structure of the APV1 population in Hainan.
## 2.3. The APV1 CP Gene Is Under Strong Purifying Selection with Episodic Adaptation
We evaluated selective pressures on the CP gene by calculating the ratio of nonsynonymous to synonymous substitution rates (dN/dS). The overall ratio was substantially less than one, indicating that strong purifying selection is the dominant evolutionary force acting on this gene (Figure 3). This finding is consistent with the critical functional constraints on the CP, which is essential for virion assembly, vector transmission, and host interactions.
To detect potential episodic adaptation, we performed a site-specific selection analysis using CODEML4.7. Likelihood ratio tests revealed that models incorporating positive selection (M2a and M8) provided a significantly better fit to the data than their neutral counterparts (M1a and M7) (p < 0.001). This result indicates the presence of specific codons evolving under positive selection. Under model M8, six codons (positions 9, 53, 87, 121, 149, and 188) were identified as positively selected, with the strongest evidence (posterior probability > 0.95) for these sites (Figure 3, Table 2). Taken together, these results suggest that while the CP gene is overwhelmingly constrained by purifying selection, a few specific sites may undergo positive selection, potentially as a result of adaptive pressures related to host interaction.
## 2.4. APV1 Populations Exhibit Significant Geographic Structuring
To evaluate the spatial distribution of genetic variation, we performed a Discriminant Analysis of Principal Components (DAPC), which resolved three well-supported genetic clusters with distinct geographic distributions (Figure 4). Cluster 1 was widespread across the island, suggesting it represents a broadly disseminated lineage. In contrast, Cluster 2 was more localized, primarily restricted to Ledong, Lingshui, Ding'an, and Qionghai. Cluster 3 was highly geographically constrained, found exclusively in Chengmai (Figures 1 and4). LG, Lingao; LS, Lingshui; QH, Qionghai; QZ, Qiongzhong; SY, Sanya; TC, Tunchang; WC, Wenchang; WN, Wanning; WZS, Wuzhishan. This population subdivision was further supported by fixation index (F ST ) analysis. Although pairwise F ST values among clusters were low in magnitude-a consequence of the overall limited nucleotide diversity of APV1-they were nevertheless statistically significant, indicating restricted gene flow among these regional populations (Table 3). We further tested this phylogeographic structure using association index (AI), parsimony score (PS), and monophyletic clade (MC) statistics (Table 4). Both AI and PS values were significantly lower than expected under a null hypothesis of no geographic association (p < 0.001), confirming a strong correlation between viral phylogeny and geographic origin. Moreover, the MC statistic for each of the three clusters was significant (p < 0.01), indicating that they represent distinct, geographically cohesive lineages that are more phylogenetically related than expected by chance. In summary, these complementary analyses provide robust evidence that the APV1 population in Hainan is not panmictic but is instead characterized by significant geographic differentiation. This structuring is likely shaped by a combination of localized vectordriven transmission, limited vector dispersal, and human-mediated movement of planting materials, which collectively maintain distinct viral subpopulations.
## 3. Discussion
In this study, we generated a high-resolution dataset by sequencing 364 APV1 CP genes from samples collected across Hainan Island (Figure 1, Table S1). This represents a substantial increase in sampling density compared with previous studies, which, despite relying on whole-genome sequencing, included only 20 isolates [8]. The expanded dataset provides a more comprehensive view of APV1 genetic diversity, population structure, and evolutionary dynamics, enabling more robust inferences about its epidemiology and adaptive patterns.
A notable observation from our analysis concerns the origin and demographic history of APV1 in Hainan. The combined patterns of high haplotype diversity, low nucleotide diversity, negative neutrality statistics (Table 1), and a unimodal mismatch distribution (Figure 2) are not consistent with long-term, stable viral endemicity. Instead, these features collectively point toward a scenario in which the APV1 population may have undergone a relatively recent demographic expansion from a limited ancestral pool, likely shaped by a founder effect in which a small number of founding genomes established local populations that subsequently expanded. Comparable demographic signatures have been reported in geographically isolated virus populations, where introduction events followed by rapid growth produce elevated haplotype richness but low overall sequence divergence [11][12][13][14].
Despite the limited genetic diversity expected under a founder effect and the relatively isolated nature of host populations in Hainan, APV1 displays unexpectedly widespread occurrence, a pattern that may in part reflect the wide distribution and ecological plasticity of its mealybug vectors. Nevertheless, this genetic signature of recent expansion presents an apparent paradox when considered alongside historical reports of YLD in Hainan dating back to 1985 [15]. Several non-exclusive hypotheses could reconcile this discrepancy. First, the now-dominant APV1 lineage may have circulated at a very low prevalence for decades before a recent surge in incidence, potentially triggered by intensified cultivation or shifts in vector ecology. Second, this lineage could have competitively displaced older, more divergent APV1 strains that are no longer detectable. Third, YLD symptoms are highly complex and may result from multiple biotic and abiotic factors [1]. Early YLD-like symptoms might have been caused by other pathogens, with APV1 emerging later as the principal etiological agent. Although further data are needed to distinguish among these scenarios, our results are consistent with the interpretation that most present-day APV1 infections derive from a relatively recent common ancestor, underlining the dynamic nature of viral emergence in agroecosystems.
Our analysis of the evolutionary forces shaping the CP gene provides further mechanistic insight. The gene is overwhelmingly constrained by strong purifying selection (Figure 3), reflecting stringent functional requirements for virion assembly, vector transmission, and host interaction. However, superimposed on this highly conserved background, our sitespecific analyses identified discrete codons under positive selection (Table 2, Figure 3). This dual-mode evolution-wherein the structural integrity of the protein is rigorously maintained while specific sites undergo adaptive change-suggests a sophisticated viral strategy. These adaptive "hotspots" may allow APV1 to fine-tune its interface with host defenses or vector components, facilitating persistence and spread without compromising core functions.
The significant geographic structuring of APV1 (Figure 4) adds a crucial layer of nuance to the expansion narrative. Although the overall genetic diversity is low, the three distinct clusters reveal that the virus's spread has not been a simple, uniform wave (Figure 4). Instead, this nascent differentiation points to a dynamic interplay between island-wide expansion and local-level drivers, such as restricted vector dispersal and regional patterns of human-mediated transport of planting materials. This suggests that even as APV1 rapidly colonizes Hainan, it is already beginning to differentiate in response to local selective pressures, a process that may presage the emergence of distinct regional lineages in the future.
From an applied perspective, our findings have critical implications for disease management. The genetic signature of rapid expansion highlights APV1 ′ s high invasive potential within intensively managed agroecosystems. Furthermore, the evidence for positive selection, though limited to a few sites, signals that the virus possesses the capacity to adapt, potentially leading to the evolution of altered virulence, transmissibility, or host range. These realities underscore the urgent need for proactive genomic surveillance to monitor the emergence of new variants and track adaptive changes in real time. Integrating this molecular intelligence with on-farm strategies-such as screening planting material, developing resistant cultivars, and implementing targeted vector control-is essential for devising robust and sustainable solutions to mitigate the devastating impact of YLD.
A primary limitation of this study is its focus on a single gene. While the CP gene is highly informative for population genetic analyses, it represents only a fraction of the viral genome. Consequently, our analysis cannot capture evolutionary dynamics in other functionally important regions, nor can it definitively detect recombination events. Future research employing large-scale, whole-genome sequencing will be crucial for building a more complete picture of APV1 ′ s evolution and adaptation. Nevertheless, the depth of sampling in our study provides a foundational understanding of APV1 population dynamics that can guide these future genomic efforts.
## 4. Materials and Methods
## 4.1. Sample Collection and Processing
APV1-infected areca palm (Areca catechu L.) leaf samples were collected between 2020 and 2022 from multiple cultivation areas across the Hainan Province (Table S1). Symptomatic leaves were excised using sterile scissors and gloves to prevent cross-contamination. Samples were immediately placed in labeled plastic bags, transported on ice to the laboratory, and stored at -80 • C until processing.
## 4.2. Virus Isolates and Genetic Diversity Analysis
To obtain the APV1 CP sequences, total RNAs was extracted using an RNAprep Pure Plant Plus Kit (Polysaccharides and Polyphenolics-rich), and cDNA was synthesized by a FastKing RT Kit (both from TIANGEN, Beijing, China), according to the manufacturer's instructions. RT PCR amplifications were performed in a total volume of 50 µL composed of 25 µL × PCR taq mix, 2.5 µL of forward primer (10 µmol/L, APV1CP_F1: 5 ′ -CCACTCTTCTGGTAGTATCAAGG-3 ′ ), 2.5 µL of reverse primer (10 µmol/L, APV1CP_R1: 5 ′ -CAGAAGCATAAGATTGTGACATTTTTACCG-3 ′ ), 15 µL of ddH 2 O, and 5.0 µL of template cDNA. The thermal profile included an initial denaturation at 94 • C for 3 min, followed by 34 cycles of 94 • C for 30 s, 55 • C for 30 s, and 72 • C for 1 min, with a 10 min extension at 72 • C. Following the completion of PCR, the amplified product was recovered and ligated into the cloning vector PMD-18T. Culture solution was then carried out to identify successful results, after which the solution was sequenced in both directions by Sangon Biological Co., Ltd. (Shanghai, China).
A codon-based alignment of the APV1 CP sequence was performed using the MAFFT algorithm [16], as recommended in PhyloSuite 1.31 [17]. To assess APV1 populations' genetic diversity, haplotype diversity and nucleotide diversity were calculated using DnaSP 6.12.03 [18]. For comparative purposes, the same analysis was performed on the CP genes sequences of two other velariviruses, GLV7 and LCV1 (Tables S2 andS3).
## 4.3. Demographic History of the APV1 Population
All mismatch distribution analyses and parameter calculations were performed using Arlequin 3.5.2.2 [19]. To test recent demographic expansion, we calculated Tajima's D [20] and Fu's F S [21] statistics, where significantly negative values suggested potential population expansion or selective sweeps. The expansion model was further evaluated through Harpending's raggedness index (H rag ) and sum of squared deviations (SSD), with their corresponding p-values [22]. Non-significant p-values (p > 0.05) for both H rag and SSD supported the population expansion hypothesis, while unimodal mismatch distributions provided additional evidence for this demographic scenario.
## 4.4. Detecting Natural Selection
To quantify patterns of natural selection acting on APV1 CP, the ratio of nonsynonymous (dN) to synonymous (dS) substitution rates (ω = dN/dS) was calculated using the CODEML algorithm implemented in EasyCodeML 1.41 [23,24]. This widely used metric for evaluating selective pressures on protein-coding genes was obtained through sitemodel comparisons of four nested model pairs: M0 (one ratio) versus M3 (discrete), M1a (neutral) versus M2a (selection), M7 (beta) versus M8 (beta and ω > 1), and M8 versus M8a. Positively selected amino acid sites in APV1 CP were identified through Bayes Empirical Bayes analysis when likelihood-ratio tests showed statistical significance (p < 0.05) [25].
## 4.5. Testing Population Differentiation
Genetic differentiation among populations was assessed using K ST , Z, and S nn statistics in DnaSP 6.12.03 [18], with significance tested through 1000 permutations of the original dataset. In addition, pairwise F ST values were also computed in Arlequin 3.5 [19] and interpreted as follows: 0.05-0.15 indicated moderate differentiation, 0.15-0.25 represented high differentiation, and values above 0.25 suggested very strong genetic divergence [26]. These F ST estimates also provided insight into gene flow patterns as well, with values below 0.33 indicating substantial gene flow and higher values suggesting more restricted genetic exchange between populations [27].
## References
1. Khan, Zhao, Wang et al. "The recent advances of the causal agent of yellow leaf disease (YLD) on areca palm (Areca catechu L.)"
2. Wang, Yin, Luo et al. (1679) "Spatiotemporal evolution and impact mechanisms of areca palm plantations in China"
3. Yu, Qi, Chang et al. (2015) "Complete genome sequence of a novel velarivirus infecting areca palm in China" *Arch. Virol*
4. Nayar, Seliskar (1978) "Mycoplasma like organisms associated with yellow leaf disease of Areca catechu L" *Eur. J. For. Pathol*
5. Wang, Zhao, Zhang et al. "Prevalence of yellow leaf disease (YLD) and its associated areca palm velarivirus 1 (APV1) in betel palm (Areca catechu) plantations in Hainan"
6. Zhang, Zhao, Cao et al. (2022) "Transmission of areca palm velarivirus 1 by mealybugs causes yellow leaf disease in betel palm (Areca catechu)" *Phytopathology*
7. Fuchs, Bar-Joseph, Candresse et al. (2020) *ICTV Virus Taxonomy Profile: Closteroviridae. J. Gen. Virol*
8. Cao, Zhao, Wang et al. (2021) "Genomic diversity of areca palm velarivirus 1 (APV1) in areca palm (Areca catechu) plantations in Hainan"
9. Niu, Xu, Tian et al. (1025) "A putative ormycovirus that possibly contributes to the yellow leaf disease of areca palm" *Forests*
10. Khan, Cao, Zhao et al. (2022) "Effect of temperature on yellow leaf disease symptoms and its associated areca palm velarivirus 1 titer in areca palm" *Areca catechu L.). Front. Plant Sci*
11. García-Andrés, Accotto, Navas-Castillo et al. (2007) "Founder effect, plant host, and recombination shape the emergent population of begomoviruses that cause the tomato yellow leaf curl disease in the Mediterranean basin" *Virology*
12. Delatte, Holota, Moury et al. (2007) "Evidence for a founder effect after introduction of tomato yellow leaf curl virus-mild in an insular environment" *J. Mol. Evol*
13. Lin, Rubio, Smythe et al. (2004) "Molecular population genetics of cucumber mosaic virus in California: Evidence for founder effects and reassortment" *J. Virol*
14. Zwart, Elena (2015) "Matters of size: Genetic bottlenecks in virus infection and their potential impact on evolution" *Annu. Rev. Virol*
15. Jin, Sun, Chen et al. (1995) "Yellows disease of betel nut palm in Hainan" *China. Sci. Silvae Sin*
16. Katoh, Standley (2013) "MAFFT Multiple Sequence Alignment Software Version 7: Improvements in Performance and Usability" *Mol. Biol. Evol*
17. Zhang, Gao, Jakovlic et al. (2020) "PhyloSuite: An integrated and scalable desktop platform for streamlined molecular sequence data management and evolutionary phylogenetics studies" *Mol. Ecol. Resour*
18. Rozas, Ferrer-Mata, Sánchez-Delbarrio et al. (2017) "DnaSP 6: DNA sequence polymorphism analysis of large data sets" *Mol. Biol. Evol*
19. Excoffier, Lischer (2010) "Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows" *Mol. Ecol. Resour*
20. Tajima (1989) "Statistical method for testing the neutral mutation hypothesis by DNA polymorphism" *Genetics*
21. Fu (1997) "Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection" *Genetics*
22. Harpending (1994) "Signature of ancient population growth in a low-resolution mitochondrial DNA mismatch distribution" *Hum. Biol*
23. Gao, Chen, Arab et al. (2019) "EasyCodeML: A visual tool for analysis of selection using CodeML" *Ecol. Evol*
24. Yang (2007) "PAML 4: Phylogenetic analysis by maximum likelihood" *Mol. Biol. Evol*
25. Yang, Wong, Nielsen (2005) "Bayes empirical Bayes inference of amino acid sites under positive selection" *Mol. Biol. Evol*
26. Balloux, Lugon-Moulin (2002) "The estimation of population differentiation with microsatellite markers" *Mol. Ecol*
27. Gao, Jin, Zou et al. (2016) "Geographically driven adaptation of chilli veinal mottle virus revealed by genetic diversity analysis of the coat protein gene" *Arch. Virol*
28. Jombart, Devillard, Balloux (2010) "Discriminant analysis of principal components: A new method for the analysis of genetically structured populations" *BMC Genet*
29. Parker, Rambaut, Pybus (2008) "Correlating viral phenotypes with phylogeny: Accounting for phylogenetic uncertainty" *Infect. Genet. Evol. J. Mol. Epidemiol. Evol. Genet. Infect. Dis*
30. "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=PMC12793018&blobtype=pdf | # Abstract citation ID: ofaf695.2332 P-2169. Integrative scRNA-seq and Transcriptomic Analysis Reveals Monocyte/ Macrophage Activation Drives EV-A71-Induced Immune Dysregulation and Neural Injury in Severe HFMD Muqi Wang, Doctor's degree 1 ; Meng Zhang, Doctor's degree 1 ; huiling Deng, n/a 2 ; Yufeng Zhang, Doctor's degree 2 ; Chenrui Liu, Doctor's degree 3 ; Yuan Chen, Doctor's degree 2 ; Chuting Zhang, Doctor's degree 3 ; Wen Zhang, Doctor's degree 3 ; Xiaoli Jia, Doctor's degree 3 ; Shuangsuo Dang, PhD 4 ; Yaping Li, Doctor's degree 3
## Abstract
Background. Enterovirus 71 (EV-A71) is a major pathogen of severe hand, foot and mouth disease (HFMD) in children, but the mechanism by which it develops into severe HFMD remains unclear, especially the role of macrophage-mediated immune dysregulation.
Results. Single-cell RNA sequencing (scRNA-seq) revealed that EV-A71 infected severe HFMD patient had higher monocyte and macrophage ratio (18.50% vs. 8.85%), especially classical (64.59% vs. 57.24%) and non-classical (32.23% vs. 23.90%) monocytes, and a lower pDC (1.19% vs. 12.01%) and monoDC (1.98% vs. 6.80%) in EV-A71 infected severe HFMD patient. Dynamic analysis of PBMCs infected with EV-A71 isolates (mild, moderate and severe) and cell trajectory analysis indicated during infection, monocyte/macrophages were initially activated, followed by three groups of T cells and NK and B cells, M1 macrophage. High concentration of EV-A71 infected macrophage supernatant inhibited SH-SY5Y cell proliferation. ENSG00000285779, TICAM2, RPL13AP26 and HNF4G are significantly different in EV-A71 or inactivated EV-A71 infected macrophages than in control. ENSG00000264324, ENSG00000260643, ISLR2, CCR7, TENM4, INO80B-WBP1, BLOC1S5-TXNDC5 are potential genes about direct virus damage or viral RNA recognition in macrophages. GO annotation and KEGG analysis indicate that EV-A71 infection cause the changes of neural receptor-ligand binding pathway, activation of specific immunity, calcium signaling pathway, and cell aggregation.
Conclusion. Macrophages are activated early during EV-A71 infection, thus initiating specific immunity, which is closely related to the severe HFMD. The nerve damage pathway and calcium signaling pathway caused by EV-A71 virus infection of macrophages deserve to more attention.
Disclosures. All Authors: No reported disclosures |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793421&blobtype=pdf | Roper St, Francis, Cassandra Simonich, Teagan Mcmahon, Jesse Bloom
Background. Respiratory syncytial virus (RSV) is a leading cause of lower respiratory tract infections of infants and older adults. Recent advances in RSV prevention include new vaccines for adults including one to protect infants via passive transfer of maternal antibodies and a new monoclonal antibody (mAb) for infants that target the fusion protein (F). However, efforts to combat RSV must deal with antigenic variability in F which can lead to escape from prophylactic measures. Viral escape as a vulnerability of mAb prophylaxis has been demonstrated, and RSV surveillance efforts have identified circulating strains that contain mutations in key antigenic sites of F. Here we perform deep mutational scanning (DMS) of RSV F to comprehensively map the functional and antigenic effects of mutations. S1112 • OFID 2026:13 (Suppl 1) • Poster Abstracts Graphical overview of RSV F deep mutational scanning for measuring cell entry and antibody escape (A) Strategy to make genotype-phenotype-linked pseudoviruses. A library of barcoded and mutagenized F variants in a lentiviral vector are transfected into 293T cells with additional lentiviral helper plasmids and a plasmid encoding VSV-G to create pseudoviruses that lack a genotype-phenotype link. To establish a link between the F on the virion surface and the genotype, we then infect LentiX cells at a low multiplicity of infection (MOI<0.01) to ensure a single integration per cell, and cells with integrated provirus are selected with puromycin creating a cell-stored library of F mutants in a lentiviral backbone. Transfection of helper plasmids into the cell-stored libraries plus either VSV-G or unmutated RSV G generates virions that have a single F protein variant on the surface and encode an identifying barcode in their genome. PacBio sequencing of the F gene and barcode from the library of singly-integrated cells is used to link variants in F with specific barcodes. For DMS experiments, Illumina sequencing of the short barcode region is used to identify the F variant. Methods. To investigate RSV antigenic evolution and map the evolutionary space accessible to RSV F, we utilize a high-throughput, pseudovirus platform to perform DMS of RSV F to measure how nearly all amino-acid mutations affect cell entry and antibody neutralization (Figure 1A-C). We made duplicate mutant libraries targeting all single amino-acid mutations of the F ectodomain.
Nirsevimab escape mutations identified by DMS (A) Nirsevimab escape mutations mapped onto the structure of the RSV F trimer (PDB 5c6b). Sites where mutations cause strong escape are in darker red. (B) Key sites of escape for nirsevimab. The height of each letter in the logo plot is proportional to the escape caused by that amino-acid mutation, and letters are colored by the effect of that mutation on cell entry (dark green indicates well-tolerated, light yellow indicates impaired cell entry).
Results. The two independent RSV F libraries each contain ∼9,500 amino-acid mutations and ∼47,000 unique barcoded F variants that each cover >99% of all possible mutations. Most F variants carried a single amino-acid mutation. Our DMS reveals regions of mutational constraint in RSV F likely due to protein folding and conformational change requirements for function (Figure 2A-B). Our study also identifies mutations at known antigenic regions that do not impact cellular entry, suggesting the possibility of viral escape from antibodies targeting these regions (Figure 2B-C). We map escape mutations for clinically relevant mAbs including nirsevimab (Figure 3). Nirsevimab escape mutations include mutations not previously identified and mutations that are functionally tolerated. These mutations exist among natural RSV sequences at low frequency.
Conclusion. The DMS maps of mutational effects further our understanding of RSV F function and facilitate interpretation of the functional and antigenic consequences of F mutations observed by RSV surveillance efforts in the setting of more widespread use of vaccines and mAbs.
Disclosures. Jesse Bloom, PhD, Apriori Bio: Advisor/Consultant|Apriori Bio: I am inventor on patents licensed to Apriori Bio by the Fred Hutch|GSK: Advisor/ Consultant|Invivyd: Advisor/Consultant|Pfizer: Advisor/Consultant |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645963&blobtype=pdf | # Lumpy skin disease virus LSDV087 positively regulates innate immune response by promoting oligomerization of MITA/STING
Zhen-Zhen Li, Yu-Lin Yang, Meng-Yao Sun, Hong-Bing Shu, Li-Bo Cao
## Abstract
Lumpy skin disease (LSD), caused by the lumpy skin disease virus (LSDV), is a contagious disease in cattle that poses a major threat to the global cattle industry. The functions of most LSDV-encoded proteins remain poorly characterized, particularly regarding their roles in regulating innate immunity. In this study, we show that the LSDV-encoded protein LSDV087 positively regulates innate immune response independ ently of its decapping enzymatic activity. LSDV087 interacts with the adaptor protein MITA (also called STING) in the innate immune pathway, inhibits its degradation by reducing K48-linked polyubiquitination, and promotes its oligomerization and subse quent activation of downstream signaling events, leading to enhanced innate immune response. Consistently, LSDV087-deficient virus (LSDV∆087) exhibits an attenuated ability to activate cGAS-MITA-mediated innate immune response. Collectively, our study reveals regulatory mechanisms of LSDV-triggered innate immune response and points to the possibility of targeting LSDV087 for rational design of live-attenuated LSDV vaccines. IMPORTANCE Lumpy skin disease virus (LSDV), which causes a contagious disease in cattle, poses a significant threat to the global cattle industry. Despite its impact, the functions of most LSDV-encoded proteins remain poorly understood. In this study, we report that LSDV087 plays dual roles in both promoting the cGAS-MITA-mediated innate immune response and downregulating host gene transcription. LSDV087 interacts with the adaptor protein MITA in the innate immune pathway, inhibits its degradation by reducing K48-linked polyubiquitination, and promotes its oligomerization, leading to the subsequent activation of downstream signaling events and an enhanced innate immune response. Additionally, as an immediate-early protein, LSDV087 functions as a decapping enzyme, preferentially targeting host transcripts with multiple exons to facilitate viral replication. This dual functionality underscores the complex interplay between LSDV immune evasion strategies and host defense mechanisms and may inform the rational design of live-attenuated LSDV vaccines. KEYWORDS lumpy skin disease virus, LSDV087, cGAS, MITA/STING, innate immunity L umpy skin disease (LSD) is an acute, subacute, or chronic infectious disease caused by lumpy skin disease virus (LSDV), which mainly affects cattle and water buffaloes (1). Upon infection with LSDV, diseased cattle may experience reduced milk production, hide damage, abortions, temporary or permanent infertility, and even death, causing huge economic losses to the cattle farming industry (2, 3). LSD has been listed by the World Organization for Animal Health as a notifiable poxviral disease of significant impact that crosses national borders due to its widespread prevalence around the world (4, 5). Current studies on LSDV mainly focus on its epidemiology, etiology, diagnostic methods, and isolation (6-8). The LSDV genome is approximately 150 kilobases in length and encodes 156 predicted open reading frames (9). The functions of most LSDV-encoded proteins and the mechanisms by which LSDV interacts with the host
immune system remain poorly understood, which limits the development of therapeu tic drugs and effective vaccines for LSD. Innate immunity is crucial for the initial monitoring of invading viruses and primes subsequent adaptive immunity, which influences the intensity and quality of long-term protective immune responses against pathogens (10). Therefore, investigating the mechanisms of interaction between LSDV and the host innate immunity would provide potential strategies for the prevention and control of LSD.
The host innate immune system senses the invasion of pathogens via pattern recognition receptors (11)(12)(13). As a double-stranded DNA virus replicating in the cytoplasm of host cells, it is expected that LSDV genomic DNA is sensed by the cytosolic DNA sensor cGAS, a ubiquitously expressed cytoplasmic sensor for DNA (14)(15)(16)(17). Upon sensing viral DNA, cGAS utilizes ATP and GTP to synthesize the second messenger cyclic GMP-AMP (cGAMP), which then binds to the endoplasmic reticulum (ER) asso ciated adaptor protein MITA (also known as STING). This causes the conformational changes, oligomerization, and cellular trafficking of MITA. In this process, MITA recruits the downstream kinase TBK1 and transcription factor IRF3, leading to activation of IRF3, induction of type I interferons (IFNs), and other antiviral effector genes (18)(19)(20)(21).
As a central adaptor protein in innate antiviral signaling, MITA is tightly regulated by post-translational modifications, including phosphorylation, sumoylation, and ubiquiti nation (22). Among them, ubiquitination plays a versatile role in regulating MITA activity (23). K63-linked polyubiquitination of MITA facilitates its oligomerization and ER-to-Golgi trafficking, enhancing downstream signaling and antiviral gene expression. Additionally, K63-linked polyubiquitination of MITA is implicated in its lysosomal degradation, serving as a feedback mechanism to avoid excessive immune activation (24,25). Conversely, K48-linked polyubiquitination of MITA targets it for proteasomal degradation, thereby limiting its signaling activity and dampening the innate immune response (26)(27)(28).
Given the critical roles of the cGAS-MITA pathway in antiviral defense, identifying LSDV-encoded proteins that modulate this signaling axis is of great importance. In this study, we identified that LSDV-encoded LSDV087 protein positively regulated cGAS-MITA-mediated innate immune response independently of its decapping enzymatic activity. Mechanistically, LSDV087 interacted with MITA, reduced its K48-linked poly ubiquitination and proteasomal degradation, and enhanced its oligomerization and signaling activity. Furthermore, deletion of LSDV087 attenuated LSDV-triggered innate immune response. Our findings reveal regulatory mechanisms of LSDV-triggered innate immune response, which may help in the rational design of live-attenuated LSDV vaccines.
## RESULTS
## LSDV087 positively regulates cGAS-MITA-mediated innate immune response
During the long-term coevolution between LSDV and its hosts, the virus may evolve effector proteins that suppress the host innate immune response to promote viral replication. Conversely, some viral proteins may inadvertently activate host immunity, either as an unintended consequence or as part of a dynamic balance between host immune activation and viral immune evasion (29,30). To identify LSDV proteins that regulate innate immune response, we constructed mammalian expression clones for 156 LSDV-encoded proteins and assessed their effects on activation of the IFN-β promoter induced by cGAS-MITA in reporter assays. This screen identified LSDV087 as a viral protein that enhances cGAS-MITA-induced IFN-β promoter activation (Fig. 1A). Further experiments indicated that overexpression of LSDV087 promoted cGAS-MITA-induced activation of the IFN-β promoter in a dose-dependent manner in HEK293 cells (Fig. 1B).
To investigate whether LSDV087 broadly regulates the expression of antiviral genes, we established a RAW264.7 cell line stably expressing LSDV087 through lentiviral transduction. Upon transfection with herring testis DNA (HT-DNA), RNA-seq analysis was performed to compare the transcriptional profiles between LSDV087-expressing and control cells. Differentially expressed genes were further analyzed through the Kyoto Encyclopedia of Genes and Genomes to explore their functions. Compared to the control cell line, 21 pathways were enriched in the LSDV087-expressing cell line, including the Cytosolic DNA-Sensing Pathway and the TNF Signaling Via NF-κB (TNF-NF-κB) pathways (Fig. 1C). Additionally, gene set enrichment analysis revealed that both the DNA-sensing and TNF-NF-κB pathways were upregulated in the LSDV087-expressing cell line com pared to the control (Fig. 1D). These results suggest that LSDV087 broadly enhances DNA-induced transcription of antiviral genes. RT-qPCR analysis demonstrated that ectopically expressed LSDV087 promoted HT-DNA but not the double-stranded RNA analog poly(I:C)-triggered transcriptional induction of Ifnb1 gene in RAW264.7 cells (Fig. 1E). Consistently, LSDV087 also promoted LSDV and another DNA virus herpes simplex virus 1 (HSV-1)-induced transcription of Ifnb1 gene in RAW264.7 cells (Fig. 1F). Given that phosphorylation of MITA and IRF3 is critical in innate immune response to DNA, we further examined the effects of LSDV087 on these events. Immunoblot analysis indicated that LSDV087 promoted phosphorylation of MITA and IRF3 in response to HT-DNA (Fig. 1G). These results suggest that LSDV087 promotes LSDV-triggered innate immune signaling in human and mouse cell lines.
## LSDV087 deficiency attenuates LSDV-triggered innate immune response
Given that overexpression of LSDV087 promotes LSDV-triggered innate immune response, we next investigated the effects of LSDV087 deficiency on LSDV-triggered innate immune response. RT-qPCR and immunoblot analysis indicated that LSDV087 mRNA and protein were detected at 0.5 hour post-infection (h.p.i.) of LSDV (Fig. 2A andB). The protein synthesis inhibitor cycloheximide (CHX) did not affect the early transcription of LSDV087 (Fig. 2C), and the DNA synthesis inhibitor arabinoside C (Ara-C) had no marked effect on its protein level at the early phase of infection (Fig. 2D). These results suggest that LSDV087 is an immediate-early gene of LSDV.
To investigate the roles of LSDV087 in the regulation of innate antiviral response, we generated a recombinant LSDV strain with deletion of the LSDV087 gene (LSDVΔ087) derived from the highly pathogenic LSDV/China/Hainan/2021 strain by homologous recombination (Fig. 2E). The purity of LSDVΔ087 was confirmed by PCR analysis of the viral genome (Fig. 2F), and the deficiency of LSDV087 in LSDV was verified by immuno blot analysis of LSDV087 level in cells infected with EGFP-tagged wild-type LSDV and LSDVΔ087 (Fig. 2G). The replication rates of LSDV∆087 were approximately twofold lower than those of wild-type LSDV at 36 and 48 h.p.i (Fig. 2H), suggesting a relatively weak effect of LSDV087 on LSDV replication, which may be attributed to its decapping activity.
To investigate whether LSDV087 promotes DNA-triggered innate immune signaling in the natural host cells of LSDV, we performed experiments with Madin-Darby bovine kidney (MDBK) cells that have been shown to be infected by LSDV (31). In these cells, stimulation with HT-DNA or cGAMP induced transcription of Ifnb1 and Ifi44 genes, while knockout of MITA impaired it in MDBK cells (Fig. 3A). These results suggest that the cGAS-MITA pathway is functional in MDBK cells. Consistently, wild-type LSDV but not LSDVD087 induced transcription of interferon-stimulated genes (ISGs), including Ifnb1, Mx2, and Il6 in MDBK cells at 48 h.p.i. In these experiments, the replication of LSDV structural genes LSDV031 and LSDV063 was not affected in MDBK cells at 48 h.p.i (Fig. 3B), suggesting that LSDV087 plays an important role in promoting LSDV-triggered induction µg) by lipofectamine for 6 hours before reverse transcription-quantitative PCR (RT-qPCR) analysis for mRNA levels of the Ifnb1 gene. (F) Effects of LSDV087 on LSDV-or herpes simplex virus 1 (HSV-1)-triggered transcription of Ifnb1 gene. LSDV087-expressing and control RAW264.7 cells (2 × 10 5 ) were infected with LSDV for 9 hours or HSV-1 for 6 hours before RT-qPCR analysis for mRNA levels of the Ifnb1 gene. (G) Effects of LSDV087 on HT-DNA-induced phosphorylation of MITA and IRF3. LSDV087-expressing and control RAW264.7 cells (2 × 10 5 ) were transfected with HT-DNA for the indicated times before immunoblot analysis with the indicated antibodies. Band intensities were quantified by densitometry using ImageJ software and normalized to β-actin levels. Data shown in panels B, E, and F are mean ± SD (n = 3) from one representative experiment. All the experiments were repeated at least two times with similar results. ns, not significant; **P < 0.01 (unpaired t-test). an MDBK cell line stably expressing LSDV087, LSDV-induced transcription of Ifnb1 and Ifi44 genes was markedly enhanced compared to that of control MDBK cells (Fig. 3D). Together, these results suggest that LSDV087 enhances MITA-dependent antiviral innate immune response in bovine cells.
## LSDV087 regulates the cGAS-MITA signaling pathway independently of its decapping enzymatic activity
LSDV087 is annotated in the NCBI database as a homolog of the vaccinia virus (VACV) D10R, which downregulates the mRNA levels of multi-exonic genes in host cells through its decapping enzymatic activity (32,33). Sequence alignment analysis suggests that LSDV087 shares 48.2% amino acid sequence identity with D10R and contains conserved decapping enzyme active residues at E147 and E148 (Fig. 4A). To investigate whether LSDV087 possesses decapping enzymatic activity similar to that of D10R, we examined the effects of LSDV087 and its E147Q/E148Q mutant (LSDV087 E147Q/E148Q ) on the mRNA levels of multi-exonic genes in HEK293 cells. RT-qPCR analysis indicated that overexpres sion of LSDV087 reduced the mRNA levels of multi-exonic genes, such as Gapdh, Ercc8, Gbe1, and Eef1a1, but barely affected the transcription of single-exon genes such as Ddx28 and Znf830. In these experiments, LSDV087 E147Q/E148Q had no marked effects on mRNA levels of either single-or multi-exonic genes (Fig. 4B). These results suggest that LSDV087 functions similarly to D10R, selectively downregulating the transcription of multi-exonic genes.
To investigate whether the functions of LSDV087 in regulating antiviral innate immune response are dependent on its decapping enzymatic activity, we examined the effects of LSDV087 E147Q/E148Q on cGAS-MITA-triggered signaling. In reporter assays, both wild-type LSDV087 and LSDV087 E147Q/E148Q , but not D10R, promoted activation of the IFN-β promoter induced by cGAS-MITA (Fig. 4C). We established RAW264.7 and MDBK cell lines stably expressing LSDV087 or LSDV087 E147Q/E148Q through lentiviral transduction. RT-qPCR experiments indicated that both wild-type LSDV087 and LSDV087 E147Q/E148Q enhanced HT-DNA-induced transcription of Ifnb1 and Isg56 genes in both RAW264.7 and MDBK cells (Fig. 4D). These results suggest that LSDV087 promotes cGAS-MITA signaling through a decapping activity-independent mechanism.
## LSDV087 positively regulates innate antiviral response by interacting with MITA
To elucidate the mechanisms by which LSDV087 regulates the cGAS-MITA axis, we examined the effects of LSDV087 on activation of the IFN-β promoter mediated by overexpression of various components in the cGAS-MITA signaling pathway in reporter assays. The results indicated that LSDV087 promoted activation of the IFN-β promoter mediated by overexpression of cGAS and MITA but not by TBK1 or IRF3 (Fig. 5A). These results suggest that LSDV087 functions at the level of cGAS or MITA. We next investigated whether LSDV087 affects cGAS activity by measuring cGAMP production following HT-DNA stimulation. cGAMP synthesis assays showed that LSDV087 had no marked effects on HT-DNA-induced cGAMP synthesis (Fig. 5B), indicating that LSDV087 did not affect cGAS activity. However, LSDV087 promoted cGAMP-induced transcription of Ifnb1 and Isg56 genes in RAW264.7 and MDBK cells (Fig. 5C). Consistently, immuno blot analysis indicated that stable expression of LSDV087 enhanced cGAMP-induced phosphorylation of MITA and IRF3 in RAW264.7 cells (Fig. 5D). These results suggest that LSDV087 promotes antiviral innate immune response by modulating MITA activity.
Co-immunoprecipitation experiments indicated that LSDV087 interacted with MITA but not cGAS or TBK1, while D10R did not interact with any of them in a mammalian overexpression system (Fig. 5E). Furthermore, stably expressed LSDV087 but not D10R constitutively interacted with endogenous MITA in RAW264.7 cells (Fig. 5F). Confocal microscopy showed that the majority of LSDV087 colocalized with MITA in the ER under resting conditions (Fig. 5G). Collectively, these results suggest that LSDV087 regulates the cGAS-MITA axis by interacting with MITA.
## LSDV087 promotes MITA oligomerization
To investigate the molecular mechanisms by which LSDV087 regulates the activity of MITA, we examined the effects of LSDV087 on the binding of cGAMP to MITA. To do this, RAW264.7 cells stably transduced with an empty vector or LSDV087 were transfected with HT-DNA for 6 hours, and then cell lysates were immunoprecipitated with anti-MITA antibody. cGAMP bound to immunoprecipitated MITA was measured by ELISA. The results indicated that LSDV087 had no marked effects on the amounts of MITA-bound cGAMP (Fig. 6A). We next treated the LSDV087-expressing and control RAW264.7 cells with cGAMP and then examined the induction of MITA oligomerization by semi-denaturing detergent agarose gel electrophoresis (SDD-AGE) assays. The results indicated that LSDV087 promoted cGAMP-induced MITA oligomerization, which is an early step for its activation of downstream signaling events (Fig. 6B). In agreement with these biochemical results, confocal microscopy demonstrated that LSDV087 enhanced cGAMP-induced aggregation of MITA to form perinuclear punctate structures, which is a hallmark of activation of MITA-mediated downstream signaling (Fig. 6C). These results suggest that LSDV087 does not affect cGAMP binding to MITA but rather promotes MITA oligomerization and trafficking to form perinuclear punctate structures. In our earlier experiments, we noticed that ectopic expression of LSDV087 inhibi ted the downregulation of MITA induced by HT-DNA or cGAMP (Fig. 1G, 5D and6B). It has been reported that MITA undergoes K48-linked polyubiquitination for protea somal degradation or K63-linked polyubiquitination for lysosomal degradation (24)(25)(26)(27)(28). Therefore, we next investigated the effects of LSDV087 on polyubiquitination of MITA. The results indicated that LSDV087 inhibited K48-linked but not K63-linked polyubiquitination of MITA in a mammalian overexpression system (Fig. 6D). Endogenous ubiquitination assays confirmed that LSDV087 inhibited HT-DNA-induced K48-linked polyubiquitination and downregulation of MITA (Fig. 6E). These results suggest that LSDV087 inhibits DNA-induced K48-linked polyubiquitination and degradation of MITA, which may contribute to its oligomerization and activation of downstream signaling.
## DISCUSSION
LSD is a significant transboundary disease that causes substantial economic losses to the livestock industry (34). Currently, there are no safe and effective commercial vaccines available for LSD. Although the attenuated vaccines derived from goat poxvirus and sheep poxvirus show some effectiveness in the prevention and control of LSD, their intradermal inoculation method requires a relatively high level of technical expertise from the operators, and there is also a potential risk of recombination with wild-type strains (35,36). Therefore, elucidating the mechanisms by which LSDV interacts with the host immune system may help to develop new strategies to fight against LSD. LSDV encodes 156 proteins, most of whose functions remain unclear. Through an unbiased screen by luciferase reporter assays, we discovered that LSDV087 promoted cGAS-MITA-mediated activation of the IFN-β promoter in HEK293 cells. Further stud ies revealed that overexpression of LSDV087 enhanced phosphorylation of MITA and IRF3 and transcription of downstream antiviral genes in response to HT-DNA and LSDV in mouse macrophage RAW264.7 cells. Overexpression of LSDV087 also increased LSDV-induced transcription of downstream antiviral genes in bovine MDBK cells. These findings suggest that LSDV087 positively regulates the cGAS-MITA signaling pathway.
RT-qPCR and immunoblot analysis indicated that LSDV087 is expressed at the early phase of LSDV infection. The protein synthesis inhibitor cycloheximide and the DNA synthesis inhibitor arabinoside C had no marked effects on its mRNA and protein levels at the early phase of infection. These results suggest that LSDV087 is an immediate-early gene of LSDV. To confirm the regulatory effects of LSDV087 on the innate immune response, we generated LSDVΔ087, an LSDV strain with deletion of the LSDV087 gene. Experimental results showed that deletion of LSDV087 had minimal (approximately twofold) effects on the production of progeny viruses and had no marked effects on the replication of viral DNA at 48 h.p.i. However, deficiency of LSDV087 fully impaired LSDV-induced transcription of downstream antiviral genes in bovine MDBK cells. These findings suggest that LSDV087 plays a positive regulatory role in LSDV-induced innate immune response. Poxviruses are well known for evolving various strategies to evade host innate immune response. However, our findings suggest that viral proteins do not always act solely as immune antagonists. During long-term host-pathogen coevolution, while viruses evolve immune evasion strategies, hosts may, in turn, evolve the ability to recognize and respond to essential viral proteins as part of their defense strategy.
LSDV087 is predicted to be a homolog of VACV D10R, and both proteins share conserved active residues of a decapping enzyme. Our experiments indicated that LSDV087 possessed decapping enzymatic activity similar to that of D10R. However, our experiments indicated that LSDV087 promoted the activation of the cGAS-MITA signaling pathway independently of its decapping enzymatic activity. Additionally, our results indicated that D10R did not interact with MITA and lacked the ability to positively regulate MITA-mediated signaling. These results suggest that the ability of LSDV087 in promoting innate antiviral response is not shared by homologs encoded by other poxviruses.
MITA is a key adaptor in DNA-induced innate immune response. Our results indicated that LSDV087 did not affect the synthesis of cGAMP or its binding to MITA upon DNA stimulation. However, LSDV087 interacted with MITA and enhanced cGAMP-induced oligomerization of MITA. LSDV087 inhibited DNA-triggered K48-linked polyubiquitina tion and degradation of MITA, which may contribute to its effects on MITA oligomeriza tion and activation. Regrettably, due to the lack of a high-quality antibody against bovine MITA, we were limited in further biochemical experiments in LSDV-infected bovine cells.
Nevertheless, our study supports a model in which LSDV087 plays dual roles during LSDV infection. As an early viral protein, LSDV087 functions as a decapping enzyme, preferentially targeting host transcripts with multiple exons to facilitate viral replication. Additionally, LSDV087 interacts with MITA to promote its oligomerization and activation, leading to amplification of the host's innate antiviral response. This dual functionality highlights the complex interplay between LSDV immune evasion strategies and host defense mechanisms.
## MATERIALS AND METHODS
## Reagents, cells, and viruses
Fetal bovine serum (SA211.02, Cellmax), penicillin and streptomycin (SV30010, HyClone), puromycin (AMR-J593,VWR), protein A + G agarose (P2055, Beyotime), PMSF (P7626, Sigma), Dual-Specific Luciferase Assay Kit (E1980, Promega), SYBR Green supermix (Q312, Vazyme), HiScript II Select RT SuperMix for qPCR (R323, Vazyme), isopropyl β-D-1-thiogalactopyranoside (IPTG) (ST098, Beyotime), HT-DNA (D6898, Sigma), Poly(I:C) (tlrl-pic, Invivogen), cGAMP ELISA Kit (CAY-501700-96S, Cayman), cGAMP (tlrl-nacga23-02, Invivogen), ATP (9804S, Cell Signaling Technology), GTP (51120, Sigma), digitonin (T2721, TargetMol), and aprotin, leupeptin, β-glycerophosphate disodium salt, Ara-C, CHX, and sodium orthovanadate (HY-P0017, HY-18234A, HY-126304, HY-13605, HY-12320, and HY-D0852, respectively, MCE) were purchased from the indicated manufacturers.
Mouse monoclonal antibody against HA (66006, Proteintech); rabbit monoclonal antibody against HA (H6908, Sigma); mouse monoclonal antibody against Flag (F3165, Sigma) and HRP-Flag (ZB15939, Servicebio); IgG (I5381, Sigma); antibodies against p-MITA (S366; 50907S, Cell Signaling Technology), MITA (A21051, ABclonal), p-IRF3 (S396; 4947S, Cell Signaling Technology), IRF3 (A19717, ABclonal), β-actin (A2228, Sigma-Aldrich); anti-mouse IgG (H + L), F(ab')2 Fragment (Alexa Fluor® 594 Conjugate; 8890, Cell Signaling Technology); Alexa Fluor 488 goat anti-rabbit IgG (H + L; A11008, Invitrogen); and 4′,6-diamidino-2-phenylindole (DAPI) (C1002, Beyotime) were purchased from the indicated manufacturers.
HEK293, RAW264.7, and HeLa cells were obtained from ATCC. MDBK cells were purchased from CCTCC. BEF cells, LSDV/China/Hainan/2021 strain, and EGFP-tagged wild-type LSDV were gifted by Mr. Huaijie Jia of Lanzhou Veterinary Research Institute. HSV-1 was previously described (37).
## Plasmids
The LSDV protein expression clones were synthesized by GenScript. Expression plasmids for Flag-tagged LSDV087, LSDV087 E147Q/E148Q , LSDV087 (60-150), and VACV D10R were constructed by standard molecular biology techniques. Expression plasmids for HA-tagged cGAS, MITA, TBK1, and IRF3 were previously described (38).
## Generation of stable cell lines
A cDNA corresponding to LSDV087-coding sequence was cloned into the pCDH vector, which was co-transfected with packaging plasmids psPAX2 and pCMV-VSV-G into HEK293 cells. Thirty-six hours after transfection, the viruses were harvested for infection of RAW264.7 or MDBK cells. The infected cells were selected with puromycin (4 µg/mL) for 6 days to establish stable cell lines.
## Gene knockout by the CRISPR-Cas9 methods
Double-stranded oligonucleotides targeting Mita sequence were inserted into the lenti-CRISPR-V2 vector, which was co-transfected with packaging plasmids psPAX2 and pCMV-VSV-G into HEK293 cells. Thirty-six hours after transfection, the viruses were harvested for infection of MDBK cells. The infected cells were selected with puromycin 5′-CCACCATCTCATGGGAGAGC-3′; Bovine Mita: 5′-CCCAAAAGGCAGCCTTGGTC-3′ and 5′-CCAGACTGCAGATTCCCTTG-3′; LSDV031: 5′-ACAGTTGAATGTGATGGCGA-3′ and 5′-TGGGGATGAAGCTCTTGCAG-3′; LSDV063: 5′-TGTTCATTCACCATCCGCATC-3′ and 5′-GGTTCTTGTAATGGCTTGTTGC-3′; LSDV087: 5′-AAACTGTCTTCGTCAGACCAT-3′ and 5′-CGTTTCTAATTACCCCACCTGGA-3′.
## Transfection and reporter assays
HEK293 cells were transfected with the indicated plasmids by the calcium phos phate precipitation method. To normalize for transfection efficiency, a pRL-TK (Renilla luciferase) reporter plasmid was added to each transfection. Luciferase assays were performed using a dual-specific luciferase assay kit. Firefly luciferase activities were normalized on the basis of Renilla luciferase activities.
## Fluorescent confocal microscopy
HeLa cells were seeded on coverslips in 24-well plates and transfected with the indicated plasmids for 24 hours. LSDV087-expressing and control RAW264.7 cells were seeded and treated with cGAMP for 30 minutes. The cells were fixed with 4% paraformaldehyde for 20 minutes and then permeabilized with 0.1% Triton X-100 for 15 minutes. Subsequently, the cells were blocked in 1% bovine serum albumin (BSA) and stained with the indicated antibodies. Fluorescence imaging was performed using a Zeiss confocal microscope.
## Co-immunoprecipitation and immunoblot analysis
Cells were lysed with lysis buffer (20 mM Tris-HCl pH 7.4, 1% NP-40, 150 mM NaCl, and 1 mM EDTA and protease inhibitors) at 4°C for 10 minutes and sonicated for 1 minute. The lysates were centrifuged at 13,000 rpm for 10 minutes at 4°C. The supernatants were immunoprecipitated with the indicated antibodies. Then the beads were washed three times with washing buffer (750 mM NaCl and 50 mM Tris-HCl pH 7.4). The bound proteins were separated by SDS-PAGE, followed by immunoblot analysis with the indicated antibodies.
## Semi-denaturing detergent agarose gel electrophoresis
RAW264.7 cells were lysed in NP-40 lysis buffer, and the cell lysates were mixed in sample buffer (0.5× Tris-borate-EDTA [TBE] buffer, 10% glycerol, 2% SDS, and 0.0025% bromo phenol blue) and loaded onto a vertical 2% agarose gel (Bio-Rad). After electrophoresis in the running buffer (1× TBE and 0.1% SDS) for about 2 hours with a constant voltage of 100 V at 4°C, the proteins were transferred to an Immobilon membrane (Millipore) for immunoblot analysis.
## Preparation of LSDV087 antibody
The prokaryotic expression vector pET-30c-LSDV087 (60-150) was transformed into E. coli BL21 (DE3) competent cells. A 10 mL Luria-Bertani (LB) medium starter culture was inoculated from a single colony and grown overnight at 37°C with shaking at 220 rpm. The culture was then diluted 1:50 into 500 mL LB medium and incubated until reaching mid-log phase (OD 600 = 0.6-0.8). Protein expression was induced with 0.5 mM IPTG at 16°C for 10 hours. The recombinant protein was purified under denaturing conditions following inclusion body isolation protocols. The purified protein was used as an antigen to immunize 8-week-old BALB/c mice via multi-site subcutaneous injections. Mice were immunized at 14-day intervals for a total of five doses. Polyclonal antibodies were harvested from serum samples collected 1 week after the final immunization.
## cGAMP treatment
cGAMP was delivered into RAW264.7 or MDBK cells using a digitonin-based permeabi lization method. Briefly, cells were pretreated with digitonin permeabilization buffer (50 mM HEPES, pH 7.0; 100 mM KCl; 3 mM MgCl 2 ; 0.1 mM dithiothreitol (DTT); 85 mM sucrose; 0.2% BSA; 1 mM ATP; 0.1 mM GTP; and 10 µg/mL digitonin) at 37°C for 20 min, followed by incubation with cGAMP.
## cGAMP activity assay
Raw264.7 cells were either untreated or transfected with HT-DNA for 6 hours. Cell lysates were then prepared and heated at 95°C for 10 minutes to denature proteins, followed by centrifugation at 20,000 g for 25 minutes at 4°C. The supernatants containing cGAMP were collected and analyzed using a cGAMP ELISA kit. The cGAMP level was measured as an indicator of cGAS activity.
## Viral plaque assay
MDBK cells were infected with wild-type LSDV and LSDVΔ087 (MOI = 0.01) for the indicated times. Both cells and supernatants were harvested and freeze-thawed to obtain viral suspensions. Serial dilutions of these suspensions were used to infect MDBK cells for 2 hours at 37°C. The cells were then overlaid with 1.5% methylcellulose and incubated for 96 hours before plaque counting.
## Facility biosafety statement
All experiments involving live LSDV were conducted in an ABSL-3 laboratory at the Lanzhou Veterinary Research Institute of the Chinese Academy of Agricultural. Envi ronmental and equipment decontamination were performed after each experiment according to institutional biosafety protocols. The experiments were approved by the Ministry of Agriculture and Rural Affairs (approval number: 07140020250302) and the China National Accreditation Service for Conformity Assessment (approval number: CNAS BL0098). We acknowledge the ABSL-3 staff for their operational and logistical support.
## Statistics analysis
Unpaired Student's t-test was used for statistical analysis with GraphPad Prism Software. The number of asterisks represents the degree of significance with respect to P values. Statistical significance was set at P < 0.05 (*) and P < 0.01 (**).
## References
1. "days to establish MITA-deficient cell lines. The following sequences were targeted for the bovine Mita gene. 5′-ACAGGCACTTAGCAGGACCA-3′. Generation of LSDVΔ087 LSDVΔ087 was generated by homologous recombination in MDBK cells. The recombi nant plasmid (pUC18-LSDV087L-EGFP-LSDV087R) was first constructed using pUC18 used as the backbone. The cassette contains the left and right homology arms of LSDV087 gene and the fluorescent gene EGFP under the control of the LSDV pA7L promoter, which was inserted into the middle of the reading frame of LSDV087 gene. MDBK cells were infected with wild-type LSDV for 12 hours and then transfected with the homologous recombinant plasmid pUC18-LSDV087L-EGFP-LSDV087R using Lipo3000 transfection reagent. After culturing for 3 days, the cells were frozen-thawed and then seeded with MDBK cells in 96-well plates by limiting dilution. After several rounds of dilution screening and amplification, the purified LSDVΔ087 was obtained. The purity of LSDVΔ087 was determined by PCR and immunoblot analysis"
2. "RT-qPCR and qPCR Total RNAs and DNA were isolated for RT-qPCR or qPCR analysis to measure mRNA or viral DNA levels of the indicated genes, respectively. Data shown were the relative abundance of the indicated mRNA normalized to that of Gapdh"
3. "Human Gbe1: 5′-AGTCGCTGGCATTTTGGTTG-3′ and 5′-AAGCCCATGCGTAATGAGTC-3′"
4. "Human Eef1a1: 5′-TGTCGTCATTGGACACGTAGA-3′ and 5′-ACGCTCAGCTTTCAGTTTATCC"
5. "Human Ercc8: 5′-AAGGCAGTTTGCGCTAATGC-3′ and 5′-TGAGGCAGGATTTGGTCATACC-3′"
6. "Human Ddx28: 5′-TCGTGACAGAGCAGAAAGGAC-3′ and 5′-TCATCAAGGCTGGCATTTGC-3′"
7. "Human Znf830: 5′-AAGAATTGCGGCGGTTAATGA-3′ and 5′-CCCAAACGGTTGTACTTCGC-3′"
8. "Mouse Gapdh: 5′-ACGGCCGCATCTTCTTGTGCA-3′ and 5′-ACGGCCAAATCCGTTCACACC-3′; Mouse Ifnb1: 5′-TCCTGCTGTGCTTCTCCACCACA-3′ and 5′-AAGTCCGCCCTGTAGGTGAGGTT-3′; Mouse Isg56: 5′-ACAGCAACCATGGGAGAGAATGCTG-3′ and"
9. "Bovine Gapdh: 5′-AGGTCGGAGTGAACGGATTC-3′ and 5′-ATGGCGACGATGTCCACTTT-3′; Bovine Ifnb1: 5′-TCCAGCACATCTTCGGCATT-3′ and 5′-AGACGATTCATCTGCCAATAGAGT-3′; Bovine Mx2: 5′-CTACCGCAACATTACGCAGC-3′ and"
10. "Bovine Il-6: 5′-ACAAGCGCCTTCACTCCATT-3′ and 5′-GCGCTTAATGAGAGCTTCGG-3′; Bovine Isg56: 5′-TCACAGCAACCATGAGTTATAAAG-3′ and 5′-ATCTCCTCCAAGACCCTGTT-3′; Bovine Ifi44: 5′-ACAGTCTGCCCATTGCTGAA-3′ and REFERENCES"
11. Woods (1988) "Lumpy skin disease-a review" *Trop Anim Health Prod*
12. Yadav, Rao, Paliwal et al. (2024) "Cracking the code of lumpy skin disease: identifying causes, symptoms and treatment options for livestock farmers" *Infect Disord Drug Targets*
13. (2174)
14. Lubinga, Tuppurainen, Coetzer et al. (2014) "Transovarial passage and transmission of LSDV by Amblyomma hebraeum, Rhipicephalus appendiculatus and Rhipicephalus decoloratus" *Exp Appl Acarol*
15. Nesterov, Mazloum, Byadovskaya et al. (2022) "Experimen tally controlled study indicates that the naturally occurring recombinant vaccine-like lumpy skin disease strain Udmurtiya/2019, detected during freezing winter in northern latitudes, is transmitted via indirect contact" *Front Vet Sci*
16. Prabhu, Malmarugan, Rajagunalan et al. (2024) "Isolation and molecular characterization of lumpy skin disease virus from Tamil Nadu, India during the outbreaks from 2020 to 2022" *Virus Genes*
17. Jabbar, Atif, Kashif et al. (2025) "Molecular epidemiology and phylogenetic insights of lumpy skin disease in cattle from diverse agro-ecological regions of Punjab, Pakistan" *PLoS One*
18. Moudgil, Chadha, Khullar et al. (2024) "Lumpy skin disease: Insights into current status and geographical expansion of a transboundary viral disease" *Microb Pathog*
19. Liang, Yao, Wang et al. (2022) "Understanding the research advances on lumpy skin disease: a comprehensive literature review of experimental evidence" *Front Microbiol*
20. Xie, Cui, Liao et al. (2024) "Genomic analysis of lumpy skin disease virus asian variants and evaluation of its cellular tropism" *NPJ Vaccines*
21. Medzhitov, Iwasaki (2024) "Exploring new perspectives in immunol ogy" *Cell*
22. Kumar, Kawai, Akira (2011) "Pathogen recognition by the innate immune system" *Int Rev Immunol*
23. Olive (2012) "Pattern recognition receptors: sentinels in innate immunity and targets of new vaccine adjuvants" *Expert Rev Vaccines*
24. Chen, Gao (2024) "Innate immune signal transduction pathways to fungal infection: components and regulation" *Cell Insight*
25. Tulman, Afonso, Lu et al. (2001) "Genome of lumpy skin disease virus" *J Virol*
26. Sun, Wu, Du et al. (2013) "Cyclic GMP-AMP synthase is a cytosolic DNA sensor that activates the type I interferon pathway" *Science*
27. Li, Wu, Gao et al. (2013) "Pivotal roles of cGAS-cGAMP signaling in antiviral defense and immune adjuvant effects" *Science*
28. Liu, Xu "2025. cGAS, an innate dsDNA sensor with multifaceted functions" *Cell Insight*
29. Zhong, Yang, Li et al. (2008) "The adaptor protein MITA links virus-sensing receptors to IRF3 transcription factor activation" *Immunity*
30. Mukai, Konno, Akiba et al. (2016) "Activation of STING requires palmitoylation at the golgi" *Nat Commun*
31. Song, Xi, Dai et al. (2024) "STING guides the STX17-SNAP29-VAMP8 complex assembly to control autophagy" *Cell Insight*
32. (2025) *Full-Length Text Journal of Virology*
33. Zhang, Zhong (2022) "Regulation and function of the cGAS-MITA/ STING axis in health and disease" *Cell Insight*
34. Zhang, Xu, Ablasser (2025) "Regulation of the cGAS-STING pathway" *Annu Rev Immunol*
35. Deng, Chen, Yang et al. (2025) "The role of cGAS-STING pathway ubiquitination in innate immunity and multiple diseases" *Front Immunol*
36. Kuchitsu, Mukai, Uematsu et al. (2023) "STING signalling is terminated through ESCRT-dependent microautophagy of vesicles originating from recycling endosomes" *Nat Cell Biol*
37. Balka, Venkatraman, Saunders et al. (2023) "Termination of STING responses is mediated via ESCRT-dependent degradation" *EMBO J*
38. Zhong, Zhang, Lei et al. (2009) "The ubiquitin ligase RNF5 regulates antiviral responses by mediating degradation of the adaptor protein MITA" *Immunity*
39. Zhang, Zhang, Zhang et al. (2016) "USP18 recruits USP20 to promote innate antiviral response through deubiquitinating STING/MITA" *Cell Res*
40. Li, Lin, Tong et al. (2018) "TRIM29 negatively controls antiviral immune response through targeting STING for degradation" *Cell Discov*
41. Kaján, Doszpoly, Tarján et al. (2020) "Virus-host coevolution with a focus on animal and human DNA viruses" *J Mol Evol*
42. Linville, Rico, Teague et al. (2022) "Dysregulation of cellular VRK1, BAF, and innate immune signaling by the vaccinia virus B12 pseudokinase" *J Virol*
43. Fay, Cook, Wijesiriwardana et al. (2020) "Madin-Darby bovine kidney (MDBK) cells are a suitable cell line for the propagation and study of the bovine poxvirus lumpy skin disease virus" *J Virol Methods*
44. Ly, Burgess, Shah et al. (2022) "Vaccinia virus D10 has broad decapping activity that is regulated by mRNA splicing" *PLoS Pathog*
45. Parrish, Resch, Moss (2007) "Vaccinia virus D10 protein has mRNA decapping activity, providing a mechanism for control of host and viral gene expression" *Proc Natl Acad Sci*
46. Akther, Akter, Sarker et al. (2023) "Global burden of lumpy skin disease, outbreaks, and future challenges" *Viruses*
47. Whittle, Chapman, Williamson (2023) "Lumpy skin disease-an emerging cattle disease in" *Europe and Asia. Vaccines (Basel)*
48. Datten, Chaudhary, Sharma et al. (2023) "An extensive examination of the warning signs, symptoms, diagnosis, available therapies, and prognosis for lumpy skin disease" *Viruses*
49. Hu, Yang, Xie et al. (2016) "Sumoylation promotes the stability of the DNA sensor cGAS and the adaptor STING to regulate the kinetics of response to DNA virus" *Immunity*
50. Lian, Wei, Zang et al. (2018) "ZCCHC3 is a co-sensor of cGAS for dsDNA recognition in innate immune response" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12481871&blobtype=pdf | # From monolayer to spheroid: assessing influenza a virus infection in 2D and 3D cell culture of A549 and HEK293
Mahsa Sisakht, Hadiseh Shokouhi, Fatemeh Gholizadeh, Hosnieh Sokhandan, Parvaneh Mehrbod
## Abstract
Background and aim Three-dimensional (3D) culture models better mimic cell-to-cell interactions compared to traditional two-dimensional (2D) cultures, providing more physiologically relevant alternative for virus infection studies. This study aimed to explore the effectiveness of 3D culture models for studying viral propagation using A549 and HEK293 cell lines in spheroid form with two different matrices: alginate (Alg) and a combination of alginate with methylcellulose (Alg + MC).
MethodsThe 3D cultures of A549 and HEK293 cells were generated in 2 matrices. The cultures were characterized by proliferation assay and size assessment. The matrices were further analyzed by scanning electron microscopy (SEM) and immunofluorescence microscopy. Influenza A virus/PR/8/34 (H1N1) was propagated in MDCK cell and virus infectious dose was determined. A549 and HEK293 cells were grown in 2D form and virus was adapted to these 2 cell lines in serial passages. The best yields were inoculated to 2D and 3D forms. The supernatants and cells were collected in 48 h and subjected to qPCR to determine and compare the virus propagation in 2D and 3D formats.
ResultsSpheroids derived from A549 and HEK293 cell lines were successfully developed in 2 different matrices and characterization confirmed assembly of the cells together with considerable growth rate and viability. In case of HEK293, in dissolved patterns, external supernatant of Alg + MC and in undissolved ones, external supernatant in Alg and external and internal supernatants in Alg + MC showed the lowest decrement in viral load. Regarding A549, among dissolved ones, internal supernatants in Alg and Alg + MC and in undissolved samples, external and internal supernatants in Alg and internal supernatants in Alg + MC showed the least reduction. In both cell samples reduction was observed in all matrices, which was significant in A549 cell (P<0.05).
## Introduction
Infectious diseases have been major challenges to human wellbeing and survival for centuries [35] despite significant progresses made in the prevention and control of infectious diseases. Therefore, it is very important to focus on improved mechanisms of viral diseases diagnosis and treatment to respond more effectively. The use of in vitro models to study virus-host interactions is critical for understanding pathogenesis and developing antiviral therapies. Our current conception of many human viral diseases and antiviral drugs is based on traditional twodimensional (2D) cell culture. This way, cells are grown on flat plastic material such as petri-dishes and multiwell plates [6,32,38]. Actually, 2D cell cultures have greatly contributed to the research of infectious disease etiology, the immune mechanisms used to defend versus human viruses [7,38], and especially the biochemistry and molecular biology of the viral replication [4]. However, 2D monolayer culture has significant limitations and often fails in mimicking physiological complexity of in vivo microenvironment. This model cannot precisely reproduce the natural and actual infection process [7].
In addition, the use of animal models is expensive and involves ethical issues [39]. A large number of pathogens are species-specific, and pathogenesis mechanisms cannot be captured by animal models [22].
To overcome these limitations associated with 2D monolayers and animal models, a variety of three-dimensional (3D) cell culture models have been developed [16].
These culture models, including spheroids, organoids, and scaffold-based systems, better mimic tissue structure, cell-to-cell interactions, and gradients of nutrients and oxygen, providing more physiologically relevant platforms for virus infection studies. The 3D cell models bridge the gap between 2D cell cultures and animal models by providing an in vitro cell model system that more closely mimic in vivo microenvironment, which contribute understanding the virus-host interactions and underlying mechanisms of human viruses [5,50]. Despite these advantages, 3D models can be more technically challenging, time-consuming, and costly [37].
The 3D models have been used to study the most common respiratory viruses including influenza, parainfluenza, respiratory syncytial virus (RSV), and coronaviruses and other viruses such as zika virus, oncolytic viruses, and poxviruses [23].
3D cultures are increasingly becoming indispensable tools in research. Some hydrogels are hydrophilic polymers that retain large amounts of water after gelation. Hydrogels have a high water storage capacity due to their microscopic pores, which facilitate the transport of nutrients, O 2 , growth factors, metabolic waste, etc. Alginate (Alg) is a biopolymer derived from brown algae and is also a natural polymer with the ability to form hydrogels. It is known for its properties, such as its ability to diffuse nutrients and other environmental substances, including drugs [43].
Methylcellulose (MC) also appears to be an interesting water-soluble and heat-sensitive material. MC is methylated cellulose in which some hydrophilic hydroxyl groups (-OH) are replaced by hydrophobic methoxy groups (-OCH3). Methoxy groups lead to the solubility of MC in water and prevent crystallization. MC is also a viscosityincreasing polymer that is widely used in food and color industry. MC-based polymers are able to self-assemble due to temperature changes and exhibit a thermally reversible sol-gel transition [3]. However, it has several limitations such as lack of extracellular matrix (ECM) components, limited mechanical tenability, poor nutrient and oxygen diffusion, weak cell-substrate interactions, batch variability, and limited suitability for long-term cultures. Despite these limitations, MC remains valuable for specific 3D culture applications, such as spheroid formation assays, colony formation assays, and viral infection models when combined with other materials such as Alg or supplemented with ECM components [42].
A549 cells are human alveolar epithelial cells and in terms of origin they are close to the human body conditions [51]. They have been used successfully in influenza virus studies [11,19,49,53]. A549 cells have been used in 3D culture of RSV [42]. These cells have been applied for influenza studies using 3D bioprinting technology [8].
HEK293 cells, which are human embryonic kidney cells, have also been widely used in influenza studies [20]. These cells have been used in both adherent and suspended forms [31]. HEK293SF cells were examined in serum-free medium and suspended in perfusion culture model (continuous change of medium) using 3-liter bioreactors, and the kinetics of influenza virus production in this cell culture was successfully tracked [40,41]. Researchers recently suggested that HEK293T cells cultured on uncoated polystyrene dishes may be useful for analyzing properties of 3D-MSC multicellular spheroids
## Conclusion
We conclude that Alg + MC matrix, with its increased porosity and lower cohesion compared to Alg alone, was easier to dissolve but more difficult to re-solidify. One possible explanation for the observed higher viral replication in this matrix is that it may have facilitated improved viral access to the cells. Future modifications that increase virus-cell interaction time in this system could further enhance infection efficiency. Keywords A549, Alginate, HEK293, Influenza a virus, Methylcellulose, Three-dimension [26]. While they are not natural lung cells, their versatility outweighs tissue origin limitations for many applications in research laboratories [15].
In this study, we tried to set-up spheroid formation from A549 and HEK293 cells, and characterize them in order to propagation for influenza A virus (A/PR/8/34).
## Materials and methods
## 2D cell culture
HEK293 cells were obtained from the Innovation Center for Stem Cell and Regenerative Medicine, Tehran University of Medical Sciences and A549 cells were obtained from the Cell Bank, Pasteur Institute of Iran. The cells were cultivated in Dulbecco's Modified Eagle's Medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS) (Gibco) and 1% penicillin/streptomycin (Pen/Strep) (Sigma). The cells were incubated at 37 °C, 5% CO 2 incubator (Memmert/Thermo Scientific). Cells were then sub-cultured using 0.25% trypsin-EDTA solution when the confluency reached 80-90%.
MDCK cells received from the Influenza and Respiratory Viruses Department, Pasteur Institute of Iran were also cultured in culture flasks containing DMEM and 10% FBS and 1% Pen/Strep. When the cells reached to 80% confluency, they were sub-cultured and frozen at -80 °C to prepare the stock.
## Hydrogel preparation
In order to make sodium alginate hydrogel, sodium Alg powder (300 mg) was dissolved in DMEM high glucose (10 mL) to make a 3% (w/v) solution. Then, sodium alginate powder (150 mg) mixed with MC (150 mg) was dissolved in DMEM high glucose (10 mL) to make a 3% (w/v) solution of Alg + MC. Both preparations were kept on magnetic hot plate stirrer (300 RPM) for 2 h to make homogeneous hydrogels [45].
## Matrices characterization by scanning electron microscopy
Hydrogels were frozen at -80 °C and lyophilized by freeze-dryer (Pishtaz engineering, FD6, Iran) for 18 h. The samples were then coated with gold before observation under Scanning Electron Microscopy (SEM) (MIRA3 TESCAN) [44].
## Spheroid formation
A549 and HEK293 cells were blended in Alg and Alg + MC hydrogels with densities of 1 × 10 6 cells/ mL. The Alg-cell and Alg + MC-cell suspensions were dropped (30 µL) into CaCl 2 solution (40 mM) and maintained for 10 min at room temperature to complete the cross-linking by CaCl 2 . Subsequently, CaCl 2 was removed and DMEM with 10% FBS (200 µL) was added to each well. The drops were then incubated for 10 days with media changed every two days [47]. In order to dissolve the spheroids, in case of virus propagation, the spheroids from Alg were incubated at room temperature with phosphate-buffered saline (PBS) for 2 min along with gentle pipetting (up and down), and for Alg + MC time decreased into 30 s.
## Spheroid viability
Viability and proliferation rate of spheroids were determined by Alamar Blue (Invitrogen DAL1025) after embedding into the matrices. The assay is based on fluorescence reading of resorufin converted by cell enzymes from resazurin and allows measurement of the signal from spheroids [17]. Absorbance was read on days 0, 3, 5, 7, and 10 after 3D culture at 570 and 600 nm (emission and excitation filters, respectively). The experiment was repeated at least three times for 4 groups (Alg-A549, Alg-HEK293, Alg-MC-A549, and Alg-MC-HEK293) as compared to 2D cells and cell-free hydrogels.
## Spheroid size measurement
Spheroid images were taken by bright field microscopy (Labomed, USA) on each day of spheroids culture untill day 10. Each condition was repeated at least three times and readings were done in duplicate. Images were analyzed by Image J software. Diameters of spheroids were measured in three images of each condition [47]. Data was represented as mean ± SD.
## Immunofluorescent staining
Spheroids were selected on the best day based on the result of proliferation experiment as compared with matrices (for A549/Alg on day 5 and for HEK293/ Alg + MC on day 7). Then, they were fixed with 4% paraformaldehyde (PFA; Sigma Aldrich) prepared in PBS for 30 min at room temperature, followed by 3 times washing by blocking buffer (0.5% FBS in PBS). The spheroids were kept in glycine 0.1 M for 30 min at room temperature. Then, they were permeabilized with 0.5% triton-X100 in PBS and incubated for 30 min at room temperature. Followed by 3 times washing by PBSTD washing solution (PBS + 0.3% triton + 1% DMSO + 0.5% FBS), goat serum primary antibody (F-actin, life technologies) was added to cells (0.5%) and incubated overnight at 4 °C. The day after, secondary antibody (Alexa Fluor 594 goat anti-rabbit, 2268327) was added and incubated for 2 h at room temperature on a shaker. Followed by 3 times washing, mounting solution containing DAPI (Abcam) was applied on spheroids and covered by coverslip. The samples were visualized using fluorescence microscopy (Zeiss/Nikon/ Leica/Olympus) [9].
## Virus inoculation and infectious dose determination on MDCK cells
Influenza virus type A (H1N1, PR/8/34) was grown in MDCK cells in media containing Trypsin-TPCK enzyme (M + T) (1 µg/mL) and the viral titer was determined using the HA method [28,46]. Virus stock was then collected and stored aliquoted at -80 °C. The virus infectious dose (TCID 50 ) was then determined. Briefly, 10-fold serial dilutions of the stock virus were exposed to MDCK cells in 96-well plate and M + T (1 µg/mL) was added. Following 48 h incubation at 37 °C, TCID 50 test was performed using the HA method and Karber formula [27,46].
## Adaptation of influenza virus to A549 and HEK293 2D cells
A549 (560-600 × 10 3 cell/well) and HEK293 (450-470 × 10 3 cell/well) cells were cultured in 6-well plates. After washing with PBS, cell line alone and/or in co-culture with MDCK were exposed to PR8 virus harvested from MDCK cells for 1-1.5 h. Then, media containing TPCK (1 µg/mL) was added and incubated for 24 to 48 h at 37 °C and 5% CO 2 . The collected supernatants were inoculated in consecutive passages (direct or 1:10 dilution) and harvested after 24 to 48 h depending on cytopathic effect (CPE) observed (with or without HA test result). Some volumes of supernatants were stored at -80 °C to perform qPCR test and some were used for another round of the cell inoculation at the same time. The consecutive cycle of inoculation continued until the HA test result obtained and increased. The infectious dose of the last supernatants with good HA titers and low Ct values were determined by TCID 50 assay as mentioned before.
## Molecular evaluation
To confirm the adaptation and propagation of the virus in consecutive inoculations, the level of viral gene expression was checked. For this purpose, the supernatants were collected and viral RNAs were extracted using High Pure Viral Nucleic Acid Kit (Roche, Germany) according to the manufacturer's protocol. RNA samples were stored at -80 °C for further use. Then, RNA samples were exposed to cDNA synthesis using Transcriptor First Strand cDNA Synthesis kit (Roche, Germany). The samples were incubated for 10 min at 25 °C for the primers annealing. Then, they were placed at 55 °C for 30 min and finally at 85 °C for 5 min and stored at -20 °C. The concentration of cDNA samples was measured using Picodrop Spectrophotometer system (Alpha Biotech, UK). Real-time PCR reactions were performed using the Light Cycler FastStart DNAMaster SYBR Green I kit (Roche, Germany) in Corbett Rotor-Gene Q 6000 machine, (Australia), according to the manufacturer's instructions. The thermal cycling program was as follows: 95 °C for 10 min, 35 cycles (95 °C for 10 s, 55 °C for 10 s, 72 °C for 30 s), and 72 °C for 5 min. The primers were used against M gene of influenza A virus synthesized by Pishgam Co (Iran). Primers specifications are as follows:
All PCR reactions were performed in duplicate along with a negative control.
$$F.PR8-G A C C A A T C C T G T C A C C T C T G A C and R.PR8-A G G G C A T T T T G G A C A A A G C G T C T A.$$
## Inoculation of the adapted viruses into 3D cultures
The viruses were inoculated on 3D cultures of A549 and HEK293 cells and simultaneously on 2D cultures. Inoculation in 3D cell culture was divided into two groups: in Alg and Alg + MC. For 1 set of 3D cell experiment, drops containing cells were first dissolved in PBS with several gentle pipetting, inoculated with virus for 1 h, and then centrifuged. Drops of Alg were prepared in CaCl 2 (40 mM) as mentioned above and M + T (1 µg/mL) was added. For the second set of 3D cell experiment, drops were not dissolved. They were inoculated with virus for 1 h in intact form, centrifuged, the supernatant was removed, and then M + T (1 µg/mL) was added.
They were all incubated for 48 h at 37 °C. The morphology of the cells was monitored at 0, 24, and 48 h incubation times. Following 48 h, cells and supernatants were separated and stored at -80 °C. In the 3D settings, the supernatants outside droplets were considered as the external supernatant (S1) and the supernatants inside droplets were considered as the internal supernatant (S2).
To determine the infectivity of the virus in two biomaterials matrices, and compare the results of two sets for the infectivity of the virus in 2D and 3D cells, TCID 50 test was performed. All the cells and supernatants collected from the 2D and 3D tests at this stage were also subjected to qPCR and obtained Ct values were normalized to Ct values of virus inoculation in 2D format.
## Statistical analysis
Graphpad prism Version 10.2.2 was used for the statistical analysis. Proliferation and size of spheroids were analyzed by two-way ANOVA test, repeated measure. The results of 3D tests were analyzed using One-way ANOVA, LSD post-hoc test to compare all investigated groups with 2D viral sample. P value less than 0.05 was considered significant.
## Results
## Matrices characterization by SEM
SEM imaging was performed to evaluate the structure of prepared hydrogel matrices (Alg and Alg + MC). Figure 1 shows hydrogels in two different magnifications. Hydrogels have a porous structure with interconnected porosities to allow flow of nutrient and removal of waste from cell, although the Alg + MC showed more porosity than Alg matrix.
## Spheroid formation and characterization
Spheroids were observed from day 3 after culture. The ability of cells to form spheroids is influenced by the expression of surface adhesion proteins such as E-cadherin, N-cadherin, and integrins that mediate cell-cell and cell-matrix interactions [33], along with the cell proliferation rate, with epithelial-origin cells like HEK293 typically exhibiting greater sphericity [16]. Figure 2 represents different morphologies of spheroids in different days after culture in both groups. Specific features of spheroids such as round shape, distinct assembly, and the absence of necrotic core during culture were observed.
As Figs. 2 and 3A show, the biggest spheroid sizes were observed in HEK293 cells. One possible explanation for this observation is the epithelial origin of HEK293, as epithelial cells often display strong cell-cell adhesion and self-assembly capacity in 3D culture [21]. However, this interpretation is based on known cell-type characteristics and was not directly focused in our study.
In both cell lines, size decreased significantly after day 10, which highlights reduction in proliferation rate as well (Fig. 3B andC).
Quantitative comparison of spheroid size across both cell lines and matrices was performed, and the results are presented in the Fig. 3A. Data showed that HEK293 cells formed significantly larger spheroids in the Alg + MC matrix compared to A549, particularly on day 7 (P < 0.0001). In contrast, under Alg-only condition, spheroid size trends between A549 and HEK293 were more similar, with A549 spheroids surpassing HEK293 in size on day 10 (P < 0.05).
These findings indicate that HEK293 cells exhibit enhanced 3D self-assembly and spheroid growth in Alg + MC, while A549 cells show relatively better growth in Alg alone over time. Statistical analysis confirmed that HEK293 spheroids in Alg + MC were significantly larger than A549 spheroids on day 7 (P < 0.0001), while in Alg-only conditions, A549 spheroids were modestly but significantly larger than HEK293 spheroids on day 10 (P < 0.05). These results strengthen the morphological observations and demonstrate that both cell types and matrix compositions significantly influence the spheroid characteristics.
The proliferation rates of A549 and HEK293 cells embedded in two matrices; Alg and Alg + MC were monitored over 10 days and compared to baseline (day 0), as shown in Fig. 3B andC. A549 cells exhibited a lower proliferation rate in the Alg + MC matrix compared to Alg, with a statistically significant decrease observed after day 5 (P < 0.05).
For HEK293 cells, normalized proliferation data relative to 2D cultures demonstrated a more stable trend, although a gradual decline was noted from day 3 to day 10. On day 3, a significant increase in proliferation was observed in HEK293 cells within the Alg matrix (P = 0.0001), whereas the increase in Alg + MC group on the same day was not statistically significant. Nonetheless, cell viability remained above 90%, indicating acceptable metabolic activity. A549 cells embedded in the Alg matrix also showed a significant increase in proliferation on day 5 (P = 0.0309), supporting the suitability of this matrix for short-term culture. Figure 3C illustrates normalized proliferation data across both matrices, highlighting overall consistency in cell viability and growth dynamics.
These findings Helped identify the optimal time window for downstream applications. Based on combined proliferation and viability data, days 3 to 5 were selected as the most appropriate time points for influenza virus inoculation.
Supporting this, immunofluorescence staining of fixed spheroids revealed well-formed 3D structures in both cell lines: A549 spheroids cultured in Alg on day 5, and HEK293 spheroids in Alg + MC on day 7 (Fig. 4). As shown in Fig. 4, staining with DAPI and F-actin confirmed the presence of organized nuclei and actin cytoskeletons, indicating successful spheroid assembly.
## Virus adaptation result on HEK293 and A549 cells
The results of HA and qPCR tests from the consecutive inoculations of the virus onto HEK293 and A549 cells (either alone or in co-culture with MDCK) are given in Supplementary File 1. As shown in Fig. 5, the 8th inoculate from HEK293 cells and 5th inoculate from A549 cells with HA titers 32 and Ct values 18.65 and 19.70, respectively were collected for infectious dose determination. The highest TCID 50 values from the selected samples were obtained at 10 3.5 and 10 2.5 for the HEK293 and A549 inoculates, respectively. Morphology of the cells in consecutive passages and inoculations are shown in Fig. 5. The consecutive inoculations of virus on both cells caused gradual increasing CPE compared to negative control cells.
## 3D cell cultures inoculation outcome
Morphology of the cells at 0, 24, and 48 h incubation times was monitored and CPE was recorded. The images are depicted in Supplementary File 2. Our limitation here for not being able to depict details and morphological changes goes to the low quality of camera.
All cells and supernatants (external (S1) and internal (S2)) collected from 2D and 3D samples were subjected to qPCR and the changes in viral expression in terms of Ct values, ΔCt, and 2 ˄ ΔCt were calculated (Data are shown in Supplementary File 3). In case of HEK293 cells, among dissolved ones, the external supernatant in Alg + MC showed the lowest reduction in virus gene expression (11.70 fold). Among un-dissolved ones, external supernatant in Alg (14.48 fold) and external (36.91 fold) and internal (19.66 fold) supernatants in Alg + MC showed the lowest reduction. In the cell samples, the fold reduction was also calculated and reduction was observed in all dissolved and un-dissolved samples.
In case of A549 cells, among the dissolved ones, internal supernatants in Alg (23.57 fold) and Alg + MC (5.31 fold) showed the lowest reduction. Among un-dissolved ones, external (53.86 fold) and internal (51.58 fold) supernatants in Alg and internal supernatant in Alg + MC showed the lowest reduction (3.30 fold). In the cell samples, same as HEK293 cells, reduction was observed in all dissolved and un-dissolved samples.
A comprehensive evaluation of Ct values obtained from 3D cultures' cells and supernatants was conducted by normalizing Ct values to 2D samples Ct value (Fig. 6).
Significant differences were observed in viral gene expression between several 3D treatments and 2D treatment. The results were different based on cell type.
In HEK293 cells, none of the 3D treatments showed significant reduction in viral gene expression (P˃0.05), but in case of A549 cells, all 4 dissolved and un-dissolved treatments in both Alg and Alg + MC matrices showed significant reduction compared to 2D (P˂0.05). Analysis of the supernatants showed another outcome. In HEK293 cells, the external and internal supernatants dissolved in Alg and un-dissolved internal supernatant in Alg showed significant lower viral levels compared to 2D (P˂0.05). In A549 cells, external supernatant dissolved in Alg showed the most significant reduction compared to 2D (P˂0.05).
Overall, although viral genome expression as measured by Ct values did not show profound increase in our 3D setting design as expected, major reduction was not observed either that demonstrates the high potential of this 3D design to be optimized for antiviral performance improvement.
## Discussion
In this study, we successfully established matrix-based 3D culture systems for A549 and HEK293 cells using two hydrogel compositions: Alg and combination of Alg + MC to study influenza virus replication in 3D cell format. We observed that Alg + MC matrix supported higher viral titers compared to Alg alone, suggesting it provides more favorable microenvironment for infection.
The novelty of our approach lies in using 3D spheroid cultures for viral propagation in two different culture conditions. 3D culture systems are generally considered to better mimic in vivo tissue microenvironments compared to traditional 2D cultures, as reported in previous studies. In our study, Alg and Alg + MC spheroid models were used as exploratory systems for viral propagation.
. Most virology studies still rely on 2D cultures, despite their inability to replicate the complexity of tissue environments. Incorporating MC into the Alg matrix improves viscosity and structural stability [48], leading to more robust constructs with better nutrient diffusion and cell-matrix interactions [1]. These low-cost, biocompatible substrates offer a practical alternative for virological studies, with potential implications for the vaccine development and antiviral screening. Shift from 2D to 3D culture in virology represents more than a technical adjustment-it enhances the biological relevance of the experimental models. 3D systems can more accurately replicate infection dynamics, heterogeneity in viral spread, and host responses observed in vivo [13]. Several viruses-including SARS-CoV-2 [14], Zika virus [10], and hepatitis viruses [34,36] have been studied using spheroids, organoids, or bio printed tissues [5,16,18,24].
Spheroids were characterized by SEM, DAPI/F-actin staining, and proliferation/morphology analyses. Optimal spheroid structures were observed on days 3-5, which we identified as the most suitable window for the virus inoculation. The combination of Alg + MC produced more porous and less cohesive scaffolds, likely facilitating greater viral access to the cells. These structural properties may have contributed to the enhanced viral replication observed in Alg + MC spheroids [30,52].
While we initially attempted to quantify infectious dose using TCID 50 assay, titers remained low-likely due to limited expression of sialic acid (SA) α2, 6 receptors in A549 and HEK293 cells, which are crucial for influenza virus entry. This limitation may be overcome by extending viral adaptation periods or optimizing matrix stiffness. Although the indirect immunofluorescence confirmed virus expression in MDCK cells (Supplementary file 4), clear and publishable images were not obtained for A549 or HEK293 cells. Again, this also could be attributed to low SA receptor expression or matrixrelated environmental factors that impair infectivity and visualization. Despite this defect, qPCR analysis provided reliable data for comparing viral loads across conditions consistent with prior studies [25].
Our findings suggest that Alg + MC, by promoting better diffusion and cell-virus interaction, supports improved viral replication compared to Alg alone. However, no major differences were observed between dissolved and undissolved spheroids, indicating that matrix composition was more influential factor than physical state at the time of 3D structure preparation. The incorporation of MC into Alg matrix has been shown to improve the structural integrity and functionality of 3D cultures. MC enhances the viscosity of Alg solution, facilitating the formation of stable, high-fidelity constructs suitable for the extended cell culture [2].
Compared to the established 3D systems such as Matrigel, collagen-Matrigel scaffolds, and organoids, our Algbased and Alg + MC models offer notable advantages in terms of cost, reproducibility, and ease of handling. Matrigel, although rich in extracellular matrix (ECM) proteins, and biologically relevant, suffers from high cost, batch-to-batch variability, animal-derived origin, and poor scalability factors [47] that limit its suitability for large-scale or translational virology applications [12] [29]. In contrast, Alg and Alg + MC are inexpensive, chemically defined, and easy to manipulate, making them attractive alternatives for such studies.
Despite these benefits, our 3D spheroid models did not consistently outperform 2D cultures in terms of viral yield, emphasizing the need for further optimization. This limitation could potentially be related to the factors such as compact spheroid structure, restricted virus diffusion, or the inherently low expression of influenza-specific receptors (e.g., SA α2,6) in A549 and HEK293 cells. To enhance infectivity in such systems, several strategies could be explored: extending virus-cell contact time, modulating matrix stiffness, functionalizing hydrogels with ECM ligands, or testing alternative materials such as collagen-Matrigel blends and synthetic commercial hydrogels like TrueGel3D. Alg-based hydrogels, bioinert by nature, can be engineered to support more physiologically relevant virus-host interactions.
In conclusion, our 3D models provide an important foundation for studying influenza infection in a more physiologically relevant context. However, additional refinements are necessary to enhance their utility for virus propagation, evaluation of specific steps of the viral lifecycle within the 3D system, and downstream applications like antiviral drug screening and vaccine development. Future work should focus on modulating 3D microenvironment to better recapitulate in vivo conditions and improve viral yield.
## Conclusion
This study presents the first comparative analysis of H1N1 influenza virus propagation in A549 and HEK293 cells cultured in alginate and alginate-methylcellulose 3D matrices versus traditional 2D systems. Our findings underscore the importance of 3D microenvironments in supporting viral infection and replication, offering a more physiologically relevant and scalable platform for virology and antiviral research. Our results suggest that addition of methylcellulose contributed to improved spheroid structural integrity and stability, especially in HEK293 cells; however, the impact on influenza A virus propagation was variable and dependent on the cell type and matrix condition. The optimal choice of hydrogel for 3D virus culture depends on multiple factors, including virus type, cell surface properties, hydrogel composition, culture duration, and the biological goals of the experiment. As 3D culture technologies continue to evolve, such models will enable more predictive virology studies and facilitate antiviral drug development. Expanding this platform to other influenza strains or respiratory viruses could offer deeper insight into the virus-host interactions and improve screening strategies. While this study did not specifically dissect individual stages of the viral life cycle within the 3D matrices, our model lays the groundwork for such future investigations. Further optimization and mechanistic studies will be essential to fully validate and refine this system for in-depth virological applications.
## References
1. Ahlfeld, Cubo-Mateo, Cometta et al. (2020) "A novel plasma-based bioink stimulates cell proliferation and differentiation in bioprinted, mineralized constructs" *ACS Appl Mater Interfaces*
2. Ahlfeld, Guduric, Duin et al. (2020) "Methylcellulose -a versatile printing material that enables biofabrication of tissue equivalents with high shape fidelity" *Biomater Sci*
3. Altomare, Cochis, Rimondini et al. (2016) "Thermo-responsive methylcellulose hydrogels as temporary substrate for cell sheet biofabrication" *J Mater Sci Mater Med*
4. Andrei (2006) "Three-dimensional culture models for human viral diseases and antiviral drug development" *Antiviral Res*
5. Antoni, Burckel, Josset et al. (2015) "Three-dimensional cell culture: a breakthrough in vivo" *Int J Mol Sci*
6. Baker, Chen (2012) "Deconstructing the third dimension-how 3D culture microenvironments alter cellular cues" *J Cell Sci*
7. Barrila, Radtke, Crabbé et al. (2010) "Organotypic 3d cell culture models: using the rotating wall vessel to study host-pathogen interactions" *Nat Rev Microbiol*
8. Berg, Hiller, Kissner et al. (2018) "Optimization of cell-laden bioinks for 3D bioprinting and efficient infection with influenza A virus" *Sci Rep*
9. Bergdorf, Phifer, Bechard et al. (2021) "Immunofluorescent staining of cancer spheroids and fine-needle aspirationderived organoids" *STAR Protoc*
10. Berglund, Lennon, Badu et al. (2024) "Zika virus infection in a cell culture model reflects the transcriptomic signatures in patients" *bioRxiv*
11. Bhowmick, Derakhshan, Liang et al. (2018) "A three-dimensional human tissue-engineered lung model to study influenza A infection" *Tissue Eng Part A*
12. Caliari, Burdick (2016) "A practical guide to hydrogels for cell culture" *Nat Methods*
13. De Dios-Figueroa, Aguilera-Marquez, Camacho-Villegas et al. (2021) "cell culture models in COVID-19 times: a review of 3D technologies to understand and accelerate therapeutic drug discovery" *Biomedicines*
14. De Melo, Benincasa, Cruz et al. (2021) "3D culture models to study SARS-CoV-2 infectivity and antiviral candidates: from spheroids to bioprinting" *Biomed J*
15. Durocher, Perret, Kamen (2002) "High-level and high-throughput recombinant protein production by transient transfection of suspension-growing human 293-EBNA1 cells" *Nucleic Acids Res*
16. Edmondson, Broglie, Adcock et al. (2014) "Three-dimensional cell culture systems and their applications in drug discovery and cell-based biosensors" *Assay Drug Dev Technol*
17. Eilenberger, Kratz, Rothbauer et al. (2018) "Optimized alamarblue assay protocol for drug dose-response determination of 3D tumor spheroids" *MethodsX*
18. Friedl, Sahai, Weiss et al. (2012) "New dimensions in cell migration" *Nat Rev Mol Cell Biol*
19. Fujioka, Nishide, Ose et al. (2018) "A sialylated voltage-dependent Ca2 + channel binds hemagglutinin and mediates influenza A virus entry into mammalian cells" *Cell Host Microbe*
20. Graham, Smiley, Russell et al. (1977) "Characteristics of a human cell line transformed by DNA from human adenovirus type 5" *J Gen Virol*
21. Green, Fu, Ndahiro et al. (2025) "Development of an HEK293 suspension cell culture medium, transient transfection optimization workflow, and analytics for batch rAAV manufacturing" *Biotechnol Bioeng*
22. Griffith, Swartz (2006) "Capturing complex 3d tissue physiology in vitro" *Nat Rev Mol Cell Biol*
23. Harb, Fakhreddine, Zaraket et al. (2021) "Three-dimensional cell culture models to study respiratory virus infections including COVID-19" *Biomimetics*
24. He, Chen, Zeng (2016) "Three-dimensional cell culture models for investigating human viruses" *Virol Sin*
25. Hussain, Cordeiro, Sevilla et al. (2010) "Comparison of egg and high yielding MDCK cell-derived live attenuated influenza virus for commercial production of trivalent influenza vaccine: in vitro cell susceptibility and influenza virus replication kinetics in permissive and semi-permissive cells" *Vaccine*
26. Iuchi, Oya, Hosoya et al. (2020) "Different morphologies of human embryonic kidney 293T cells in various types of culture dishes" *Cytotechnology*
27. Karber (1931) "50% end point calculation" *Archive Experimental Pathol Pharmacol*
28. Khodsiani, Kianmehr, Brycki et al. (2023) "Evaluation of the antiviral potential of gemini surfactants against influenza virus H1N1" *Arch Microbiol*
29. Knight, Przyborski (2015) "Advances in 3D cell culture technologies enabling tissue-like structures to be created in vitro" *J Anat*
30. Lawko, Plaskasovitis, Stokes et al. (2021) "3D tissue models as an effective tool for studying viruses and vaccine development" *Front Mater*
31. Ru, Jacob, Transfiguracion et al. (2010) "Scalable production of influenza virus in HEK-293 cells for efficient vaccine manufacturing" *Vaccine*
32. Li, Cui (2014) "Three-dimensional perfused cell culture" *Biotechnol Adv*
33. Lin, Chang (2008) "Recent advances in three-dimensional multicellular spheroid culture for biomedical research" *Biotechnol J*
34. Liu, Chen, Hagedorn (2014) "Direct visualization of hepatitis C virus-infected Huh7. 5 cells with a high titre of infectious chimeric JFH1-EGFP reporter virus in three-dimensional matrigel cell cultures" *J Gen Virol*
35. Morens, Folkers, Fauci (2004) "The challenge of emerging and reemerging infectious diseases" *Nature*
36. Murakami, Inoue, Hmwe et al. (2008) "Dynamic behavior of hepatitis C virus quasispecies in a long-term culture of the three-dimensional radial-flow bioreactor system" *J Virol Methods*
37. Ngo, Kashani, Imbalzano et al. (2018) "Additive manufacturing (3D printing): A review of materials, methods, applications and challenges" *Compos Part B: Eng*
38. Nickerson, Richter, Ott (2007) "Studying host-pathogen interactions in 3-D: organotypic models for infectious disease and drug development" *J Neuroimmune Pharmacol*
39. Page, Flood, Reynaud (2013) "Three-dimensional tissue cultures: current trends and beyond" *Cell Tissue Res*
40. Petiot, Amine (2013) "Real-time monitoring of influenza virus production kinetics in HEK293 cell cultures" *Biotechnol Prog*
41. Petiot, Jacob, Lanthier et al. (2011) "Metabolic and kinetic analyses of influenza production in perfusion HEK293 cell culture" *BMC Biotechnol*
42. Saleh, Harb, Soudani et al. (2020) "A three-dimensional A549 cell culture model to study respiratory syncytial virus infections" *J Infect Public Health*
43. Samimi, Sohi, Irani et al. (2021) "Alginate-based 3D cell culture technique to evaluate the half-maximal inhibitory concentration: an in vitro model of anticancer drug study for anaplastic thyroid carcinoma" *Thyroid Res*
44. Santana, Nedel, Ferrúa et al. (2015) "Comparing different methods to fix and to dehydrate cells on alginate hydrogel scaffolds using scanning electron microscopy" *Microsc Res Tech*
45. Gajić, Savić, Svirčev (2023) "Preparation and characterization of alginate hydrogels with high water-retaining capacity" *Polymers*
46. Targhi, Mehrbod, Fotouhi et al. (2022) "In vitro anti-influenza assessment of anionic compounds ascorbate, acetate and citrate" *Virol J*
47. Sisakht, Gholizadeh, Hekmatirad et al. (2025) "Cost-reduction strategy to culture patient derived bladder tumor organoids" *Sci Rep*
48. Sadeghzade, Yang, Li et al. (2024) "Highly stretchable alginate/methylcellulose hydrogels for 3D bioprinting: photopolymerization approach enhancing structural integrity" *Giant*
49. Ujie, Takada, Kiso et al. (2019) "Long-term culture of human lung adenocarcinoma A549 cells enhances the replication of human influenza A viruses" *J Gen Virol*
50. Urzì, Gasparro, Costanzo et al. (2023) "Threedimensional cell cultures: the Bridge between in vitro and in vivo models" *Int J Mol Sci*
51. Wu, Wang, Liu et al. (2017) "Characterization of air-liquid interface culture of A549 alveolar epithelial cells" *Braz J Med Biol Res*
52. Xie, Pal, Yu et al. (2024) "A comprehensive review on 3D tissue models: biofabrication technologies and preclinical applications" *Biomaterials*
53. Yang, Zhang, Yang et al. (2021) "Differential mitochondrial proteomic analysis of A549 cells infected with avian influenza virus subtypes H5 and H9" *Virol J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12830800&blobtype=pdf | # Molecular mechanisms of receptor recognition and antibody neutralization of coxsackievirus A6
Xianliang Ke, Xue Li, Zeyu Liu, Kexin Liu, Weichi Liu, Xingyu Yan, Bo Shu, Chao Zhang
## Abstract
Coxsackievirus A6 (CVA6), a major cause of hand, foot, and mouth disease, lacks approved vaccines or drugs. KRM1 is its only known receptor, but its precise role remains unclear. This study investigates CVA6's entry mechanism and antibody neutralization. Cryo-EM shows CVA6 clinical strain HeB primarily exists as mature virions. KRM1 binding within the canyon triggers conversion to uncoating intermediate, defining KRM1 as an uncoating receptor for CVA6. However, KRM1 knockout reduces CVA6 infectivity without affecting attachment. Conversely, disrupting heparan sulfate proteoglycan (HSPG) impairs both viral attachment and infectivity, and CVA6 virions bind heparin directly. These results support a two-receptor entry model for CVA6: HSPG mediates viral attachment, while KRM1 induces uncoating. Additionally, we develop two CVA6-specific protective antibodies (1F4 and 3H7), targeting a new antigenic site near the three-fold axis of the viral capsid. These antibodies sterically block KRM1 binding and function post-attachment, consistent with KRM1's role. The findings elucidate CVA6 entry and offer a basis for antibody interventions.
Hand, foot, and mouth disease (HFMD), a highly contagious viral illness, has become a significant global public health challenge, due to frequent large-scale epidemics across Asia and other regions 1 . HFMD is mainly caused by species A enteroviruses (EV-A) within the Picornaviridae family. Historical HFMD outbreaks have been predominantly associated with enterovirus A71 (EV-A71) and coxsackievirus A16 (CVA16). However, recent epidemiological surveys indicate that coxsackievirus A6 (CVA6) has rapidly emerged as the leading causative agent of HFMD in many countries, surpassing EV-A71 and CVA16 [2][3][4][5][6] . Unlike classic HFMD, CVA6 infections are frequently associated with atypical clinical features, such as widespread vesiculobullous eruptions, onychomadesis, and severe complications like meningitis, myocardial injury, and pulmonary edema 1,7 . This highlights the urgent need for research on CVA6's pathogenic mechanisms and effective countermeasures.
CVA6 is a nonenveloped RNA virus with a ~30 nm icosahedral capsid composed of 60 protomers. Cryo-electron microscopy (cryo-EM) studies reveal that CVA6 has three distinct particle forms: mature virions, uncoating intermediate A (altered) particles, and empty particles 8,9 . CVA6 virions contain the infectious RNA genome within a compact capsid composed of VP1, VP2, VP3, and VP4. A-particles have viral RNA and an expanded capsid lacking VP4. Empty particles have an expanded capsid of VP1, VP3, and uncleaved VP0, but no viral RNA 8,9 . All three particle forms share key structural features: a star-shaped protrusion at the five-fold axis, surrounded by a deep depression (canyon), and a protrusion at the three-fold axis. In mature virions, VP1 hydrophobic pocket beneath the canyon contains a lipid molecule termed pocket factor, which stabilizes the capsid 8 . This pocket is empty in A-particles and empty particles 8,9 . Unlike most enteroviruses where mature virions are predominant 10,11 , studies of purified CVA6
## Results
## Development and characterization of neutralizing MAbs against CVA6
To generate CVA6-specific neutralizing MAbs, hybridomas were produced from mice immunized with purified CVA6-TW-141 (CVA6-141) virions. These hybridomas were subsequently screened for neutralizing activity against the CVA6-141 strain. Ultimately, two hybridoma clones (1F4 and 3H7; IgG2a isotype) demonstrating potent neutralizing activity were isolated (Fig. 1a). MAbs 1F4 and 3H7 were purified, and their neutralizing activity against both CVA6-141 and CVA6-HeB strains was quantified using two methods: (1) the minimal neutralizing concentration (MNC), defined as the lowest antibody concentration that completely prevented cytopathic effect (CPE) in microscopic observation assays, and (2) the half-maximal inhibitory concentration (IC50), calculated from dose-response curves in cell viability assays. As shown in Fig. 1a-c, MAb 1F4 exhibited potent neutralization against both the CVA6-141 strain (MNC: 0.313 μg/mL; IC50: 0.052 μg/mL) and the CVA6-HeB strain (MNC: 0.625 μg/mL; IC50: 0.193 μg/mL). Similarly, MAb 3H7 demonstrated strong neutralizing activity against CVA6-141 (MNC: 0.625 μg/mL; IC50: 0.103 μg/mL) and CVA6-HeB (MNC: 1.25 μg/mL; IC50: 0.387 μg/mL). In contrast, the IgG control antibody 3A2 19 showed no neutralization at any tested concentration. Together, these results indicate that both MAbs (1F4 and 3H7) display comparable neutralization potency against CVA6 strains. Furthermore, neither MAb neutralized CVA10 strain S0148b even at the highest concentration tested (10 µg/mL) (Fig. 1a), confirming their specificity for CVA6.
The MAbs were tested by ELISA for their ability to recognize purified CVA6-HeB virions. The HeB strain of CVA6 was selected for subsequent experiments due to its higher proportion of mature virions compared to the 141 strain, which predominantly produces A-particles as documented 9 . Both anti-CVA6 MAbs 1F4 and 3H7 specifically bound to CVA6-HeB virions, whereas the IgG control antibody showed no reactivity (Fig. 1d). Notably, MAb 3H7 exhibited significantly stronger binding to CVA6-HeB virions compared to MAb 1F4. To quantify this difference, we measured the binding affinity of the MAbs to CVA6-HeB virions using bio-layer interferometry (BLI). Consistent with the ELISA results, BLI revealed that MAb 3H7 displayed a higher binding affinity (KD = 0.40 nM) than MAb 1F4 (KD = 3.43 nM) (Fig. 1e).
The protective efficacy of MAbs 1F4 and 3H7 was evaluated in the established neonatal mouse models of CVA6 infection, using two distinct highly lethal strains: CVA6-HeB 20 and CVA6-S0087b 21,22 . Note that CVA6-S0087b exhibited high virulence in neonatal mice, despite failing to produce detectable CPE in vitro. For the prophylactic evaluation, one-day-old ICR mice (n ≥ 11 per group) were randomly assigned to receive PBS, 10 mg/kg of anti-CVA6 MAbs (1F4 or 3H7), or 10 mg/kg of control IgG 3A2 19 . Twenty-four hours after treatment, the mice were challenged with either CVA6-HeB or CVA6-S0087b. Survival rates were monitored daily and are summarized in Supplementary Fig. 1. Following CVA6-HeB challenge, PBS-and control IgG-treated groups developed progressive limb weakness and paralysis, culminating in 100% mortality within 10 dpi. Strikingly, 10 of 11 mice administered 1F4 and all 3H7-treated mice survived the infection. A parallel trend was observed in the CVA6-S0087b challenge model: all PBS-and control IgG-treated mice succumbed to infection by 7 dpi, whereas both 1F4and 3H7-treated mice remained fully protected. These results conclusively demonstrate the robust in vivo prophylactic activity of 1F4 and 3H7 against divergent CVA6 strains.
To evaluate therapeutic efficacy, two-day-old ICR mice (n ≥ 11 per group) were infected with CVA6-HeB or CVA6-S0087b and 24 hours later treated with PBS or anti-CVA6 MAbs (1F4 or 3H7; 10 mg/kg). Survival rates were monitored daily (Fig. 1f). Following CVA6-HeB infection, all PBS-treated mice succumbed to infection within 8 dpi, while 10/13 mice in the 1F4 group and all 3H7-treated mice survived until 14 dpi. A parallel protective trend was observed against CVA6-S0087b: both 1F4 and 3H7 MAbs conferred 100% survival in treated mice, whereas 100% mortality was observed in the PBS-treated control group. These findings highlight the significant therapeutic efficacy of 1F4 and 3H7 MAbs against distinct CVA6 strains. Taken together, these results demonstrate robust in vivo efficacy of these MAbs in both preand post-exposure scenarios.
## Neutralization mechanisms of MAbs 1F4 and 3H7
To investigate the neutralization mechanisms of MAbs 1F4 and 3H7 against CVA6, time-of-addition assays were performed to identify the specific viral lifecycle stage targeted by these antibodies. In this assay, MAbs were assessed under: (1) Pre-attachment: antibody-virus (CVA6-HeB) premixes were adsorbed onto cells at 4 °C prior to 37 °C infection; (2) Post-attachment: antibodies were administered at 0 h or 0.5 h after shifting virus-bound cells to 37 °C. Viral RNA levels were quantified at 6 hpi by real-time quantitative reverse transcription PCR (RT-qPCR). As shown in Fig. 2a, MAb 1F4 exhibited potent inhibition in both pre-attachment (Pre, 33% residual viral RNA) and immediate post-temperature shift (Post-0h, 33% residual RNA) regimens, but efficacy sharply diminished at Post-0.5 h (82% residual RNA), defining a narrow temporal window for blocking early postattachment entry. MAb 3H7 showed robust inhibition at preattachment (1% residual viral RNA) and Post-0h (17% residual RNA), with reduced activity at Post-0.5 h (87% residual RNA), demonstrating neutralization capacity spanning both viral attachment and early post-attachment entry steps.
To determine whether MAbs 1F4 or 3H7 interfere with the initial attachment phase of CVA6 entry, we pre-incubated CVA6-HeB virions with antibodies prior to adsorption onto pre-chilled RD cells at 4 °C. After washing, cell-bound viral RNA was measured by RT-qPCR. 1F4 and control IgG showed no inhibition at any tested dose, while the 3H7 CVA6-S0087b challenge antibody demonstrated weak suppression of viral attachment only at the highest concentration (100 ng/well) (Fig. 2b). These observations suggest that the neutralizing mechanisms of both 1F4 and 3H7 predominantly target post-attachment steps. Human KRM1 mediates CVA6 infection 15 . To investigate whether MAbs 1F4 and 3H7 interfere with CVA6-KRM1 interactions, we developed a competitive receptor-binding ELISA (Fig. 2c). Immobilized CVA6-HeB virions were incubated with biotinylated KRM1-Fc fusion protein in the presence of IgG control, 1F4, or 3H7. Bound KRM1-Fc was detected using HRP-streptavidin (SA-HRP). Compared with IgG control, both 1F4 and 3H7 significantly inhibited receptor-virus interaction, with 3H7 showing stronger blockade (Fig. 2d). These data suggest 1F4 3H7 Ctr Fig. 2 | Mechanisms of action of CVA6-neutralizing MAbs. a Time of addition assay. CVA6-HeB was incubated with MAbs 1F4 or 3H7 before (Pre) or at indicated time points after (Post) viral attachment to RD cells. Viral RNA was quantified by RT-qPCR at 6 hours post-infection (hpi) and is expressed as a percentage of the virusonly control. Data are mean ± SD (n = 4 for virus-only; n = 3 for others). p = 0.0055 (1F4-Post-0.5 h); p = 0.0336 (3H7-Post-0.5 h). b Attachment inhibition assay. CVA6-HeB was incubated with serially diluted anti-CVA6 MAbs (1F4, 3H7) or a control antibody for 1 h, followed by adsorption to pre-chilled RD cells at 4 °C for 2 h. Cellassociated viral RNA was analyzed via RT-qPCR. Viral RNA levels are expressed as percentages relative to the virus-only group. and is expressed as a percentage of the virus-only control. Data are mean ± SD (n = 3). p = 0.0112 (1F4-1ng), p = 0.0030 (1F4-10ng), p = 0.0181 (1F4-100ng), p = 0.0152 (3H7-1ng), p = 0.0126 (3H7-100ng).
Statistical note for a, b: Significance was determined by two-tailed Student's t-test versus the virus-only control: ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ****, p < 0.0001. c Schematic representation of competitive binding ELISA. (Left) KRM1 binding: Biotinylated KRM1-Fc protein binds to immobilized CVA6 virions, detected by HRPconjugated streptavidin (SA-HRP). (Right) Antibody blockade: Anti-CVA6 MAbs bind to virions, sterically hindering KRM1-Fc binding and reducing detection signal. d KRM1 competition ELISA. Serially diluted anti-CVA6 MAbs (1F4, 3H7) or control antibody were tested for blocking biotinylated KRM1-Fc binding to immobilized CVA6-HeB virions. Bound KRM1-Fc was detected using HRP-streptavidin. Data represent mean ± SEM of triplicate wells. Source data are provided as a Source Data file.
that 1F4 and 3H7 may sterically hinder CVA6 engagement with membrane-associated KRM1.
Both 1F4 and 3H7 MAbs effectively blocked CVA6-KRM1 binding but had little effect on viral attachment to cells. Notably, both antibodies potently neutralized the virus post-attachment (Fig. 2). These results support the hypothesis that KRM1 is not essential for initial CVA6 attachment but is critical for postattachment entry steps.
## KRM1 binding triggers uncoating of mature CVA6 virions
Previous studies identified KRM1 as essential for CVA6 infection 15 . To validate the infection dependency, we infected wild-type RD cells and their KRM1-knockout (ΔKRM1) counterparts (generated via CRISPR-Cas9 16 ) with CVA6-HeB. Quantification of viral titers at 24 hours postinfection (hpi) revealed a 36-fold reduction in ΔKRM1 cells compared to wild-type RD cells (Supplementary Fig. 2a), confirming KRM1's essential role in productive CVA6 infection. Importantly, ELISA-based binding assay showed that immobilized CVA6-HeB virions exhibited dose-dependent binding to human KRM1-Fc, whereas ACE2-Fc controls showed no binding activity (Supplementary Fig. 2b), providing the first experimental evidence of direct physical interaction between CVA6 and KRM1. These findings established KRM1 as both a physical binding partner and functional receptor, prompting us to investigate its structural impact.
To determine whether KRM1 binding induces structural reorganization in CVA6, we compared cryo-EM structures of CVA6-HeB virus particles before and after KRM1 receptor engagement (Supplementary Tables 1 and2). For virus-only analysis, UV-inactivated CVA6-HeB viral particles were analyzed by cryo-EM, revealing two populations: (1) mature virions (93.0%, 20,204 particles); (2) empty particles (7.0%, 1531 particles) (Supplementary Fig. 3). The CVA6-HeB mature virion structure (2.52 Å resolution) exhibits a compact capsid (~310 Å diameter) with closed two-fold axis channels, pocket factor (modeled as stearic acid) in VP1 hydrophobic pocket, and retained genomic RNA (Fig. 3a,c,e,g). The CVA6-HeB empty particle structure (3.47 Å resolution) shows expanded architecture (~317 Å diameter), open two-fold channels, loss of pocket factor, and no viral RNA (Fig. 3b,d,f,h). In both maps, densities corresponding to the residue backbones and side chains-particularly the bulky ones-were well resolved and readily identifiable (Supplementary Fig. 4). Structural alignment with CVA6 prototype strain Gdula 8 confirmed structural conservation, with overall root-mean-square deviation (RMSD) values of 0.3 Å for mature virions and 0.5 Å for empty particles (Supplementary Fig. 5). Notably, no A-particles were detected in CVA6-HeB samples (Supplementary Fig. 3), contrasting earlier reports of A-particle-dominated CVA6-Gdula and CVA6-141 preparations 8,9 .
For receptor interaction studies, purified CVA6-HeB virus particles were incubated with recombinant His-tagged human KRM1 ectodomain at a 1:120 molar ratio (35 min at room temperature) and subjected to cryo-EM analysis. Three distinct populations were resolved: (1) KRM1-bound mature virions (28.7%, 24,506 particles); (2) A-particles (69.8%, 59,654 particles); (3) empty particles (1.5%, 1290 particles) (Supplementary Fig. 6). The KRM1-virion complex structure (2.55 Å resolution) retains mature virion architecture with closed twofold axis channels, VP1 pocket factor and intact RNA, while displaying clear KRM1 density on the capsid surface (Fig. 3i,l,o,r). The A-particle structure (2.49 Å resolution) displays expanded capsids with open twofold channels, an ejected pocket factor, and retained RNA, with only residual KRM1 density (Fig. 3j,m,p,s). The clear and well-fitted densities of both the residue backbones and side chains enabled the determination of the structures of viral particles at atomic or nearatomic resolution (Supplementary Fig. 7). Structural alignment with A-particles of CVA6 prototype strain Gdula 8 and clinical strain 141 9 confirmed high similarity, with overall RMSD values of 0.5 and 1.0 Å, respectively (Supplementary Fig. 8). The empty particle structure (3.17 Å resolution) remains structurally identical to the untreated empty particle (Fig. 3k, n, q, t).
Compared to virus-only preparations, KRM1 treatment induced a striking redistribution of particle populations: mature virion proportions decreased from >90% to <30%, while A-particlesabsent in untreated virus samplesemerged as the dominant species (70%). Empty particle levels remained consistently low ( < 10%) throughout the process (Fig. 3). This receptor-dependent shift demonstrates that A-particles originate from structural remodeling of mature virions upon KRM1 binding. Crucially, KRM1 drives this conversion with high efficiency under neutral pH at room temperature (35-min incubation), establishing it as the physiological uncoating receptor for CVA6.
## Structural analysis of CVA6-KRM1 complex
The cryo-EM structure of KRM1-bound CVA6 virion (Fig. 3i) reveals the interaction interface between CVA6 and KRM1. KRM1 binds near each five-fold vertex within the canyon, spanning the VP1, VP2, and VP3 proteins of two adjacent protomers (Fig. 4a,b). Each KRM1 molecule engages both protomers, burying 741.7 Å 2 on protomer 1 and 775.6 Å 2 on protomer 2, resulting in a total interface area of 1517.3 Å 2 (Fig. 4b).
The KRM1 ectodomain comprises three structural domains: the N-terminal Kringle (KR) domain, the WSC domain, and the CUB domain. The interaction with CVA6 is mediated through the KR and WSC domains of KRM1 (Fig. 4c,d). CVA6 protomer 1 forms extensive contacts with the WSC domain (Supplementary Table 3). Specifically, the R90 residue in VP3 CD loop forms a hydrogen bond with WSC residue T138 (Fig. 4e); the VP3 C-terminal residues D231 and Q234 form hydrogen bonds with WSC residues N128 and Y178, respectively (Fig. 4f); A dense hydrogen-bond network links VP1 C-terminal residues D284, A286, D292, and E294 to the WSC region T135-K141 (Fig. 4g). CVA6 protomer 2 interacts with both the KR and WSC domains of KRM1 (Supplementary Table 3). Specifically, the K157 and D159 residues in VP1 EF-loop form salt bridges and hydrogen bonds with WSC residues D201 and H126, respectively (Fig. 4h); the Q211 residue in VP1 GH-loop forms a hydrogen bond with WSC residue G192 (Fig. 4i); the K140 and N142 residues in VP2 EF-loop establish hydrogen bonds with KR residues G89 and D90; VP2 K140 additionally forms salt bridges with KR residues D88 and D90; VP2 K140 also forms π-cation interactions with the aromatic KR residues W94 and W106 (Fig. 4j). Notably, VP2 K140 is fully conserved in KRM1-dependent enteroviruses (including CVA6) and is critical for receptor binding and infectivity 16 . The interaction of K140 with KRM1 residues D88, G89, D90, W94, and W106 is conserved not only in CVA6 and CVA10 (Supplementary Fig. 9) 17 , but also potentially in other KRM1-dependent enteroviruses.
The KRM1-bound CVA6 retains a mature virion architecture nearly identical to its unbound (apo) state, with a global RMSD value of 0.4 Å (Supplementary Fig. 10a). Detailed structural comparisons reveal only minor deviations in specific regions: VP2 EF-loop (RMSD = 0.4 Å), VP3 C-terminus (0.4 Å), and VP3 GH-loop (0.5 Å) (Supplementary Fig. 10b). Prior studies on KRM1-bound CVA10 17 reported a similarly mature virion structure but with larger conformational changes in the same regions: VP2 EF-loop (RMSD = 1.4 Å), VP3 C-terminus (0.7 Å), and VP3 GH-loop (2.3 Å) (Supplementary Fig. 10c,d). These differences in structural variability may stem from inherent variations between the viral serotypes or from differences in the experimental methods used.
## Cryo-EM structure of CVA6 in complex with 1F4 Fab reveals neutralization mechanism
To elucidate the neutralizing mechanism of the 1F4 antibody against CVA6, we determined the cryo-EM structure of CVA6 complexed with 1F4 Fab (Supplementary Table 4). Incubation of purified CVA6-HeB viral particles with excess 1F4 Fab yielded two distinct populations: (1) 1F4 Fab-bound mature virions (98.9%); (2) empty particles (1.1%) (Supplementary Fig. 11). The well-fitted backbones and side chains (Supplementary Fig. 13a-c) allowed us to determine the high-resolution structure of the 1F4-virion complex at 2.11 Å, which preserves characteristic features of mature virions, including closed two-fold axis channels, pocket factor in VP1, and intact genomic RNA. Notably, clear Fab density is observed on the capsid surface (Fig. 5a,c,e,g). Structural comparison with unbound CVA6 virions reveals remarkable similarity, with an overall RMSD of 0.4 Å. The empty particle structure, resolved at 3.12 Å resolution, is identical to untreated empty particle (Fig. 3b), exhibiting capsid expansion without detectable pocket factor, RNA or Fab density (Fig. 5b,d,f,h). Notably, 1F4 Fab treatment does not significantly alter the population distribution between mature virions ( >90%) and empty particles ( <10%) compared to virus-only preparations (Fig. 3a,b). These findings demonstrate that 1F4 selectively binds to mature virions rather than empty particles, and that 1F4 Fab binding does not induce significant conformational changes in the viral capsid. The cryo-EM structure of the 1F4-virion complex shows symmetric antibody binding, with three 1F4 Fab molecules arranged around each three-fold axis of viral capsid (Fig. 5a). Each 1F4 Fab binds to a single protomer by interacting with the VP3 AB-loop (also termed "knob", residues 56-65) and the VP1 C-terminus (Fig. 5i-j, Supplementary Table 5). The VP3 knob inserts into a cleft formed by the heavy chain complementarity-determining region 1 (CDR1) and 3 (CDR3) loops, as well as the light chain CDR1 of 1F4 (Fig. 5j). Key interactions include hydrogen bonds between: (1) VP3 knob residues G60, T62, and S65 and heavy chain CDR1 residues T30, S31, and Y33; (2) VP3 residues N56 and T58 with heavy chain CDR3 residue N101; (3) VP3 residue T59 with light chain CDR1 residue N32 (Fig. 5j). The VP1 C-terminus further stabilizes the interface through hydrogen bonds: residue I288 interacts with heavy chain CDR3 residue N101, and S287/T289 engage light chain framework 3 (FR3) residue N53 (Fig. 5j). The interaction interface between CVA6 and 1F4 covers 842.4 Å 2 per protomer, with the 1F4 heavy and light chains contributing 61.8% and 38.2% of the buried areas, respectively (Supplementary Table 6).
To explore the spatial relationship between the 1F4 epitope and the KRM1 receptor-binding site, we conducted comparative footprint mapping on virion surfaces using RIVEM analysis (Fig. 5k). The analysis revealed that the VL (light chain variable region) domain of 1F4 directly overlaps with the KRM1 binding region, indicating a potential mechanism of receptor blockade mediated by the VL domain. Additionally, structural superposition of the KRM1-CVA6-virion complex and the 1F4-CVA6-virion complex demonstrated steric clashes between the 1F4 VL domain and the WSC and CUB domains of KRM1 (Fig. 5l). These structural analyses demonstrate that 1F4 can sterically hinder KRM1 from binding to the CVA6 virion, thereby elucidating the structural basis for its receptor-blocking activity (Fig. 2d).
## Cryo-EM structure of CVA6 in complex with 3H7 Fab
To determine the structural basis for CVA6 neutralization by the 3H7 antibody, we determined the cryo-EM structure of the CVA6-3H7 Fab complex (Supplementary Table 7). After incubation of purified CVA6 particles with excess 3H7 Fab, two major populations were observed:
(1) 3H7 Fab-bound CVA6 virions (98.1%, resolved at 2.59 Å); (2) 3H7 Fab-decorated empty particles (1.9%, resolved at 3.26 Å) (Fig. 6a,b, Supplementary Fig. 12). Notably, 3H7 Fab binds both mature virions and empty particles, as evidenced by clear Fab densities on both particle types (Fig. 6a,b). The well-resolved backbones and side chains allowed us to determine the structures of both complexes (Supplementary Fig. 13d-i). The 3H7 Fab-bound CVA6 virions retain key features of mature virions, including a compact capsid with closed twofold axis channels, the VP1 pocket factor, and well-resolved genomic RNA (Fig. 6a,c,e,g). When compared to unbound CVA6 virions, the overall structure shows remarkable similarity, with an RMSD of 0.3 Å. In contrast, the 3H7 Fab-decorated empty particles display an expanded architecture with open two-fold channels and lack both the VP1 pocket factor and internal RNA densities (Fig. 6b,d,f,h). In addition, the 3H7 Fab treatment did not change the population ratio of mature virions (>90%) to empty particles ( < 10%) compared to the untreated virus (Figs. 3a, b and6a, b). These results indicate that 3H7 Fab binds to the virus without inducing significant conformational changes in the viral capsid.
Similar to 1F4, the 3H7 Fab binds close to the icosahedral threefold axis on the viral capsid (Fig. 6a,b), with each Fab engaging a single protomer (Fig. 6i). Both heavy and light chains of 3H7 mediate interactions with VP1, VP2, and VP3 proteins within the same protomer (Fig. 6i). The 3H7 Fab buries 1124.4 Å 2 of surface area per protomer. Specifically, the 3H7 heavy and light chains account for 70.1% and 29.9% of the binding interactions, respectively (Supplementary Table 6). The VH (heavy chain variable region) domain establishes an extensive network of hydrogen bonds and salt bridges with the capsid (Supplementary Table 8). The key interactions are as follows: (1) the VP2 BC-loop residues T73 and E74 form hydrogen bonds with the heavy chain CDR1 (residue S30) and CDR2 (residue S52 and S53) loops;
(2) the VP2 HI-loop residue K225 forms one hydrogen bond with heavy chain CDR1 residue T31, and also creates two salt bridges with heavy chain CDR3 residue D103; (3) the VP3 βI residue N211 forms a hydrogen bond with heavy chain CDR3 residue D105 (Fig. 6j). The 3H7 VL domain recognized two regions on the viral capsid that partially overlap with the 1F4 epitope, namely the VP3 knob (AB-loop) and VP1 C-terminus (Fig. 6k). However, these interactions are less extensive than those of 1F4 (Figs. 5j and6k). The key interactions involve hydrogen bonds between: (1) the VP3 knob residue T61 and light chain CDR1 residue R32; (2) VP1 residue A286 with light chain CDR1 residue S31 (Fig. 6k). Notably, 3H7 maintains its binding mode in both CVA6 mature virions and expanded empty particles, with most interactions conserved and only some differing (Supplementary Fig. 14), indicating the 3H7 epitope remains accessible after capsid expansion.
To evaluate spatial relationships between the 3H7 epitope and KRM1 receptor-binding site, we compared the footprints of 3H7 and KRM1 on the CVA6 capsid using RIVEM analysis. The analysis shows that the 3H7 VL domain partially overlaps with the KRM1 binding site, whereas the VH domain does not (Fig. 6l). Structural superposition of the 3H7-CVA6-virion and KRM1-CVA6-virion complexes reveals that the 3H7 VL domain spatially blocks the KRM1 WSC domain from engaging the viral capsid (Fig. 6m). Together, these findings provide structural evidence that 3H7 can sterically block KRM1 from binding to CVA6, consistent with its functional blockade of KRM1 engagement (Fig. 2d).
## Comparison of the epitopes and binding kinetics of 1F4 and 3H7
Structural superposition of the 1F4-virion and 3H7-virion complexes shows that the two immune complexes adopt highly similar overall architectures. Both antibodies bind to the viral capsid near the threefold axis and have overlapping binding footprints (Fig. 7a-c). However, a key difference is that 3H7's binding sites are closer to the three-fold axis than 1F4's, resulting in a more compact footprint and reduced steric interference with the adjacent KRM1 receptor-binding site (Fig. 7a-c). This structural observation appears contradictory to the Fig. 3 | Cryo-EM structures of CVA6 particles before and after KRM1 binding. a, b Cryo-EM density maps of UV-inactivated CVA6-HeB mature virion (a) and empty particle (b). All maps are radially colored (scale in Å) and viewed along the two-fold axis. One icosahedral asymmetric unit is marked by a black triangle. Pentagons and triangles denote five-fold and three-fold symmetry axes, respectively. 2D classification revealed 93.0% mature virions and 7.0% empty particles. c, d Twofold axis channel structures in the CVA6 mature virion (c) and empty particle (d). VP1 (blue), VP2 (green), VP3 (red), and VP4 (yellow) are shown; this color scheme applies to all panels. Ellipses represent the two-fold axis. Conformational differences are outlined by yellow dashed rectangles. e, f Atomic models of the VP1 hydrophobic pocket in the CVA6 mature virion (e) and empty particle (f). A pocket factor (magenta stick) is fitted into the VP1 hydrophobic pocket of the CVA6 virion, with corresponding cryo-EM density (magenta mesh). g, h Central cross-sectional views (Z-axis) of CVA6 mature virion (g) and empty particle (h). i-k Cryo-EM analysis of the KRM1-treated CVA6 sample revealed a shifted composition: 28.7% KRM1-bound virions (i), 69.8% A-particles (j), and 1.5% empty particles (k). l-n Twofold axis channel structures in the KRM1-bound CVA6 virion (l), A-particle (m), and empty particle (n). o-q VP1 pocket models for KRM1bound CVA6 virion (o), A-particle (p), and empty particle (q). The pocket factor (magenta) is present in KRM1-bound virion but absent in A-particle and empty particle. KRM1 density is omitted for clarity. r-t Central cross-sectional views of the KRM1-bound CVA6 virion (r), A-particle (s), and empty particle (t) density maps. Black arrows indicate KRM1 receptor density. Central sections are displayed without masking or sharpening to preserve the native density information. results from the KRM1 receptor competition ELISA assay, which shows that MAb 3H7 blocks KRM1 binding to CVA6 virion more effectively than MAb 1F4 (Fig. 2d). To address this discrepancy, we conducted a detailed comparison of the epitopes and binding affinities of the two antibodies.
Comparative epitope analysis reveals that the 3H7 antibody recognizes a broader array of structural elements, including the VP1 C-terminus, the VP2 BC and HI loops, and the VP3 AB loop and βI. In contrast, 1F4 focuses on a more restricted area, mainly consisting of the VP1 C-terminus and the VP3 AB-loop (Fig. 7c,d). Consistently, 3H7 exhibits a greater buried surface area (1124.4 Å 2 ) than 1F4 (842.4 Å 2 ) (Fig. 7d), suggesting stronger binding. In line with its broader interface, 3H7 exhibits superior virion-binding kinetics, with 8. dissociation rate (Kdis = 1.04 × 10 -4 s -1 vs. 1F4's Kdis = 2.08 × 10 -3 s -1 ) (Fig. 1e). 3H7's broader interface and superior binding characteristics should allow it to maintain more persistent viral particle occupancy than 1F4. To experimentally validate this, we conducted a BLI-based competitive binding assay. In this assay, immobilized CVA6-HeB virions were first incubated with either kinetics buffer (reference control) or the first antibody, followed by sequential exposure to the second antibody (Fig. 7e). The binding signals of secondary MAbs were analyzed, with results presented in Fig. 7f,g. The results confirmed asymmetric interference: when virions were pre-incubated with 3H7, subsequent 1F4 binding was completely blocked (Fig. 7f), whereas prebound 1F4 only partially inhibited later 3H7 binding (Fig. 7g). These findings indicate that the epitopes of 1F4 and 3H7 are overlapping but distinct, and also demonstrate that 3H7 has stronger retention on viral particles under competitive conditions.
The apparent contradiction is resolved by considering antibody binding kinetics. Although 3H7's compact binding creates less direct physical obstruction to KRM1, its broad epitope recognition and slow dissociation enable sustained occupancy, which effectively excludes receptor access. In contrast, 1F4's more extensive spatial interference with KRM1 is offset by its faster dissociation from the virion. These findings indicate that effective receptor blockade depends on both spatial blocking effects and binding kinetics rather than steric interference alone.
Heparan sulfate mediates initial attachment of CVA6 to host cells followed by KRM1-dependent post-attachment entry Both 1F4 and 3H7 MAbs block CVA6-KRM1 binding (Figs. 2d, 5k-l and 6l-m), but have little effect on viral attachment (Fig. 2b). This suggests KRM1 is not essential for CVA6's initial attachment. So, another receptor is likely involved in CVA6 attachment. Heparan sulfate proteoglycans (HSPG), widely expressed on host cell surfaces, serve as key attachment receptors for several enteroviruses, such as EV-A71 and CVA16 23,24 . However, whether HSPGs mediate attachment for CVA6 remains unknown. Structural analysis shows CVA6 virions have a conserved, positively charged patch at the five-fold axis (Fig. 8a), a region that mediates HSPG binding in EV-A71 and CVA16 23,24 . So, we hypothesized that HSPG may also help CVA6 attach to the cell surface. To test this, we systematically examined HSPG's role in CVA6 attachment and infection using genetic and biochemical methods.
Solute carrier family 35 member B2 (SLC35B2) encodes the 3′phosphoadenosine 5′-phosphosulfate (PAPS) transporter 1 (PAPST1), a Golgi-localized transmembrane protein that transports PAPS from cytosol to Golgi for HSPG sulfation (Fig. 8b). To assess HSPG dependency in CVA6 infection, wild-type and CRISPR-engineered SLC35B2 knockout (ΔSLC35B2) RD cells 13 were infected with CVA6-HeB. Quantification of viral titers at 48 hpi showed a 35-fold reduction in ΔSLC35B2 cells compared to wild-type controls, though less pronounced than the 132-fold decrease observed in ΔKRM1 cells (Fig. 8c). These results demonstrate that CVA6 infection exhibits dual dependency, requiring both the KRM1 receptor and HSPG sulfation for productive infection.
To determine whether CVA6 directly interacts with heparan sulfate, CVA6-HeB culture supernatants were applied to a heparin-immobilized agarose chromatography column. Following extensive PBS washes, bound virus particles were eluted with 2 M NaCl. Western blot analysis of column fractions using an anti-VP0 polyclonal antibody revealed striking enrichment of CVA6 capsid proteins (VP0 and VP2) in the high-salt eluate (Fig. 8d). In contrast, control experiment using empty agarose beads demonstrated no detectable binding of CVA6 particles (Fig. 8d). These results demonstrate that CVA6 virions specifically interact with heparin in vitro.
To assess the impact of heparin on CVA6 attachment, we pretreated CVA6-HeB with soluble heparin prior to its exposure to RD cells at 4 °C. Cell-bound viral RNA levels were quantified at 6 hpi by RT-qPCR. We found that heparin potently inhibited CVA6-HeB binding to RD cells in a concentration-dependent manner. Specifically, pretreatment with heparin at concentrations ≥500 μg/mL significantly inhibited viral attachment, whereas no inhibitory effect was observed at 50 μg/mL (Fig. 8e). These findings strongly suggest that HSPG act as attachment receptors for CVA6.
To directly compare the potential roles of KRM1 and HSPG in the CVA6 entry process, we evaluated their relative contributions to viral attachment and internalization. For attachment assays, pre-cooled wild-type RD, ΔKRM1, or ΔSLC35B2 cells were incubated with CVA6-HeB at 4 °C to allow viral binding without internalization. Quantification of cell-associated viral RNA revealed no significant difference in attachment between ΔKRM1 and wild-type cells (Fig. 8f), indicating that KRM1 is not essential for initial binding. In contrast, ΔSLC35B2 cells exhibited an 82% reduction in viral attachment (Fig. 8f), demonstrating HSPG's dominant role in this step. For internalization assays, cell surface-bound virions (4 °C) were allowed to enter cells by shifting to 37 °C for 1 h, followed by trypsin treatment to remove residual surface virions. Quantification of internalized viral RNA revealed a 76% reduction in ΔKRM1 cells and a 94% decrease in ΔSLC35B2 cells compared to wild-type (Fig. 8g). Notably, ΔKRM1 cells exhibited no attachment deficit but showed a marked 76% reduction in internalized virus (Fig. 8f,g), directly demonstrating that KRM1 is essential for postattachment internalization independent of initial binding. The severe internalization defect observed in ΔSLC35B2 cells (94% reduction compared to wild-type) likely results from two factors: impaired initial viral attachment (82% fewer surface-bound virions; Fig. 8f), which drastically reduces the pool of virus available for entry, and potential contributions of HSPG to post-attachment entry processes. Collectively, these data establish a two-receptor entry mechanism for CVA6: HSPG mediate the initial attachment of viral particles to the cell surface, while KRM1 facilitates their subsequent entry into the cytoplasm (Fig. 8h).
## Discussion
In this study, we systematically investigated the molecular mechanisms of CVA6 receptor recognition and antibody neutralization. We showed that purified CVA6-HeB primarily exists as mature virions. We identified KRM1 as the uncoating receptor for CVA6, as its binding induces A-particle formation. We propose a two-receptor entry model for CVA6: HSPG mediates attachment, and KRM1 facilitates uncoating. Additionally, we developed two protective MAbs, 1F4 and 3H7, which target a novel antigenic site near the viral capsid's three-fold axis. The Fig. 4 | Molecular interactions between CVA6 and KRM1. a Surface representation of a CVA6 capsid pentamer bound to KRM1, viewed along the five-fold symmetry axis. CVA6 pentamer is colored gray. KRM1 is magenta. One icosahedral asymmetric unit is marked by a triangle. Density maps displayed are unsharpened. b One KRM1 molecule interacts with two adjacent CVA6 protomers. KRM1 is shown as magenta ribbon, superimposed on cryo-EM density. CVA6 subunits are shown as ribbons and color-coded: Protomer 1 (VP1.1: light blue; VP2.1: light green; VP3.1: light red). Protomer 2 (VP1.2: blue; VP2.2: green; VP3.2: red). c Interaction of KRM1's KR and WSC domains with adjacent CVA6 protomers. KRM1 is drawn as a cartoon with the domains color-coded: KR, cyan; WSC, orange; CUB, black. CVA6 protomers are drawn as surface. d Enlarged view of the CVA6-KRM1 interaction interface, highlighting key residues on KRM1. e-j Binding interfaces between KRM1 and specific capsid regions: e VP3.1 CD-loop, f VP3.1 C-terminus, g VP1.1 C-terminus, h VP1.2 EF-loop, i VP1.2 GH-loop, j VP2.2 EF-loop. Hydrogen bonds are indicated by gray dashed lines; salt bridges by black dashed lines; regions with both hydrogen bonds and salt bridges are marked with red dashed lines.
## MAbs sterically block KRM1-virion interactions and function primarily at post-attachment steps, consistent with KRM1's uncoating role.
Research on viral structures is vital for developing effective vaccines and antiviral drugs. CVA6 is now the main global cause of HFMD. Yet, its structural characterization is limited, largely due to viral culturing difficulties 25 . Previously, only two CVA6 structural studies existed, analyzing the Gdula and 141 strains 8,9 . Both reported A-particles as the predominant form, with mature virions being rare 8,9 . This led to the view that A-particles are CVA6's main infectious form and ideal vaccine target 9 . However, our study showed that purified CVA6-HeB particles are mostly mature virions, with no A-particles detected (Fig. 3). Thus, A-particle dominance isn't universal across all CVA6 strains. This discrepancy is likely attributable to a combination of strain-specific characteristics and methodological variations. Intrinsic differences between isolates, such as capsid stability and susceptibility to uncoating, may account for the divergent particle composition. Furthermore, extrinsic factors including cell culture systems, purification protocols, and the criteria for selecting gradient fractions could also contribute to the observed variance. Future studies employing standardized methods to comparatively analyze these strains will be crucial to delineate the underlying causes. Notably, CVA6 virions bind efficiently to the KRM1 receptor, while A-particles bind poorly (Fig. 3). This indicates that mature virions are CVA6's main infectious form. In addition, the neutralizing antibody 1F4 binds effectively only to compact virions, not to expanded empty particles (Fig. 5), indicating its epitope is specific to the compact conformation. In summary, compact CVA6 virions offer more neutralizing epitopes than expanded particles, making them better candidates for vaccine development.
Enterovirus receptors are generally classified as attachment or uncoating receptors. KRM1 is the sole identified receptor for both CVA6 and CVA10 15 . Previous research indicated that KRM1 acts as both the attachment and uncoating receptor for CVA10 15,17,18 . In our study, we discovered that KRM1 binds to mature CVA6 virions and triggers their conversion to A-particles (Fig. 3), suggesting its role in uncoating. However, KRM1 knockout reduces viral infectivity without affecting attachment (Fig. 8). In contrast, disrupting HSPG by knocking out SLC35B2 impairs both attachment and infectivity (Fig. 8). Moreover, CVA6 virions directly bind to heparin-beads, and soluble heparin can potently inhibit CVA6 binding to RD cells (Fig. 8). These results indicate that HSPG, rather than KRM1, serves as the primary attachment receptor for CVA6. Additionally, studies on neutralizing antibodies, such as 1F4, have shown that while the antibody can sterically block KRM1-virion interactions, it does not prevent viral attachment (Fig. 2, Fig. 5), further indicating that KRM1 functions in post-attachment steps. These findings support a tworeceptor entry mechanism for CVA6: HSPG mediates viral attachment, while KRM1 facilitates viral uncoating. This mechanism is analogous to that of EV-A71, where HSPG mediates attachment and SCARB2 acts as the uncoating receptor 12,13 .
Neutralizing MAbs are promising antiviral candidates. The only previously reported CVA6-specific neutralizing MAb, 1D5, targets the five-fold vertex and blocks viral attachment 9 . In this study, we developed two novel CVA6-specific neutralizing MAbs, 1F4 and 3H7, which showed excellent preventive and therapeutic efficacy against lethal CVA6 infection in mice (Fig. 1, Supplementary Fig. 1). Cryo-EM studies revealed that both MAbs bind to a novel antigenic site near the viral capsid's three-fold axis. This site includes the VP1 C-terminus, VP2 BC and HI loops, and VP3 AB loop and βI (Figs. 567). Notably, the VP1 residues A286 to T289 in this site are adjacent to the VP1 C-terminal peptide P59 (residues 291-305), a known linear B-cell epitope recognized by CVA6 virus-like particle (VLP) immune sera 26 . Furthermore, this antigenic site corresponds to EV-D68 site III, which consists of EV-D68 VP1 C-terminal residues 285 and 293 27 . Biochemical and structural analyses show that MAbs 1F4 and 3H7 sterically block KRM1 receptor binding and function primarily at post-attachment steps (Fig. 2 and5-6). Thus, our study reveals a novel CVA6 antibody epitope and a distinct neutralization mechanism.
Our data demonstrate that MAbs 1F4 and 3H7 are highly specific for CVA6. Neutralization assays confirmed that neither antibody exhibited cross-neutralizing activity against a panel of related viruses, including CVA10 (Fig. 1a) and other KRM1-using enteroviruses such as CVA2, CVA3, CVA4, CVA5, CVA8, and CVA12, even at the highest concentration of 10 µg/mL (Supplementary Fig. 15a). The molecular basis for this high specificity is revealed by our sequence analysis, which shows that the key capsid residues constituting the 1F4 and 3H7 epitopes are not conserved among other KRM1-using enteroviruses (Supplementary Fig. 15b). While this indicates that 1F4 and 3H7 are not broad-spectrum therapeutics, this high specificity is a crucial finding. It definitively shows that the epitopes targeted by our antibodies are unique to CVA6, making them highly valuable as precise tools for CVA6-specific diagnosis and pathogenesis studies.
Collectively, our findings establish a two-receptor entry mechanism for CVA6, with HSPG mediating attachment and KRM1 facilitating uncoating. We also reveal the structural basis for therapeutic antibody intervention. This study enhances our understanding of CVA6 infection and pathogenesis, and offers key insights for developing anti-CVA6 vaccines and drugs.
## Methods
## Cells and viruses
Human rhabdomyosarcoma (RD) wild-type and knockout cells were cultured in DMEM (Gibco, Thermo Fisher Scientific, USA) supplemented with 10% fetal bovine serum (FBS) at 37 °C under standard culture conditions. CVA6 strain 54203/HeB/CHN/2012 (HeB; GenBank ID: MK106189) 20 and CVA10 strain S0148b (GenBank ID: KX094564) were propagated in RD cells. CVA6 strain TW-2007-00141 (141; GenBank ID: KR706309) was successfully rescued from infectious clone 28 . Viral titers were quantified by 50% tissue culture infectious dose (TCID 50 ) assays using RD cells.
## Proteins and antibodies
CVA6 virions were purified from RD cells infected with CVA6-HeB or CVA6-141. At 72-96 hpi, supernatants and cell lysates were clarified via high-speed centrifugation and then precipitated with 10% PEG 8000 and 200 mM NaCl. Pelleted virions were resuspended in 0.15 M PBS, purified sequentially through 20% sucrose cushion at 112,700 × g for 5 h and 10-50% sucrose gradients at 270,000 × g for 3 h. Virus-containing fractions were confirmed by SDS-PAGE and quantified by Bradford assay.
The His-tagged hKRM1-Fc fusion protein, comprising the human KRM1 ectodomain (residues A23-G373) fused to the human IgG1 Fc region, was engineered and purified from HEK293F cells via nickel affinity chromatography 29 . Similarly, the His-tagged hKRM1 ectodomain (residues A23-G373) was also prepared following the same protocol.
Polyclonal antibodies against CVA6 VP0 were generated by immunizing BALB/c mice with recombinant CVA6 VP0 protein Fig. 5 | Cryo-EM structures of CVA6 particles in complex with 1F4 Fab. a, b Cryo-EM density maps of CVA6-HeB-1F4 in two conformations: 1F4-bound virion (a) and empty particle (b). Both maps are radially colored and viewed along the twofold axis. An icosahedral asymmetric unit is marked by a black triangle. 2D classification resolved 98.9% 1F4-bound virions and 1.1% empty particles. c, d Density maps of twofold-related protomers in the 1F4-bound virion (c) and empty particle (d), superimposed with atomic models. Conformational differences are outlined by yellow dashed rectangles. 1F4 Fab is omitted in (c) for clarity. e, f Atomic models of the VP1 hydrophobic pocket in the 1F4-bound virion (e) and empty particle (f). The pocket factor (magenta) is present in the 1F4-bound virion but absent in empty particle. Consequently, the VP1 pocket collapses in empty particle, accompanied by a significant inward shift of the VP1 GH loop (residues 223-225; red arrowhead, RMSD = 1.4 Å). 1F4 Fab is omitted in (e) for clarity. The five-fold axis is indicated. g, h Central cross-sections of the 1F4-bound virion (g) and empty particle (h) along the z-axis. Black arrows indicate 1F4 Fab density. Central sections are displayed without masking or sharpening to preserve the native density information. i Atomic model of a CVA6 protomer bound to 1F4 Fab. VH is cyan and VL is magenta. The five-fold axis is indicated. j Enlarged views of CVA6-1F4 interaction interfaces. Hydrogen bonds are marked with gray dashed lines. k Surface footprint of 1F4 variable domains on the CVA6 virion using RIVEM analysis. Contour lines denote KRM1 (orange), 1F4 VL (white), and 1F4 VH (yellow). l The CVA6-1F4 complex structure was superimposed onto the CVA6-KRM1 complex, revealing spatial clashes between the 1F4 VL domain (magenta) and KRM1 (tomato). Protomer 1 is colored pink; Protomer 2 is colored white. expressed in Escherichia coli, emulsified in Freund's adjuvants 22 . MAb 3A2, an IgG antibody targeting SARS-CoV-2 19 , served as a negative control in this study.
## Preparation and sequencing of anti-CVA6 MAbs
To generate anti-CVA6 MAbs, adult female BALB/c mice were immunized intraperitoneally three times with 5 µg/dose of purified CVA6-141 virions in aluminum adjuvant at 2-week intervals. About two weeks after the final immunization, one mouse was intravenously boosted with 20 µg of purified CVA6-141 virions. Splenocytes were harvested three days post-boost and fused with SP2/0 myeloma cells using PEG 1450 (Sigma, USA), followed by HAT selection. Hybridoma supernatants were screened via CVA6-141 neutralization assays (described below), and positive clones were subcloned 2-4 times to ensure monoclonality. Antibody isotypes were determined using an HRPbased ELISA kit (Southern Biotech, USA). Variable region sequences of heavy and light chains were amplified with mouse Ig primers (Novagen, Germany) and analyzed via IgBLAST. MAbs were purified from ascites using protein G agarose (Yeasen, China).
## Neutralization assay
Undiluted hybridoma supernatants or serially diluted MAbs (50 µl/ well) were mixed with 100 TCID 50 of CVA6 or CVA10 in 96-well plates and incubated at 37 °C for 1 h. RD cells (20,000 cells/well) were added, and plates were incubated for ~3 days. CPE was visually assessed, and cell viability was quantified using CellCounting-Lite® 2.0 kit (Vazyme). Percent neutralization was calculated as: 100 × (RLU of sample -RLU of virus control) / (RLU of untreated cells -RLU of virus control). IC50 values were determined via nonlinear regression (GraphPad Prism).
## Antibody binding ELISA
ELISA plates were coated overnight at 4 °C with 100 ng/well of purified CVA6-HeB virions and then blocked with 5% milk. After washes, twofold serially diluted MAbs was added (50 µL/well) and incubated for 2 h at room temperature. After washing, HRP-conjugated anti-mouse IgG (1: 10,000; Proteintech, China) was added and incubated for 1 h at room temperature. Absorbance at 450 nm was measured after TMB color development.
## Bio-layer interferometry (BLI) assay
CVA6-HeB virions were biotinylated using EZ-Link Sulfo-NHS-LC-LC-Biotin (Thermo Fisher) and purified via Zeba spin desalting columns. Virus-antibody binding affinity was analyzed on an Octet RED96 system (Pall FortéBio). Briefly, biotinylated CVA6 virions were immobilized on streptavidin biosensors, exposed to serially diluted MAbs for association (500 s), and then transferred to dissociation buffer (PBS with 0.1% BSA/0.02% Tween 20) for 500 s. Equilibrium dissociation constants (KD) were calculated using Octet software.
For competition assays, CVA6 virion-coated sensors were sequentially incubated with buffer (control) or 15 µg/mL first MAb for 500 s, followed by 15 µg/mL second MAb alone or combined with first MAb (to prevent dissociation of pre-bound molecules) for 500 s. Binding levels of the second MAb were quantified via Octet software.
## In vivo protection assays
The protective efficacy of MAbs 1F4 and 3H7 was assessed in neonatal ICR mouse models infected with highly lethal CVA6 strains HeB 20 and S0087b 21,22 . Note that CVA6-S0087b lacks detectable CPE in vitro. To standardize challenge doses, the 50% lethal doses (LD50) of CVA6-HeB and CVA6-S0087b were determined in neonatal ICR mice.
For prophylactic evaluation, groups of one-day-old ICR mice received intraperitoneal (i.p.) injections of PBS, 10 mg/kg of anti-CVA6 MAbs or control antibody 19 , followed 24 h later by i.p. challenge with ~10 LD50 of CVA6-HeB or ~25 LD50 of CVA6-S0087b.
For therapeutic evaluation, groups of two-day-old ICR mice were i.p. infected with ~10 LD50 of CVA6-HeB or ~25 LD50 of CVA6-S0087b. After 24 h, mice received i.p. injections of PBS or 10 mg/kg MAbs. For both assays, all infected mice were monitored for 14 days to record survival and clinical symptoms.
## Time-of-addition assay
The assay was performed according to our previously described protocol 30,31 . Briefly, for pre-attachment inhibition, 1000 TCID 50 of CVA6-HeB was mixed with 1 μg of MAb 1F4 or 3H7 for 1 h, cooled on ice, and applied to pre-chilled RD cells in 24-well plates for viral attachment at 4 °C for 2 h. The cells were washed twice with ice-cold PBS and incubated in DMEM with 1% FBS at 37 °C.
For post-attachment inhibition, 1000 TCID 50 of CVA6-HeB was adsorbed to pre-chilled RD cells at 4 °C for 2 h. After cold PBS washes, cells were incubated at 37 °C for 0 or 0.5 h to initiate viral entry, followed by treatment with 1 μg of MAb 1F4 or 3H7 in fresh medium at 37 °C.
In both assays, total RNA was extracted 6 h post-infection using VeZol reagent (Vazyme). cDNA was synthesized using the HiScript III 1st Strand cDNA Synthesis Kit (Vazyme). Quantitative PCR was performed with SYBR Premix Ex Taq (Takara) to measure CVA6 RNA levels, which were normalized to the housekeeping gene β-actin. Specific primers for CVA6 (Forward: 5′-TACTTTGGGTGTCCGTGTTT-3′, Reverse: 5′-TGGCCAATCCAATAGCTATATG-3′) 9 and β-actin (Forward:
5′-GGACTTCGAGCAAGAGATGG-3′, Reverse: 5′-AGCACTGTGTTGGCGTACAG-3′) were used.
## Inhibition of virus attachment by the MAbs
CVA6-HeB (50,000 TCID 50 ) was pre-incubated with MAbs 1F4, 3H7, or control antibody (1, 10, or 100 ng) at 37 °C for 1 h. The mixtures were ice-cooled and added to pre-chilled RD cells in 24-well plates for 2 h adsorption at 4 °C. After ice-cold PBS washes, cell-associated viral RNA was extracted with VeZol reagent (Vazyme). cDNA synthesis and qPCR were performed as described above, with CVA6 RNA levels normalized to β-actin.
## Receptor blockade ELISA
ELISA plates were coated with purified CVA6-HeB virions (50 ng/well) overnight at 4 °C and blocked with 5% non-fat milk in PBST for 1 h at room temperature. Serially diluted anti-CVA6 MAbs or control antibody were mixed with 12.5 ng/well of biotinylated hKRM1-hFc 29 and transferred to the virus-coated plates. After 1 h incubation at room temperature and washing, bound biotinylated hKRM1-hFc was detected using HRP-conjugated streptavidin (1:5000; Proteintech) for 1 h. Absorbance at 450 nm was measured following TMB substrate development.
## Fab preparation
IgG antibodies 1F4 and 3H7 were buffer-exchanged into sample buffer (20 mM phosphate, 10 mM EDTA, pH 7.0) using ultrafiltration. The antibodies were digested with papain-agarose in PBS containing 2 mM TCEP (pH 7.0) at 37 °C with rotation. Digestion progress was monitored by periodic SDS-PAGE analysis of aliquots. After complete digestion, samples were purified using Q-column chromatography (Cytiva) to collect Fab fragments in the flow-through. These Fabs were then purified by size-exclusion chromatography (Superdex 200 Fig. 6 | Cryo-EM analysis of CVA6-3H7 interactions. a, b Cryo-EM density maps of CVA6-HeB-3H7 resolved into two states: a 3H7-bound virion (98.1%) and b 3H7associated empty particle (1.9%). c, d Structural features of the twofold axis channel in the 3H7-bound virion (c) and 3H7-associated empty particle (d). Structural differences are outlined by yellow dashed rectangles. 3H7 Fab is removed for clarity. e, f Structures of VP1 hydrophobic pocket in the 3H7-bound virion (e) and 3H7associated empty particle (f). A pocket factor (magenta) is present in the virion but absent in the empty particle, leading to pocket collapse and an inward displacement of the VP1 GH loop (red arrowhead). 3H7 Fab is omitted in both panels for clarity. The fivefold axis is indicated. g, h Central cross-sections of the 3H7-bound virion (g) and 3H7-associated empty particle (h). Black arrows indicate 3H7 Fab density. i Atomic model of a CVA6 protomer in complex with 3H7 Fab. VH is cyan, and VL is magenta. The fivefold axis is labeled. j, k Enlarged views of interaction interfaces between CVA6 and the 3H7 VH (j) or VL (k). Gray dashed lines represent hydrogen bonds. Black dashed lines indicate salt bridges. l Footprint analysis of 3H7 variable domains on CVA6 virion. Contour lines mark KRM1 in orange, 3H7 VL in white, and 3H7 VH in yellow. m Structural superposition of the CVA6-3H7 and CVA6-KRM1 complexes, demonstrating steric clashes between 3H7 VL (magenta) and KRM1 (tomato). Protomer 1 is pink, and protomer 2 is white. 3H7 binding signal (nm) Increase, Cytiva) and quantified by A 280 absorbance (Nanodrop, Thermo).
## Cryo-EM sample preparation and data collection
Purified CVA6-HeB (0.11 mg/mL) was incubated with excess KRM1 or Fab (molar ratio ~1:120) in PBS (pH 7.4) for 30-35 min at room temperature. Then, 3 µL of apo-CVA6 or CVA6 complexes with KRM1/Fab were applied to glow-discharged 200-mesh lacey carbon grids (Ted Pella, USA). Grids were blotted and vitrified by plunging into liquid ethane using a Vitrobot Mark IV (Thermo, USA) at 4 °C and 100% humidity. Apo-CVA6 datasets were acquired on a Thermo Fisher Titan Krios (300 kV) equipped with a Falcon IV detector using EPU software. Micrographs of KRM1-CVA6 and Fab-CVA6 complexes were recorded in super-resolution mode on a JEOL cryoARM 300 electron microscope equipped with a K3 direct electron detector (Gatan). Detailed data collection parameters are provided in Supplementary Tables 1,2,4, Micrographs underwent patch-based motion correction and CTF estimation in cryoSPARC (version 4.5.3) 32 , followed by particle picking targeting ~30 nm viral particles. After 2D classification, particle classes showing distinct morphologies were selected for separate model building. Initial models were generated via ab initio reconstruction in cryoSPARC or using previously reported CVA6 density maps (Supplementary Figs. 3, 6, 11, and 12). Subsequent iterative heterogeneous, homogeneous, or non-uniform refinement was performed, and the resulting density maps were used for model building, refinement, validation, and structural interpretation. The maps were further sharpened using DeepEMhancer to facilitate model refinement 33 . Atomic models of viral protomers, Fabs, and KRM1 were predicted using AlphaFold3 34 and manually fitted into density maps using UCSF Chimera (version 1.18) 35 . Models were refined in WinCoot (version 0.9.8.95) and PHENIX (version 1.17.1) 36 , and validated using MolProbity (in Phenix) and the PDBe Validation Server. Final structures were deposited in PDB. Interaction interfaces were analyzed using PDBePISA 37 . All structural figures were prepared using UCSF Chimera and ChimeraX (version 1.9) 38 .
## Virus infection assay
RD wild-type, ΔKRM1 16 , and ΔSLC35B2 13 cells in 24-well plates were infected with CVA6-HeB at an MOI of 0.15 for 1 h at 37 °C. After PBS washing, cells were incubated in fresh medium at 37 °C for 24 h or 48 h. Cells and culture supernatants were harvested, subjected to freeze-thaw cycles, and viral titers were quantified using the TCID 50 method.
## KRM1 receptor-binding ELISA
To assess KRM1 receptor-binding activity of CVA6, ELISA plates were coated overnight at 4 °C with serially diluted purified CVA6-HeB virions. After blocking with 5% milk, the plates were incubated with 100 ng/well of hKRM1-Fc or ACE2-Fc (control protein) 19 for 1 h at room temperature. Following washes, horseradish peroxidase (HRP)-conjugated anti-human IgG (Proteintech) was added and incubated for 1 h at room temperature. Absorbance was measured at 450 nm after color development.
## Binding of CVA6 to heparin-agarose beads
Chromatography columns were packed with heparin-agarose (Yeasen, China) or empty agarose beads (control) and equilibrated with 0.01 M PBS. CVA6-HeB culture supernatant (5 mL, serum-free) was applied to the columns. After loading, columns were washed with 10 column volumes of PBS. Bound viral particles were eluted with elution buffer (0.01 M PBS, 2 M NaCl). Each fraction was analyzed for the presence of CVA6 proteins by western blotting using an anti-VP0 polyclonal antibody as primary antibody (1:1000 dilution) and goat anti-mouse IgG-HRP (Proteintech) as secondary antibody (1:10,000 dilution).
## Inhibition of CVA6 attachment with soluble heparin
To evaluate the inhibitory effect of heparin on CVA6 attachment to RD cells, CVA6-HeB (1000 TCID 50 ) was mixed with serially diluted heparin sodium salt (Sanjie, Shanghai, China) and incubated at 37 °C for 1 h to allow potential interactions between heparin and viral particles. The mixture was then cooled and added to precooled RD cells in a 24-well plate, followed by incubation at 4 °C for 2 h under conditions that permitted viral attachment. After incubation, unbound virus was removed by washing the cells twice with ice-cold PBS. Cell-bound viral RNA was extracted using VeZol reagent (Vazyme, China). cDNA synthesis and qPCR were performed as described above, with CVA6 RNA levels normalized to β-actin.
## CVA6 attachment assay
The CVA6 attachment assay was conducted by adding CVA6-HeB (50,000 TCID 50 ) to precooled RD wild-type, ΔKRM1 16 , or ΔSLC35B2 13 cells in 24-well plates, followed by incubation at 4 °C for 2 h to allow viral attachment. Unbound virus was removed with ice-cold PBS washes, and cell-bound viral RNA was extracted using VeZol reagent (Vazyme). cDNA synthesis and qPCR were performed as described above, with CVA6 RNA levels normalized to β-actin.
## CVA6 internalization assay
The CVA6 internalization assay was performed by first binding CVA6-HeB to pre-cooled RD wild-type, ΔKRM1 16 , or ΔSLC35B2 13 cells in 24-well plates at 4 °C for 2 h. After washing, cells were transferred to 37 °C for 1 h to permit viral internalization. Surface-bound virions were removed by trypsin treatment and additional washes. RNA extraction, cDNA synthesis, and qPCR were carried out as described above, with CVA6 RNA levels normalized to β-actin to quantify internalized virus.
Fig. 8 | Heparan sulfate serves as the primary attachment receptor for CVA6, whereas KRM1 functions as a post-attachment entry receptor. a Electrostatic surface potential maps of CVA6-HeB viral pentamers. Color scheme: red for negative potential, white for neutral, and blue for positive potential. The black dashed circle highlights the positively charged patch. b Schematic diagram illustrating the sulfation process of heparan sulfate and its role in mediating cellular attachment of multiple enteroviruses (EV). c Wild-type RD cells, ΔKRM1 cells, and ΔSLC35B2 cells were infected with CVA6-HeB, and viral titers were quantified at 48 hpi. Data are means ± SD of triplicate biological samples. p = 0.0422 (ΔKRM1), p = 0.045 (ΔSLC35B2). d CVA6 specifically binds heparin-agarose beads. CVA6-HeB supernatant was loaded onto heparin or control columns, followed by washing and elution. Samples were analyzed by western blotting with anti-CVA6-VP0 antibody. FT, flow through. Two independent experiments were performed, with similar results. e Inhibition of CVA6 attachment to RD cells by soluble heparin. CVA6-HeB was treated with heparin and allowed to attach to cells for 2 h at 4 °C. Attached virus was quantified by RT-qPCR analysis and normalized to β-actin. Data are means ± SD (n = 4 for virus-only; n = 3 for others). p = 0.2084 (heparin-50 μg/ml), p = 0.0004 (heparin-500 μg/ml). f Viral attachment assay. CVA6-HeB was incubated with wildtype RD cells, ΔKRM1 cells, and ΔSLC35B2 cells at 4 °C for 2 h. After washing, cellbound virus was quantified via RT-qPCR. Data are means ± SD of four biological replicates. p = 0.2859 (ΔKRM1). g Viral internalization assay. After virus binding at 4 °C, cells were shifted to 37 °C. Internalized virus was quantified by RT-qPCR following trypsin removal of surface virions. Data are means ± SD of four biological replicates. Source data are provided as a Source Data file. Statistical note for (c, e, f, g): significance was determined by two-tailed t-test. ns, p ≥ 0.05; *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. h Two-step model of CVA6 entry: attachment via heparan sulfate, followed by KRM1-mediated internalization. 9VG1, respectively. The corresponding unsharpened cryo-EM density maps have been deposited in the Electron Microscopy Data Bank (EMDB) under accession codes EMD-65031, EMD-65030, EMD-65032, EMD-65034, EMD-65038, EMD-65036 and EMD-65043, respectively. The sequences of 1F4-VH, 1F4-VL, 3H7-VH, and 3H7-VL have been deposited in the DNA Data Bank of Japan (DDBJ) under accession codes LC900916, LC900917, LC900918, and LC901458, respectively. Source data are provided with this paper.
Science Foundation of China (32170948 and 82472252), National High-Level Talent Special Support Programs (10,000 Talents Program) -Young Talents, and the Shanghai Municipal Science and Technology Major Project (ZD2021CY001). B.S. received a grant from the National Key Research and Development Program (2025YFC2608600). This study was also supported by a grant from the National Natural Science Foundation of China (U23A20147) to Peng Gong. W.L. was supported by a grant from the National Natural Science Foundation of China (32470163). The funders played no role in study design, data collection, analysis, manuscript preparation, or publication decisions.
## References
1. Zhu (2023) "Current status of hand-foot-and-mouth disease" *J. Biomed. Sci*
2. Zhao (2020) "A review and meta-analysis of the epidemiology and clinical presentation of coxsackievirus A6 causing hand-footmouth disease in China and global implications" *Rev. Med. Virol*
3. Tomba Ngangas (2010) "Coxsackievirus A6 recombinant subclades D3/A and D3/H were predominant in hand-foot-and-mouth disease outbreaks in the paediatric population" *Viruses*
4. Mirand (2021) "A large-scale outbreak of hand, foot and mouth disease" *Euro. Surveill*
5. Li (2022) "Coxsackievirus A6 was the most common enterovirus serotype causing hand, foot, and mouth disease in Shiyan City, central China" *World J. Clin. Cases*
6. Puenpa (2019) "Evolutionary and genetic recombination analyses of coxsackievirus A6 variants associated with hand, foot, and mouth disease outbreaks in Thailand between" *Viruses*
7. Lu (2012) "Evolutionary diversity of coxsackievirus A6 causing severe hand, foot, and mouth disease -China" *China CDC Wkly*
8. Buttner, Spurny, Fuzik et al. (2022) "Cryo-electron microscopy and image classification reveal the existence and structure of the coxsackievirus A6 virion" *Commun. Biol*
9. Xu (2017) "Atomic structures of coxsackievirus A6 and its complex with a neutralizing antibody" *Nat. Commun*
10. Cifuente (2013) "Structures of the procapsid and mature virion of enterovirus 71 strain 1095" *J. Virol*
11. Ren (2015) "Structures of coxsackievirus A16 capsids with native antigenicity: implications for particle expansion, receptor binding, and immunogenicity" *J. Virol*
12. Nishimura (2024) "Enterovirus A71 does not meet the uncoating receptor SCARB2 at the cell surface" *PLoS Pathog*
13. Guo (2022) "SLC35B2 acts in a dual role in the host sulfation required for EV71 infection" *J. Virol*
14. Coyne, Bergelson (2006) "Virus-induced Abl and Fyn kinase signals permit coxsackievirus entry through epithelial tight junctions" *Cell*
15. Staring (2018) "KREMEN1 is a host entry receptor for a major group of enteroviruses" *Cell Host Microbe*
16. Liu (2025) "Completely conserved VP2 residue K140 of KREMEN1dependent enteroviruses is critical for virus-receptor interactions and viral infection" *mBio*
17. Cui (2020) "Molecular basis of Coxsackievirus A10 entry using the two-in-one attachment and uncoating receptor KRM1" *Proc. Natl. Acad. Sci. USA*
18. Zhao (2020) "Hand-foot-and-mouth disease virus receptor KRE-MEN1 binds the canyon of Coxsackie Virus A10" *Nat. Commun*
19. Zhang (2021) "Development and structural basis of a two-MAb cocktail for treating SARS-CoV-2 infections" *Nat. Commun*
20. Wang (2024) "Effects of glycine 64 substitutions in RNA-dependent RNA polymerase on ribavirin sensitivity and pathogenicity of coxsackievirus A6" *Virus Res*
21. Zhang (2018) "A virus-like particle-based tetravalent vaccine for hand, foot, and mouth disease elicits broad and balanced protective immunity" *Emerg. Microbes Infect*
22. Shen (2016) "Virus-like particle-based vaccine against coxsackievirus A6 protects mice against lethal infections" *Vaccine*
23. Tan, Poh, Sam et al. (2013) "Enterovirus 71 uses cell surface heparan sulfate glycosaminoglycan as an attachment receptor" *J. Virol*
24. Zhang (2017) "Coxsackievirus A16 utilizes cell surface heparan sulfate glycosaminoglycans as its attachment receptor" *Emerg. Microbes Infect*
25. Nsaibia (2007) "The difficult-to-cultivate coxsackieviruses A can productively multiply in primary culture of mouse skeletal muscle" *Virus Res*
26. Chen (2018) "A 3.0-angstrom resolution cryo-electron microscopy structure and antigenic sites of coxsackievirus A6-like particles" *J. Virol*
27. Dai, Li, Liu et al. (2023) "Identification of four neutralizing antigenic sites on the enterovirus D68 capsid" *J. Virol*
28. Yang (2015) "Construction and characterization of an infectious clone of coxsackievirus A6 that showed high virulence in neonatal mice" *Virus Res*
29. Li (2023) "VP2 residue N142 of coxsackievirus A10 is critical for the interaction with KREMEN1 receptor and neutralizing antibodies and the pathogenicity in mice" *PLoS Pathog*
30. Zhang (2021) "Functional and structural characterization of a two-MAb cocktail for delayed treatment of enterovirus D68 infections" *Nat. Commun*
31. Zhang (2022) "Molecular mechanism of antibody neutralization of coxsackievirus A16" *Nat. Commun*
32. Punjani, Rubinstein, Fleet et al. (2017) "cryoS-PARC: algorithms for rapid unsupervised cryo-EM structure determination" *Nat. Methods*
33. Sanchez-Garcia (2021) "DeepEMhancer: a deep learning solution for cryo-EM volume post-processing" *Commun. Biol*
34. Abramson (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
35. Pettersen (2004) "UCSF Chimera-a visualization system for exploratory research and analysis" *J. Comput. Chem*
36. Afonine (2018) "Real-space refinement in PHENIX for cryo-EM and crystallography" *Acta Crystallogr. D. Struct. Biol*
37. Krissinel, Henrick (2007) "Inference of macromolecular assemblies from crystalline state" *J. Mol. Biol*
38. Meng (2023) "UCSF ChimeraX: tools for structure building and analysis" *Protein Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802250&blobtype=pdf | # HIV-1 infection regulates gene expression by altering alternative polyadenylation correlated with CPSF6 and CPSF5 redistribution
Charlotte Luchsinger, Annie Dai, Hari Yalamanchili, Aiswarya Balakrishnan, Kai-Lieh Huang, Cinzia Bertelli, Bin Cui, Ramon Lorenzo-Redondo, Eric Wagner, Felipe Diaz-Griffero
## Abstract
HIV-1 viral core transport to the nucleus, an early infection event, triggers the redistribution of cleavage and polyadenylation specificity factors (CPSF) 5 and CPSF6 to nuclear speckles, forming puncta-like structures. CPSF5 and CPSF6 regulate alternative polyadenylation (APA), which governs approximately 70% of gene expression. APA alters the lengths of mRNA 3'-untranslated regions (3′-UTRs), which contain regulatory signals influencing RNA stability, localization, and function. We investigated whether HIV-1 infection-induced changes in CPSF5 and CPSF6 subcellular localization are accompanied by APA changes. Using two independent methodologies to assess APA in human cell lines and primary CD4+ T cells, we found that HIV-1 infection regulates APA, shaped by the interaction of CPSF6 with the viral capsid, recapitulating the APA phenotype observed in CPSF6 knockout cells. Our study demonstrates that HIV-1 infection leverages the interaction between the viral capsid and CPSF6 to co-opt cellular processes, alter gene expression, and potentially contribute to viral pathogenesis.
IMPORTANCEThe interaction between HIV-1 and the cellular protein CPSF6 has been known for over 15 years; however, depletion of CPSF6 does not impair productive infection. An alternative possibility is that the virus exploits this protein to modulate cellular processes. This study demonstrates that HIV-1 infection alters the cellular function of CPSF6, an essential regulator of alternative polyadenylation-a mechanism that controls 70% of gene expression. Here, we show that HIV-1 regulates gene expression by disrupting the alternative polyadenylation function of CPSF6 through direct interaction. Overall, this reveals a novel strategy employed by the virus to modulate cellular gene expression.
The role of human CPSF6 in HIV-1 replication has been extensively studied in both cell lines and human primary cells, revealing that HIV-1 infectivity is affected by either the partial or complete depletion of CPSF6 expression (8,(11)(12)(13)(14)(15). Several roles for CPSF6 during HIV-1 infection have been proposed, including (1) facilitating HIV-1 core entry into the nuclear compartment (8,(16)(17)(18)(19), (2) influencing HIV-1 integration and integration site selection (12,18,(20)(21)(22), and (3) assisting with HIV-1 core disassembly (23).
HIV-1 infection alters the immunofluorescence microscopy patterns of CPSF5 and CPSF6 from a mixture of diffuse and granular nuclear staining to easily recognizable large condensates or puncta-like structures (2,4,5,8) that colocalize with the nuclear speckle marker SC35, implying that HIV-1 infection triggers the translocation of CPSF5 and CPSF6 to bona fide nuclear speckles (2,4,5,24). However, viruses bearing the capsid protein mutations N74D, A77V, or N57S, mutations that weaken the interaction between the capsid and CPSF6, fail to induce this translocation, indicating that this event is dependent on the capsid protein (2,4,5). Although the translocation of CPSF5 and CPSF6 to nuclear speckles is widely accepted, the effects of this translocation on the cellular functions of CPSF5 and CPSF6 remain unknown.
CPSF5, CPSF6, and CPSF7 are essential components of the tetrameric cleavage factor I (CFIm) complex, which consists of two 25-kDa (CPSF5) subunits and two proteins of either 59 or 68 kDa (CPSF7 or CPSF6) (25)(26)(27). CFIm complex plays a critical role in selecting the appropriate polyadenylation signal during mRNA processing (28,29). Most genes contain more than one polyadenylation signal, suggesting that mRNAs may be regulated by alternative polyadenylation (APA), an essential mechanism that controls approximately 70% of human gene expression (30)(31)(32)(33). The CFIm complex promotes the selection of distal rather than proximal polyadenylation signals, resulting in longer 3' untranslated regions (3'UTRs) (29). The 3'UTR contains regulatory elements that influence key mRNA metabolic properties, including mRNA stability, translation efficiency, nuclear export, and cellular localization (30,34,35). When CFIm components are depleted or CFIm activity is disrupted, proximal polyadenylation signals become more commonly selected, leading to shorter 3'UTRs (25,36,37). For example, the specific depletion of CPSF5 or CPSF6, but not CPSF7, in human cells leads to the shortening of mRNA 3'UTRs (38)(39)(40)(41).
The functional consequences of HIV-1 infection-induced CPSF5 and CPSF6 transloca tion to nuclear speckles are not understood. Because CPSF5 and CPSF6 control APA, we tested the hypothesis that HIV-1 infection modulates the APA functions of CPSF5 and CPSF6. We infected the human cell line A549 with HIV-1 and measured changes in APA using two independent methodologies, followed by similar analyses in human primary CD4 + T cells. We found that the infection of human cell lines and primary T cells with wild-type HIV-1 triggered APA changes, whereas APA changes were not induced by infection with HIV-1 bearing N74D or A77V mutation in the capsid protein, which prevents the capsid protein from interacting with CPSF6. Human primary T cells infected with wild-type HIV-1 showed an increase in transcripts with shorter 3' UTRs, which ultimately altered global cellular gene expression. To model these results, we utilized the human cell line A549, a cellular model in which HIV-1 can trigger CPSF5 and CPSF6 translocation to nuclear speckles in 80-90% of infected cells at a multiplicity of infection (MOI) of 2. APA analysis revealed changes in the polyadenylation signal used by a number of genes, resulting in an increase in the number of transcripts with shorter 3'UTRs. However, infection with HIV-1 bearing the capsid mutants N74D and A77V did not induce APA changes, suggesting that HIV-1 infection-induced APA changes require the direct interaction of the capsid protein with CPSF6. Unlike HIV-1 bearing the capsid mutations N74D or A77V, wild-type HIV-1 induced the translocation of CPSF5 and CPSF6 to nuclear speckles, suggesting that the ability of HIV-1 to modulate cellular APA depends on the interaction of the capsid with CPSF6. Overall, our study demon strates that HIV-1 infection exerts control over cellular gene expression by leveraging the interaction between the viral capsid and CPSF6 to alter the subcellular localization of CPSF5 and CPSF6, impacting their functions.
## RESULTS
## HIV-1 infection induces significant changes in alternative polyadenylation
In human cells, HIV-1 infection triggers the translocation of CPSF5 and CPSF6 to nuclear speckles; however, the efficiency of this process varies across cell lines (42). In the human A549 cell line, HIV-1 infection at an MOI of 2 is sufficient to induce CPSF5 and CPSF6 translocation in 80% of cells (42), whereas translocation is only observed in 20% of HeLa cells and 1% of THP-1 cells at an MOI of 2 (42). Due to the high efficiency of HIV-1 infection-induced CPSF5 and CPSF6 translocation to nuclear speckles in human A549 cells, we initially studied HIV-1-infection-induced changes in cellular expression in this cell line.
We infected A549 cells with an HIV-1 construct modified to express green fluorescent protein (HIV-1-GFP) as a reporter of infection at an MOI of 2. After 24 h, we measured infection as the percentage of GFP-positive cells (Fig. 1A), showing that ~90% of cells were infected. We also evaluated CPSF6 translocation by quantifying the percentage of infected cells containing CPSF6 colocalized with the known nuclear speckle marker SC35 (Fig. 1B), findings that 80%-90% of infected cells contained CPSF6 in nuclear speckles (Fig. 1C). In mock-infected cells, CPSF6 was distributed throughout the nucleus, showing a combination of diffuse and granular staining that only slightly overlaps with the SC35 marker and is excluded from the nucleolus. The limited overlap observed in mock-infected cells contrasts sharply with the strong colocalization of CPSF6 with SC35 observed in infected cells. Given extensive prior reports indicating roles for CPSF5 and CPSF6 in pre-mRNA cleavage and APA, the dramatic change in subcellular localization of these factors in response to HIV-1 infection suggests that viral infection could impact polyadenylation site selection, ultimately leading to altered gene expression (6).
To evaluate whether HIV-1 infection changes gene expression, we obtained total RNA from three HIV-1-infected and three mock-infected A549 samples and performed RNA-sequencing to determine the presence and quantity of RNA molecules (Fig. 1D through F). Changes in gene expression triggered by HIV-1 infection were plotted on a volcano plot, with the magnitude (as Log 2 fold-change) of change plotted on the x-axis (positive values, increased expression; negative values, decreased expression), and the P-value (as -Log 10 P, with higher values indicating more significant changes) plotted on the y-axis (Fig. 1D). HIV-1 infection of A549 cells resulted in the significant upregulation of 334 genes and the significant downregulation of 4589 genes, indicating that HIV-1 infection dramatically changes the cellular gene expression profile.
Our studies showed that HIV-1 infection dramatically altered cellular gene expres sion, indicating that HIV-1 may employ specific mechanisms to manipulate host gene expression in support of its replication. Gene ontology analysis of activated and suppressed pathways (Fig. 1E) showed that HIV-1 infection activated genes involved in countering viral infection (response to virus, viral processes, and others) and suppressed genes involved in DNA metabolism (DNA replication, recombination, and repair). Gene expression analysis using the Reactome database revealed that HIV-1 infection activated genes involved in countering viral infection (Interferon [IFN]α/B. IFNγ, IFN signaling, and others) and suppressed genes involved in cell division (M phase, mitotic anaphase, and cell cycle checkpoints) (Fig. S1A).
We and others have described the translocation of CPSF6 to nuclear speckles (2, 4, 5); however, the effect of this translocation on CPSF6 function is not understood. This translocation to nuclear speckles may inhibit the cellular functions of CPSF6, including APA regulation. APA is a major gene expression regulatory mechanism, with 70% of human genes containing proximal and distal APA signs in the 3'UTR (30). Because CPSF5 or CPSF6 depletion is known to induce the use of proximal polyadenylation signals (12,38,39,(43)(44)(45), we tested the hypothesis that the relocalization of CPSF5 and CPSF6 to nuclear speckles during HIV-1 infection correlates with their functions and leads to APA dysregulation (6). We used regression of polyadenylation compositions (REPAC) to analyze APA in total RNA-sequencing data from HIV-1-infected and mock-infected A549 cells (46). We identified 246 genes associated with shorter 3'UTRs and 65 genes associated with longer 3'UTRs in HIV-1-infected cells than in mock-infected cells (Fig. 1F). These results showed significant 3'UTR shortening in HIV-1-infected cells compared with mock-infected cells, suggesting that HIV-1 infection induces 3'UTR shortening and the use of proximal polyadenylation sites, phenotypically resembling the APA patterns observed with the knockout (KO) of CFIm complex components, such as CPSF5 and CPSF6. These findings suggest that HIV-1 infection interferes with the functions of CPSF5 and CPSF6.
To determine whether changes in the 3' UTR length affect gene expression, we analyzed changes in the expression of genes with HIV-1 infection-induced changes in 3' UTR length (Fig. 1F). Of the 246 genes with shortened 3'UTRs following HIV-1 infection, 87 demonstrated altered transcript levels, including 82 with reduced transcript levels; however, APA may alter protein expression without altering transcript levels, as the 3'UTR can impact translation efficiency and transcript localization. By contrast, 47 of the 65 genes with longer 3' UTRs following HIV-1 infection showed altered transcript levels, including 45 with reduced transcript levels. These findings indicate that HIV-1 infection-induced changes in APA alter cellular gene expression.
We used Reactome to analyze the cellular pathways associated with genes modulated by HIV-1 infection and found that most of the genes downregulated by HIV-1 infec tion-induced changes in APA control cellular processes, such as SUMOylation, organelle biogenesis, transport, and general transcription (Fig. S1B).
## HIV-1 infection-induced changes in alternative polyadenylation are correla ted with the translocation of CPSF5 and CPSF6 to nuclear speckles
Capsid proteins bearing the N74D and A77V mutations bind CPSF6 at significantly lower levels than wild-type capsid proteins (13,47). To investigate the role played by HIV-1 infection-induced CPSF5 and CPSF6 translocation in infection-induced changes in APA, we infected A549 cells with HIV-1 bearing the capsid mutation N74D (HIV-1-N74D) or A77V (HIV-1-A77V), which fail to induce CPSF5 and CPSF6 translocation to nuclear speckles (2,4). Infections with HIV-1-WT, HIV-1-N74D, and A77V viruses were carried out at an MOI=2 for 48h, resulting in ~85% of the cells being infected.
We employed a 3'-end sequencing approach, known as poly(A)-Click-sequencing (PAC-seq), to map global mRNA 3'-end changes in HIV-1-infected and mock-infected A549 cells. The PAC-seq technique is specifically designed to capture, enrich, and sequence polyadenylation tail junctions, enabling the direct identification and quan tification of APA events rather than inferring their locations and assessing relative changes, which is the typical approach used when re-analyzing standard RNA-sequenc ing databases (48,49). We utilized our robust PolyAMiner analysis pipeline, which was custom-designed to analyze PAC-seq data by employing a vector projection-based engine to calculate intermediate polyadenylation signal usage (50,51). immunolabeled using specific antibodies against SC35 (red, marker for nuclear speckles) and CPSF6 (green). Nuclei were stained with DAPI (blue). (C) The average (and standard deviation) percentage of cells containing CPSF6 in nuclear speckles (SC-35 positive compartments) from three independent experiments, as determined by visual examination of 200 cells, is shown. Significance was determined using an unpaired t-test; ***P < 0.001. (D) Total RNA from three infected and three mock-infected samples was sequenced, and gene expression is displayed in a Volcano plot. The x-axis shows log 2 fold changes in gene expression, with positive indicating upregulation and negative indicating downregulation. The y-axis shows the statistical significance, expressed as -Log 10 P. (E) Gene ontology analysis of all genes upregulated and downregulated by HIV-1 infection. The number of genes in each pathway is expressed as the size of the sphere, and the color represents the P-adjusted value. (F) Volcano plot showing polyadenylation signal (PAS) changes. The x-axis shows compositional fold change, which represents the 3' UTR length for each transcript. A positive compositional fold change indicates the presence of a longer 3'UTR (red), and a negative compositional fold change indicates the presence of a shorter 3'UTR (green). The y-axis shows the statistical significance of the result, conveyed as the -Log 10 P. CPSF6, cleavage and polyadenylation specificity factor subunit 6; DAPI, 4'6-diamino-phenylindole; GFP, green fluorescent protein; MOI, multiplicity of infection; PAS, polyadenylation signal; UTR, untranslated region.
We first used differential gene expression analysis using DESeq2 to compare the PAC-seq results for RNA isolated from A549 cells infected with wild-type and mutant HIV-1. We found very similar results to those obtained using standard RNA-sequencing analyses (Fig. S2A through C), showing significant upregulation of antiviral genes (Fig. S2D) and significant downregulation of genes involved in chromosome maintenance and cell division (Fig. S2E). We then analyzed PAC-seq data using PolyAMiner and found similar numbers of total mapped polyadenylation sites for each virus (Table S1). HIV-1 infection of A549 cells resulted in significant APA events, including both 3'UTR shortening and lengthening, compared with mock-infected controls (Fig. 2A). Compared with wild-type HIV-1 infections, infections with HIV-1-N74D or HIV-1-A77V, which do not induce CPSF5 and CPSF6 translocation to nuclear speckles, resulted in a similar number of 3'UTR lengthening events but a reduced number of 3'UTR shortening events (Fig. 2B andC). These data suggest that HIV-1 infection partially mimics the CPSF6 KO phenotype, which results in a significant increase in the number of mRNA transcripts with shortened 3' UTRs (43). Functional gene analysis indicated an expected increase in antiviral gene expression, such as IFN regulatory factor 7 (Fig. 2D), and shortened 3'UTRs for both vascular membrane protein 21 (VMA21, Fig. 2E) and TBC1 domain family member 12 (TBC1D12, Fig. 2F), although the shifts toward proximal polyadeny lation sites for both VMA21 and TBC1D12 were less pronounced following infection with mutant HIV-1 than with wild-type HIV-1. These experiments suggest that HIV-1 infection-induced changes in APA depend on the capsid-CPSF6 interaction, and that HIV-1 infection leverages the viral capsid-CPSF6 interaction to modulate the function of the CFIm complex.
## HIV-1 infection of human primary CD4 + T cells induces changes in alternative polyadenylation mediated by the interaction between the viral capsid and CPSF6
To investigate whether HIV-1 infection of human CD4+ T cells affects CFIm complex function, we measured changes in APA in human primary CD4+ T cells infected with wild-type HIV-1, HIV-1-N74D, or HIV-1-A77V (all expressing GFP as an infection reporter) at an MOI of 2. After 48 h, ~30% of cells infected with HIV-1 wild-type or the N74D and A77V mutants were GFP-positive. These GFP-positive cells were sorted by FACS and maintained in culture for an additional 48 h. At 96 h post-infection, RNA was isolated from the sorted cells and subjected to PAC-seq and PolyAMiner analysis. Similar to the response observed in A549 cells, human primary CD4 + T cells showed significant increases in the number of mRNA transcripts with altered polyadenylation site selection in HIV-1-infected cells than in mock-infected cells (Fig. 3A; Table S2A). We also observed fewer mRNAs with shortened 3' UTRs in cells infected with HIV-1-N74D (Fig. 3B; Table S2B) or HIV-1-A77V (Fig. 3C; Table S2C). In accordance with our data from A549 cells, although of a more limited magnitude, these results show that 3' UTR shortening events induced by HIV-1 infection of primary CD4 + T cells correlate with the interaction between the viral capsid and CPSF6.
## HIV-1 infection-induced changes in alternative polyadenylation result in altered cellular protein expression
APA is a post-transcriptional regulatory mechanism that allows a single gene to produce mRNA transcripts with varying 3' UTR lengths. Even when the overall mRNA levels remain unchanged, the effects of different 3' UTR lengths become obvious at the protein level, as the 3' UTR contains regulatory elements, such as binding sites for RNA-binding proteins and microRNAs, that can significantly influence the gene expression pattern by altering mRNA stability, localization, or translational efficiency. Therefore, we sought to evaluate whether HIV-1 infection alters the expression of proteins encoded by mRNAs with varying 3' UTR lengths (Table 1), including those that displayed changes in transcript levels (Table 1).
Our analysis revealed that the transcript encoding Schlafen family member 5 (SLFN5) was upregulated upon HIV-1 infection but with a shortened 3' UTR (Table 1) due to changes in APA upon HIV-1 infection (Fig. 4A). Consequently, SLFN5 was selected as a promising candidate to explore the effects of changes in 3' UTR length on protein expression. We infected human A549 cells with HIV-1-GFP at an MOI of 2 for 48 h and assessed protein expression by western blot analysis. Across three independent HIV-1-GFP preparations, HIV-1 infection increased SLFN5 protein expression by 6-7 fold that of mock-infected cells (Fig. 4B). We also infected A549 cells with an HIV-1 expressing luciferase (HIV-1-Luc) at an MOI of 2 for 48 h, and immunoblot analysis revealed that infected cells had significantly increased SLFN5 levels that were 7-8-fold the levels in mock-infected controls (Fig. 4C). To assess whether the change in SLFN5 expression induced by HIV-1 infection also occurs in the main targets of HIV-1, we infected human primary cells, including monocytes, macrophages, and CD4+ T cells with HIV-1-GFP at an MOI of 2 for 96 h and measured SLFN5 protein expression by western blot analysis (Fig. S3A through C). Consistent with the results in A549 cells, we observed that after HIV-1 infection, SLFN5 expression levels increased in monocytes, macrophages, and to a lesser extent, in CD4+ T cells (Fig. S3A through C). To further address the importance of HIV-1 infection in the SLFN5 expression pattern, we performed additional experiments. First, we challenged A549 cells with increasing amounts of HIV-1-GFP for 48 h and determined SLFN5 expression levels by western blot analysis (Fig. S4A). We found that the SLFN5 expression levels increase with higher viral input, indicating that HIV-1 infection induces alterations in SLFN5 expression in a dose-dependent manner (Fig. S4A). Next, to assess SLFN5 expression over time, we performed a time course of HIV-1 infection in which A549 cells were infected with HIV-1-GFP (Fig. S4B) at an MOI of 2 for 0, 48, 72, or 96 h. Interestingly, we observed that the effect on SLFN5 expression was dependent on the duration of infection, with the greatest increase in SLFN5 levels occurring at 96 h post-infection (Fig. S4B). We also measured SLFN5 expression upon infection with HIV-1-N74D or HIV-1-A77V, which express a capsid protein that does not interact with CPSF6, and found that these mutant viruses failed to induce SLFN5 expression (Fig. 4D), suggesting that this induction is dependent on the interaction between the capsid protein and CPSF6. To verify whether the observed increase in SLFN5 expression was specifically due to productive HIV-1 infection, we repeated infections using heat-inactivated HIV-1-GFP viruses (95°C for 30 mins). Heat inactivation alters HIV-1 components by inducing viral protein denaturation and RNA genome degradation, rendering the virus unable to infect cells. We found that SLFN5 expression was significantly reduced in A549 cells exposed to heat-inactivated HIV-1 when compared to cells infected with replication-competent virus (Fig. 4E), indicating that active viral infection is required to induce SLFN5 expres sion. To determine whether the increased SLFN5 expression observed in HIV-1-infected cells correlates with CPSF5 and CPSF6 translocation to nuclear speckles, we analyzed SLFN5 expression in human cells infected with HIV-1 in the presence of the small molecule PF74, which prevents CPSF5 and CPSF6 translocation to nuclear speckles during infection (52). Interestingly, the use of PF74 increases expression of SLFN5 in mock-infected cells when compared to untreated cells (Fig. 4F). Similar to cells infected with HIV-1-N74D and HIV-1-A77V, cells infected with HIV-1 in the presence of PF74 failed to induce SLFN5 expression above the levels observed in mock-infected cells (Fig. 4F). To obtain additional data on the connection between capsid-CPSF6 interaction and increased SLFN5 expression, we challenged A549 cells with HIV-1-GFP at MOI=2 for 48 h in the presence or not of reverse transcriptase (RT) inhibitor Nevirapine (Nev), which doesn't prevent CPSF5 and CPSF6 translocation to nuclear speckles during infection. SLFN5 expression was evaluated by western blot analysis (Fig. S5). We found that the presence of the RT inhibitor partially increased SLFN5 expression, indicating that while the capsid-CPSF6 interaction is important for inducing the change in SLFN5 expression, other factors can affect SLFN5 expression. Together, these findings support the notion that the induction of SLFN5 expression in HIV-1-infected cells is associated with the translocation of CPSF5 and CPSF6 to nuclear speckles, highlighting a potential link between capsid-dependent nuclear events and SLFN5 expression. Because SLFN5 is a type I IFN-induced protein, we tested whether IFN-inducing contaminants in the viral preparation could explain the observed SLFN5 upregulation. A549 cells were infected with HIV-1-GFP at an MOI of 2 for 48 h, and the protein levels of myxovirus resistance protein 1 (MxA) and IFN-stimulated gene 15 (ISG15), which are both type I IFN-induced proteins, were analyzed by western blot. As a positive control, uninfected cells were treated with 1000 U/mL of IFNα or IFNβ for ~40 h. The protein expression levels of MxA and ISG15 in infected cells were comparable to the levels observed in mock-infected controls (Fig. 4G), indicating the absence of IFN-inducing contaminants. These findings demonstrate that HIV-1 infection itself drives the upregulation of SLFN5.
Overall, these experiments indicate that the protein expression levels of SLFN5 are upregulated as a result of HIV-1-induced changes in APA.
## The loss of CPSF6 expression shortens the 3'UTRs of cellular mRNAs at a global level
Our results suggest that HIV-1 infection affects APA through the interaction of viral capsid with CPSF6, resulting in the shortening of mRNA 3'UTRs, ultimately changing cellular protein expression levels. The absence of CPSF5 or CPSF6 expression in different cell types results in the shortening of cellular mRNA 3'UTRs (12,38,39,(43)(44)(45); there fore, we tested whether knocking out CPSF6 in A549 cells would also increase the number of cellular mRNAs with shorter 3'UTRs. We used the CRISPR-Cas9 system to generate CPSF6-KO cells and control A549 cells using a non-targeting clone (NT#H1). From 12 single CPSF6-KO clones, we selected three independent clones for this study (Fig. 5A), designated as CPSF6-KO#B4, CPSF6-KO#B7, and CPSF6-KO#C8, and confirmed total CPSF6 depletion in each clone by western blot. As controls, we used parental A549 cells (wild-type) and the NT#H1 cells. We also evaluated the expression levels of other CFIm complex members, such as CPSF5 and CPSF7. We found a significant decrease in CPSF5 expression levels in CPSF6-KO clones compared with the levels in control cells (Fig. 5A). We observed that CPSF7 expression levels in CPSF6-KO cell lines were comparable to those observed in wild-type cells (Fig. 5A). These experiments suggest the possibility that CPSF5 stability depends on CPSF6 expression, either because CPSF5 and CPSF6 form a heterotetramer or because any impact on the CFIm complex affects preparations at an MOI of 2 for 48 h; (D) HIV-1-GFP, HIV-1-N74D-GFP, or HIV-1-A77V-GFP at an MOI of 2 for 48 h; (E) HIV-1-GFP with or without heat-inactivation (95°C for 30 min) at an MOI of 2 for 48 h; (F) HIV-1-GFP at an MOI of 2 for 48 h with or without 10 µM PF74; and (G) HIV-1-Luc or HIV-1-GFP at an MOI of 2 for 48 h. Cells were lysed, and proteins were analyzed by western blot using the indicated antibodies. Virus presence was assessed using anti-p24 antibodies, and anti-GAPDH antibodies were used as a protein loading control. Each experiment was repeated at least three times, and a representative image is shown. Graphs show the average densitometry quantification of three replicates, with standard deviations, is shown. Significance was determined using unpaired t-test (B and C) or ANOVA multiple comparisons tests (D-F); *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant. GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GFP, green fluorescent protein; ISG15, interferon-stimulated gene 15; Luc, luciferase; MOI, multiplicity of infection; MxA, human myxovirus resistance protein 1; p24, viral capsid; SLFN5, Schlafen family member 5. CPSF5 expression. Together, these results suggest that a regulatory mechanism may exist among the expression levels of different CFIm complex members.
We also used PAC-seq to analyze changes in APA in CPSF6-KO cells (Fig. 5B; Table S3A through C). Consistent with the known role of CPSF5 in APA, we observed a global shortening of mRNA 3' UTRs in CPSF6-KO cells compared with those in wild-type cells. Because CPSF5 and CPSF6 work together in the CFIm complex, these findings align with previous reports that the reduction or complete depletion of CPSF5 in human cells increases the prevalence of mRNAs with shorter 3'UTRs (28,43,53). Consistently, we observed that the expression of VMA21, a well-established reporter of changes in alternative polyadenylation (APA) (54), was increased in CPSF6-KO cells compared to wild-type cells, suggesting that CPSF6-KO leads to a complete deregulation of cellular APA (Fig. S6).
Next, we tested the effects of CPSF6 depletion on HIV-1 infectivity in A549 cells. We challenged CPSF6-KO and control cells with increasing amounts of HIV-1-GFP for 24 or 48 h and determined infectivity by measuring the percentage of GFP-positive cells using flow cytometry (Fig. 5C). These results showed that CPSF6 depletion does not significantly affect HIV-1 infection in A549 cells. Previously, we and others described that HIV-1 infection induces the translocation of both CPSF6 and CPSF5, a CFIm complex member that works together with CPSF6, to nuclear speckles (5,8). Therefore, we next examined whether HIV-1 infection can induce CPSF5 translocation to nuclear speckles in CPSF6-KO cells. CPSF6-KO A549 cells were infected with HIV-1-GFP at an MOI of 2 for 48 h, fixed, permeabilized, and subjected to immunofluorescence analysis using anti-SC35, anti-CPSF5, and anti-lens epithelium-derived growth factor/p75 (LEDGF/p75) antibodies. Unlike in wild-type cells, HIV-1 infection did not induce CPSF5 translocation to nuclear speckles in CPSF6-KO cells, suggesting that CPSF5 translocation to nuclear speckles is dependent on CPSF6 translocation (Fig. 5Di andE). However, the localization of the nuclear speckle marker SC35 was unchanged between wild-type and CPSF6-KO A549 cells, indicating that nuclear speckles were not affected (Fig. 5Dii andE). We previously showed that the HIV-1 integration cofactor LEDGF/p75 surrounds nuclear speckles containing CPSF5 and CPSF6 upon HIV-1 infection5; therefore, we tested whether CPSF6 was required for this process to occur and found that LEDGF/p75 does not surround nuclear speckles upon HIV-1 infection in CPSF6-KO cells (Fig. 5Diii andiv). As a control, we quantified the percentage of cells containing CPSF5, CPSF6, and CPSF7 in nuclear speckles upon HIV-1 infection of wild-type and CPSF6-KO cells (Fig. 5E). Overall, these results indicate that the translocation of CPSF6 to nuclear speckles is important for the recruitment of CPSF5 and LEDGF/p75 to nuclear speckles upon HIV-1 infection.
## HIV-1 infection partially mimics the loss of CPSF6 expression
The analysis of protein expression levels encoded by mRNA transcripts undergoing APA revealed that SLFN5 protein levels are upregulated upon HIV-1 infection. Because HIV-1 standard deviation. Significance was determined using ANOVA multiple comparisons tests; ***P < 0.001; ns, not significant. (B) Total RNA from CPSF6-KO and WT A549 cells was prepared, and polyadenylated transcripts were identified by PAC-seq, followed by APA analysis using PolyAMiner. (C) A549 WT, NT#H1, CPSF6-KO#B4, CPSF6-KO#B7, and CPSF6-KO#C8 cells were infected with increasing amounts of HIV-1-GFP for 24 or 48 h. Infection was assessed as the percentage of GFP-positive cells by flow cytometry. (D) WT and CPSF6-KO#B7 A549 cells were infected with HIV-1-GFP at an MOI of 2 for 48 h. Cells were fixed, permeabilized, and stained using the following antibodies: (i) anti-CPSF5 (red) and anti-CPSF6 (green); (ii) anti-SC-35 (red) and anti-CPSF6 (green);
(iii) anti-LEDGF/p75 (red) and anti-CPSF6 (green); and (iv) anti-LEDGF/p75 (red) and anti-SC35 (green). Nuclei were stained with DAPI (blue). Scale bar, 10 µm. (E) Percentage of A549 cells containing CPSF6, CPSF5, or CPSF7 in nuclear speckles (condensates) upon HIV-1 infection (average of three independent experiments with standard deviation). Cells containing CPSF6, CPSF5, or CPSF7 in nuclear speckles were determined by visual examination of 200 cells. infection mimics the phenotype observed in CPSF6-KO cells, we investigated whether CPSF6 depletion influences SLFN5 expression. CPSF6-KO cells exhibited high SFLN5 levels, up to 20-fold the levels observed in wild-type or NT#H1 control cells (Fig. 6A), similar to the upregulation observed during HIV-1 infection. These experiments suggest that SLFN5 expression levels are regulated by APA.
Because we found that CPSF6-KO cells exhibit a consistent decrease in CPSF5 protein expression compared with wild-type cells, we tested whether HIV-1 infection decreases CPSF5 expression. Similar to the phenotype observed in CPSF6-KO cells, we observed that HIV-1 infection decreases CPSF5 protein expression levels (Fig. 6B). To test the specificity of HIV-1 infection-induced decreases in CPSF5 protein expression, we measured CPSF5 levels upon infection with HIV-1-A77V, which does not induce the translocation of CPSF5 and CPSF6 to nuclear speckles. Unlike the decrease in SLFN5 levels observed with wild-type HIV-1 infection, CPSF5 expression levels in cells infected with HIV-1-A77V were similar to those in mock-infected cells (Fig. 6C), suggesting that HIV-1 infection specifically alters cellular expression patterns through the interaction of the viral capsid with CPSF6. Finally, we tested whether HIV-1 infection in CPSF6-depleted A549 cells (CPSF6-KO) had an effect on SLFN5 expression, a readout of APA regulation. To do this, we challenged A549 CPSF6-KO#B4, CPSF6-KO#B7, and CPSF6-KO#C8 cells with HIV-1-GFP at an MOI=2 for 48h (Fig. S7). In CPSF6-KO cells, the presence of HIV-1 infection did not produce a significant increase in SLFN5 levels compared to uninfected CPSF6-KO cells, indicating that in CPSF6-KO cells, the expression level of SLFN5, and consequently also the APA, reached its threshold. These data suggest that CPSF6 is necessary for HIV-1mediated APA remodeling: in the absence of CPSF6, the viral effect is not observed.
## DISCUSSION
We and others have shown that HIV-1 infection triggers the translocation of CPSF6 to nuclear speckles (2,4). We also demonstrated that the HIV-1 core must enter the nuclear compartment for this translocation to occur and that this process is dependent on the viral capsid protein. Interestingly, CPSF5 also translocates to nuclear speckles upon HIV-1 infection (5,8,55). We recently speculated that HIV-1 induces the translocation of CPSF6 to nuclear speckles as a strategy for regulating cellular gene expression (6). Because both CPSF5 and CPSF6 are involved in the regulation of APA, a key cellular mechanism that controls gene expression, we tested the hypothesis that HIV-1 infection alters the nuclear localization of CPSF5 and CPSF6, thereby dysregulating APA and ultimately modifying cellular gene expression.
To understand the ability of HIV-1 to modulate cellular gene expression in human cells, we first used RNA sequencing to define all gene expression changes triggered by HIV-1 infection in human primary cells and cell lines. Using different databases, we identified the activation of genes countering viral infection (response to virus, viral processes, and others) and the suppression of genes involved in DNA metabolism (DNA replication, recombination and repair, and cell division). HIV-1 infection of A549 cells led were analyzed by western blot using anti-SLFN5, anti-CPSF6, and anti-CPSF5. Virus presence was assessed using anti-p24 antibodies, and anti-GAPDH antibodies were used as a protein loading control. (C) A549 cells were challenged with three different HIV-1-A77V-GFP preparations using an MOI of 2 for 48 h. Cells were lysed and analyzed by western blot using anti-SLFN5 and anti-CPSF5. Virus presence was assessed using anti-p24 antibodies, and anti-GAPDH antibodies were used as a protein loading control. (A-C) All experiments were repeated at least three times, and a representative image is shown.
Graphs show the average densitometry quantification of at least three replicates with standard deviation. Significance was determined using ANOVA multiple comparisons tests; ***P < 0.001; ns, not significant. CPSF5, cleavage and polyadenyla tion specificity factor subunit 5; CPSF6, cleavage and polyadenylation specificity factor subunit 6; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GFP, green fluorescent protein; KO, knockout; Luc, luciferase; MOI, multiplicity of infection; NT, non-targeting; p24, viral capsid; SLFN5, Schlafen family member 5; WT, wild-type.
to more genes being significantly downregulated than upregulated, which is consistent with previous RNA-sequencing studies of HIV-1-infected cells (56).
To test the hypothesis that HIV-1 may impact gene expression via APA modulation, we measured APA in HIV-1-infected human A549 cells and human primary CD4 + T cells using two independent, state-of-the-art methodologies: PAC-seq, which maps global mRNA 3'-end changes by specifically sequencing polyadenylated transcripts (48), followed by PolyAMiner (50); and REPAC (46), an algorithm that uses total RNA-sequenc ing data to identify changes in 3'UTRs. Our analyses revealed that HIV-1 infection globally modulates APA, leading to a significant increase not only of mRNAs with shorter 3'UTRs but also, to a lesser extent, longer 3'UTRs. The mRNA 3'UTR contains regula tory elements that determine mRNA metabolic features, including stability, translation, nuclear export, and cellular localization (30). Therefore, changing the 3' UTR length through the selection of APA sites can dramatically affect gene expression and function. These results demonstrate that HIV-1 uses APA as at least one mechanism for controlling gene expression. We also showed that infection by HIV-1 bearing a capsid mutation that prevents the interaction of the viral capsid with CPSF6 was unable to modulate APA, suggesting that the ability of the viral capsid to interact with CPSF6 is essential for modulating cellular APA.
This study suggests that HIV-1 infection-induced translocation to nuclear speckles affects the cellular functions of CPSF5 and CPSF6. In support of this finding, CPSF5 depletion in human cells alters APA regulation (43,53), ultimately impacting cellular gene expression. Thus, the interaction of the viral core with CPSF6 within the nuclear compartment likely represents an upstream event that leads to changes in cellular gene expression. In agreement with these results, similar to CPSF5 depletion, CPSF6 KO in human A549 cells significantly increases the number of mRNAs with short 3' UTRs. The ability of viruses to modulate host gene expression to facilitate viral replication is well established, with numerous viruses from various families known to manipulate cellular gene expression through viral proteins that target transcription, mRNA processing, and translation. For example, HSV-1 inhibits splicing via the viral protein ICP27, poliovirus blocks mRNA export via the protein 2A, influenza virus suppresses polyadenylation via NS1, adenovirus inhibits transcription initiation via E1A, rubella virus impedes translation via its capsid protein, and Bunya viruses inhibit translation via their NS proteins (57)(58)(59). These examples illustrate that viruses have evolved diverse strategies for controlling the gene expression profiles of infected cells. In most cases, the viral regulation of host gene expression ultimately serves to create a permissive environment for viral replication. We propose that HIV-1 infection alters cellular expression by modulating APA through the interaction of the viral capsid with CPSF6.
Depleting CPSF5 expression in human cells significantly increases the number of cellular mRNAs with shorter 3' UTRs (43,53). To determine whether CPSF6 depletion has similar effects, we generated CPSF6-KO A549 cells and analyzed changes in APA. Consistent with the results of CPSF5 depletion (12,43), CPSF6-KO A549 cells displayed a marked increase in mRNAs with shorter 3' UTRs. CPSF5 and CPSF6 function together within the CFIm complex, and the loss of either protein is likely to disrupt the integrity of this complex, severely impacting APA. Our observations also revealed that CPSF5 and LEDGF/p75 translocation to nuclear speckles is dependent on CPSF6 expression, suggesting that the interaction of the viral capsid with CPSF6 occurs upstream from the recruitment of CPSF5 and LEDGF/p75 to nuclear speckles. HIV-1 infection modulates the CFIm complex through direct interaction between the viral capsid and CPSF6, and these experiments suggest that the sequestration of CPSF6 to nuclear speckles by the viral capsid affects the cellular function of the CFIm complex.
To identify a protein that could potentially be used to monitor changes in APA during HIV-1 infection, we examined several proteins encoded by transcripts that undergo APA in HIV-1-infected cells. SLFN5 expression increases during HIV-1 infection in A549 cells and primary cells, which is in agreement with previous observations in peripheral blood mononuclear cells (60). Furthermore, this report also shows that SLFN5 overexpression results in decreased HIV-1 infection, while decreased SLFN5 promotes HIV-1 replication, demonstrating a connection between SLFN5 expression and HIV-1 infection (60). We also found similar alterations in SLFN5 expression levels in CPSF6-KO cells, possibly due to changes in APA. However, the APA changes induced by HIV-1 infection were more modest than those observed after CPSF6 knockdown. This difference likely reflects the biology of infection, where capsid-CPSF6 interactions modulate CFIm function transiently and only to a partial extent, in contrast to the complete disruption pro duced by genetic depletion of CPSF6. To further examine the relevance of these infection-induced changes, we measured SLFN5 expression, a readout of APA regula tion, under different experimental conditions. Infection at increasing MOIs revealed a clear dose-dependent increase in SLFN5 expression, demonstrating that the extent of APA remodeling correlates with viral input. Moreover, time-course experiments showed that SLFN5 expression increased progressively with infection time, consistent with a temporal window during which capsid-CPSF6 interactions are active in the nucleus. Taken together, these results indicate that although HIV-1-induced APA remodeling is moderate in magnitude, it is reproducible and dependent on both the dose and timing of infection. These findings also suggest that SLFN5 may serve as a useful tool for monitoring HIV-1-induced changes in APA. We observed that both HIV-1 infection and CPSF6-KO reduce CPSF5 expression levels in human cells. Overall, this study suggests that HIV-1 infection mimics the phenotype observed in CPSF6-KO cells, indicating that HIV-1 sequesters CPSF5 and CPSF6 in nuclear speckles to alter their function, represent ing a novel regulatory mechanism.
## MATERIALS AND METHODS
## Cell culture and generation of cell lines
Human lung carcinoma A549 and human embryonic kidney HEK293T cells were obtained from the American Type Culture Collection (A549: Cat # CCL-1885; HEK293T: Cat #CRL-3216) and maintained in Dulbecco's modified Eagle medium supplemented with 10% heat-inactivated fetal bovine serum (FBS), 100 U/mL penicillin, 100 µg/mL streptomycin, 29.2 mg/mL L-glutamine (Life Sciences), and 5 µg/mL plasmocin (Invivo Gen, San Diego, CA), in a humidified incubator with 5% CO 2 at 37°C. Human lymphocytes were isolated from peripheral blood mononuclear cells obtained from three healthy donors using the Pan T Cells Isolation kit (Miltenyi Biotec, Cat#130-096-535). T cells were activated by culturing in complete Roswell Park Memorial Institute (RPMI-1640) medium supplemented with 4 µg/mL phytohemagglutinin-L (PHA-L; Sigma-Aldrich Cat#11-249-738-001) and 20 IU/mL interleukin-2 (IL-2; Miltenyi Biotec Cat#130-097-743). Monocytes were isolated from peripheral blood mononuclear cells (PBMCs) from a healthy donor using the Pan monocytes Isolation kit (Miltenyi Biotec, Cat#130-096-537). Monocytes were differentiated into macrophages by culturing in RPMI supplemen ted with 10% human serum and 40ng/mL Macrophage colony-stimulating factor (M-CSF; BioLegend Cat#574804) for 7 days. CPSF6-KO A549 cells were generated using the CRISPR-Cas9 ribonucleoprotein (crRNP) gene-editing system, according to the manufacturer's instructions (Synthego). The sequences of the synthetic guide RNAs (sgRNAs) targeting CPSF6 were: sgRNA-1: 5′-CUUUUUAGGUUUGCCCUUGU -3' , sgRNA-2: 5′-UUACCUAAAAGAGAACUUCA-3′, and sgRNA-3: 5′-UGAGGACUGCUUACUUUUCC -3' . Cells were electroporated with Amaxa 4D nucleofector according to the manufacturer's instructions (Lonza). After 72 h, cells were subjected to limiting dilution in 96-well plates. Clones were assessed for KO by immunofluorescence and immunoblot analysis with rabbit polyclonal CPSF6 antibody. Control cells were nucleofected with crRNPs carrying a non-targeting sgRNA (NT#H1): 5′-GCACUACCAGAGCUAACUCA-3′.
## Antibodies and cell reagents
We used mouse monoclonal antibodies targeting the following proteins: SC35 (clone SC-35; Cat# ab11826, Abcam), CPSF5 (clone 3F8; Cat# H00011051-M12, Novus Biologicals), CPSF7 (clone A-9; Cat# sc-393880, Santa Cruz), CPSF6 (clone F-3; Cat# sc-376228, Santa Cruz), ISG15 (clone F-9; Cat# sc-166755, Santa Cruz), and HIV-1 p24 (clone 183-H12-5C; Cat# ARP-3537, NIH AIDS Reagent Program). We used rabbit polyclonal antibodies against the following proteins: CPSF6 (Cat# ab99347, Abcam), CPSF7 (Cat#A301-359A, Bethyl Laboratories, Inc), LEDGF/p75 (Cat# A300-847A, Bethyl Laboratories, Inc), SLFN5 (Cat#PA5-53638, Invitrogen), and MxA (Cat#PA5-22101, Invitrogen) and VMA21 (Cat#21921-1-AP, Proteintech). The fluorescent nuclear stain 4' ,6-diamidino-2-phenylindole (DAPI), and the following fluorescently labeled antibodies were from Life Technologies: Alexa Fluor 488-conjugated, Alexa Fluor 594-conjugated, and Alexa Fluor 647-conjugated donkey anti-rabbit IgG and anti-mouse IgG. The human IFNαA and 1βA (Cat#IF007 and IF014, respectively) were from Millipore. TRIzol reagent (Cat#15596026) was from Invitrogen. The inhibitor nevirapine (Nev) (Cat#4666) was from NIH AIDS Reagent Program. Dimethyl sulfoxide (DMSO) (Cat# D2438) and PF74 (Cat#SML0835) were from Sigma-Aldrich.
## Production of HIV-1
Wild-type and mutant HIV-1 (HIV-1-N74D and HIV-1-A77V) expressing GFP or Luc as reporters were produced by co-transfecting HIV-1-gag-pol, LTR-GFP-LTR/LTR-Luc-LTR, tat, rev, and VSV-G in HEK293T/17 cells, as described (61). Viruses were collected 48 h after transfection, filtered, concentrated, tittered, aliquoted, and stored at -80°C. Viral stocks were titrated by limiting dilution in A549 cells by measuring the percentage of GFP-pos itive cells at 48 h. Heat-inactivation of HIV-1 was performed by incubating the viral preparations at 95°C for 30 min.
## Analysis of viral infectivity by flow cytometry
The infectivity and titer of HIV-1-Luc were measured using TZM-bl GFP-reporter cells, in which HIV-1 infection induces GFP, as described (62). We infected human A549 cells and determined the percentage of GFP-positive cells using a flow cytometer (BD Celesta). Virus titer calculation was performed according to the following equation: Infectious units (IU)/mL = (cell number) × (% of GFP-positive cells) × (dilution factor), where the dilution factor = 1,000 µL ÷ viral input (µL). To calculate the volume of virus used at a specific MOI, use the following equation: MOI = [(virus stock IU/mL) × (volume of virus used)] ÷ (number of cells in infection).
## Immunofluorescence microscopy image acquisition and deconvolution
Samples for immunofluorescence analysis were prepared as described previously, with some modifications (2). Briefly, cells were seeded on 12-mm, round, glass coverslips in a 24-well plate and maintained in a complete culture medium. After HIV-1 infection, the coverslips were rinsed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde (Cat#BM-1155, Boston BioProducts) in PBS for 15-30 min at room temperature. Subsequently, cells were incubated in 0.1 M glycine-PBS for 10 min at room temperature. Cells were permeabilized using 0.5% Triton X-100 (Cat#T-9284, Sigma-Aldrich) for 5 min at room temperature. Non-specific binding was prevented using a blocking solution (3% bovine serum albumin in PBS) for 60 min at room temperature. Samples were incubated with primary antibodies in a blocking solution for 60 min in a dark room at room temperature. Subsequently, the coverslips were rinsed with PBS and incubated with the appropriate secondary antibodies against mouse or rabbit IgG in a blocking solution for 30 min at room temperature. Finally, the coverslips were washed with PBS and mounted using FluorSave reagent (Cat#345789, Sigma-Aldrich). Fluorescence microscopy images were acquired with an AxioObserver.Z1 microscope equipped with a PlanApo 63× oil immersion objective (NA 1.4) and an AxioCam MRm digital camera (Carl Zeiss). Image acquisition was performed using a Zeiss Z1 Observer inverted microscope and ZEN 3.3 (blue edition) software. Image deconvolution was performed with the ZEN 3.3 software using an acquired point spread function. Images for figures were processed with Adobe Photoshop CS5 software (Adobe Systems, Mountain View, CA).
## Protein extracts and immunoblotting
Preparation of protein extracts from cultured cells was performed using a method that we have described elsewhere (2). Briefly, cells were washed twice with ice-cold PBS and incubated at 4°C for 1 h in whole-cell extract buffer (50 mM Tris-HCl, 280 mM NaCl, 0.5% IGEPAL, 0.2 nM EDTA, 2 mM EGTA, 10% glycerol, and 1 mM DTT, pH 8) supplemented with a cocktail of protease inhibitors (Cat#P8340, Sigma-Aldrich) benzonase (Cat#sc-391121B, ChemCruz), and 50 µg/mL ethidium bromide (Cat#E3050, Teknova). Lysates were cleared by centrifugation at 14,000×g for 60 min at 4°C. Protein concentration was determined with a protein assay dye reagent (Bio-Rad Laboratories, Hercules, CA, USA). Depend ing on the protein analyzed, samples of protein extracts were denatured at 37°C or 65°C for 10 or 60 min, respectively, in 4× Nu PAGE LDS sample buffer (Cat#NP0007, Invitrogen) and separated by Nu-PAGE. After electrophoresis, proteins were transferred to a nitrocellulose membrane and incubated sequentially with corresponding primary and secondary antibodies overnight at 4°C and 1 h at room temperature, respec tively. Detection of proteins was performed using the LI-COR Odyssey Imaging System (LI-COR, Biotech). As an internal gel loading control, we analyzed the levels of glyceralde hyde 3-phosphate dehydrogenase (GAPDH). The immunoblot signal from images with non-saturated pixels was estimated using ImageJ software (version 1.47h, Bethesda, MD). For each condition, protein bands were quantified from three independent experiments.
## RNA extraction
Total RNA from human primary CD4 + T and A549 mock-infected and HIV-1-infected cells (in triplicate) was extracted with TRIzol reagent (Cat#15596026, Invitrogen) according to the manufacturer's protocol. Briefly, TRIzol Reagent was directly added to the culture dish to lyse the cells (~5 × 10 6 cells). The homogenized content was transferred to a 1.5-mL microtube and incubated for 5 min at room temperature. Chloroform (Cat#C298-1, Sigma-Aldrich) was added to the sample, and the sample was shaken for 30 s and incubated for 3 min at room temperature, followed by centrifugation at 12,000×g for 15 min at 4°C. The aqueous phase containing the RNA was transferred to a new 1.5-mL microtube. For RNA precipitation, an equal volume of isopropanol (Cat#383920025, Thermo Scientific) was added to the sample, which was incubated at 4°C for 10 min. Next, the sample was centrifuged at 12,000×g for 10 min at 4°C, and the supernatant was discarded. The RNA pellet was washed using 75% (v/v) ethanol (Sigma-Aldrich), followed by centrifugation at 10,000×g for 5 min at 4°C. The supernatant was discarded, and the pellet was dried at room temperature. RNA was solubilized in RNase-free ultrapure water. The A230/260 and A260/280 ratios were measured by a NanoDrop spectrophotometer to determine the purity of the isolated total RNA. Isolated RNA was stored at -80°C until analysis.
## RNA-sequencing data processing
The RNA-sequencing data processing pipeline was adapted from published protocols. Briefly, we first trimmed the paired-end reads in FASTQ files using Trimmomatic (version 0.39), targeting known adapters (TruSeq3-SE.fa:2:30:10). We also trimmed the first three and last three base pairs if their quality fell below set thresholds. Additionally, we employed a 4-base sliding window technique, trimming reads when the average window quality dropped below 15. Reads shorter than 36 base pairs were excluded. Next, the trimmed reads were aligned to the human reference genome GRCh38 (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_000001405.26/) using HISAT2 (PMID: 31375807, v 2.1.0) and converted into BAM format using sam tools (version 1.6) (63). The alignment data were then processed by StringTie (ver sion 2.1.3) to estimate gene expression levels. The StringTie merge function merged samples' gene structures to create a unified set of transcripts across all samples. Finally, read counts were obtained using prepDE.py (https://ccb.jhu.edu/software/string tie/index.shtml?t=manual) to prepare the data for DESeq2.
## Differential gene expression analysis
The raw read counts were processed in R (version 4.2.3) using DESeq2 (version 1.38.3) to investigate differences across conditions, accounting for donor variability and batch effects. Genes with a minimum count of 10 in at least three samples were retained for analysis. To visualize treatment effects, counts were transformed using variance stabilizing transformation function from DESeq2 and subjected to principal components analysis, visualizing the first two principal components. Differential gene expression between infected and mock-infected cells was assessed using the DESeq function from DESeq2, which estimates size factors and dispersions, fitting a negative binomial generalized linear model. P-values were adjusted using the Benjamini-Hochberg method (https://www.jstor.org/stable/2346101). Adjusted P-values below 0.05 were considered significant. Gene enrichment analysis was subsequently performed using gene set enrichment analysis (GSEA) to identify specific Gene Ontology, Kyoto Encyclopedia of Genes and Genomes, and Reactome pathways associated with HIV infection. We performed GSEA using clusterProfiler v.4.10.0 in R, with all gene lists ranked by the corresponding log2 fold change. For these analyses, all genes whose gene symbols could be mapped to ENTREZ IDs using the org.Hs.eg.db v.3.18.0 Bioconductor annotation package were included.
## APA analysis using REPAC
APA analysis was conducted by analyzing RNA-sequencing data using the REPAC pipeline to determine changes in 3' UTR length in treated (HIV-1-infected) versus control (mock-infected) conditions9. 3' UTRs were designated as shortened if they had compositional fold-change (cFC)<0.25 and -log10 (adjusted P-value) <0.05 and as lengthened if they had cFC>0.25 and -log10 (adjusted P-value) <0.05.
## PAC-Seq library construction and sequencing
First, 2 µg total RNA was mixed with 1 µL of 100 µM 3' Illumina_4N_21T_VN primer (5′-GT GACTGGAGTTCAGACGTGTGCTCTTCCGATCTNNNNTTTTTTTTTTTTTTTTTTTTTVN-3′), 2 µL of 5 mM AzNTP (Baseclick, #BCT-25 to -28)/dNTP mixture, at a ratio of 1 to 5, and the total volume was raised to 13 µL by the addition of H 2 O. The mixture was heated at 65°C/5 min and snap-cooled on ice for 3 min. Reverse transcription was performed using SuperScript III (Cat#18080093, ThermoFisher). In brief, 7 µL master mix containing 4 µL 5× Superscript First Strand buffer, 1 µL of 0.1 M DTT, 1 µL RNaseOUT (Cat#10777019, ThermoFisher), and 1 µL SuperScript III were mixed with the RNA sample. Heating was performed in a thermocycler sequentially as follows: 25°C/10 min, 50°C/40 min, and 75°C/15 min, followed by RNase H (Cat#AM2293, ThermoFisher) treatment using 1U per reaction for 37°C/30 min and then 80°C/10 min. The resulting cDNA was purified using Sera-Mag Speedbeads (Cat#65152105050250, Cytiva). Speedbeads working solution was made by washing 1 mL bead slurry twice in 1×TE buffer and then resuspending beads in 50 mL of 1×TE buffer containing 9 g PEG-8000, 1 M NaCl, and 0.05% Tween-20. Speedbeads, 1.8× reaction vol, were mixed with cDNA followed by a 5 min incubation at room temperature. The beads were then pelleted using a magnetic bead collector, and the supernatant was discarded. Two washes with 80% ethanol were performed while the beads were pelleted on the magnet. The beads were then dried, and the cDNA was eluted by resuspending the beads in 22 µL of 50 mM HEPES pH 7.2 for 2 min at room temperature. Then, 20 µL cDNA was mixed with 11 µL Click Mix (2.5 M NaCl, 30% EtOH in H 2 O) and 4 µL of 5 µM UMI-click-adapter (5′-Hexynyl-NNNNNNNNNNNNA GATCGGAAGAGCGTCGTGTAGGGAAAGAGTGT-3′, HPLC purified). To activate the reaction, 4 µL of 50 mM vitamin C and 1 µL of Click Catalyst (4 mM CuSO 4 in H 2 O and 20 mM THPTA [Baseclick, #BCMI-006] in H 2 O) were pre-mixed before being added into the cDNA solution for 60 min incubation at room temperature in the dark. Clicked cDNA was then purified in the same manner as post-RT purification but eluted in 22 µL of 10 mM Tris pH 7.4. Half of the eluted cDNA was subjected to PCR amplification with indexing primers i5 (5′-AATGATACGGCGACCACCGAGATCTACAC[Index]ACACTCTTTCCCTACACGAC GCTCTTCCGATC*T-3′) and i7 (5′-CAAGCAGAAGACGGCATACGAGAT[Index]GTGACTGGAGT TCAGACGTGTGCTCTTCCGAT*C-3′). Both primers have a phosphorothioate bond added at the 3' end (denoted as *) to increase stability. The PCR reaction was conducted in a 50 µL volume, including 25 µL 2× OneTaq Master Mix (Cat#M0482, NEB), 2 µL each of 5 µM i5 and i7, and cDNA in H 2 O. A total of 13 amplification cycles were conducted on 2 µg input RNA. The program was run at 94°C/4 min; 53°C/30 sec; 68°C/10 min; 15× [94°C/30 sec, 53°C/30 sec; 68°C/2 min]; 68°C/5 min; hold at 4°C. The library was then purified and size-selected with Speedbeads. In brief, 45 µL beads (0.9× volume of PCR reaction) were mixed with the reaction to remove larger fragments. Supernatant was collected and mixed with 10 µL beads (0.2× volume of PCR reaction) to remove smaller fragments and primer dimers. Two washes with 100% ethanol were then performed while the beads were on the magnet. Beads were dried before being resuspended in 12 µL of 10 mM Tris pH 7.4. The library was then collected and subjected to quality control on Agilent 2200 TapeStation and analysis by Illumina NovaSeq 6000 at the Genomics Research Center at the University of Rochester.
## Pre-processing the raw data derived from PAC-Seq
Samples were sequenced on the 10B-300 NovaSeq X-Plus machine with a paired-end 150 bp configuration. For alternative polyadenylation (APA) analysis by PolyA-Miner, only forward reads were considered (51). Raw fastq reads were subjected to pre-processing steps as described here: Fastp was used to remove the Illumina adapters (using -a option), trim low-quality bases in the start of the read (using -f 4), and to discard reads that are shorter than forty bases (using -l 40) (64). The quality-trimmed reads were then aligned to the reference genome (GRCh38, version 33 downloaded from GENCODE portal), using bowtie2 version 2.3.5.1 in single-end mode. Bowtie2 was run in very sensitive local mode to account for soft clipped regions towards the end of the read. Bowtie2 parameters (65): (-p 16 --very-sensitive-local). The alignment files were sorted by genomic coordinates and indexed with samtools version 1.9 (63). To address PCR duplicates, Unique Molecular Identifiers (UMI) bar codes were used in the library construction to identify unique transcripts. UMI tools extract version 1.1.5 was used to extract the UMI bar codes from the reads and append to the read names (66). Umitools dedup function was used on mapped reads to discard the duplicated reads that share the same UMI (66). This will ensure that each unique read is represented once. We then generated the genome-wide coverage tracks for these de-duplicated bams in bigWigs format using bamCoverage version 3.5.5 (67). Replicates were combined to a single bigWig to be hosted in UCSC Browser to make the gene tracks (68). Raw counts were generated from sorted BAMs using featureCounts version v2.0.1 (69).
## PolyA-miner analysis
PolyA-miner 2022 version was used to quantify alternative polyadenylation (51). Parameters used are: -mode bam -s 0 -expNovel 0 -p 20 -ip 50 -a 0.65 -pa_p 0.6 -pa_a 8 -pa_m 4 -gene_min 10 -apa_min 0.05 -t BB. We used bam mode in polyA-miner to process the pre-processed and aligned BAM files (-mode bam). PolyA-miner was run on the following comparisons: HIV-WT infected cells vs Mock cells, HIV-N74D infected cells vs Mock cells, HIV-A77V infected cells vs Mock cells, in A549 cells and T cells and three CPSF6 KO lines vs control cells. Only known polyadenylation sites identified in the polyA database were considered for the analysis (-expNovel 0). A polyA site was only considered in the analysis if 60% (-pa_p 0.6) replicates have eight reads (-pa_a 8) aligned to the site with the third replicate having a minimum of 4 reads (-pa_m 4). A minimum read count of 10 reads per gene is used (-gene_min 10). To avoid sites which were picked due to internal priming, a 20-nucleotide window (-p 20) was set to slide 50 nucleotides (-ip 50) to the downstream of the site. And if in any of the sliding windows, there is more than 65% of adenine (-a 0.65), then the site was discarded. All polyA sites retained after mis-priming and other quality filters were included in the downstream APA analysis, regardless of their genomic location (e.g., gene body, UTR exon, or intron). To infer the differential polyA usage variations, a beta binomial test was performed using proportions of individual features (polyA signals) to overall gene counts (69). For each of the comparisons, polyA-miner generates an output file which includes genes that undergo APA changes. Volcano plots were generated in GraphPad Prism to visualize the changes.
Differential Gene Expression and Gene Ontology Analysis. Gene level counts were calculated as the sum of all the reads mapped to individual polyA signals in each gene. These read counts were normalized and subjected to differential gene expression analysis using DESeq2 version 1.44.0 in R version 4.4.1 (https://www.R-project.org/)
Gene ontology analysis was performed using enrichGO function from clusterProfiler 4.12.6 (70,71).
## Statistical analysis
The mean and standard deviation values were calculated using GraphPad Prism 8. Statistical analysis was performed using unpaired t-tests or ANOVA multiple comparisons tests as indicated. *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant.
## References
1. Dharan, Bachmann, Talley et al. (2020) "Nuclear pore blockade reveals that HIV-1 completes reverse transcrip tion and uncoating in the nucleus" *Nat Microbiol*
2. Selyutina, Persaud, Lee et al. (2020) "Nuclear import of the HIV-1 core precedes reverse transcription and uncoating" *Cell Rep*
3. Burdick, Li, Munshi et al. (2020) "HIV-1 uncoats in the nucleus near sites of integration" *Proc Natl Acad Sci*
4. Francis, Marin, Singh et al. (2020) "Publisher correction: HIV-1 replication complexes accumulate in nuclear speckles and integrate into speckle-associated genomic domains" *Nat Commun*
5. Luchsinger, Lee, Mardones et al. (2023) "Formation of nuclear CPSF6/CPSF5 biomolecular condensates upon HIV-1 entry into the nucleus is important for productive infection" *Sci Rep*
6. Bialas, Diaz-Griffero (2024) "HIV-1-induced translocation of CPSF6 to biomolecular condensates" *Trends Microbiol*
7. Cardinale, Cisterna, Bonetti et al. (2007) "Subnuclear localization and dynamics of the Pre-mRNA 3' end processing factor mammalian cleavage factor I 68-kDa subunit" *Mol Biol Cell*
8. Bejarano, Peng, Laketa et al. (2019) "HIV-1 nuclear import in macrophages is regulated by CPSF6-capsid interactions at the nuclear pore complex"
9. Ilık, Aktaş (2022) "Nuclear speckles: dynamic hubs of gene expression regulation" *FEBS J*
10. Lamond, Spector (2003) "Nuclear speckles: a model for nuclear organelles" *Nat Rev Mol Cell Biol*
11. Selyutina, Simons, Kirby et al. (2022) "TRIM5α restriction of HIV-1-N74D viruses in lymphocytes is caused by a loss of cyclophilin A protection" *Viruses*
12. Sowd, Serrao, Wang et al. (2016) "A critical role for alternative polyadenylation factor CPSF6 in targeting HIV-1 integration to transcriptionally active chromatin" *Proc Natl Acad Sci*
13. Saito, Henning, Serrao et al. (2016) "Capsid-CPSF6 interaction is dispensable for HIV-1 replication in primary cells but is selected during virus passage in vivo" *J Virol*
14. Rasaiyaah, Tan, Fletcher et al. (2013) "HIV-1 evades innate immune recognition through specific cofactor recruit ment" *Nature*
15. Albanese, Ruhle, Mittermaier et al. (2022) "Rapid, efficient and activation-neutral gene editing of polyclonal primary human resting CD4 + T cells allows complex functional analyses" *Nat Methods*
16. Zhong, Ning, Boggs et al. (2021) "Cytoplasmic CPSF6 regulates HIV-1 capsid trafficking and infection in a cyclophilin Adependent manner" *mBio*
17. Zila, Müller, Laketa et al. (2019) "Analysis of CA content and CPSF6 dependence of early HIV-1 replication complexes in SupT1-R5 cells" *mBio*
18. Achuthan, Perreira, Sowd et al. (2018) "Capsid-CPSF6 interaction licenses nuclear HIV-1 trafficking to sites of viral DNA integration" *Cell Host Microbe*
19. Chin, Perreira, Savidis et al. (2015) "Direct visualization of HIV-1 replication intermedi ates shows that capsid and CPSF6 modulate HIV-1 intra-nuclear invasion and integration" *Cell Rep*
20. Rasheedi, Shun, Serrao et al. (2016) "The cleavage and polyadenylation specificity factor 6 (CPSF6) subunit of the capsid-recruited premessenger RNA cleavage factor I (CFIm) complex mediates HIV-1 integration into genes" *J Biol Chem*
21. Ingram, Kline, Hughson et al. (2024) "Spatiotemporal binding of cyclophilin A and CPSF6 to capsid regulates HIV-1 nuclear entry and integration"
22. Li, Singh, Sowd et al. (2020) "CPSF6-dependent targeting of speckle-associated domains distinguishes primate from nonprimate lentiviral integration" *mBio*
23. Ning, Zhong, Fischer et al. (2018) "Truncated CPSF6 forms higher-order complexes that bind and disrupt HIV-1 capsid" *J Virol*
24. Spector, Landon, Keefe (1993) "Organization of RNA polymer ase II transcription and pre-mRNA splicing within the mammalian cell nucleus" *Biochem Soc Trans*
25. Gruber, Martin, Keller et al. (2012) "Cleavage factor Im is a key regulator of 3' UTR length" *RNA Biol*
26. Rüegsegger, Beyer, Keller (1996) "Purification and characterization of human cleavage factor Im involved in the 3' end processing of messenger RNA precursors" *J Biol Chem*
27. Yang, Gilmartin, Doublié (2010) "Structural basis of UGUA recognition by the Nudix protein CFI m 25 and implications for a regulatory role in mRNA 3' processing" *Proc Natl Acad Sci*
28. Tan, Li, Zhang et al. (2018) "NUDT21 negatively regulates PSMB2 and CXXC5 by alternative polyadenylation and contributes to hepatocellular carcinoma suppression" *Oncogene*
29. Hardy, Norbury (2016) "Cleavage factor Im (CFIm) as a regulator of alternative polyadenylation" *Biochem Soc Trans*
30. Mitschka, Mayr (2022) "Context-specific regulation and function of mRNA alternative polyadenylation" *Nat Rev Mol Cell Biol*
31. Tian, Hu, Zhang et al. (2005) "A large-scale analysis of mRNA polyadenylation of human and mouse genes" *Nucleic Acids Res*
32. Shepard, Choi, Lu et al. (2011) "Complex and dynamic landscape of RNA polyadenylation revealed by PAS-Seq" *RNA*
33. Gallicchio, Olivares, Berry et al. (2023) "Regulation and function of alternative polyadenylation in development and differentiation" *RNA Biol*
34. Tian, Manley (2013) "Alternative cleavage and polyadenylation: the long and short of it" *Trends Biochem Sci*
35. Elkon, Ugalde, Agami (2013) "Alternative cleavage and polyadeny lation: extent, regulation and function" *Nat Rev Genet*
36. Martin, Gruber, Keller et al. (2012) "Genome-wide analysis of pre-mRNA 3' end processing reveals a decisive role of human cleavage factor I in the regulation of 3' UTR length" *Cell Rep*
37. Kim, Yamamoto, Chen et al. (2010) "Evidence that cleavage factor Im is a heterotetrameric protein complex controlling alternative polyadenylation" *Genes Cells*
38. Ge, Huang, Chen et al. (2024) "Downregulation of CPSF6 leads to global mRNA 3' UTR shortening and enhanced antiviral immune responses" *PLoS Pathog*
39. Liu, Wu, Chen et al. (2023) "CPSF6 regulates alternative polyadenylation and proliferation of cancer cells through phase separation" *Cell Rep*
40. Kubo, Wada, Yamaguchi et al. (2006) "Knock-down of 25 kDa subunit of cleavage factor Im in Hela cells alters alternative polyadenylation within 3'-UTRs" *Nucleic Acids Res*
41. Wuttinontananchai, Yamamoto, Sakamoto et al. (2025) "Genome-wide CRISPR screen for human factors involved in alternative polyadenylation based on differential localization of CD47" *Sci Rep*
42. Luchsinger, Diaz-Griffero (2024) "Detection of CPSF6 in biomolecular condensates as a reporter of HIV-1 nuclear import" *Methods Mol Biol*
43. Witkowski, Lee, Lee et al. (2022) "NUDT21 limits CD19 levels through alternative mRNA polyadenylation in B cell acute lymphoblastic leukemia" *Nat Immunol*
44. Masamha, Xia, Yang et al. (2014) "CFIm25 links alternative polyadenylation to glioblastoma tumour suppression" *Nature*
45. Shi, Ding, Zhao et al. (2021) "Suppression of CPSF6 enhances apoptosis through alternative polyadenylationmediated shortening of the VHL 3′UTR in gastric cancer cells" *Front Genet*
46. Imada, Wilks, Langmead et al. (2023) "REPAC: analysis of alternative polyadenylation from RNA-sequencing data" *Genome Biol*
47. Lee, Ambrose, Martin et al. (2010) "Flexible use of nuclear import pathways by HIV-1" *Cell Host Microbe*
48. Jensen, Elrod, Yalamanchili et al. (2021) "Application and design considerations for 3'-end sequencing using click-chemistry" *Methods Enzymol*
49. Routh, Ji, Jaworski et al. (2017) "Poly(A)-ClickSeq: click-chemistry for next-generation 3΄-end sequencing without RNA enrichment or fragmentation" *Nucleic Acids Res*
50. Jonnakuti, Wagner, Maletić-Savatić et al. (2024) "PolyAMiner-Bulk is a deep learning-based algorithm that decodes alternative polyadenylation dynamics from bulk RNA-seq data" *Cell Rep Methods*
51. Yalamanchili, Alcott, Ji et al. (2020) "PolyA-miner: accurate assessment of differential alternative polyadenylation from 3'Seq data using vector projections and non-negative matrix factorization" *Nucleic Acids Res*
52. Selyutina, Hu, Miller et al. (2022) "GS-CA1 and lenacapavir stabilize the HIV-1 core and modulate the core interaction with cellular factors"
53. Chu, Elrod, Wang et al. (2019) "Nudt21 regulates the alternative polyadenylation of Pak1 and is predictive in the prognosis of glioblastoma patients" *Oncogene*
54. Zhu, Wang, Forouzmand et al. (2018) "Molecular mechanisms for CFIm-mediated regulation of mRNA alternative polyadenylation" *Mol Cell*
55. Guedán, Burley, Caroe et al. (2024) "HIV-1 capsid rapidly induces long-lived CPSF6 puncta in non-dividing cells, but similar puncta already exist in uninfected T-cells" *Viruses*
56. Coelho, Gratton, Melo et al. (2021) "HIV-1 infection transcrip tomics: meta-analysis of CD4+ T cells gene expression profiles" *Viruses*
57. Nacken, Mayr, Schreiber et al. (2025) "Influenza A virus NS1 suppresses nuclear speckles promoted gene expression by inhibition of transcription" *NPJ Viruses*
58. Steegenga, Van Laar, Riteco et al. (1996) "Adenovirus E1A proteins inhibit activation of transcription by p53" *Mol Cell Biol*
59. Duboeuf, Legrand, Lozach et al. (2025) "NSs: the multiface ted bunyavirus virulence factor" *NPJ Viruses*
60. Ding, Wang, Wang et al. (2022) "Schlafen 5 suppresses human immunodeficiency virus type 1 transcription by commandeering cellular epigenetic machinery" *Nucleic Acids Res*
61. Diaz-Griffero, Perron, Mcgee-Estrada et al. (2008) "A human TRIM5α B30.2/SPRY domain mutant gains the ability to restrict and prematurely uncoat B-tropic murine leukemia virus" *Virology (Auckl)*
62. 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)*
63. Li, Handsaker, Wysoker et al. "Genome Project Data Processing S. 2009. The sequence alignment/map format and SAMtools"
64. Chen, Zhou, Chen et al. (2018) "fastp: an ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
65. Langmead, Salzberg (2012) "Fast gapped-read alignment with Bowtie 2" *Nat Methods*
66. Smith, Heger, Sudbery (2017) "UMI-tools: modeling sequencing errors in Unique Molecular Identifiers to improve quantification accuracy" *Genome Res*
67. Ramírez, Ryan, Grüning et al. (2016) "deepTools2: a next generation web server for deep-sequencing data analysis" *Nucleic Acids Res*
68. Perez, Barber, Benet-Pages et al. (2025) "The UCSC Genome Browser database: 2025 update" *Nucleic Acids Res*
69. Tan, Yalamanchili, Park et al. (2016) "Extensive cryptic splicing upon loss of RBM17 and TDP43 in neurodegeneration models" *Hum Mol Genet*
70. Love, Huber, Anders (2014) "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2" *Genome Biol*
71. Yu, Wang, Han et al. (2012) "clusterProfiler: an R package for comparing biological themes among gene clusters" *OMICS* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548418&blobtype=pdf | # Pseudorabies virus DNA polymerase processivity factor pUL42 inhibits type I IFN production by negatively regulating cGAS-STING signaling pathway
Guangqiang Ye, Jiaxiu Gao, Haoxuan Cao, Xiaohong Liu, Hongyang Liu, Shanghui Wang, Yunfei Liu, Longfei Han, Qiongqiong Zhou, Yandong Tang, Jin Tian, Liping Huang, Zhaoxia Zhang, Changjiang Weng
## Abstract
Pseudorabies (PR), caused by the PR virus (PRV), is an acute infectious disease in livestock and various wild animals. PRV has developed several immune evasion mechanisms to antagonize the host's antiviral immune response. However, the precise role of PRV-encoded proteins in regulating the cyclic GMP-AMP synthase (cGAS)-stimula tor of interferon (IFN) gene (STING) signal for immune evasion remains unclear. In this study, we demonstrate that the PRV UL42 protein (pUL42) inhibited cGAS-mediated antiviral signaling by modulating cGAS recognition of dsDNA. Mechanistically, pUL42 interacts with the DNA-binding domain of cGAS, thereby inhibiting its recognition of double-stranded DNA, leading to the inhibition of its dimerization and oligomerization activation. Furthermore, knocking down the expression of the Ul42 gene in the PRV genome diminishes the antagonistic effect on type I IFN production and inhibits PRV replication. Ultimately, we have established that pUL42 targets cGAS-mediated signaling, thereby inhibiting the production of type I IFN and facilitating viral replication. Overall, our findings suggest that PRV pUL42 functions as an antagonist to evade the host's antiviral response by targeting the cGAS-STING axis.IMPORTANCE Cyclic GMP-AMP synthase (cGAS)-stimulator of interferon gene (STING) axis is essential for host resistance to DNA virus infections by regulating type I interferon production. However, whether pseudorabies virus (PRV) antagonizes the cGAS-STING signaling pathway to immune evasion is not fully investigated. In this study, we clearly demonstrated that the PRV pUL42 protein inhibits the recognition of double-stranded DNA of cGAS, leading to inhibiting the oligomerization and activation of cGAS, thereby suppressing the cGAS-mediated host antiviral immune responses. Taken together, our results reveal a novel strategy employed by PRV to evade host defenses, which will provide theoretical support for the development of anti-PRV drugs for the prevention and control of PRV. KEYWORDS pseudorabies virus, DNA polymerase processivity factor, pUL42, cGAS-STING, interferon T he innate immune system is the host's first line of defense against invading pathogens. Upon the invasion of pathogenic microorganisms, pattern recognition receptors (PRRs) in host cells detect pathogen-associated molecular patterns (PAMPs) and rapidly initiate a series of signaling events. This cascade ultimately leads to the production of type I interferon (IFN) and the expression of hundreds of other IFNinduced genes (ISGs) (1). The cyclic GMP-AMP synthase (cGAS) has been identified as a universal sensor for recognizing double-stranded DNA (dsDNA) in intracellular environ ments during DNA viral infection (2). cGAS detects dsDNA to induce the synthesis of
the second messenger cyclic GMP-AMP (cGAMP), which then binds to and activates the stimulator of IFN gene (STING) factor, resulting in facilitating its translocation from the endoplasmic reticulum to the Golgi apparatus (3). Subsequently, STING recruits TANK-binding kinase (TBK1) and activates its phosphorylation (4). The active TBK1 phosphorylates the transcription factor IFN regulatory factor (IRF3), leading to its entering the nucleus and initiating type I IFN production (5).
Pseudorabies virus (PRV), one of the members of the alpha herpesvirus subfamily, is the pathogen of Aujeszky's disease, which also threatens most mammals, including domestic pigs and wild boars (6,7). PRV primarily invades the peripheral nervous system first and causes severe clinical and neurological symptoms, which can lead to the acute death of piglets (8). Although PRVs have not been found to spread from person to person, under certain conditions, PRVs may infect humans and cause intense neurolog ical symptoms (9). Recent studies have shown that PRVs have evolved various mecha nisms to counteract host immune responses and achieve effective infection (10)(11)(12)(13)(14)(15)(16)(17). For example, PRV pUS2 interacts with the ligand-binding domain of STING and recruits TRIM21 to degrade STING (10). PRV pUL13 suppresses the expression of RIG-I and MDA5 by inhibiting the activation of the transcription factor NF-κB, thereby suppressing the host's antiviral immunity (13). PRV EP0 inhibits the IFN-induced antiviral innate immune response by downregulating the basal expression of IRF9 via transcriptional repression (16). However, it is still unclear whether the PRV proteins target the cGAS-STING axis for immune evasion.
The DNA polymerase processing factor encoded by the Ul42 gene is one of the components of DNA polymerase in PRV and HSV-1, which is an essential protein for virus replication (18,19). PRV pUL42 is a continuous synthesis factor of DNA polymerase, which can activate its ability to continuously synthesize viral genomes by binding to the catalytic subunit pUL30 of DNA polymerase (19). Recently, PRV pUL42 was found to induce ubiquitination degradation of p65 by upregulating cytokine signaling inhibitor 1 (SOCS1) to prevent excessive inflammatory response during PRV infection (20). In addition, PRV pUL42 binds to IFN-stimulated response element (ISRE) through four conserved DNA-binding sites, which prevents the binding of IFN-stimulated gene factor 3 (ISGF3) to ISRE and leads to IFN-induced transcriptional disruption (21). However, it is still unknown whether pUL42 regulates the cGAS-STING axis. In this study, we found that PRV pUL42 interacts with the DNA-binding domain of cGAS to inhibit its binding to dsDNA, which subsequently inhibits the dimerization and oligomerization activation of cGAS. We also found that pUL42 interacts with cGAS, which subsequently leads to the suppression of type I IFN production and promotes viral replication. These findings support the notion that pUL42 is a pivotal inhibitor of the cGAS-STING axis, revealing a new mechanism by which PRV-encoded protein pUL42 plays a role in antagonizing host antiviral responses.
## RESULTS
## PRV pUL42 inhibits the cGAS-STING signaling pathway
The cGAS-STING signaling pathway plays an important role in sensing cytoplasmic DNA to induce type I IFN production to trigger host antiviral immune responses (1). Previ ous research results have shown that PRV infection inhibits cGAS-STING-mediated IFN production (22). To identify which PRV-encoded protein can inhibit the promoter activity of IFN-β induced by cGAS-STING, we conducted an unbiased screening to evaluate the effects of 44 PRV-encoded proteins on the regulation of type I IFN production. We found that pUL24, pUL42, and pUL32 exhibited the most significant inhibitory effects (Fig. 1A). To further validate the effectiveness of the screening system, three PRV proteins (pUL24, pUL48, and pUL49.5) were selected for further evaluation of their inhibitory effects. HEK293T cells were transfected with an IFN-β-luciferase reporter, an internal control Renilla-TK Luc, two plasmids expressing cGAS and STING, together with different doses of a plasmid expressing pUL24, pUL48, or pUL49.5, respectively. As shown in Fig. 1B through D, the ectopic expression of pUL24 and pUL48 inhibited cGAS-STING-induced IFN-β promoter activity in a dose-dependent manner, while pUL49.5 had no effect (Fig. 1B through D). These results demonstrated that our unbiased screening assay is effective.
In our impartial screening system, PRV pUL24 exhibits the most pronounced influence on the activity of the IFN-β promoter induced by cGAS-STING, followed by pUL42. Furthermore, numerous studies have documented the inhibitory role of pUL24 on the cGAS-STING signaling pathway (23)(24)(25). A previous study showed that PRV pUL42 can inhibit virus infection-induced type I IFN signaling pathway (21), and HSV-1 pUL42 inhibits the production of type I IFN by inhibiting phosphorylation of IRF3 (26). However, the inhibition effect of PRV pUL42 on the production of type I IFN has not been studied. To test the effect of PRV pUL42 on the production of type I IFN, HEK293T cells were transfected with a plasmid encoding pUL42 or pUL49.5 as a negative control, together with an IFN-β-luciferase reporter, a pRL-TK, along with plasmids expressing cGAS and STING for 24 h. As shown in Fig. 2A, ectopic expression of pUL42 inhibited cGAS-STINGinduced IFN-β promoter activity, while pUL49.5 had no effect (Fig. 2A). Additionally, we noticed that pUL42 inhibited the activities of these promoters of IFN-β (Fig. 3A), ISG56 (Fig. 3B), ISG54 (Fig. 3C), ISRE (Fig. 3D), and nuclear factor kappa B (NF-κB; Fig. 3E). To further analyze the inhibitory effect of pUL42 on type I IFN production, HeLa cells were transfected with a plasmid-encoding pUL42 or pUL49.5. At 24 h post-transfection (hpt), the cells were stimulated with poly(dA:dT), a simulation dsDNA sequence to activate the cGAS-STING signaling pathway, for another 12 h, and then the mRNA level of Ifnb1 was detected by qPCR. As shown in Fig. 2B, ectopically expressed pUL42 significantly decreased the mRNA levels of Ifnb1 induced by poly(dA:dT), but not pUL49.5. Consis tently, we also found that pUL42 inhibited the mRNA levels of Ifnb1 (Fig. 3F), Isg56 (Fig. 3G), and Isg54 (Fig. 3H) induced by cGAS-STING. These data suggest PRV pUL42 is a potent inhibitor of type I IFN production. Phosphorylation of TBK1 and IRF3 is an important marker of cGAS-STING signaling pathway activation. To test whether pUL42 affects the phosphorylation of TBK1 and IRF3, HeLa cells were transfected with a plasmid encoding pUL42 or pUL49.5 for 24 h, and then transfected with poly(dA:dT) for another 12 h. The phosphorylation levels of IRF3 (Ser396) and TBK1 (Ser172) were detected. As shown in Fig. 2C andD, pUL42 significantly inhibited the phosphorylation of IRF3 and TBK1 induced by poly(dA:dT), while pUL49.5 had no effect.
## Knockdown of the expression of Ul42 promotes type I IFN production induced by PRV
To investigate whether Ul42 affects the production of type I IFN during PRV infection, we used CRISPR-Cas9 and homologous recombination technology to construct an Ul42-deficient PRV, but due to the important role of Ul42 in viral replication, we were unable to obtain Ul42-deficient PRV. Then, porcine alveolar macrophages (PAMs) were transfected with small interfering RNA (siRNA) control (sicon) and three siRNAs targeting Ul42 (si385, si520, and si856) for 48 h and then infected with PRV-JM (multiplicity of infection [MOI] of 1) for 24 h (Fig. 4A). The results showed that the knockdown efficiency of si520 is the highest for pUL42. PAMs were transfected with sicon and si520 for 48 h and PAMs (M) and HeLa cells (N) were transfected with si520 or sicon (100 nM/each) for 24 h, and then PAMs (M) or HeLa cells (N) were infected with PRV ( then infected with PRV-JM (1 MOI) for 12 h or 24 h. As shown in Fig. 4B and C, we found that knockdown of the expression of Ul42 resulted in increased mRNA levels of Ifnb1 and Isg56 during PRV infection. Consistent with these results, si520 markedly inhibited the mRNA level of Ul42 during PRV infection (Fig. 4D). The si520 resulted in a decrease of approximately two logs in the viral titers of PRV in PAMs (Fig. 4E).
Previous studies have shown that PRV spreads among populations of different species (27). We further investigated the function of pUL42 in HeLa cells. HeLa cells were transfected with sicon and si520 for 24 h and then infected with PRV-JM (1 MOI) for 0, 4, 8, 12, or 24 h. As shown in Fig. 4F through H, we found that knockdown of Ul42 resulted in increased mRNA levels of Ifnb1, Ifnα4, and Isg56 during PRV infection. Consistent with the results on PAMs, knocking down Ul42 resulted in a significant decrease of virus titers (Fig. 4I andJ). Meanwhile, we found that knockdown of Ul42 expression significantly facilitated PRV-induced phosphorylation of TBK1 and IRF3 in HeLa cells (Fig. 4K andL). Further research found that knockdown Ul42 inhibited the protein expression of gB, gE, and EP0 in the process of PRV infection in PAMs (Fig. 4M) and HeLa cells (Fig. 4N). These findings are consistent with previous results that highlight the critical role of Ul42 in viral replication (19,28). In summary, we have demonstrated that pUL42 is necessary for virus replication, which also inhibits the production of type I IFN by targeting the cGAS-STING signaling pathway.
## PRV pUL42 interacts with cGAS
To elucidate the potential molecular mechanism of pUL42 negatively regulating type I IFN production, immunoprecipitation assays were performed. As shown in Fig. 5A and B, we found that pUL42 interacted with cGAS in HEK293T cells, but not STING, TBK1, or IRF3. In addition, we found that endogenous cGAS interacted with pUL42 in PRV-infected PAMs at 12 h and 24 h, but not gB (Fig. 5C). Previous studies have shown that pUL42 relies on nuclear location signals (NLS) to transport to the nucleus and participates in DNA replication along with other viral proteins (29), and cGAS mainly recognizes viral-dsDNA in the cytoplasm. Western blotting results clearly demonstrated that pUL42 exists in the cytoplasm within 8-24 h of PRV infection (Fig. 5D). We also found pUL42 colocalized with cGAS in the cytoplasm when the two proteins were co-expressed in HeLa cells, with Pearson coefficients of about 0.8 (Fig. 5E). In addition, we noticed that endogenous cGAS colocalized with pUL42 in the cytoplasm in PRV-infected HeLa cells, with Pearson coefficients of about 0.4 at 24 h (Fig. 5F). These results indicate that pUL42 targets cGAS to inhibit type І IFN production in cytoplasm.
The cGAS protein consists of an unstructured N-terminus and a highly conserved C-terminus (30). The N-terminus of cGAS has been predicted to bind to DNA (31). The C-terminal domain of cGAS contains two highly conserved motifs, including a nucleotide transferase (NTase) core domain and a Mab21 domain with zinc-ribbon insertion (32,33). To investigate which domain is required for the interaction between pUL42 and cGAS, three plasmids expressing three truncated mutants of cGAS (cGAS-C1, cGAS-C2, and cGAS-C3) were constructed (Fig. 6A). Co-immunoprecipitation (Co-IP) results showed that the DNA-binding domain at the N-terminus of cGAS was required for its interaction with pUL42 (Fig. 6B). To elucidate the functional domain by which pUL42 negatively regulates type I IFN production, seven truncated mutants of pUL42, including pUL42-P1 (aa 1-371), pUL42-P2 (aa 1-353), pUL42-P3 (aa 116-371), pUL42-P4 (aa 255-371), pUL42-P5 (aa 1-255), pUL42-P6 (aa 1-116), and pUL42-P7 (aa 116-255) were constructed (Fig. 6C). We confirmed that the 116-353 aa of pUL42 interacted with cGAS (Fig. 6D). As shown in Fig. 6E andF, ectopically expressed pUL42, pUL42-P1, pUL42-P2, pUL42-P3, pUL42-P4, pUL42-P5, and pUL42-P7 significantly inhibited the IFN-β-promoter activation and mRNA level of Ifnb1 induced by cGAS and STING, but not pUL42-P6 and pUL49.5. We also found that the 116-255 aa and the 255-371 aa of pUL42 independently interact with cGAS. These interactions result in truncated mutants of pUL42 exhibiting differential impairments in IFN-β-promoter activation and mRNA level of Ifnb1 induced by cGAS and STING. Notably, pUL42-P4 demonstrates an inhibitory effect comparable to that of pUL42-FL and pUL42-P1, whereas pUL42-P5 and pUL42-P7 do not exhibit such inhibitory activity (Fig. 6D through F). These results suggested that 116-353 aa of pUL42 is required for its inhibition of type I IFN production. Taken together, we demonstrated that the 116-353 aa of pUL42 interacts with the N-terminal of cGAS, which is dependent on the 116-353 aa of pUL42 to inhibit type I IFN production.
Previous studies have identified that pUL42 contains a functional and transferable bipartite NLS at amino acids 354-370 (34). To investigate whether pUL42 exerts its primary inhibitory function in the cytoplasm, a plasmid expressing a truncated mutant of pUL42 lacking NLS (aa 353-371) (pUL42-∆NLS) was constructed. We found that wild-type pUL42 was primarily located in both the nucleus and cytoplasm, while pUL42-∆NLS was completely localized in the cytoplasm of HeLa cells (Fig. 6G). As shown in Fig. 6H andI, ectopically expressed pUL42 and pUL42-∆NLS significantly inhibited the activation of the IFN-β-promoter and the mRNA level of Ifnb1 induced by cGAS-STING, we noticed that pUL42-∆NLS has the capacity to mitigate some of the effects of pUL42 on the promoter activity of IFN-β induced by the cGAS and STING. This suggests that pUL42 may interact with additional nuclear proteins to exert an inhibitory influence on type І IFN production. Furthermore, these findings corroborate the notion that pUL42 is involved in the suppression of type І IFN production within the cytoplasmic compartment. HEK293T cells were transfected with an IFN-ꞵ luciferase reporter (100 ng), a Renilla-TK reporter (10 ng), and plasmids expressing pUL42 and its truncation mutants (400 ng/each) along with plasmids expressing cGAS (50 ng) and STING (50 ng) for 24 h. The cells were collected to detect luciferase activity (upper panels). GAPDH, Flag-tagged cGAS and STING, and HA-tagged pUL42 and its truncation mutants were verified by Western blotting (lower panels). (F) Ifnb1 mRNA levels in HEK293T cells induced by cGAS and STING upon overexpression of pUL42 or its truncation mutants. HEK293T cells were transfected with plasmids expressing pUL42 or its truncation mutants (400 ng/each) along with plasmids expressing cGAS (50 ng) and STING (50 ng) for 24 h. The mRNA levels of Ifnb1 in the HEK293T cells were analyzed by qPCR (upper panels). GAPDH, Flag-tagged cGAS and STING, and HA-tagged pUL42 and its truncation mutants were verified by Western blotting (lower panels). (G) The subcellular localization of overexpressed pUL42 or pUL42 lacking NLS (pUL42-∆NLS). HeLa cells were transfected with 1 µg of a plasmid expressing pUL42 or pUL42-∆NLS. At 24 hpt, the subcellular localization of pUL42 or pUL42-∆NLS in the HeLa cells was detected by confocal microscopy. Scale bars, 20 µm. (H) The effect of the pUL42 or pUL42-∆NLS on the luciferase activity of IFN-β promoter reporter in HEK293T cells induced by cGAS and STING. HEK293T cells were transfected with an IFN-β luciferase reporter (100 ng) and a Renilla-TK reporter (10 ng), and plasmids expressing Flag-cGAS (50 ng) and Flag-STING (50 ng), together with different amounts (0, 100, 200, and 400 ng) of a plasmid expressing HA-UL42 or HA-UL42-∆NLS for 24 h, and then luciferase activities were analyzed (upper panels). The Flag-tagged cGAS and STING, HA-tagged UL42 or HA-UL42-∆NLS proteins, and GAPDH were detected by Western blotting (lower panels). (I) The effect of the pUL42 or pUL42-∆NLS on the Ifnb1 mRNA level in HEK293T cells induced by cGAS and STING. HEK293T cells were transfected with plasmids expressing Flag-cGAS (50 ng) and Flag-STING (50 ng), together with different amounts (0, 100, 200, and 400 ng) of a plasmid expressing HA-UL42 or HA-UL42-∆NLS for 24 h, and then qPCR was performed to analyze Ifnb1 mRNA level (upper panels). The Flag-tagged cGAS and STING, HA-tagged UL42 or HA-UL42-∆NLS proteins, and GAPDH were detected by Western blotting (lower panels). NS, not significant (P > 0.05), ***P < 0.001 (two-tailed Student's t-test). The DNA-binding sites of PRV pUL42 are required for inhibiting type I IFN production
Previous research results have shown that the four conserved arginine residues of HSV-1 pUL42 are crucial for interacting with DNA, whereas pUL42 binds to DNA through several positively charged amino acids on its surface (21,35). We performed sequence alignment and found that the Lys124, Arg196, Gln279, and Arg280 in PRV pUL42, corresponding to the four arginine residues of HSV-1 pUL42 (Fig. 7A). To examine whether the DNAbinding sites of pUL42 are necessary for interaction with cGAS, four mutants of pUL42 DNA-binding sites (K124A, R196A, Q279A/R280A, and 4M) were constructed (Fig. 7A).
Co-IP results showed that pUL42-4M (lacking all four DNA-binding sites K124A/R196A/ Q279A/R280A) completely lost the ability to interact with cGAS, while other mutants lacking one to three DNA-binding sites retained or partially retained the ability to interact with cGAS (Fig. 7B). Consistently, pUL42-4M completely lost the ability to inhibit the activation of IFN-β-promoter and increasing mRNA levels of Ifnb1 induced by cGAS-STING, while other mutants of pUL42 retained or partially retained the ability to inhibit the activation of IFN-β-promoter and increasing mRNA levels of Ifnb1 induced by cGAS-STING (Fig. 7C andD). pUL49.5 served as a negative control. To test whether four DNA-binding sites of pUL42 affect the ability to inhibit the phosphorylation of TBK1 and IRF3 induced by poly(dA:dT), HeLa cells were transfected with a plasmid encoding pUL42, pUL42-4M, or pUL49.5 for 24 h, and the cells were then transfected with poly(dA:dT) for another 12 h. The phosphorylation levels of IRF3 (Ser396) and TBK1 (Ser172) were detected. As shown in Fig. 7E-F, pUL42-4M completely lost the ability to inhibit the phosphorylation of IRF3 and TBK1 induced by poly(dA:dT), pUL49.5 as a negative control. As a second messenger, cGAMP can be catalyzed by cGAS in the cytoplasm to directly bind and activate the STING-mediated type I IFN production, playing an important role in host antiviral immune responses. To investigate whether pUL42 has the effect on the cGAMP production induced by poly(dA:dT), HeLa cells were transfected with a plasmid encoding pUL42, pUL42-K124A, pUL42-R196A, pUL42-Q279A/R280A, pUL42-4M, or pUL49.5 for 24 h, and the cells were then transfected with poly(dA:dT) for another 12 h. The cell supernatant and cell lysates were collected, and the content of cGAMP was detected. As shown in Fig. 7G andH, pUL42 significantly inhibited the cGAMP production induced by poly(dA:dT), whereas pUL42-R196A and pUL42-4M completely lost the ability to inhibit the cGAMP production induced by poly(dA:dT). As a negative control, pUL49.5 did not inhibit the cGAMP production induced by poly(dA:dT). These results have demonstrated that pUL42 significantly inhibited the cGAMP production induced by poly(dA:dT) and associated with its DNA-binding sites. Taken together, we have demon strated that pUL42 interacts with cGAS through its DNA-binding sites, which is important for pUL42 function to inhibit the cGAS-STING-mediated type I IFN production.
## PRV pUL42 binds to cGAS and prevents the association of cGAS with DNA
The recognition of DNA by cGAS is the first step in activating the cGAS-STING-TBK1 cascade signal. To test whether pUL42 interacts with its DNA-binding region of cGAS to inhibit its recognition of dsDNA, DNA pull-down analysis was performed to measure whether pUL42 has the ability to affect the dsDNA recognition of cGAS. Biotin-labeled PRV genomic DNA was added to HeLa cell lysate expressing Flag-cGAS, HA-pUL42, or HA-pUL42-4M, and a pull-down assay was performed with streptavidin beads. As shown in Fig. 8A, HA-UL42-4M has lost the ability to bind to the PRV genomic DNA. We also found that Flag-cGAS was pulled down by biotin-labeled PRV genomic DNA, but not by biotin-unlabeled PRV genomic DNA. PRV pUL42 significantly inhibited the binding of cGAS to PRV genomic DNA, while pUL42-4M did not (Fig. 8B). Consistent with these results, we also found that Flag-cGAS was pulled down by biotin-labeled poly(dA:dT), but not by biotin-unlabeled poly(dA:dT). pUL42 significantly inhibited the binding of cGAS to poly(dA:dT), while pUL42-4M exhibited a complete loss of this inhibitory capacity (Fig. 8C). To further confirm these results, HeLa cells were transfected with sicon and si520 for for luciferase activities (upper panels). The Flag-tagged cGAS and STING, HA-tagged pUL42, pUL42-K124A, pUL42-R196A, pUL42-Q279A/R280A, pUL42-4M, or pUL49.5 proteins and GAPDH were detected by Western blotting (lower panels). (D) The effect of the DNA-binding sites of PRV pUL42 on the Ifnb1 mRNA levels in HEK293T cells induced by cGAS and STING. HEK293T cells were co-transfected with 400 ng of empty vector or a plasmid expressing pUL42, pUL42-K124A, pUL42-R196A, pUL42-Q279A/R280A, pUL42-4M, or pUL49.5, along with 50 ng of plasmids expressing cGAS and STING for 24 h, and then qPCR was performed to analyze Ifnb1 mRNA levels (upper panels). The Flag-tagged cGAS and STING, HA-tagged pUL42, pUL42-K124A, pUL42-R196A, pUL42-Q279A/R280A, pUL42-4M, and pUL49.5 proteins and GAPDH were detected by Western blotting (lower panels). (E) The effect of the DNA-binding sites of PRV pUL42 for pUL42-mediated inhibition of phosphorylation of TBK1 and IRF3 in HeLa cells induced by poly(dA:dT). HeLa cells were transfected with 2 µg of empty vector, a plasmid expressing PRV pUL42, pUL42-4M, or PRV pUL49.5 for 24 h, and then the cells were transfected with 1 µg/mL of poly(dA:dT) for 12 h. The protein levels of phosphorylated TBK1, TBK1, phosphorylated IRF3, IRF3, HA-pUL42, pUL42-4M, pUL49.5, and GAPDH were analyzed by Western blotting. (F) The ratio of the intensity value of the TBK1-p/TBK1 and IRF3-p/IRF3 immunoblotting result was quantified by ImageJ. (G and H) HeLa cells were transfected with 2 µg of empty vector, a plasmid expressing PRV pUL42, pUL42-K124A, pUL42-R196A, pUL42-Q279/R280A, pUL42-4M, or PRV pUL49.5 for 24 h, and then the cells were transfected with 1 µg/mL of poly(dA:dT) for 12 h. The cell supernatant and cell lysate were collected, and the content of cGAMP was detected. The data represent three independent experiments with three biological replicates or represent three independent experiments with similar results. NS, not significant (P > 0.05), **P < 0.01, ***P < 0.001 (one-way ANOVA). Subsequently, the cells were transfected with si520 or sicon (100 nM/each) for 24 h, and then HeLa cells were infected with PRV (1 MOI) for 24 h. Co-IP analysis of the interactions of Flag-cGAS and HA-cGAS. The ratio of the intensity value of the HA-cGAS immunoblotting result was quantified by ImageJ. (G) The effect of pUL42 on the oligomerization of cGAS in HeLa cells treated with or without poly(dA:dT). HeLa cells were transfected with 2 µg of a plasmid expressing Flag-cGAS alone, or together with the plasmid expressing HA-pUL42 or HA-pUL42-4M (2 µg/each) for 24 h, and then the cells were untreated or treated with 1 µg/mL of poly(dA:dT) for 12 h. The oligomerization of cGAS and the expression of Flag-cGAS, HA-pUL42, HA-pUL42-4M, and GAPDH were detected by Western blotting.
The ratio of the intensity value of the Flag-cGAS immunoblotting result was quantified by ImageJ. (H) The effect of pUL42 knockdown on the oligomerization of cGAS in HeLa cells induced by PRV. HeLa cells were transfected with si520 or sicon (100 nM/each) for 24 h. HeLa cells were infected with PRV (1 MOI) for 24 h, and then the cells were untreated or treated with 1 µg/mL of poly(dA:dT) for 12 h. The oligomerization of cGAS, and the expression of cGAS, pUL42, and GAPDH were detected by Western blotting. The ratio of the intensity value of the cGAS immunoblotting result was quantified by ImageJ.
PRV infection in HeLa cells. These data indicate that PRV pUL42 binds to cGAS specifically to prevent the association of cGAS with DNA.
## PRV pUL42 inhibits cGAS oligomerization
cGAS oligomerization is the hallmark of cGAS activation, which is a key step for producing cGAMP (36). Therefore, we analyzed whether PRV pUL42 is required for the inhibition of the dimerization and oligomerization of cGAS. As shown in Fig. 8E, overexpressed pUL42 inhibited the dimerization of cGAS, but pUL42-4M did not. To further confirm this result, HeLa cells were transfected with Flag-cGAS and HA-cGAS for 12 h and then transfected sicon and si520 for 24 h, finally infected with PRV-JM (1 MOI) for 24 h. We also found that PRV infection significantly inhibits the dimerization of cGAS, while knocking down pUL42 completely lost the ability to inhibit the dimeriza tion of cGAS during PRV infection in HeLa cells (Fig. 8F). Consistent with these results, we found that overexpressed pUL42 inhibited the oligomerization of cGAS induced by poly(dA:dT), while pUL42-4M lost the ability to inhibit the oligomerization of cGAS (Fig. 8G). Interestingly, we also found that the oligomerization of cGAS induced by poly(dA:dT) was significantly inhibited during PRV infection in HeLa cells, while knocking down pUL42 lost the ability to inhibit the oligomerization of cGAS during PRV infection in HeLa cells (Fig. 8H). Taken together, these results demonstrated that pUL42 inhibits the dimerization and oligomerization of cGAS during PRV infection.
## PRV pUL42 targets the cGAS-mediated signaling pathway to promote PRV replication
To examine the inhibitory influence of pUL42 on the cGAS-STING signaling pathway and to assess its implications for efficient viral replication, we conducted the transfection of either sicon or si520 into WT (PK15-WT) or cGAS knockout PK15 (PK15-cGAS -/-) cells, followed by infection with PRV to assess the production of type І IFN and the viral replication. In PK15-WT cells, the knockout of cGAS markedly reduces both the mRNA levels of Ifnb1 and its overall production (Fig. 9A andB), thereby indicating that cGAS is a critical component in the type І IFN production pathway activated by PRV. Additionally, we also found that knocking down Ul42 significantly promotes the mRNA levels of Ifnb1 and the production of type І IFN in PK15-WT during PRV infection. However, in PK15-cGAS -/-cells, knocking down Ul42 has no effect on the production of type І IFN induced by PRV (Fig. 9A andB). This suggests that pUL42 targets the cGAS-mediated signaling pathway to inhibit type І IFN production. Subsequently, we found that in PK15-WT cells, si520 resulted in a reduction of PRV titer by approximately 2 logs, which reflects the overall performance of pUL42 in viral replication and its function in inhibiting type І IFN production. However, in PK15-cGAS -/-cells, si520 resulted in a decrease in PRV titer by about 1 log, which serves as an evaluation of pUL42's role in viral replication, particularly given that the knockout of cGAS nearly eliminated type І IFN signaling (Fig. 9C). At the same time, we also observed that the protein levels of pUL42 and gB in PK15-cGAS -/-cells were significantly higher than those in PK15-WT cells (Fig. 9D). In summary, we found that pUL42 targets the cGAS-mediated signaling pathway to enhance viral replication, which is equally significant to pUL42's role in the viral replication process.
## DISCUSSION
cGAS-STING axis is the core pathway for host antiviral immune responses during viral infection. Previous studies have shown that PRVs have evolved multiple strategies to counteract host antiviral responses (1,37,38). How PRV targets the cGAS-STING signaling pathway to evade host antiviral response is worth further investigation. In this study, we found that PRV pUL42 is a negative regulator of type I IFN by targeting cGAS. PRV pUL42 interacts with the DNA-binding domain of cGAS and inhibits its recognition of dsDNA during PRV infection, thereby suppressing the oligomerization of cGAS, leading to inhibiting the production of cGAMP and type I IFN. Using siRNA to knock down the expression of the Ul42 gene, we also revealed that PRV pUL42 is beneficial for PRV replication and evades host antiviral responses by antagonizing the cGAS-mediated innate antiviral immune responses (Fig. 10A).
Upon DNA virus infection, PAMPs in virus-infected cells are sensed by PRRs, such as cGAS, to induce the production of type I IFN (39). The secreted type I IFN binds to IFNAR1/2 on the cell membrane to induce the expression of hundreds of ISGs and exert their antiviral function (40). In this study, we found that PRV infection inhibits the production of type I IFN to promote PRV infection. However, PRV has the ability to antagonize the host's antiviral immune response, which is crucial for its replication and maintenance of infection in vivo. Recently, researchers have shown that several PRV proteins can negatively regulate the production of type I IFN. For example, PRV pUL24 interacts with IFN regulatory factor 7 (IRF7) through the proteasome pathway and degrades IRF7, significantly reducing the production of type I IFN (24). PRV pUL13 recruits E3 ligase RING finger protein 5 (RNF5) to promote K27-/K29-linked ubiquitination of STING, which promotes the degradation of STING, inhibiting type I IFN production (37). Peroxidase 1 (PRDX1) binds to TBK1 and IκB kinase ε (IKKε), actively regulating the production of type I IFN. PRV pUL13 interacts with the antiviral regulatory factor PRDX1 through the ubiquitin proteasome pathway and promotes its degradation, thereby inhibiting-mediated antiviral immune response (41). All these PRV proteins inhibit the production of type I IFN by targeting the downstream components of the cGAS-STING signaling pathway. We proposed that certain PRV proteins may be responsible for targeting cGAS, thereby obstructing the interaction between cGAS and viral DNA. This mechanism is believed to inhibit type I IFN production at an upstream point in the cGAS-STING pathway. Based on unbiased screening, we found that PRV pUL42 is capable of inhibiting the interaction between cGAS and viral genomes, as well as the interaction between cGAS and poly(dA:dT) during PRV infection.
cGAS is activated upon binding to dsDNA, which triggers the synthesis of the second messenger cGAMP. This molecule subsequently activates the STING, thereby initiating an antiviral immune response. HSV-1 possesses mechanisms to inhibit the enzymatic activity of cGAS. Specifically, the HSV-1 pVP22 can directly interact with cGAS, leading to a suppression of its enzymatic function and consequently inhibiting the production of type I IFN (42). Additionally, HSV-1 pUL37 can impede cGAS activation through its deamidase activity, resulting in a reduction of cGAMP synthesis and subsequent blockage of downstream signaling pathways (43). Recent findings have also demonstra ted that HSV-1 pUL41 significantly decreases the accumulation of cGAS mRNA, thereby inhibiting the activation of the cGAS-STING-mediated type I IFN signaling pathway (44). Both HSV-1 and PRV pUL21 have been shown to degrade cGAS, facilitating viral infection (45). Furthermore, the envelope protein ORF9 of varicella-zoster virus acts as an antagonist to cGAS, by binding to DNA and phase-separated together with DNA, which may disrupt the cGAS-DNA oligomer and inhibit the activation of the cGAS-STING signaling pathway (46). Our research indicates that PRV pUL42 inhibits the binding of cGAS to dsDNA, thereby suppressing type I IFN production and promoting PRV replication. These findings illustrate that herpesviruses employ various strategies to target cGAS, effectively inhibiting type I IFN production and enhancing viral replication.
DNA sensor cGAS recognizes and binds to viral dsDNA in the cytoplasm to catalyze the production of cGAMP from the substrates GTP and ATP (47). The cGAMP then binds and activates STING to activate host antiviral responses by mediating the production of type I IFN (48). The binding of cGAS to dsDNA forms a 2:2 complex, which is crucial for the activation of cGAS (36,49). Previous studies have shown that nucleosomes can bind to cGAS and inhibit dsDNA-mediated cGAS activation. The interaction between nucleosomes and cGAS blocks the recognition and binding between cGAS and DNA, disrupting cGAS dimerization and thus "hijacking" it in an inactive monomeric state (50). However, previous studies have shown that cGAS does not have sequence specificity for DNA recognition. In our study, we confirmed that PRV pUL42 interacts with the DNAbinding domain of cGAS and inhibits the dimerization and oligomerization activation of cGAS in vitro and in vivo.
Alpha herpesvirus pUL42 is a highly conserved DNA polymerase that is an important processing factor for viral DNA replication (18). Functional analysis of HSV-1 pUL42 revealed three main biochemical functions, including binding to DNA, stable binding to viral DNA polymerase pUL30, and acting on increasing the length of DNA chains synthesized by pUL30 (18). Previous studies have shown that the PRV pUL42 can enhance the catalytic activity of DNA polymerase, which is crucial for virus replication. Consistently, we found that knocking down the expression level of the Ul42 gene significantly inhibited PRV replication, inducing higher levels of type I IFN in PRV-infec ted cells. It is well known that both PRV pUL42 inhibit the IFN signaling pathway by suppressing the interaction of ISRE-ISGF3 (21). Consistently, we also found that ectopically expressed pUL42 significantly decreased the mRNA levels of Isg56 and Isg54 induced by cGAS-STING or by poly(dA:dT), suggesting that PRV pUL42 is a potent inhibitor of both type I IFN production and type I IFN signaling pathway.
Taken together, our findings indicate that the Ul42 gene has a critical role in PRV replication, and pUL42 can also facilitate PRV replication by inhibiting type I IFN production and the IFN signaling pathway. Even though pUL42 has been shown to inhibit the JAK-STAT signaling pathway induced by IFN-α (21), cGAS functions as an upstream component in the cGAS-STING signaling pathway. So, the inhibition of cGAS-STING signaling through the disruption of the interaction between dsDNA and cGAS by pUL42 is important in the process of PRV infection. Our findings contribute to understanding the function of PRV pUL42 and its role in viral infection, providing new clues to designing anti-PRV or attenuated live vaccines by targeting the PRV Ul42 gene.
## MATERIALS AND METHODS
## Cells and viruses
PK15 cells, HEK293T cells, and HeLa cells were cultured in Dulbecco's modified Eagle's medium (DMEM). PK15-cGAS -/-cells from Professor Beibei Chu of Henan Agricultural University (51). PAMs were isolated from the lung lavage fluid of 4-week-old healthy specific pathogen-free piglets (without African swine fever virus, Classical swine fever virus, Porcine reproductive and respiratory syndrome virus, PRV, and other 28 patho gens), and they were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), 100 U penicillin, and 100 µg/mL streptomycin at 37°C with 5% CO 2 . PRV-JM was isolated from the aborted piglet samples of PRV-positive pig farms at Jinmen, Guang dong Province of China, as previously described (52).
## Reagents, plasmids, and antibodies
poly(dA:dT) (P0883-10UN) and anti-Flag (M2) beads (M8823) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Protease inhibitor cocktail (4693132001) was purchased from Roche (Basel, Switzerland). DMEM (C11995500CP) and FBS (10091-148) were purchased from GIBCO (Grand Island, NE, USA). The Dual-luciferase Reporter Assay System (E1910) was purchased from Promega (Madison, MI, USA). PrimeScript RT Reagent Kit (RR037A) and SYBR Premix Ex Taq II (RR820A) were purchased from Takara (Shiga, Japan). Rabbit anti-Flag (F7425-2MG) (final dilution 1:1,000), mouse anti-Flag (F1804-1MG) (final dilution 1:1,000), rabbit anti-HA (SAB4300603) (final dilution 1:1,000), and mouse anti-HA (B9183) (final dilution 1:2,000) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Mouse anti-GAPDH (60004-1-Ig) (final dilution 1:10,000), rabbit anti-Lamin B (12987-1-AP) (final dilution 1:1,000), and rabbit anti-TBK1 (28397-1-AP) (final dilution 1:1,000) were purchased from Proteintech (Wuhan, China). Rabbit anti-Phos pho-TBK1 (5483) (final dilution 1:1,000), rabbit anti-IRF3 (11904) (final dilution 1:1,000), and rabbit anti-Phospho-IRF3 (29047) (final dilution 1:1,000) were purchased from Cell Signaling Technology (Danvers, MA, USA). Mouse anti-cGAS was from Professor Yong Huang of Northwest A&F University (53). Mouse anti-UL42 was from Professor Liping Huang of Harbin Veterinary Research Institute (54). The IRDye 800CW goat anti-rabbit IgG (H + L) (925-32211) and IRDye 800CW goat anti-mouse IgG (H + L) (925-32210) were purchased from LI-COR (Lincoln, NE, USA). Alexa Fluor 488 goat anti-Rabbit IgG(H + L) (A11008) and Alexa Fluor 594 goat anti-Mouse IgG(H + L) (A11032) were purchased from Thermo Fisher Scientific (Waltham, MA, USA).
The IFN-β-, ISG54-, ISG56-, NF-қB-, ISRE-reporters, and Renilla-TK reporter were obtained from Professor Hong Tang. The 44 cDNAs corresponding to PRV-encoded proteins were synthesized based on the genome of the PRV-JM (52) isolate and cloned into the pCAGGS-Flag (pFlag) vector from GenScript (Nanjing, China). To construct plasmids expressing Flag-tagged or HA-tagged proteins involved in the cGAS-STING signaling pathway, the cDNAs corresponding to these swine genes were amplified by qPCR using total RNA extracted from PAMs as templates and were then cloned into the pCAGGS-Flag or pCAGGS-HA vector, respectively. All constructs were validated by DNA sequencing. The primers used in this study are listed in Table 1.
## Cell transfection
HEK293T cells and HeLa cells were transfected with plasmids using PEI or Lipofectamine 2000 transfection reagent. The ratio of plasmid amount to transfection reagent was 1:2 (e.g., 1 µg of plasmid to 2 µL of transfection reagent). Taking one well of a 6-well or 48-well cell culture plate as an example, when the cells grew to 80%-90% confluence, 100 µL of Opti-MEM was added to two sterile 1.5 mL EP tubes, followed by the addi tion of plasmid and the corresponding transfection reagent to the two EP tubes. The mixtures were gently vortexed to ensure even distribution in Opti-MEM, incubated at room temperature for 5 minutes, and then the Opti-MEM containing the transfection reagent was mixed with the Opti-MEM containing the plasmid. The combined solution was gently vortexed and allowed to sit at room temperature for 20 minutes. Finally, the mixture was slowly added dropwise to the cell culture medium. Samples were collected for subsequent experiments 24-36 hours after transfection.
## Confocal microscopy
HeLa cells were transfected with plasmids expressing HA-tagged or Flag-tagged proteins for 24 h. These cells were fixed with 4% paraformaldehyde and permeabilized with 0.1% Triton X-100. After blocking with 10% FBS, the cells were incubated with anti-Flag and anti-HA antibodies for 2 h. Samples were visualized with a Leica SP2 confocal system (Carl Zeiss AG, Oberkochen, Germany). HeLa cells were infected with PRV (1 MOI) for 0, 4, 8, 12, or 24 h, and the protein level of pUL42 or cGAS was examined with mouse anti-pUL42 or rabbit anti-cGAS antibodies, respectively. The subcellular localization of these proteins was visualized with a Leica SP2 confocal system (Carl Zeiss AG, Oberkochen, Germany).
## Co-immunoprecipitation and Western blot assay
For Co-IP, the cells were lysed in lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM MgCl 2 , 1 mM EDTA, 1% Triton X-100, and 10% glycerol) containing 1 mM PMSF and a 1 × protease inhibitor cocktail (Roche, Basel, Switzerland). Then, cell supernatants were incubated with anti-Flag (M2) agarose or with protein G Plus-Agarose immunoprecipita tion reagent (Sigma-Aldrich, St. Louis, MO, USA) together with 1 µg of the indicated antibodies at 4°C overnight on a roller. The pellets were washed five times with cell lysis buffer. For Western blot analysis, 20% amounts of cell lysates and immunoprecipitants were resolved by 10%-12% sodium dodecyl sulfate polyacrylamide gel electrophoresis and were then transferred to a polyvinylidene difluoride membrane (Sigma-Aldrich, St. Louis, MO, USA). After incubation with primary and secondary antibodies, the membranes were visualized by enhanced chemiluminescence (Thermo Fisher Scientific, Waltham, MA, USA) or an Odyssey two-color infrared fluorescence imaging system (LI-COR).
## Dual-luciferase reporter assay
HEK293T cells were co-transfected with the indicated plasmids. After 24 h, the cells were harvested and lysed in lysis buffer, and luciferase activities of IFN-β-, ISG54-, ISG56-, NF-қB-, ISRE-reporters, and TK-Renilla reporter were measured with a Dual-luciferase Reporter Assay System (Promega, Madison, MI, USA) according to the manufacturer's instructions. The data were normalized to the transfection efficiency by dividing the firefly luciferase activity by the Renilla luciferase activity. Each experiment was conducted three times independently, and the representative results are shown.
## Quantitative PCR
To detect the mRNA level of Ifnb1, Ifnα4, Isg56, Isg54, or Ul42, total RNA was extrac ted using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA), and reverse transcription was performed with a PrimeScript RT Reagent Kit (Takara, Tokyo, Japan). Reverse transcription products were amplified using an Agilent-Strata gene Mx Real-Time qPCR system with SYBR Premix Ex Taq II (Takara, Tokyo, Japan) according to the manufacturer's instructions. Data were normalized to the level of β-actin expression in individual sample. qPCR was carried out on a QuantStudio5 system (Applied Biosystems, USA) according to the OIE-recommended procedure. All the qPCR primers are listed in Table 2.
## RNA interference
The siRNAs targeting the Ul42 gene are listed in Table 3. The transfection of siRNA was performed with HiPerFect Transfection Reagent (QIAGEN, Germantown, MD) following the manufacturer's instructions. Forty-eight hours after siRNA transfection, the knock down efficiency of pUL42 was assessed by Western blotting.
## DNA pulldown
HEK293T cells were co-transfected with HA-UL42, HA-UL42-4M, or Flag-cGAS for 24 h. Then, the cells were lysed in lysis buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 5 mM MgCl 2 , 1 mM EDTA, 1% Triton X-100, and 10% glycerol) containing 1 mM PMSF and a 1 × protease inhibitor cocktail (Roche, Basel, Switzerland). Cell supernatants were incubated with biotinylated-PRV-DNA or poly(dA:dT) (Sigma-Aldrich, St. Louis, MO, USA) for 12 h at 4°C, and then incubated with streptavidin beads (Sigma-Aldrich, St. Louis, MO, USA) for another 4 h at 4°C. The beads were washed three times with lysis buffer and analyzed by immunoblotting with the indicated antibodies.
## References
1. Ye, Liu, Zhou et al. (2022) "A tug of war: pseudorabies virus and host antiviral innate immunity" *Viruses*
2. Chen, Sun, Chen (2016) "Regulation and function of the cGAS-STING pathway of cytosolic DNA sensing" *Nat Immunol*
3. Ishikawa, Ma, Barber (2009) "STING regulates intracellular DNAmediated, type I interferon-dependent innate immunity" *Nature*
4. Zhou, Zhang, Xu (2020) "TBK1, a central kinase in innate immune sensing of nucleic acids and beyond" *Acta Biochim Biophys Sin (Shanghai)*
5. Liu, Cai, Wu et al. (2015) "Phosphorylation of innate immune adaptor proteins MAVS, STING, and TRIF induces IRF3 activation" *Science*
6. (2025) *Full-Length Text Journal of Virology*
7. Cheng, Kong, Liu et al. (2020) "Natural infection of a variant pseudorabies virus leads to bovine death in China" *Transbound Emerg Dis*
8. Minamiguchi, Kojima, Sakumoto et al. (2019) "Isolation and molecular characterization of a variant of Chinese gC-genotype II pseudorabies virus from a hunting dog infected by biting a wild boar in Japan and its pathogenicity in a mouse model" *Virus Genes*
9. Pomeranz, Reynolds, Hengartner (2005) "Molecular biology of pseudorabies virus: impact on neurovirology and veterinary medicine" *Microbiol Mol Biol Rev*
10. Guo, Chen, Zhang (2021) "Human PRV infection in China: an alarm to accelerate eradication of PRV in domestic pigs" *Virol Sin*
11. Kong, Chen, Gong et al. (2024) "Pseudorabies virus tegument protein US2 antagonizes antiviral innate immunity by targeting cGAS-STING signaling pathway" *Front Immunol*
12. Yan, Yue, Zhang et al. (2024) "Pseudorabies virus VHS protein abrogates interferon responses by blocking NF-κB and IRF3 nuclear translocation" *Virol Sin*
13. Jansens, Olarerin-George, Verhamme et al. (2023) "Alphaherpesvirus-mediated remodeling of the cellular transcriptome results in depletion of m6A-containing transcripts"
14. Zhao, Wang, Kong et al. (2022) "Pseudorabies virus tegument protein UL13 suppresses RLR-mediated antiviral innate immunity through regulating receptor transcription" *Viruses*
15. Zhao, Ding, Zhu et al. (2024) "Pseudorabies virus infection triggers mitophagy to dampen the interferon response and promote viral replication" *J Virol*
16. Verhamme, Jansens, Liu et al. (2025) "The pseudorabies virus UL13 protein kinase triggers phosphorylation of the RNA demethylase FTO, which is associated with FTO-dependent suppression of interferonstimulated gene expression" *J Virol*
17. Wang, Liu, Qin et al. (2022) "Pseudorabies virus EP0 antagonizes the type i interferon response via inhibiting IRF9 transcription" *J Virol*
18. Yin, Romero, Favoreel (2021) "Pseudorabies virus inhibits type I and type III interferon-induced signaling via proteasomal degradation of janus kinases" *J Virol*
19. Digard, Chow, Pirrit et al. (1993) "Functional analysis of the herpes simplex virus UL42 protein" *J Virol*
20. Gottlieb, Marcy, Coen et al. (1990) "The herpes simplex virus type 1 UL42 gene product: a subunit of DNA polymerase that functions to increase processivity" *J Virol*
22. Bo, Li, Wang et al. (2023) "Suppression of NF-κB signaling by pseudorabies virus DNA polymerase processivity factor UL42 via recruiting SOCS1 to promote the ubiquitina tion degradation of p65" *Vet Microbiol*
23. Zhang, Chen, Zhang et al. (2021) "Pseudorabies virus DNA polymerase processivity factor UL42 inhibits type I IFN response by preventing ISGF3-ISRE interaction" *J Immunol*
24. Lv, Cao, Bai et al. (2020) "PRV-encoded UL13 protein kinase acts as an antagonist of innate immunity by targeting IRF3-signaling pathways" *Vet Microbiol*
25. Chen, Kong, Xu et al. (2021) "Pseudorabies virus UL24 antagonizes OASL-mediated antiviral effect" *Virus Res*
26. Liu, Zhang, Ye et al. (2021) "Inhibition of the DNA-Sensing pathway by pseudorabies virus UL24 protein via degradation of interferon regulatory factor 7" *Vet Microbiol*
27. Wang, Yang, Feng et al. (2020) "Pseudorabies virus UL24 abrogates tumor necrosis factor alphainduced NF-κB activation by degrading P65" *Viruses*
28. Chapon, Parvatiyar, Aliyari et al. (2019) "Comprehen sive mutagenesis of herpes simplex virus 1 genome identifies UL42 as an inhibitor of type I interferon induction" *J Virol*
29. Liu, Wang, Xie et al. (2021) "A novel human acute encephalitis caused by pseudorabies virus variant strain" *Clin Infect Dis*
30. Berthomme, Monahan, Parris et al. (1995) "Cloning, sequencing, and functional characterization of the two subunits of the pseudorabies virus DNA polymerase holoenzyme: evidence for specificity of interaction" *J Virol*
31. Alvisi, Avanzi, Musiani et al. (2008) "Nuclear import of HSV-1 DNA polymerase processivity factor UL42 is mediated by a C-terminally located bipartite nuclear localization signal" *Biochemistry*
32. Kuchta, Knizewski, Wyrwicz et al. (2009) "Comprehensive classification of nucleotidyltransferase fold proteins: identification of novel families and their representatives in human" *Nucleic Acids Res*
33. Kranzusch, Vance (2013) "cGAS dimerization entangles DNA recognition" *Immunity*
34. Yu, Liu (2021) "Cytosolic DNA sensing by cGAS: regulation, function, and human diseases" *Signal Transduct Target Ther*
35. Civril, Deimling, De Oliveira Mann et al. (2013) "Structural mechanism of cytosolic DNA sensing by cGAS" *Nature*
36. Wang, Du, Huang et al. (2016) "The pseudorabies virus DNA polymerase accessory subunit UL42 directs nuclear transport of the holoenzyme" *Front Microbiol*
37. Thornton, Chaudhuri, Monahan et al. (2000) "Analysis of in vitro activities of herpes simplex virus type 1 UL42 mutant proteins: correlation with in vivo function" *Virology (Auckl)*
38. Li, Shu, Yi et al. (2013) "Cyclic GMP-AMP synthase is activated by double-stranded DNAinduced oligomerization" *Immunity*
39. Kong, Yin, Wang et al. (2022) "Pseudorabies virus tegument protein UL13 recruits RNF5 to inhibit STING-mediated antiviral immunity" *PLoS Pathog*
40. Bo, Miao, Xi et al. (2020) "PRV UL13 inhibits cGAS-STING-mediated IFN-β production by phosphorylating IRF3" *Vet Res*
41. Zhu, Zheng (2020) "The race between host antiviral innate immunity and the immune evasion strategies of herpes simplex virus 1" *Microbiol Mol Biol Rev*
42. Schneider, Chevillotte, Rice (2014) "Interferon-stimulated genes: a complex web of host defenses" *Annu Rev Immunol*
43. Lv, Bai, Gao et al. (2021) "Peroxiredoxin 1 interacts with TBK1/IKKε and negatively regulates pseudorabies virus propagation by promoting innate immunity" *J Virol*
44. Huang, You, Su et al. (2018) "Herpes simplex virus 1 tegument protein VP22 abrogates cGAS/STING-mediated antiviral innate immunity"
45. Zhang, Zhao, Xu et al. (2018) "Species-specific deamidation of cGAS by herpes simplex virus UL37 protein facilitates viral replication" *Cell Host Microbe*
46. Su, Zheng (2017) "Herpes simplex virus 1 abrogates the cGAS/STINGmediated cytosolic DNA-sensing pathway via its virion host shutoff protein, UL41" *J Virol*
47. Ma, Bai, Jiang et al. (2023) "Tegument protein UL21 of alpha-herpesvirus inhibits the innate immunity by triggering CGAS degradation through TOLLIP-mediated selective autophagy" *Autophagy*
48. Hertzog, Zhou, Fowler et al. (2022) "Varicella-zoster virus ORF9 is an antagonist of the DNA sensor cGAS" *EMBO J*
49. Sun, Wu, Du et al. (2013) "Cyclic GMP-AMP synthase is a cytosolic dna sensor that activates the type I interferon pathway" *Science*
50. Tong, Song, Zhang et al. (2024) "When DNA-damage responses meet innate and adaptive immunity" *Cell Mol Life Sci*
51. Zhang, Wu, Du et al. (2014) "The cytosolic DNA sensor cGAS forms an oligomeric complex with DNA and undergoes switch-like conformational changes in the activation loop" *Cell Rep*
52. Cao, Han, Fan et al. (2020) "Structural basis for nucleosome-mediated inhibition of cGAS activity" *Cell Res*
53. Wang, Wang, Ming et al. (2020) "Porcine IFITM1 is a host restriction factor that inhibits pseudorabies virus infection" *Int J Biol Macromol*
54. Zhou, Zhang, Liu et al. (2022) "Isolation and characterization of two pseudorabies virus and evaluation of their effects on host natural immune responses and pathogenicity" *Viruses*
55. Wang, Chen, Wu et al. (2021) "PCV2 targets cGAS to inhibit type I interferon induction to promote other DNA virus infection" *PLoS Pathog*
56. Wang, Wei, Wu et al. (2022) "Specific inhibition of the interaction between pseudorabies virus DNA polymerase subunits UL30 and UL42 by a synthetic peptide" *Vet Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802151&blobtype=pdf | # Molecular basis for occlusion of the jeilongvirus receptorbinding site by the elongated C-terminus
Alice Stelfox, Airah Javorsky, Robert Stass, Geoff Sutton, Kamel Omari, Thomas Bowden
## Abstract
Paramyxoviral transmission between hosts may be, in part, attributed to the ability of the viral envelope-displayed receptor-binding protein (RBP) to bind to cell surface receptors of different host species. We sought to elucidate the architecture of the receptor-binding head region of the RBPs presented by jeilongviruses, a group of emerging and genetically unique paramyxoviruses belonging to the genus Jeilongvi rus, family Paramyxoviridae. Structure determination of J and Beilong jeilongvirus RBPs reveals that the proteins exhibit a prototypical six-bladed β-propeller fold, present a binding site with residues associated with sialic acid recognition and hydrolysis, and bear a close structural relationship with sialic acid binding hemagglutinin-neuraminidase (HN)-type paramyxoviral RBPs. Additionally, unlike other paramyxoviruses, jeilongviruses encode an RBP with an unusually long C-terminal extension. In our dimeric Beilong virus RBP structure, we find that the C-terminal extension exchanges a hat-like domain with the central region of the β-propeller of the opposing protomer through domain-swap ping. The hat-like domain occludes residues putatively associated with sialic acid binding and hydrolysis, providing a structural rationale for the absence of observed hemadsorp tion and neuraminidase activity. The insights gleaned from this analysis expand our appreciation of the structural palette available to the plastic paramyxoviral RBP and how their architectures may be adapted to regulate host-cell interactions at the cell surface.
IMPORTANCEThe paramyxovirus receptor-binding protein (RBP) plays a primary role in determining cell and species tropism. Here, we study the RBPs of jeilongviruses, a group of paramyxoviruses that present a distinctive RBP that encodes an elongated C-terminal region. While the jeilongviral RBP structurally categorizes with paramyxoviral RBPs that interact with sialic acid during host-cell entry, the unusually long C-terminal domain was found to sterically occlude the associated binding site, suggesting that the molecule has developed strategies for autoinhibition of receptor interactions. These data expand our understanding of the architectural space occupied by paramyxoviral RBPs and the structural elaborations that may be incorporated into the paramyxovirus genome to modulate native functionality. KEYWORDS paramyxovirus, Jeilongvirus, virus-host interactions, structure, glycopro tein, viral attachment L ike other paramyxoviruses, viruses within the recently established genus, Jeilongvi rus, family Paramyxoviridae, encode a negative-sense, non-segmented RNA genome. J paramyxovirus (JPV) (1-3) and Beilong paramyxovirus (BeiV) (4-8) are founding members of a genus that currently includes thirty-two members (3, 9) and genetically divides into two clusters, reflecting the differential use of small mammals or bats and felines as putative host reservoirs (10). Viral surveillance and discovery efforts have revealed that this group of pathogens exhibits a near worldwide distribution. Further more, although other paramyxoviruses are of substantial threat to human health and
animal husbandry, little is known about the potential threat that jeilongviruses pose due to a paucity of knowledge about their pathobiology, tropism characteristics, and inter-species transmission potential.
Like other paramyxoviruses, jeilongviruses display two major proteins on the viral envelope surface, a receptor-binding protein (RBP) and fusion (F) glycoprotein, which function in concert to facilitate viral entry into a host cell (11). The RBP is presented on the viral surface as a dimer-of-dimers (12,13), with each protomer formed of an N-terminal intraviral (IV) region, transmembrane TM domain, α-helical stalk, and a C-terminal receptor-interacting six-bladed β-propeller receptor-binding head domain. Depending on the paramyxovirus, the membrane distal six-bladed β-propeller domain interacts with proteinaceous or glycan host-cell surface receptors during host-cell entry. While there remains much to be elucidated about the range of host-cell surface receptors utilized by paramyxoviruses, three types of RBP classes are currently recog nized: hemagglutinin-neuraminidase (HN), hemagglutinin (H), and glycoprotein (G). The protein-binding H-type RBPs from morbilliviruses interact with nectin-4 and signaling lymphocytic activation molecule F1, and the G-type RBPs from henipaviruses differen tially interact with ephrins. Conversely, viruses within the Respirovirus, Orthoavulavirus, Metaavulavirus, Paraavulavirus, and Orthorubulavirus genera present HN-type RBPs, which bind and hydrolyze sialic acid, a carbohydrate (11). Paramyxoviral HN-type RBPs retain the seven conserved sialidase residues (14) and a hexapeptide motif (15), which all surround the sialic acid binding pocket (11). Interestingly, the Jeilongvirus RBP retains the majority of the residues responsible for sialic acid binding and hydrolysis, which are highly conserved across HN-type RBPs (14,15). However, investigations of JPV indicate that the virus lacks the expected neuraminidase activity associated with HN-mediated viral egress (1,3).
The genomes of viruses within the genus Jeilongvirus are uncharacteristically long (up to 20 kb) when compared to other viruses in the order Mononegavirales (3,8,16). This is attributed, in part, to the presence of two unique open reading frames: a small hydrophobic (SH) protein and a transmembrane (TM) protein, which are thought to play roles in virulence (17) and cell-cell fusion (18), respectively. Additionally, Jeilongvirus RBPs feature a unique C-terminal extension of unknown structure and variable length (140-1,015 amino acids) (Fig. 1A; Fig. S1A). While little is known about the structure of this part of the RBP, the N-terminal region of the C-terminus encodes conserved cysteine residues, which may form local structure through disulfide bonding (Fig. S1B). Addition ally, the C-terminal region of the C-terminus may exhibit O-linked type glycosylation and a high level of predicted overall structural disorder due to the high prevalence of proline, serine, and threonine (S/T/P) amino acid residues (19). Further reflective of the genomic variability of the jeilongvirus RBP, JPV-RBP, and a strain of BeiV-RBP isolated from human mesangial cells (HMC), lack the variable S/T/P-rich region and instead encode "RBP-associated" open reading frames enriched with S/T/P residues, termed "X" ORFs (19). Interestingly, variants of BeiV isolated from rats lack a stop codon at position 2,205 of the BeiV(rat)-RBP gene, resulting in an RBP of 1,046 amino acids, as opposed to the 734 observed from the laboratory-isolated BeiV(HMC) (20) (Fig. S1A).
Cellular and tissue tropism characteristics of a paramyxovirus are, in part, dictated by the interaction between an RBP and host-cell surface receptor (11,22). Thus, a molecularlevel understanding of paramyxoviral RBP architecture is essential when considering the determinants of interspecies transmission and zoonotic potential (22)(23)(24). Here, to address the paucity of knowledge about Jeilongvirus RBP function, we describe the crystal structures of the JPV-RBP β-propeller head domain (JPV-RBP β ) and the BeiV(HMC)-RBP β-propeller head domain and previously undescribed C-terminal hat-like domain [BeiV(HMC)-RBP β+ ] to 2.2 and 3.5 Å resolution, respectively. This work broadens our appreciation of the pathobiological and architectural diversity of this understudied group of paramyxoviruses.
## RESULTS
## Structure determination and overview of JPV-RBP β and BeiV(HMC)-RBP β+
Given the importance of paramyxoviral RBPs in dictating cellular tropism (11,22), we sought to elucidate the structure of jeilongviral RBPs. To this end, constructs encoding the C-terminal six-bladed β-propeller head domain and extended C-terminus of JPV-RBP (Glu149-Asn709; termed "JPV-RBP β+ ") and BeiV(HMC)-RBP [Asn145-Glu734; termed "BeiV(HMC)-RBP β+ "] were recombinantly produced in the presence of the alpha-manno sidase inhibitor, kifunensine (25), and purified from human embryonic kidney (HEK) 293T cells (Fig. S2). Purified JPV-RBP β+ and BeiV(HMC)-RBP β+ were treated with endoglycosi dase F1 (endoF1) (25), and crystallized using the vapor diffusion method (26). X-ray diffraction data were collected to 2.2 and 3.5 Å resolution, respectively. The JPV-RBP β+ structure was solved by molecular replacement using the structure of human parain fluenza virus 3 RBP (PDB: 1V21) as a search model. Subsequently, BeiV(HMC)-RBP β+ was solved using the refined structure of JPV-RBP β+ as a search model, where the majority of the BeiV(HMC)-RBP β+ C-terminal extension was visible (Lys576-Asn687).
Both JPV-RBP β+ and BeiV(HMC)-RBP β+ exhibit a six-bladed β-propeller fold charac teristic of paramyxoviral RBP proteins, with each blade (β1 to β6) composed of four antiparallel β-strands (Fig. 1B andC, respectively). When superposed, the six-bladed β-propellers of JPV and BeiV RBPs align closely, with a calculated root-mean-square deviation (RMSD) of ~0.8 Å over 366 Cα atoms (Fig. S3A), with variation in the structures mostly contained in the loops and the flexible termini (Fig. S3B). In both BeiV(HMC)-RBP β+ and JPV-RBP β+ structures, electron density corresponding to the majority of the crystallized N-terminal stalk region (Pro146-Gln162 and Glu149-Met160, respectively) and β-propeller (Cys163-Cys575 and Cys161-Cys573, respectively) was well ordered. For BeiV(HMC)-RBP β+ , a large portion of the C-terminal extension was visible (Lys576-Asn687). In contrast, many residues constituting the C-terminal extension of JPV-RBP β+ could not be built (Gly585-Asn709). The remaining C-terminal residues for both RBPs were disordered and directed toward the solvent channels in their respective crystals.
As the C-terminal extension of JPV-RBP β+ was not well ordered, the N-linked glycosylation was not observed at any of the three predicted N-linked glycosylation sequons within this region (Fig. 1B). Likewise, we do not observe a shift upon EndoF1 digestion (Fig. S2A andB), but it is not possible to know whether this is due to incom plete digestion or no occupancy at predicted sequons. Within the ordered region of the β-propeller of BeiV(HMC)-RBP β+ , there are two predicted N-linked glycosylation sequons, Asn515 and Asn579 (Fig. 1C). We observe a partially ordered N-linked glycan at residue Asn515 BeiV(HMC)-RBP β+ on both chains (Fig. S7A), which is consistent with our SEC and SDS-PAGE analysis following EndoF1 treatment (Fig. S2C andD), where we did not observe a shift in protein mass, indicating that the chitobiose core of the glycan was likely inaccessible. This N-linked glycosylation is positioned at the pinnacle of loop β5L34 and is packed against the β-sheet formed by residues of the N-terminus and C-terminal extension, which potentially have a role in stabilizing and protecting the glycan from the enzyme (Fig. S7A). Furthermore, electron density at residue Asn579 in one chain of the asymmetric unit enabled the partial building of N-linked glycosylation at the beginning of the C-terminal extension (Fig. S7A).
Of the previously characterized paramyxoviral RBP β-propellers, JPV-RBP β+ and BeiV(HMC)-RBP β+ display a pattern of disulfide bonding comparable to and most closely structurally aligned with HN-type RBPs, including Newcastle disease virus (NDV)-RBP β (Fig. 2A andB). In contrast, JPV-RBP β+ and BeiV(HMC)-RBP β+ exhibit a more distant structural relationship with henipaviral, henipa-like, morbilliviral, and the narmoviral RBPs (Fig. 2A andB). Furthermore, both JPV-RBP β+ and BeiV(HMC)-RBP β+ β-propellers show no conservation with morbillivirus and the henipavirus RBPs at known receptorbinding sites (Fig. S4), suggesting that they are unlikely to exhibit shared receptor usage.
## BeiV-RBP and JPV-RBP encode residues required for sialic acid entry and hydrolysis
Although previous studies indicate that JPV-RBP lacks hemagglutinin and hemadsorp tion activity (1,3), Jeilongviral RBPs display many of the conserved residues required to interact with sialic acid that are found in other HN-type paramyxoviral RBPs (27,32,35,43), including six of the seven sialidase residues: Arg 1 , Glu 4 , Arg 4 , Arg 5 , Tyr 6 , and Glu 6 (Fig. 3A andB) (subscript refers to location on blades 1-6 of the β-propeller fold) (14,27,44). Both JPV-RBP and BeiV-RBP (HMC and rat origin) lack Asp 1 , where this residue is replaced with Ser or Glu, respectively. Interestingly, however, Asp 1 may not be essential, as observed in the interaction between 3′-sialylactose (3′-SL) and the RBP of mumps virus (MuV) RBP (35). Additionally, JPV-RBP and BeiV-RBP (HMC and rat origin) exhibit a variant hexapeptide motif (Asn-Arg-Lys-Ser-Cys-Ser), where an Arg is presented at the third position as opposed to a Lys. In both our JPV-RBP β+ and BeiV(HMC)-RBP β+ crystal structures, residues from both the sialidase and hexapeptide motifs adopt conformations that largely match those of NDV-RBP (Fig. 3A). Additionally, JPV-RBP β+ and BeiV-RBP β+ CedV, Cedar virus (6THB) (38); MeV (2ZB5) (30); LANV, Lanya virus (8K80) (39); CDV, Canine distemper virus (7ZNY) (40); NarV, Nariva virus (7ZM6) (31); and MosV, Mossman virus (7ZM5) (31). The Structural Homology Program (41) was utilized to calculate evolutionary distance matrices by pairwise superpo sition of RBP structures. The resultant matrices were used to plot an unrooted tree in PHYLIP (42).
BeiV(HMC)-RBP β+ is displayed in cartoon representation and colored as a rainbow from N-terminus (blue) to C-terminus (red). The remaining RBPs are displayed as cartoon putty, where RMSD with respect to the structure of BeiV(HMC)-RBP is represented by a color scale and the thickness of the chain, with blue/thin illustrating the least RMSD and red/thick the largest RMSD between equivalent Cα atoms of the two structures. Arcs are displayed to highlight the genusspecific groupings of the RBPs. The arcs of avula, respiro, and orthorubulaviruses are colored dark blue to reflect their sialic acid receptor usage. retain either identical or physicochemically similar residues associated with a secondary sialic acid binding site implicated in triggering fusion for NDV-RBPs (Fig. S5A) (45)(46)(47)(48). Further, this site is unobstructed by the C-terminal extension. However, despite the presence of these residues and consistent with previous findings (1,3), HEK 293T cells displaying either full-length (IV-TM-Stalk-head-hat) JPV-RBP, BeiV(HMC)-RBP, and BeiV(rat)-RBP lacked hemadsorptive and neuraminidase activities under the conditions tested, when compared to an NDV-RBP control (Fig. 3C) (49,50). Interestingly, this second sialic acid site has not been observed in all HN-type RBP structures (12,13,43,51), indicating that paramyxoviral sialic acid functionality is not dependent upon its exis tence.
## Extension of the homodimeric interface provides a structural basis for the absence of observed hemadsorption and neuraminidase activity
Two near-identical β-propeller head domains form dimers in the asymmetric units of JPV-RBP β+ (RMSD of ~0.2 Å over 435 Cα atoms) and BeiV(HMC)-RBP β+ (RMSD of ~0.7 Å over 527 Cα atoms). Similar to other paramyxovirus RBPs (31,33,37), the interfaces between JPV-RBP β+ and BeiV(HMC)-RBP β+ monomers are mediated through the interaction between blades β1 and β6 and occlude ~1,230 and ~1,310 Å 2 of solvent accessible surface area (52), respectively (Fig. 4). The angle of association between JPV-RBP β+ and BeiV(HMC)-RBP β+ monomers is ~65° and ~72°, respectively (Fig. 4), which is similar to the HN-type RBPs, with angles ranging from ~53° to 64° (31,33).
While the majority of the JPV-RBP β+ extended C-terminus (Gly585-Asn709) is disordered, the elongated C-terminus of BeiV(HMC)-RBP β+ is largely observable in the crystal (Lys576-Asn687) (Fig. 1C) and increases the extent of the homodimeric inter action to ~4,250 Å 2 . Similar to the RBP from NDV Ulster strain (NDV-RBP Ulster ) (Fig. S5B), C-terminal residues of BeiV(HMC)-RBP β+ occlude residues implicated in sialic acid recognition and hydrolysis in other HN RBPs (Fig. 5D andE), providing a structural basis for the absence of observed hemadsorptive and neuraminidase activity (Fig. 3C). The BeiV(HMC)-RBP β+ C-terminal extension exchanges a hat-like domain with the adjoining protomer of the homodimer (Fig. 5A), a process termed "domain-swapping, " where identical monomers switch domains with one another (53). Although reminiscent, the occlusion of the potential receptor-binding site in the BeiV(HMC)-RBP β+ homodimer differs from that observed in the avirulent NDV-RBP Ulster . Indeed, the hat-like domain of BeiV(HMC)-RBP β+ forms a structure that resembles a blade from the six-bladed β-propeller fold. However, instead of a hat-like domain, NDV-RBP Ulster bears a shorter, 45 amino acid C-terminal peptide (Fig. S1A andB), which extends along the outside of the β-propeller before forming an α-helix that is stabilized at the top of the dimeric interface through an interchain disulfide bond (Fig. S5B) (48). Following the α-helix, the chain terminates in the sialic acid binding site of the same protomer, providing a mechanism of autoinhibition (48). In contrast, the C-terminal extension of BeiV(HMC)-RBP β+ does not encode an interchain disulfide, and it forms the much more substantial hat-like domain that recognizes the putative sialic acid binding site of the adjacent protomer.
## The jeilongvirus hat-like domain
The BeiV(HMC)-RBP β+ hat-like domain is unique amongst all reported paramyxoviral RBP structures and consists of an α-helix, a β-sheet composed of four antiparallel β-strands, and a long anchoring arm, which buries into the putative sialic acid binding cavity (Fig. 5A). Residues Leu573-Gly577 of the C-terminus form a β-sheet with residues Met168-Leu172 of the N-terminus and are conformationally reinforced by a disulfide bond (Cys163-Cys575) (Fig. 5B). A long linker composed of residues Lys578-Glu604, which is observed to be glycosylated at position Asn579 on one chain in asymmetric unit (Fig. S7A), traverses the dimer interface. A well-conserved intra-domain disulfide bond Cys605-Cys648 secures the position of the hat-like domain (Fig. S1B; Fig. 5C). Several salt bridges and an extensive hydrogen bonding network form between the linker and residues of the C-terminus, N-terminus, and the pinnacles of loops β6L12 and β1L34, further comprising the interaction between the hat-like domain and the β-propeller (Fig. S6).
The hat-like domain of BeiV(HMC)-RBP β+ displaces loop β2L23 of the propeller (Fig. S3B andS5), enabling projection of the C-terminal hook of the hat-like domain into the top center of the β-propeller (Fig. 5D andE). This interaction is stabilized by hydrogen bonds and salt bridges formed with residues of loops β2L01, β3L01, and β3L23 (Fig. S6), as well as interactions between the main chain carbonyl groups and the positively charged surfaces of the RBP. For example, Arg685 of the hat-like domain forms a salt loop β2L23 interacts with the hat-like domain, is rendered as a surface (pink). The "hat-like" domain is shown in cartoon representation (gray) with residues involved in H-bonding and salt-bridge formation shown as sticks (residues identified using PDBePISA analysis). (F) The dimeric head domain of BeiV(HMC)-RBP β+ is rendered as a surface (pink). The linker of the C-terminal extension (gray), represented using a cartoon, traverses across the dimeric interface. Analysis using the PDBePISA server (52) reveals many residues of the C-terminal extension are involved in H-bonding and salt-bridge formation, including residues within this linker region. The main chain and side chains of relevant residues are shown as sticks.
bridge with Glu413 (4) , a residue associated with sialidase activity (Fig. 5D). Additionally, Arg430 (4) of the triarginyl motif forms a hydrogen bond with the main chain carbonyl of Leu684 of the "hook" shaped C-terminus.
Although not well ordered enough to model, electron density putatively correspond ing to residues C-terminal to Asn687 of BeiV(HMC)-RBP β+ of the hat-like domain is observed. Indeed, electron density observed between the BeiV(HMC)-RBP β+ hat-like domain of one protomer and the BeiV(HMC)-RBP β+ hat-like domain of its symmetry mate (Fig. S7B) provides evidence for further, albeit possibly weaker and more transient interactions between the C-terminus and the rest of the BeiV(HMC)-RBP β+ . In the case of JPV-RBP β+ , although the C-terminal extension was not ordered in the structure, a short portion of electron density potentially corresponding to the Cα backbone (Fig. S7C [53]) is bound within the putative sialic acid site. This density is proximal to Arg191 (1) , Glu408 (4) , Arg425 (4) , Arg501 ( 5) , Tyr529 (6) , Glu548 (6) , and Arg557, and Arg253 of the variant hexapeptide motif, and coincides with the electron density visible within the BeiV(HMC)-RBP β+ putative active site (Fig. S7D). Consistent with the low level of order and occupancy of residues in this region of the C-terminal extension, loop β2L23 of JPV-RBP β+ is not displaced as it is in the BeiV(HMC)-RBP β+ structure (Fig. S3B), and instead forms a well-ordered α-helix proximal to the putative sialic acid binding site, as observed in other paramyxoviral HN-type RBPs (27,29,30,32,35,36,38,43,54). Furthermore, the superposition of the hat-like domain from BeiV(HMC)-RBP β+ onto JPV-RBP β+ reveals that loop β2L23 would sterically obstruct the positioning of the ordered β-sheet of the hat-like domain.
While we do not observe ordered electron density for the majority of the C-terminal extension of JPV-RBP β+ in the crystal structure, an AlphaFold3 (55) prediction of the C-terminal extension of JPV-RBP, generated from a prediction of two copies of the full-length sequence, bears a striking resemblance to the BeiV(HMC)-RBP β+ hat-like domain (Fig. S8A). Indeed, despite the low sequence identity across the C-terminal extension (Fig. S1A andB), the AlphaFold3 prediction of the JPV-RBP hat-like domain aligns moderately well, with an RMSD of 2.2 Å (over 70 aligned Cα residues). Further more, the model exhibits a similar placement of the hat-like domain at the top central region of the six-bladed β-propeller to that observed in BeiV(HMC)-RBP β+ , suggestive of similar modes of interaction. A search with Foldseek (56) reveals that similarly, the BeiV(HMC)-RBP β+ hat-like domain exhibits the greatest expected structural similarity to C-terminal regions of other jeilongvirus receptor-binding proteins, as predicted by Colabfold (57) in the Big Fantastic Virus Database (58). Additionally, this search revealed a lowconfidence similarity with an uncharacterized monkey poxvirus protein, B15L. Furthermore, the hat-like domain of BeiV(HMC)-RBP β+ had numerous lowconfidence matches with the ATP-binding N-terminal lobe of protein kinase domain-containing (PKDC) proteins from diverse species (59). For example, the BeiV(HMC)-RBP β+ hat-like domain has the same topology as the N-terminal lobe of a PKDC protein from Glycine max (Fig. S8B andC). However, α-helix 1 of the hat-like domain differs in its position relative to the β-sheet, where the BeiV(HMC)-RBP β+ hat-like domain α-helix rests on the concave side of the β-sheet. While interesting, the functional relevance of similarity to PKDCs is likely limited as the hat-like domain of BeiV(HMC)-RBP β+ lacks the GxGxxG motif (ATP-binding loop) and the b3 Lys present in PKDC proteins (59). Thus, although it is possible that BeiV(HMC)-RBP β+ may exhibit some ATP-binding functionality, as has been shown previously for pseudokinases that have degraded canonical kinase ATP-binding motifs, the RBP is unlikely to exhibit kinase or ATP-catalysis functionality. While the evolutionary basis for how the hat-like domain became a part of the paramyxovirus RBP repertoire remains a mystery, these findings provide new insights into the possible relationship of the RBP with ancestral proteins.
## DISCUSSION
Jeilongviruses constitute a group of emerging and poorly understood paramyxoviruses. Here, we provide molecular-level insights into the unique architecture of the jeilongvirus RBP, a cellular and species-tropism-dictating protein. Our molecular-level analysis of JPV-RBP β+ and BeiV(HMC)-RBP β+ clarifies the structural relationship of these jeilongviral proteins with other paramyxoviral RBPs. Indeed, consistent with genetic analysis and the near-conservation of residues required for sialic acid binding and hydrolysis, both at the primary (14, 15) (Fig. 3A andB) and putative secondary (45)(46)(47) binding sites (Fig. S5A), we find that both JPV-RBP β+ and BeiV(HMC)-RBP β+ bear the closest structural relation ship with HN-type RBPs, including respiroviruses, orthoavulaviruses, and rubulaviruses (Fig. 2A andB).
Intriguingly, we were unable to observe sialic acid binding and hydrolysis activity for these glycoproteins under the conditions tested (Fig. 3C). Our structures provide a molecular-level rationale for the absence of this observed activity. Indeed, we find that the elongated C-terminus of the jeilongvirus RBP (Fig. S1A andB), which may have possibly arisen from gene duplication of a β-propeller blade or acquisition of a host protein (Fig. S8B andC), extends toward the membrane distal region of the adjacent protomer of the β-propeller homodimer and forms a unique hat-like domain that sterically impedes the putative sialic acid-binding site (Fig. 5). The presence of the unique C-terminus, combined with the high level of sialidase and hexapeptide motif conservation (Fig. 3A andB), suggests that jeilongviral RBPs are likely able to regulate sialic acid binding and hydrolysis activity. Furthermore, the observation that the hat-like domain is differentially ordered in the crystal structures of JPV-RBP β+ and BeiV(HMC)-RBP β+ supports that this region of the protein may regulate activity by only being transiently associated with the sialic acid-binding site (Fig. 1; Fig. S7C andD). While the cell type, host-cell factors, or virus replication stage that may modulate this putative autoinhibitory activity remain unknown, such a hypothesis is consistent with our observation that the polypeptide corresponding to the otherwise disordered C-terminus of JPV-RBP β+ appears to occupy the putative sialic acid-binding site (Fig. S7C). As it is likely that our hemadsorption and neuraminidase assays do not replicate the environment relevant to native rodent host-cell infection, further studies are required to fully dissect the restriction factors and conditions necessary to reconstruct jeilongvirus RBP activity. To assess whether removal of the C-terminal region hat-like domain may regulate jeilongvirus RBP sialic acid-associated activity (e.g., via proteolytic cleavage by an unknown protease), we attempted to produce constructs of BeiV-RBP β and JPV-RBP β that lacked the C-terminal hat-like domain. However, unlike our construct containing the C-terminal extension, our truncated constructs were heterogeneous in solution (JPV-RBP β ) or did not express to a level necessary for protein purification (BeiV-RBP β ), indicative that further construct screening is required to produce truncated jeilongvirus RBPs suitable for functional studies. Future interrogation of this hypothesis would benefit from assessment of jeilongvirus RBP maturation in a range of rodent cells and determina tion of the physicochemical conditions necessary to (e.g., pH and temperature) modulate RBP functionality, in vitro. We also demonstrate that the region of the jeilongvirus RBP that aligns both structurally (Fig. S5A) and by sequence to the position of the secon dary binding site in the RBP of NDV is accessible and not occluded by the C-terminal extension. Given a lack of observed hemadsorption activity, it appears that this region, which has also been implicated in fusion activation in NDV (46), also does not interact with sialic acid interactions under the conditions tested.
How obstruction of the putative sialic acid-binding site by the C-terminal hat-like domain could regulate the jeilongvirus RBP activity remains a mystery. In the case of the avirulent Newcastle disease virus Ulster strain RBP (NDV-RBP Ulster ), the C-terminal extension modulates pathogenicity (48, 60) (Fig. S5B). The extended C-terminus of NDV-RBP Ulster both occludes the receptor-binding site of its own protomer in the dimer and exhibits an inter-subunit disulfide bond that regulates HN activity and dimerization. While the C-terminal extensions of BeiV(HMC)-RBP and NDV-RBP Ulster bear no apparent relationship in sequence, length, or structure (Fig. 5A; Fig. S1A,B, and S5B), it is possi ble that, similar to NDV-RBP Ulster , jeilongvirus RBPs may also require proteolysis at the C-terminus to recognize and hydrolyze sialic acid (61). Although we did not identify a conserved proteolysis motif, such as that found in NDV-RBP Ulster (i.e., 572 Lys-Glu-Ala-Lys 575 ) (61), analysis with the EXPASY PeptideCutter tool (62) indicates that the JPV-RBP or BeiV(HMC)-RBP C-terminal extension may encode multiple potential cleavage sites for proteases, including trypsin, chymotrypsin, elastase, and thermolysin (63). Future investigations into whether the hat-like domain is shed via such a mechanism to establish sialic acid binding and hydrolysis activity would allow assessment of whether jeilongviruses utilize a universal approach toward autoinhibition, as well as whether such protease sensitivity is a restriction factor specific to the host rodent reservoir species. Moreover, the distinctiveness of the structure and length of the extended jeilongviral C-termini, combined with the observation that only a subset of NDV strains bear an elongated C-terminus on their RBP that modulates activity (Fig. 5A; Fig. S1A, B, andS5B) (64), suggests that the extended C-termini of jeilongviruses and avirulent NDV strains arose independently over the course of their evolution from common sialic acid-binding ancestral paramyxoviruses.
Emerging and re-emerging viruses continue to make a profound impact on human health and the economy (65,66). Although the economic, environmental, and biomedi cal importance of jeilongviruses remains unclear, this does not diminish the potential impact of these or related paramyxoviruses in the future. Furthermore, although much work is needed to understand the tropism and replication characteristics of jeilong viruses, the detailed understanding of RBP structure and function delivered by this study constitutes an essential building block that will help us decipher pathobiological determinants of interspecies transmission at a molecular level. Indeed, by increasing our understanding of the structural repertoires available to paramyxovirus RBPs, this work strengthens our preparedness for the potential emergence of these and other paramyxoviruses from native reservoirs into new species, including humans.
## MATERIALS AND METHODS
## Protein production
JPV-RBP β+ and BeiV(HMC)-RBP β+ constructs were generated from human codon-opti mized genes of the full-length RBPs (Genbank NC_007454.1 and YP_512253.1, respec tively). JPV-RBP β+ (Glu149-Asn709) and BeiV(HMC)-RBP β+ (Asn145-Pro1046) were cloned into the pHLsec mammalian expression vector, which encodes a C-terminal hexahisti dine tag (67). Proteins were produced by transient transfection of HEK 293T cells in the presence of 5 µM kifunensine (25), and secreted protein was harvested after 72 h incubation at 37°C, 5% CO 2 . Cell supernatant was exchanged into 10 mM Tris (pH 8.0), 150 mM NaCl, and concentrated using an ÄKTA Flux diafiltration system (Cytiva). Immobilized metalaffinity chromatography was then performed with a HisTrap HP (Cytiva) column and eluted using 250 mM imidazole. Endoglycosidase F1 (EndoF1) was used to cleave (10 µg/mg protein, 12 h, 21°C) any existing N-linked glycans at the di-N-acetylchitobiose core. Size-exclusion chromatography was performed in 10 mM Tris, pH 8.0, 150 mM NaCl buffer using a Superdex 200 10/30 column (Cytiva).
## Crystallization and X-ray diffraction data collection
JPV-RBP β+ crystals were grown using the nanoliter-scale sitting-drop vapordiffusion method at room temperature, using 100 nL protein (3.5 mg/mL) and 100 nL precipi tant (26). Crystals grew in 0.1 M carboxylic acid, 0.1 M tris/bicine pH 8.5, 6% sucrose, 0.2 M ammonium sulfate, 37.5% Morpheus (Molecular dimensions) precipitant mix 4, consisting of 25% vol/vol 2-methyl-2,4-pentanediol (MPD), 25% wt/vol PEG 1000 (P1k) and 25% wt/vol polyethylene glycol 3350 (PEG 3350). The crystal was immersed in a solution consisting of the precipitant with an additional 10% MPD, prior to cryo-cooling. X-ray diffraction data were collected at a wavelength of 0.9282 Å on beamline I04-1, Diamond Light Source (DLS), Didcot, UK. Reflections were indexed, integrated, and scaled using the xia2 package (68) (Table S1), and molecular replacement was performed using PHASER (69), the human parainfluenza virus 3 RBP (PDB: 1V21) (32) as a search model. COOT (70) was used for model building, Phenix (71) for refinement. Structure validation was performed using Molprobity (72).
BeiV(HMC)-RBP β+ crystals were grown using nanoliter-scale sitting-drop vapordiffu sion at room temperature, using 100 nL protein (3.2 mg/mL) and 100 nL precipitant (26). Crystals grew in 0.2 M potassium thiocyanate, 0.1 M sodium cacodylate, pH 6.5, 8% wt/vol poly-γ-glutamic acid polymer, 6% 1,5-diaminopentane di-HCl. Crystal cryo-cooling was performed in a solution consisting of reservoir solution with an additional 20% glycerol, prior to X-ray data collection, which was performed at a wavelength of 0.9282 Å on beamline I04-1, DLS. Reflections were indexed, integrated, and scaled using the xia2 package (68) (Table S2) and molecular replacement using PHASER (69) with the JPV-RBP β structure as a search model. COOT (70) was used for model building, Phenix (71) for refinement. Structure validation was performed using Molprobity (72). Density modification using Parrot (73) was performed to aid in the building of this structure.
## Hemadsorption and neuraminidase assays
Full-length (IV-TM-Stalk-head-hat) JPV-RBP (Met1-Asn709, Genbank: YP_338084.1) and BeiV(HMC)-RBP (Met1-Pro1046, Genbank: YP_512253.1), and BeiV(rat)-RBP (Met1-Glu734, Genbank: AOV81769.1) were cloned into the pHLsec vector with a C-terminal hexa-his tidine tag (67) and transfected with Lipofectamine 2000 (ThermoFisher, product no. 11668030) into HEK 293T cells. The hemadsorption method to determine sialic acid binding was adapted from Morrison and McGinnes (50). HEK 293T cell monolayers were washed with phosphatebuffered saline (PBS), pH 7.4 (with MgCl 2 and CaCl 2 ), 24 h following transfection, prior to incubation with 2% sheep blood (Thermo Scientific Oxoid, 12967755) at 4°C for 30 min. Cells were gently washed to remove unadsorbed erythrocytes, prior to lysing absorbed erythrocytes using 50 mM Tris, pH 7.4, 5 mM ethylenediaminetetraacetic acid, 150 mM NaCl, and 0.5% Nonidet P-40. Absorbance at 540 nm was measured using a CLARIOStar plate reader (BMG Labtech). Assays were normalized to the NDV-RBP positive control and the relative level of protein cell surface expression, as assessed by enzyme-linked immunosorbent assay.
Neuraminidase activity was determined through hydrolysis of the substrate 2′-(4methylumbelliferyl)-α-d-N-acetylneuraminic acid (MU-Neu5Ac; Sigma-Aldrich, product no. M8639), as described previously (49,74). Twenty-four hours following transfection, HEK 293T monolayers were washed with PBS, pH 7.4, counted, and seeded in a 96-well black nontransparent plate at a density of 25,000 cells per mL. Cells were pelleted by spinning at 1,500 rpm for 5 min, and the supernatant was replaced with 0.1 M sodium acetate, pH 6.0 containing 1 mM MU-Neu5Ac. The plate was incubated for 1 h at 37°C prior to the addition of 0.25 M glycine buffer, pH 10.7 to stop the reaction. The amount of free 4-methylumbelliferone was fluorometrically determined at 365 nm for excitation and 450 nm for emission using a CLARIOStar plate reader (BMG Labtech).
## Structure-based phylogenetic analysis
To perform structural phylogenetic analysis of the paramyxoviral RBP β-propellers, water molecules, ligands, and protein residues outside of the canonical fold were removed from the six-bladed β-propeller monomers. Pairwise distances between RBP structures were calculated using the Structural Homology Program (41,75,76). In PHYLIP, pairwise evolutionary distance matrices were used to generate an unrooted phylogenetic tree (42).
## References
1. "The Center for Human Genetics was supported by Wellcome Center"
2. Jun, Karabatsos, Johnson (1977) "A new mouse paramyxovirus (J virus)" *Aust J Exp Biol Med Sci*
3. Mesina, Campbell, Glazebrook et al. (1974) "The pathology of feral rodents in North Queensland" *Tropenmed Parasitol*
4. Jack, Boyle, Eaton et al. (2005) "The complete genome sequence of J virus reveals a unique genome structure in the family Paramyxoviridae" *J Virol*
5. Woo, Lau, Wong et al. (2012) "Novel variant of Beilong paramyxovirus in rats" *China. Emerg Infect Dis*
6. Schomacker, Collins, Schmidt (2004) "In silico identification of a putative new paramyxovirus related to the Henipavirus genus" *Virology (Auckl)*
7. Liang, Zhang, Zhou et al. (2003) "AngRem104, an angiotensin IIinduced novel upregulated gene in human mesangial cells, is potentially involved in the regulation of fibronectin expression" *J Am Soc Nephrol*
8. Basler, García-Sastre, Palese (2005) "A novel paramyxovirus?" *Emerg Infect Dis*
9. Li, Yu, Zhang et al. (2006) "Beilong virus, a novel paramyxovirus with the largest genome of non-segmented negative-stranded RNA viruses" *Virology (Auckl)*
10. (2024) "International Committee on the Taxonomy of Viruses (ICTV)"
11. Ch'ng, Low, Borthwick et al. (2023) "Evolution and ecology of Jeilongvirus among wild rodents and shrews in Singapore" *One Health Outlook*
12. Bowden, Crispin, Jones et al. (2010) "Shared paramyxoviral glycoprotein architecture is adapted for diverse attachment strategies" *Biochem Soc Trans*
13. Yuan, Leser, Demeler et al. (2008) "Domain architecture and oligomerization properties of the paramyxovirus PIV 5 hemagglutinin-neuraminidase (HN) protein" *Virology (Auckl)*
14. Yuan, Swanson, Leser et al. (2011) "Structure of the Newcastle disease virus hemagglutininneuraminidase (HN) ectodomain reveals a four-helix bundle stalk" *Proc Natl Acad Sci*
15. Langedijk, Daus, Van Oirschot (1997) "Sequence and structure alignment of Paramyxoviridae attachment proteins and discovery of enzymatic activity for a morbillivirus hemagglutinin" *J Virol*
16. Mirza, Deng, Iorio (1994) "Site-directed mutagenesis of a conserved hexapeptide in the paramyxovirus hemagglutininneuraminidase glycoprotein: effects on antigenic structure and function" *J Virol*
17. Ictv (2024) "Subfamily: Orthoparamyxovirinae, genus: Jeilongvirus"
18. Abraham, Arroyo-Diaz, Li et al. (2018) "Role of small hydrophobic protein of J paramyxovirus in virulence" *J Virol*
19. Li, Hung, Paterson et al. (2015) "Type II integral membrane protein, TM of J paramyxovirus promotes cell-to-cell fusion" *Proc Natl Acad Sci*
20. Vanmechelen, Bletsa, Laenen et al. (2018) "Discovery and genome characterization of three new Jeilongviruses, a lineage of paramyxovi ruses characterized by their unique membrane proteins" *BMC Genomics*
21. Woo, Wong, Wong et al. (2016) "Comparative genome and evolutionary analysis of naturally occurring Beilong virus in brown and black rats" *Infect Genet Evol*
22. Dosztányi (2018) "Prediction of protein disorder based on IUPred" *Protein Sci*
23. Bowden, Jones, Stuart (2011) "Cells under siege: viral glycopro tein interactions at the cell surface" *J Struct Biol*
24. Zeltina, Bowden, Lee (2016) "Emerging Paramyxoviruses: receptor tropism and zoonotic potential" *PLoS Pathog*
25. Thibault, Watkinson, Moreira-Soto et al. (2017) "Zoonotic potential of emerging paramyxoviruses: knowns and unknowns" *Adv Virus Res*
26. Chang, Crispin, Aricescu et al. (2007) "Glycoprotein structural genomics: solving the glycosylation problem" *Structure*
27. Walter, Diprose, Mayo et al. (2005) "A procedure for setting up high-throughput nanolitre crystallization experiments. Crystallization workflow for initial screening, automated storage, imaging and optimization" *Acta Crystallogr D Biol Crystallogr*
28. Crennell, Takimoto, Portner et al. (2000) "Crystal structure of the multifunctional paramyxovirus hemagglutinin-neuraminidase" *Nat Struct Biol*
29. Bowden, Crispin, Harvey et al. (2008) "Crystal structure and carbohydrate analysis of Nipah virus attachment glycoprotein: a template for antiviral and vaccine design" *J Virol*
30. Rissanen, Ahmed, Azarm et al. (2017) "Idiosyncratic Mòjiāng virus attachment glycoprotein directs a host-cell entry pathway distinct from genetically related henipaviruses" *Nat Commun*
31. Hashiguchi, Kajikawa, Maita et al. (2007) "Crystal structure of measles virus hemagglutinin provides insight into effective vaccines" *Proc Natl Acad Sci*
32. Stelfox, Oguntuyo, Rissanen et al. (2023) "Crystal structure and solution state of the C-terminal head region of the narmovirus receptor binding protein"
33. Lawrence, Borg, Streltsov et al. (2004) "Structure of the haemaggluti nin-neuraminidase from human parainfluenza virus type III" *J Mol Biol*
34. Stelfox, Bowden (2019) "A structure-based rationale for sialic acid independent host-cell entry of Sosuga virus" *Proc Natl Acad Sci*
35. Welch, Yuan, Bose et al. (2013) "Structure of the parainfluenza virus 5 (PIV5) hemagglutinin-neuramini dase (HN) ectodomain" *PLoS Pathog*
36. Kubota, Takeuchi, Watanabe et al. (2016) "Trisaccharide containing α2,3-linked sialic acid is a receptor for mumps virus" *Proc Natl Acad Sci*
37. Lee, Pernet, Ahmed et al. (2015) "Molecular recognition of human ephrinB2 cell surface receptor by an emergent African henipavirus" *Proc Natl Acad Sci*
38. Bowden, Crispin, Harvey et al. (2010) "Dimeric architecture of the Hendra virus attachment glycoprotein: evidence for a conserved mode of assembly" *J Virol*
39. Pryce, Azarm, Rissanen et al. (2020) "A key region of molecular specificity orchestrates unique ephrin-B1 utilization by Cedar virus" *Life Sci Alliance*
40. Wang, Li, Wang et al. (2024) "Structural insights into the Langya virus attachment glycoprotein" *Structure*
41. Kalbermatter, Jeckelmann, Wyss et al. (2023) "Structure and supramolecular organization of the canine distemper virus attachment glycoprotein" *Proc Natl Acad Sci*
42. Stuart, Levine, Muirhead et al. (1979) "Crystal structure of cat muscle pyruvate kinase at a resolution of 2.6 A" *J Mol Biol*
43. Felsenstein (1989) "PHYLIP -phylogeny inference package (Version 3.2)" *Cladistics*
44. Yuan, Thompson, Wurzburg et al. (2005) "Structural studies of the parainfluenza virus 5 hemagglutininneuraminidase tetramer in complex with its receptor, sialyllactose" *Structure*
45. Iorio, Field, Sauvron et al. (2001) "Structural and functional relationship between the receptor recognition and neuraminidase activities of the Newcastle disease virus hemagglutinin-neuraminidase protein: receptor recognition is dependent on neuraminidase activity" *J Virol*
46. Zaitsev, Von Itzstein, Groves et al. (2004) "Second sialic acid binding site in Newcastle disease virus hemagglutinin-neuraminidase: implications for fusion" *J Virol*
47. Porotto, Salah, Devito et al. (2012) "The second receptor binding site of the globular head of the Newcastle disease virus hemagglutinin-neuraminidase activates the stalk of multiple paramyxovirus receptor binding proteins to trigger fusion" *J Virol*
48. Porotto, Fornabaio, Kellogg et al. (2007) "A second receptor binding site on human parainfluenza virus type 3 hemaggluti nin-neuraminidase contributes to activation of the fusion mechanism" *J Virol*
49. Yuan, Paterson, Leser et al. (2012) "Structure of the ulster strain newcastle disease virus hemagglutinin-neuraminidase reveals auto-inhibitory interactions associated with low virulence" *PLOS Pathog*
50. Potier, Mameli, Bélisle et al. (1979) "Fluorometric assay of neuraminidase with a sodium (4-methylumbelliferyl-α-D-Nacetylneuraminate) substrate" *Anal Biochem*
51. Morrison, Mcginnes (1989) "Avian cells expressing the newcastle disease virus hemagglutinin-neuraminidase protein are resistant to newcastle disease virus infection" *Virology (Auckl)*
52. Streltsov, Pilling, Barrett et al. (2015) "Catalytic mechanism and novel receptor binding sites of human parainfluenza virus type 3 hemagglutinin-neuraminidase (hPIV3 HN)" *Antiviral Res*
53. Krissinel, Henrick (2007) "Inference of macromolecular assemblies from crystalline state" *J Mol Biol*
54. Bennett, Choe, Eisenberg (1994) "Domain swapping: entangling alliances between proteins" *Proc Natl Acad Sci*
55. Bowden, Aricescu, Gilbert et al. (2008) "Structural basis of Nipah and Hendra virus attachment to their cell-surface receptor ephrin-B2" *Nat Struct Mol Biol*
56. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
57. Van Kempen, Kim, Tumescheit et al. (2024) "Fast and accurate protein structure search with Foldseek" *Nat Biotechnol*
58. Mirdita, Schütze, Moriwaki et al. (2022) "ColabFold: making protein folding accessible to all" *Nat Methods*
59. Kim, Karin, Mirdita et al. (2025) "BFVD-a large repository of predicted viral protein structures" *Nucleic Acids Res*
60. Sheetz, Lemmon (2022) "Looking lively: emerging principles of pseudokinase signaling" *Trends Biochem Sci*
61. Nagai, Klenk (1977) "Activation of precursors to both glycoproteins of Newcastle disease virus by proteolytic cleavage" *Virology (Auckl)*
62. Gorman, Nestorowicz, Mitchell et al. (1988) "Characterization of the sites of proteolytic activation of Newcastle disease virus membrane glycoprotein precursors" *J Biol Chem*
63. Gasteiger, Hoogland, Gattiker et al. (2005) "Protein identification and analysis tools on the expasy server"
64. Famutimi, Adebiyi, Akinmolu et al. (2024) "Trypsin, chymotrypsin and elastase in health and disease" *Futur J Pharm Sci*
65. Sakaguchi, Toyoda, Gotoh et al. (1989) "Newcastle disease virus evolution: I. Multiple lineages defined by sequence variability of the hemagglutinin-neuraminidase gene" *Virology (Auckl)*
66. Baker, Mahmud, Miller et al. (2022) "Infectious disease in an era of global change" *Nat Rev Microbiol*
67. Jones, Patel, Levy et al. (2008) "Global trends in emerging infectious diseases" *Nature*
68. Aricescu, Lu, Jones (2006) "A time-and costefficient system for high-level protein production in mammalian cells" *Acta Crystallogr D Biol Crystallogr*
69. Winter, Lobley, Prince (2013) "Decision making in xia2" *Acta Crystallographica Section D*
70. Mccoy, Grosse-Kunstleve, Adams et al. (2007) "Phaser crystallographic software" *J Appl Crystallogr*
71. Emsley, Lohkamp, Scott et al. (2010) "Features and development of Coot" *Acta Crystallogr D Biol Crystallogr*
72. Liebschner, Afonine, Baker et al. (2019) "Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix" *Acta Crystallogr D Struct Biol*
73. Davis, Leaver-Fay, Chen et al. (2007) "MolPro bity: all-atom contacts and structure validation for proteins and nucleic acids" *Nucleic Acids Res*
74. Cowtan (2010) "Recent developments in classical density modification" *Acta Crystallogr D Biol Crystallogr*
75. Connaris, Takimoto, Russell et al. (2002) "Probing the sialic acid binding site of the hemagglutininneuraminidase of Newcastle disease virus: identification of key amino acids involved in cell binding, catalysis, and fusion" *J Virol*
76. Bamford, Grimes, Stuart (2005) "What does structure tell us about virus evolution?" *Curr Opin Struct Biol*
77. Ng, Stelfox, Bowden (2020) "Unraveling virus relationships by structure-based phylogenetic classification" *Virus Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12721777&blobtype=pdf | # David Baltimore (1938-2025): architect of the modern molecular approach to conquer leukemia
Ruibao Ren, Tao Cheng
## Abstract
This is an open-access article distributed under the terms of the Creative Commons Attribution-Non Commercial-No Derivatives License 4.0 (CCBY-NC-ND), where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.
## 1. A LEGACY THAT TRANSCENDS TIME
The world of science mourns the passing of David Baltimore, Nobel laureate and pioneering biologist, who died on September 6, 2025, at the age of 87. While Baltimore's innovative discoveries fundamentally reshaped virology, oncology, immunology, and modern molecular biology, and his revolutionary work on reverse transcriptase earned him the highest honor in science, his laboratory's contributions to understanding and conquering leukemia represent an equally profound aspect of his scientific legacy. As we reflect on his monumental career, the blood cancer research community expresses deep gratitude for his fundamental discoveries that provided the essential molecular toolkit for understanding leukemogenesis, ultimately paving the way for targeted therapies that have saved countless lives.
Baltimore's journey in science was characterized by extraordinary insight and relentless curiosity. His research not only altered our fundamental understanding of biological processes but also generated the precise knowledge necessary for therapeutic interventions in hematologic malignancies. This editorial pays tribute to his specific, foundational contributions to leukemia research, which continue to resonate in laboratories and clinics worldwide.
## 2. LAYING THE GROUNDWORK: FROM VIRAL ONCOGENES TO CELLULAR PROTOTYPES
Baltimore's groundbreaking Nobel Prize-winning work in 1970 demonstrated that genetic information could flow from RNA to DNA, contrary to the previously unidirectional central dogma of molecular biology. 1 This discovery of reverse transcriptase fundamentally changed how scientists approached cancer research, particularly regarding retroviral involvement in oncogenesis. It was within this context that Baltimore's lab made its first decisive contributions to oncology.
His laboratory isolated and cloned the v-abl oncogene from the Abelson murine leukemia virus, which caused leukemia in mice. 2,3 This critical work directly led to the identification and cloning of its cellular counterpart, the proto-oncogene c-abl 4 (later renamed as ABL1). In a seminal achievement, Baltimore's team was the first to demonstrate that the Abl protein possessed tyrosine kinase activity. 5 This discovery was revolutionary; it identified Abl as one of the earliest known tyrosine kinases and established a direct mechanistic link between enzymatic activity and the dysregulated cell growth that characterizes cancer.
## 3. CONNECTING THE DOTS: FROM ABL1 TO THE PHILADELPHIA CHROMOSOME
The profound importance of this foundational work became clear when other laboratories discovered that the Philadelphia chromosome, the hallmark of chronic myeloid leukemia (CML), was formed by a translocation between chromosomes 9 and 22. This translocation resulted in the creation of a novel fusion gene, BCR/ABL1. Because Baltimore's lab had already characterized the normal ABL1 gene and its kinase function, the scientific community was immediately equipped to understand the sinister nature of this fusion. The BCR/ABL1 oncoprotein was quickly recognized as a constitutively active tyrosine kinase that perpetually signals cells to proliferate, effectively hijacking the normal regulatory mechanisms of blood cell production. Baltimore's prior work provided the essential reference point that allowed researchers to decipher this oncogenic mechanism with remarkable speed.
## 4. DEEPENING THE UNDERSTANDING: PATHOGENESIS, SIGNALING, AND NEW DOMAINS
Rather than resting on these foundational contributions, Baltimore's laboratory actively switched to studying the BCR/ ABL1 pathogenesis itself. His group developed critical experimental models, including mouse models for CML, 6 which allowed researchers to study the disease's progression and test new therapeutic strategies in a living system. This deep into the mechanics of the oncoprotein led to further landmark discoveries. Through meticulous study of how BCR/ABL1 transmits signals, Baltimore's lab made two pivotal contributions to all of cell signaling: the discovery of the SH3 domain binding site 7 and the PH domain. 8 These domains are crucial for protein-protein interactions and membrane targeting, respectively. Their identification in the context of ABL1 research provided fundamental insights into how signaling networks are organized within cells, insights that have resonated far beyond cancer biology into immunology, neurobiology, and developmental biology.
## 5. FROM FUNDAMENTAL DISCOVERY TO THERAPEUTIC REVOLUTION
The direct line from Baltimore's basic science to the clinic is one of the most compelling stories in modern medicine. The characterization of ABL1 as a tyrosine kinase made it a "druggable" target. The understanding that BCR/ABL1 was an active version of this kinase directly spurred the development of imatinib (Gleevec), the paradigm-shifting targeted therapy that revolutionized CML treatment.
Baltimore's laboratory did not develop imatinib, but their work created the intellectual framework that made its conception possible. Furthermore, their subsequent research into BCR/ ABL1 signaling pathways and resistance mechanisms provided critical knowledge that helped inform the development of secondand third-generation tyrosine kinase inhibitors (TKIs) when resistance emerged.
## 6. ENDURING LEGACY: THE BALTIMORE IMPRINT ON HEMATOLOGY
David Baltimore's contributions to conquering leukemia exemplify how profound, curiosity-driven basic research provides the essential foundation for clinical revolutions. He embodied a mode of scientific inquiry that connected fundamental biological mechanisms to human disease. His study on the ABL1 gene did not just describe a molecule; it provided the language and the tools to understand, target, and defeat a specific cancer.
The development of tyrosine kinase inhibitors-which turned CML from a fatal disease into a manageable condition for many patients-stands as a tribute to this approach. This achievement rests squarely on the foundational work carried out in Baltimore's laboratory.
## 7. CONCLUSION: HONORING A VISIONARY
David Baltimore's passing represents the loss of a singular figure in modern science. For the hematology community, his legacy is both specific and profound: the meticulous cloning and biochemical characterization of ABL1 created the cornerstone upon which our understanding of CML was built. His later work unraveling the pathogenesis and signaling of BCR/ABL1 continued to illuminate the path forward.
As we reflect on Baltimore's monumental contributions, we recognize that the best way to honor his memory is to champion the same spirit of rigorous, fundamental, and creative science that he exemplified throughout his career. His study is a permanent reminder that investing in basic biological inquiry, pursued with excellence and vision, yields the most practical of human benefits: the alleviation of suffering and the extension of life.
"We must be willing to challenge dogma, to take risks, and to recognize that scientific understanding is a journey rather than a destination."-David Baltimore.
The editorial board of Blood Science joins the global scientific community in mourning David Baltimore's passing while celebrating his extraordinary contributions to hematology and to the conquest of leukemia.
## References
1. Baltimore (1970) "RNA-dependent DNA polymerase in virions of RNA tumour viruses" *Nature*
2. Shields, Goff, Paskind et al. (1979) "Structure of the Abelson murine leukemia virus genome" *Cell*
3. Witte, Rosenberg, Paskind et al. (1978) "Identification of an Abelson murine leukemia virus-encoded protein present in transformed fibroblast and lymphoid cells" *Proc Natl Acad Sci*
4. Wang, Ledley, Goff et al. (1984) "The mouse c-abl locus: molecular cloning and characterization" *Cell*
5. Witte, Dasgupta, Baltimore (1980) "Abelson murine leukaemia virus protein is phosphorylated in vitro to form phosphotyrosine" *Nature*
6. Daley, Van Etten, Baltimore (1991) "Blast crisis in a murine model of chronic myelogenous leukemia" *Proc Natl Acad Sci U S A*
7. Ren, Mayer, Cicchetti et al. (1993) "Identification of a tenamino acid proline-rich SH3 binding site" *Science*
8. Mayer, Ren, Clark et al. (1993) "A putative modular domain present in diverse signaling proteins" *Cell* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12779748&blobtype=pdf | # Handling editor Massimo Turina
Roland Zell, Marco Groth, Lukas Selinka, Hans-Christoph Selinka
## Abstract
Mycoviruses have been classified into 35 virus families so far. In addition to numerous mycoviruses with proven fungal or oomycetes hosts, many newly discovered viruses from environmental water and soil samples and various animal or plant specimens exhibit significant similarity to classified mycoviruses, thereby expanding the known sequence space of fungal and related viruses. In this study, we searched for mycoviruses in two environmental water samples that had been collected from the Teltow Canal and the Havel River in Berlin, Germany. Viral sequences with similarity to members of 16 virus families were identified. The most common viruses in our samples were botourmia-like viruses with moderate similarity to members of the genus Ourmiavirus. Notably, 58 of the ourmia-like sequences from the Teltow Canal and Havel River and 10 ourmia-like sequences from other sources exhibited a dicistronic genome layout. The second open reading frame (ORF) of these viruses encoded a putative capsid protein with an S domain that showed little similarity to the structural proteins of the classified ourmiaviruses. The second-largest virus group (59 sequences) was assigned to the order Ghabrivirales, and 13 of these sequences exhibited similarity to members of the suborder Alphatotivirineae (families Orthototiviridae, Pseudototiviridae, Botybirnaviridae, and Chrysoviridae). Thirtythree sequences clustered with members of the suborder Betatotivirineae -three of them with members of the family Artiviridae. Fifteen highly divergent toti-like sequences suggest the need to establish a new higher-order taxon within the order Ghabrivirales. Other virus sequences were assigned to the families Mitoviridae (three unuamitoviruses and 10 highly divergent mitovirus-like sequences), Narnaviridae (five "alphanarnavirus"-like sequences), Amalgaviridae (two zybavirus-like sequences), Hypoviridae (one partial RdRP sequence), and Mymonaviridae (one partial RdRP sequence), and one was not classified (Sclerophthora macrospora B-like virus). Notable results include a clade of highly divergent mitovirus-like sequences with a standard translation code, three narnavirus-like sequences with a reverse-frame ORF, and a clade of four Ghabrivirales-like replicase sequences that were found to have numerous termination codons regardless of which translation table was used.
## Introduction
Mycoviruses have been detected in a wide range of host species comprising all of the major taxa of fungi and oomycetes. They exhibit great diversity in their genome structure and replication mechanisms. Moreover, mycoviruses have mostly nonlytic life cycles, as they forgo extracellular routes of infection and make no use of vectors but instead establish persistent asymptomatic infections of their fungal hosts. Mycoviruses propagate by efficient vertical and horizontal transmission routes. Typical vertical transmission routes are cell division and sporogenesis; horizontal transmission includes cell fusion mechanisms such as hyphal anastomosis and heterokaryosis [1]. Although viral persistence is common, mycoviruses with icosahedral particles have been described. The complex interactions of most mycovirus-fungus-plant and mycovirus-fungus-animal host systems are still underexplored, as is the ecological role of viral hyperparasitism [2]. Viral infection of some fungi may cause hyper-or hypovirulence of their plantpathogenic hosts, with the latter feature making them potential viral agents for controlling virulent pathogenic fungi. Furthermore, some viruses can shuttle between fungi and plants or insects [3][4][5], and fungi may serve as vectors of plant viruses [6,7].
Aquatic systems are often overlooked as fungal habitats although fungi play important roles in cycling of organic matter and food web dynamics [8]. Moreover, mycoviruses are also important players in these interactions and in ecological control [9]. In addition to their complex ecological role, the great genetic diversity of mycoviruses has attracted the attention of researchers in recent years. Hundreds of mycoviruses belonging to at least 35 virus families have been classified (see Virus Metadata Resource MSL40.v1.20250307, available for download from h t t p s : / / i c t v . g l o b a l / ) , but thousands of mycovirus-like sequences still await classification. Mycoviral genomes may consist of single-stranded DNA (Genomoviridae), segmented or unsegmented double-stranded RNA or single-stranded RNA with positive-or negative-strand polarity. Furthermore, positive-stranded RNA viruses with reverse transcriptase activity have also been described (Metaviridae, Pseudoviridae). Other oddities in the mycovirus world include singlestranded RNA viruses with circular genomes and two open reading frames in ambisense orientation (viruses of the phylum Ambiviricota) and multisegmented positive-stranded RNA viruses with a composite RNA-dependent RNA polymerase (RdRP) complex assembled from two separate domains that are encoded by different RNA segments (viruses of the family Splipalmiviridae) [10]. One mycovirus with a double-stranded DNA genome, rhizidiovirus, has been reported, but no sequence data are available [11] and its significance is unclear.
The number of published mycovirus sequences has increased exponentially with the availability of highthroughput sequencing techniques, and virus taxonomists have to cope with the continuously growing amount of sequence data and its classification. Each new published fungal metagenome and each new sequenced virome of environmental samples has revealed novel mycoviruses or mycovirus-like sequences [10,12,13]. Often, fungi may contain many different viruses. For example, the facultative plant-pathogenic fungus Rhizoctonia solani -which is in fact a complex of many related species, so-called anastomosis groups -hosts as many as 100 different viruses belonging to 18 virus families [14]. Similarly, the oomycete Bremia lactucae hosts more than 15 viruses of various families [15]. The situation is complicated by cross-kingdom infections, e.g., if fungi serve as vectors for plant viruses, or if viruses shuttle between fungi and plants or between fungi and insects [3][4][5][6][7]16].
The Teltow Canal and Havel River are two waterbodies traversing the metropolitan area of Berlin, Germany. In our previous investigations of their fluvial viromes, we identified complete or partial genomes of more than 1500 novel RNA viruses, which were reported in a series of papers [17][18][19][20][21]. Here, we add viral sequences detected in samples from these waterbodies that are related to known mycoviruses. The novel viruses exhibit similarity to members of the suborders Alpha-and Betatotivirineae and the families Botourmiaviridae, Mitoviridae, Narnaviridae, and Amalgaviridae (genus Zybavirus), as well as to the unclassified Sclerophthora macrospora B virus.
## Materials and methods
Sample collection and methods of virus enrichment, RNA extraction, Illumina sequencing, sequence data processing, and sequence analysis have been described previously [21]. They are briefly summarized here:
The 50-L Teltow Canal water sample (ID MR233-17) was collected in Berlin, Bäkebrücke (site coordinates: 52°26′03″ N, 13°18′57″ E), on July 18, 2017. The 50-L Havel River water sample (ID MR644-17) was collected in Berlin, Heerstrasse (site coordinates: 52°30'46'' N 13°12'14'' E), on June 28, 2018. Both 50-L samples were divided into five 10-L aliquots and further treated as described by Wyn-Jones et al. [22]. For this, the aliquots were stirred for 20 minutes to suspend floating detritus, acidified to pH 3.5 by addition of hydrochloric acid, and loaded on a glass-wool-packed column for adsorption of virus particles. Adsorbed virus particles were consecutively washed with hydrochloric acid (1 M), sodium hydroxide (1 M), and tap water and finally eluted with 3% beef extract/0.05 M glycine buffer, pH 9.5. Eluates were neutralized by addition of sodium hydroxide and passed through a 0.45-µm filter to remove residual detritus and bacteria. Virus particles were sedimented by ultracentrifugation (100,000 × g, 2.5 h at 4 °C). Finally, the sediment was redissolved in 0.5 ml of phosphate-buffered saline by agitation in a ball mill and stored at -80°C for later RNA extraction.
RNA was extracted using a QIAamp Viral RNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instruction and analyzed by Illumina sequencing as described by Zell et al. [21]. Briefly, 100 ng of RNA from the Havel River sample was used for library preparation employing an Illumina TruSeq Stranded Total RNA Library 1 3
Preparation Kit, whereas the library from the Teltow Canal sample was prepared from 450 ng of RNA using an Illumina TruSeq Stranded Total RNA Library Preparation Kit combined with a Ribo-Zero Gold rRNA Removal Kit according the manufacturer's instructions. The protocol was adapted to include all RNA molecules rather than only polyadenylated RNA by adding an RNA precipitation step with isopropanol and redissolving the pellet in Fragment, Prime, Finish Mix (FPF). Thereafter, library quantitation and quality checking were done using a DNA 7500 kit and a 2100 Bioanalyzer instrument (Agilent Technologies, Waldbronn, Germany). Paired-end Illumina sequencing (2 × 150 bp) was done on a HiSeq 2500 platform using the rapid run mode.
Sequence data were extracted in FastQ format using bcl-2FastQ v2. 19.1.403 (Illumina). Adapter sequences were removed using Cutadapt v1.8.3 [23]. After duplicon extraction, the remaining 70,018,635 read pairs of the Teltow Canal sample and 51,902,006 read pairs of the Havel River sample were used for de novo assembly by two methods: (i) Employing clc_assembler v5.2.1 (QIAGEN) with the parameters -p fb ss 50 500, a total of 537,529 contigs longer than 200 nucleotides were obtained from the Teltow Canal sample, and 162,082 contigs were obtained from the Havel River sample. (ii) Employing the metaSPAdes assembler v3. 15.3 [24] with standard parameters (-k auto), 1,314,849 scaffolds were obtained from with the Teltow Canal sample, and 388,367 were obtained from the Havel River sample. The final sequences were obtained by manual curation, i.e., linking of overlapping contigs and scaffolds.
We used a 2-step procedure for analysis of sequence data. First, virus-specific scaffolds and contigs were identified by comparing the assembled sequences with our in-house virus protein database compiled from all NCBI GenBank entries with the Taxonomy ID 10239, using DIAMOND v2.0.10 [25]. DIAMOND identified 66,371 virus-specific scaffolds and 41,408 contigs from the Teltow Canal sample and 16,922 scaffolds and 8810 contigs from the Havel River sample. In the second step, selected DIAMOND hits were confirmed using BLAST+ v2.13.0 ( h t t p s : / / f t p . n c b i . n l m . n i h . g o v / b l a s t / e x e c u t a b l e s / b l a s t + / 2 . 1 3 . 0 /) with the BLASTp and tBLASTx tools. If appropriate, specific BLASTn searches were performed with reference sequences downloaded from GenBank. For identification of conserved protein domains, the NCBI web search tools BLASTp suite and Pfam conserved domain database (CDD) were used ( h t t p s : / / b l a s t . n c b i . n l m . n i h . g o v / B l a s t . c g i, h t t p s : / / w w w . n c b i . n l m . n i h . g o v / S t r u c t u r e / c d d / w r p s b . c g i). ClustalW and MUSCLE implemented in MEGA X [26] were employed for protein sequence alignments. Alignments were adjusted manually if necessary. For construction of phylogenetic trees, we used IQ-TREE 2.1.3 for Windows [27]. The best-fit substitution models were identified using the automatic model selection option (ModelFinder) implemented in IQ-TREE. Branch support was assessed using the ultrafast bootstrap approximation UFBoot2 with 10,000 replications [28].
All viruses were provisionally assigned to established taxa using the current virus taxonomy (2024 release) based on the most recent Master Species List #40 ( h t t p s : / / i c t v . g l o b a l / m s l ; accessed on March 7, 2025) and the most recent Virus Metadata Resource spreadsheet MSL40.v1.20250307 ( h t t p s : / / i c t v . g l o b a l / v m r / c u r r e n t; released on March 7, 2025).
## Results
Analysis of the Teltow Canal and the Havel River viromes revealed the presence of numerous RNA viruses with similarity to fungal viruses. Genomic sequences related to members of 15 virus families known to contain mycoviruses were detected (Table 1). These included -in alphabetical order Alphaflexiviridae, Amalgaviridae, Artiviridae, Barnaviridae, Botybirnaviridae, Botourmiaviridae, Chrysoviridae, Endornaviridae, Hypoviridae, Mitoviridae, Mymonaviridae, Narnaviridae, Orthototiviridae, Partitiviridae, and Pseudototiviridae. In addition, sequences related to the unclassified Sclerophthora macrospora viruses A and B and Plasmopara halstedii virus A, as well as many unclassified viruses of the orders Ghabrivirales and Cryppavirales were detected. Furthermore, viruses with similarity to fungusassociated aspivirus-like viruses were identified (Table 1). Notably, the majority of the novel virus sequences exhibited similarity to members of the family Botourmiaviridae (positive-strand RNA viruses) or the order Ghabrivirales (double-strand RNA viruses), but most of the new sequences were highly divergent compared to those of the established members. As Teltow Canal and Havel River viruses with similarity to alphaflexiviruses, barnaviruses, endornaviruses, and partitiviruses have been described previously [19], they were excluded here. No significant sequences corresponding to mycoviruses of the families Polymycoviridae (dsRNA), Deltaflexiviridae (+RNA), Fusariviridae (+RNA), Gammaflexiviridae (+RNA), Hadakaviridae (+RNA), Yadokariviridae (+RNA), Discoviridae (-RNA), Phenuiviridae (-RNA), Tulasviridae (-RNA), Metaviridae (RT viruses), or Pseudoviridae (RT viruses) were found.
The most characteristic features of the detected mycovirus-like sequences and the high variability in their genomic layouts (Fig. 1) are outlined in the following paragraphs:
## Botourmiaviridae
The majority of the sequences obtained in this study were related to botourmiaviruses. Viruses of the family Botourmiaviridae (order Ourlivirales) infect plants, fungi, and non-encapsidated monopartite, monocistronic positivestranded RNA genomes and infect fungi. The sizes of the monopartite genomes of the classified botourmiaviruses range from 2 to 5 kb. The gene segments of the tripartite ourmiaviruses are 2.8 kb, 1.1 kb, and 1 kb long. The RNA of fungus-infecting botourmiaviruses and RNA segment 1 of ourmiaviruses encode a replicase (cd23183), whereas RNA segment 2 of ourmiaviruses encodes a movement protein, and segment 3 encodes a coat protein with an S domain (pfam00729). The RdRP sequences of all botourmiaviruses cluster with those of leviviruses (now assigned to the family Fiersviridae), mitoviruses, and narnaviruses on branch 1 of the so-called 'comprehensive RdRP tree' [30]. The viruses of this branch have recently been assigned to the phylum Lenarviricota [31].
DIAMOND identified 121 sequences from the Teltow Canal sample and 33 from the Havel River sample encoding a botourmia-like RdRP (cd23183). These sequences were named TC-botourmia-LVs and Havel-botourmia-LVs. Notably, five clades contained dicistronic virus sequences. Altogether, 68 viruses exhibited a dicistronic genome layout with ORF 1 encoding the RdRP (cd23183) and ORF 2 encoding a putative capsid protein with an S domain (pfam00729). The genome of TC-botourmia-LV-17 has a putative third ORF with unknown significance. Another sequence of 2589 nt was identified by the CDD tool as a botourmia-like virus based on a short stretch of 120 amino acids representing a partial RdRP sequence (cd23183), although with low statistical significance (1.31e-12). This RdRP sequence was located on the C-terminal side of a helicase domain, which is unusual for botourmiaviruses.
The RdRP sequences of the botourmia-like viruses were aligned with the respective sequences of classified reference viruses, including a number of mitoviruses and narnaviruses as an outgroup. The phylogenetic tree reproduced three branches corresponding to the families Botourmiaviridae, Mitoviridae, and Narnaviridae (Supplementary Fig. S1). None of the Teltow Canal or Havel River botourmia-like viruses clustered with members of any of the fungus-infecting viruses but were related to plant-infecting ourmiaviruses.
To examine the sequences of the structural proteins, we aligned these sequences with the most similar CP sequences possibly oomycetes. The family comprises 12 genera, one of which is the genus Ourmiavirus, whose members have a positive-stranded RNA genome with three monocistronic segments, produce non-enveloped bacilliform virions, and infect plants [29]. The viruses of the remaining genera have translation table is used. The significance of this finding remains to be elucidated.
In addition, DIAMOND identified three scaffolds representing alphachrysoviruses. One 278-nt scaffold corresponded to segment 1 (RdRP) and exhibited 52% aa sequence identity to Brassica campestris chrysovirus 1 (accession no. KP782031). However, this fragment was too short for a proper phylogenetic analysis. Two other scaffolds with lengths of 671 and 546 nt, respectively, showed about 30% aa sequence identity to the major capsid proteins of two different alphachrysoviruses (data not shown). No Ghabrivirales sequences with significant similarity to those of members of the Alternaviridae, Fusagraviridae, Giardiaviridae, Inseviridae, Lebotiviridae, Megabirnaviridae, Megatotiviridae, Monocitiviridae, Ootiviridae, Phlegiviridae, Pistolviridae, Quadriviridae, Spiciviridae, or Yadonushiviridae were found.
## Narnaviridae and Mitoviridae
Narnaviruses and mitoviruses possess a non-encapsidated positive-stranded RNA genome ranging from 2.1 to 5 kb in length. Its single ORF encodes an RdRP. A recent taxonomic revision removed mitoviruses from the family Narnaviridae and placed them into the newly created family Mitoviridae based on the low degree of similarity of their replicase sequences. The family Narnaviridae (order Wolframvirales) is comprised of a single genus with two species. In contrast, the family Mitoviridae (order Cryppavirales) includes four genera with 105 species [33]. Numerous related viruses from a wide spectrum of hosts await classification. Whereas the two acknowledged narnaviruses infect yeast cells, most mitoviruses use filamentous fungi as hosts and utilize translation table 4, in which UGA encodes tryptophan, indicating persistence in fungal mitochondria [34]. Only a few mitoviruses have been associated with plant hosts [35,36], and those viruses use the standard genetic code.
By comparing RdRP sequences, we identified five sequences with similarity to those of narnaviruses (cd23177) and 13 with similarity to those of mitoviruses (pfam05919). identified using BLAST, but this failed to yield a significant alignment with the ourmiavirus CP sequences. A CDD search of the polymerase sequences linked to the CP sequences shown in the phylogenetic tree in Fig. 2 revealed six protein families, i.e. cd23183 (Botourmiaviridae-RdRP, ▲), cd23173 (Nodaviridae-RdRP, ♦), cd23179 (Tolivirales-RdRP, ◼), cd23180 (Solemoviridae-RdRP, ○), cd23206 (Tombusviridae-RdRP, ◻) and cd23242 (Gammacarmovirus-RdRP, ◻). This finding might suggest that recombination events occurred frequently in the evolution of these RNA viruses.
## Ghabrivirales
The taxonomy of the order Ghabrivirales is based on the polymerase phylogeny, resulting in three suborders, Alpha-, Beta-and Gammatotivirineae, which together include 19 virus families [31]. Monopartite and multipartite doublestranded RNA genomes with variable gene layouts have been described in this order. The polymerase belongs to the protein family pfam02123 (RdRP4). Where structural data are available, the members of the Ghabrivirales have been shown to possess an icosahedral capsid with a size of 35-50 nm. The capsid protein gag of L-A virus (species Totivirus ichi, family Orthototiviridae) which is the best-studied member of the order Ghabrivirales, belongs to protein family pfam09220 and exhibits similarity to the reovirus capsid protein regarding protein folding and arrangement in a T=1/ pseudo T=2 capsid symmetry [32].
We investigated 59 sequences from the Teltow Canal and Havel River samples with similarity to members of the order Ghabrivirales. Our alignment of RdRP sequences includes reference sequences of all 19 Ghabrivirales families plus several unclassified viruses. The resulting tree supports the present taxonomy (Supplementary Fig. S2). Three branches correspond to the three suborders. All 19 families exhibit long branch lengths. Within the suborder Alphatotivirineae, two new orthototiviruses, four new pseudototiviruses, and four new botybirna-like viruses were identified. Noteably, our orthototi-and pseudototiviruses possess overlapping stop/start codons (UAAUG) at the ORF1-ORF2 junction, which would enable a coupled termination-reinitiation mechanism for the expression of the capsid protein and RdRP rather than a programmed ribosomal shift. Furthermore, 33 toti-like viruses, including three artiviruses, grouped with members of the suborder Betatotivirineae. Fifteen toti-like viruses, which clustered distinctly from the three known suborders, were also observed. This clade of divergent viruses suggests the need for another higher-order taxon within the order Ghabrivirales. Among these divergent viruses, one subclade has four RdRP-like sequences that contain many internal stop codons regardless of which Fig. 2 Phylogenetic analysis of 80 capsid protein sequences of members of dicistronic unclassified botourmia-like viruses. The CP sequences of members of five clades of botourmia-like viruses from the Teltow Canal and Havel River (listed in Supplementary Figure S1) were aligned with CP sequences from various hosts and used for construction of a maximum-likelihood tree, using IQ-TREE 2 (optimal substitution model: Q.pfam+F+R5). Presented are GenBank accession numbers, vernacular virus names, and strain designations, if available (in parentheses). Numbers at nodes indicate bootstrap support greater 65% obtained after 10,000 ultrafast replications. The bar indicates amino acid substitutions per site. Color code: blue, unclassified viruses; red, viruses from the Teltow Canal and Havel River. Symbols (♦, ◼, ○, ▲, ◻, •, ♢) indicate the CDD family of the corresponding RdRP 1 3 zyba-like virus (accession no. KX525322), the fusion protein lacks fewer than five amino acids at both the N-and C-terminal ends. The sequence of TC-zyba-like virus 2 corresponds to a partial RdRP. A phylogenetic analysis based on the RdRP sequences of all members of the Amalgaviridae plus several related but unclassified virus sequences, including both TC-zyba-like sequences, revealed three major clades corresponding to the three amalgavirus genera plus two clades with zyba-like virus sequences. One of these clades includes fungal viruses, while the other is comprised of viruses that are associated with non-fungal hosts (Fig. 3).
## Sclerophthora macrospora viruses A and B, Plasmopara halstedii virus A
The oomycete Sclerophthora macrospora is an important pathogen that causes downy mildew disease of cereal crop plants of the family Poaceae. Two viruses infecting S. macrospora have been isolated so far: Sclerophthora macrospora viruses A and B (SmVA and SmVB). A third oomycete-infecting virus is Plasmopara halstedii virus A (PhVA). These three viruses have icosahedral capsids with sizes of 32 nm (SmVA), 35 nm (SmVB), and 37 nm (PhVA), respectively [42][43][44]. SmVA and PhVA have segmented single-stranded RNA genomes with RNA segment 1 encoding noda-like methyltransferase (pfam19222) and RdRP (cd23173) domains, whereas RNA segment 2 codes for a capsid protein with similarity to tombusviruses (pfam00729) [45,46]. In addition, SmVA possesses a third satellite-like RNA. In contrast, the genome of SmVB consists of a monopartite, dicistronic single-stranded RNA (Fig. 1). The polyprotein of SmVB that is encoded by ORF 1 includes a V8-like Glu-specific endopeptidase, several VPg oligopeptides, and a polymerase that is related to the RdRPs of solemoviruses [47]. The second ORF codes for a capsid protein.
Our phylogenetic analysis of the RdRP sequences of 85 viruses, including SmVA, SmVB, PhVA, solemoviruses, tombusviruses, nodaviruses, and a number of unclassified viruses, revealed six clades of unclassified viruses that clustered on the branch with the Solemoviridae members Phylogenetic analysis of these RdRP sequences revealed that all of the narna-like viruses were novel and clustered with viruses that were recently proposed to be classified as "alphanarnaviruses" [37] (Supplementary Fig. S3). Notably, three viruses of the "alphanarnavirus" clade (TC-narna-LV-1 and Havel-narna-LV-3 and -4) contained a reverse-frame ORF (rORF, also known as an ambigrammatic sequence), as has been described previously for some narna-like viruses [37][38][39][40].
The Mitoviridae branch contains monophyletic clades of the four mitovirus genera, albeit with moderate bootstrap support (Supplementary Fig. S3). Three of our mito-like sequences clustered with unuamitoviruses (Havel-mito-LV-4, Havel-mito-LV-5, and TC-mito-LV-8). The remaining mito-like sequences from the Havel River and Teltow Canal were too divergent to assign them to an existing mitovirus genus. Two clades of the tree contain viruses which utilize the standard translation table rather than translation table 4. One of these clades comprises nine plant-infecting mitoviruses of the genus Duamitovirus. The other clade consists of a group of eight novel virus sequences from the Teltow Canal and Havel River plus six still unclassified "narnalike" viruses that have been proposed recently to be members of the order Cryppavirales [33]. These six "narna-like" viruses were associated with hosts as diverse as plants, insects, tunicates, penguins, and rabbits. The use of the standard translation table is compatible with the reported nonfungal hosts and also suggests non-fungal hosts for the eight Teltow Canal and Havel River viruses of this clade. Another novel, untypeable virus is Havel-mito-like virus 3, which also lacks internal UGA codons, suggesting the use of the standard genetic code.
## Amalgaviridae
The family Amalgaviridae (order Durnavirales) consists of three genera: Amalgavirus, with plant-infecting viruses; Unirnavirus, with viruses of filamentous ascomycetes; and Zybavirus, with viruses that infect budding yeast. Members of the Amalgaviridae have double-stranded RNA genomes with two open reading frames. ORF 1 encodes a protein of unknown function and is not conserved among the genera. The second ORF is partly overlapping with ORF 1 and is expressed as a fusion protein, either by a -1 or a +1 ribosomal frameshift [41]. The fusion protein is a polyprotein with an RdRP1 domain (pfam00680, cd01699). Members of the Amalgaviridae cause persistent infections in their hosts without a visible cytopathic effect.
DIAMOND identified two scaffolds with similarity to the zybaviruses. TC-zyba-like virus 1 has a length of 3185 nt and represents the nearly complete ORFs 1 and 2 (Fig. 1). In comparison to Antonospora locustae virus 1, an unclassified Fig. 3 Phylogenetic analysis of 52 RdRP sequences of members of the family Amalgaviridae and related viruses. The replicase sequences of members of all species of the Amalgaviridae, unclassified related viruses from the Teltow Canal, Havel River, and various hosts were aligned using MEGA, and a maximum-likelihood tree was constructed using IQ-TREE 2 with the optimal substitution model Q.pfam+F+R6. Presented are GenBank accession numbers, vernacular virus names, and strain designations, if available (in parentheses). Square brackets indicate hosts. Numbers at nodes indicate bootstrap support obtained after 10,000 ultrafast replications. The tree was arbitrarily rooted using unirnavirus sequences. The bar indicates amino acid substitutions per site. Color code: black, classified viruses; blue, unclassified viruses; red, viruses from the Teltow Canal and Havel River 1 3
activity and contain microbes associated with municipal surface water drains, discharged effluents from a wastewater treatment plant, and recreational activities. Our previous work revealed the presence of thousands of novel, highly divergent virus sequences. In recent papers, we focused on picorna-like viruses, hepeliviruses, plant viruses, and invertebrate viruses [17][18][19][20][21], and these included descriptions of mycovirus-like sequences resembling sequences of members of the families Alphaflexiviridae, Barnaviridae, Endornaviridae, Partitiviridae, and Spinareoviridae and of the order Elliovirales [19,21]. These virus families are known to include fungal viruses. Although environmental samples do not allow us to associate sequences with specific hosts, phylogenetic analysis suggested that the alphaflexiviruses, bunyaviruses, and spinareoviruses from the Teltow Canal and Havel River were not mycoviruses. In contrast, the SmVA-and PhVA-like viruses, endorna-like viruses, gammapartitiviruses, and aspi-like viruses were related to mycoviruses or still-unclassified fungus-associated viruses [19,21]. Unfortunately, the possible hosts of the barna-like viruses and many of the mycovirus-related sequences in our present study remain uncertain because the related reference virus sequences are associated with both fungal and non-fungal hosts. This illustrates once more the limitations of using environmental samples. Among the detected and analyzed viruses whose hosts are still uncertain are the following viruses:
## TC-and Havel-botourmia-like viruses
Although we characterized more than 150 botourmia-like viruses, none of them belonged to the known genera of fungus-infecting viruses with monopartite, monocistronic genomes. Even their similarity to the plant-infecting ourmia viruses was only moderate. Five clades of botourmia-like viruses exhibited a dicistronic genome layout, suggesting that there are a number of new taxa to be created. Chen et al. and Sadiq et al. have also described botourmia-like viruses (Fig. 4, upper branch). Among these are (i) the previously described solemo-like clade C viruses with a barna-like RdRP (cd23184 [19]), (ii) carascovirus and similar viruses from the Teltow Canal, (iii) SmVB and related viruses, and (iv) three clades of Chinese solemo-like viruses from invertebrate samples. The viruses of this branch were characterized by the presence of an S domain (pfam00729) in their capsid protein. Phylogenetic analysis of the CP sequences of these six clades revealed the relatedness of their structural proteins (Fig. 5). However, our attempts to align the CP sequence of the detected TC-SmVB-like virus and Solemoviridae members failed. Too little similarity was observed (data not shown) although a CDD search suggested that both capsid proteins belong to protein family pfam00729. Notably, most of the viruses of the upper branch have a serine proteinase domain, either a trypsin-like proteinase (solemoviruses, viruses of the solemo-like clade C, carascolike viruses, sobemo-like viruses from China) or a V8-like glutamyl endopeptidase, which is a characteristic of SmVB, the SmVB-like virus from the Teltow Canal, and Qianjiang sobemo-like virus 4.
The other oomycetes viruses, SmVA and PhVA, as well as 20 related viruses from the Teltow Canal clustered with nodaviruses, but in a distinct clade (Fig. 4 lower branch, see also reference 19).
## Hypoviridae
A single 264-bp scaffold related to a hypovirus was identified. It showed 47% aa sequence identity to Botrytis cinerea hypovirus 2, accession no. MN617169 (data not shown).
## Mymonaviridae
A 216-nt scaffold was assigned by DIAMOND to Lentimonavirus lentinulae, a species of negative-strand RNA viruses of the family Mymonaviridae. According to BLASTp, this sequence showed 57% aa sequence identity to the RdRP of Golovinomyces magnicellulatus associated virus and approximately 53% identity to the RdRPs of Erysiphe necator associated negative-stranded RNA virus 23 and Plasmopara viticola lesion associated mononegaambi virus 8 (data not shown).
## Discussion
In order to better understand the virome composition of riverine freshwaters, we examined the virus content of two 50-L water samples from the Teltow Canal and Havel River, which traverse the metropolitan area of Berlin, Germany. Both freshwater bodies are affected by human Fig. 4 Phylogenetic analysis of 85 RdRP sequences of oomycetes viruses, related viruses, and reference viruses. The replicase sequences of Sclerophthora macrospora viruses A and B, Plasmopara halstedii virus A, unclassified related viruses from the Teltow Canal, Havel River, and various hosts as well as reference viruses (nodaviruses, luteoviruses, and solemoviruses) were aligned using MEGA, and a maximum-likelihood tree was constructed using IQ-TREE 2 with the optimal substitution model VT+F+R5. Presented are GenBank accession numbers, vernacular virus names, and strain designations, if available (in parentheses). Square brackets indicate family names (solid lines) and protein families of polymerase and coat proteins according to the CDD and pfam databases (broken lines). Numbers at nodes indicate bootstrap support obtained after 10,000 ultrafast replications. The tree was arbitrarily rooted with sequences with a Nodaviridae RdRP (cd23173). The bar indicates amino acid substitutions per site. Color code: black, classified viruses; blue, unclassified viruses; red, viruses from the Teltow Canal and Havel River 1 3 known botybirnaviruses exhibit specificity for fungi, but our TC-botybirna-like viruses are only distantly related to them and may utilize other hosts. Our remaining viruses of the order Ghabrivirales either belong to the suborder Betatotivirineae (n = 30), which includes only a few mycoviruses, or are too divergent to be assigned to any of the existing Ghabrivirales suborders, suggesting the need to create a new higher order taxon (n = 15). Related viruses of this clade for which host data are available do not appear to be associated with fungi. Nevertheless, this clade contains a branch with four interesting sequences with numerous in-frame termination codons, regardless of which translation table is used. It remains to be determined whether these sequences belong to endogenous, inactive virus-like RNA or indicate the existence of viruses that use a genetic code for protein translation that has not yet been described. Another possible explanation for this phenomenon could be that their genomes contain unusual RNA base modifications that lead to misincorporations during the reverse transcription step of library preparation.
## TC-and Havel-mito-like viruses
Hundreds of novel mitovirus-like sequences have been published in recent years (e.g., reference 50), and most of these with dicistronic genomes [48,49]. Moreover, GenBank has released a number of similar dicistronic botourmia-like virus sequences from other unpublished metagenomic studies. The presence of a second ORF encoding a putative CP with an S domain suggests that these viruses form virions with icosahedral or hemi-icosahedral capsids. At present, Ourmiavirus is the only genus of the family Botourmiaviridae whose members are known to have a capsid. Ourmiaviruses have a tripartite genome, with the genes for the RdRP, CP, and movement protein located on separate segments. Our data do not indicate whether the dicistronic botourmia-like viruses have a segmented genome or encode a movement protein, which would suggest a plant host. Qianjiang botourmia-like virus 13 (PQ054635) and TC-boturmia-like virus 17 both have a third ORF, but because no information is available on the expression or function of these hypothetical proteins, the significance of the additional ORF is unclear.
## Ghabrivirales-like viruses from the teltow canal and havel river
The association of 59 viruses with similarity to members of the order Ghabrivirales with their possible hosts remains inconclusive. Some artiviruses, orthototiviruses, and pseudototiviruses are known to infect both fungi and plants. The 5 Phylogenetic analysis of 20 capsid protein sequences of Sclerophthora macrospora virus B and related viruses. The CP sequences of Sclerophthora macrospora virus B and unclassified related viruses from the Teltow Canal, Havel River, and various hosts were aligned using MEGA, and a maximum-likelihood tree was constructed using IQ-TREE 2 with the optimal substitution model: Q.pfam+F+G4. Pre-sented are GenBank accession numbers, vernacular virus names, and strain designations, if available (in parentheses). Numbers at nodes indicate bootstrap support greater than 65% obtained after 10,000 ultrafast replications. The bar indicates amino acid substitutions per site. Color code: blue, unclassified viruses; red, viruses from the Teltow Canal and Havel River 1 3 programmed frameshift and cluster with zyba-like viruses collected from various sources (Fig. 3). TC-zyba-like virus 1 groups closely with Antonospora locustae virus 1 and Tetrodontophora bielanensis associated virus 1, which were recently suggested on the basis of sequence identity scores to comprise a new genus of the family Amalgaviridae [51]. Other viruses of this clade were associated with unspecified diptera (KX884149 [50]), beetles (MZ209909), mosquitoes (OL700164), and insectivorous bats (OR871395). The oomycete Antonospora locustae is a pathogen of grasshoppers, and it seems plausible that the remaining zyba-like viruses of this clade also infect oomycetes, which are pathogenic to insects. Even the bat-associated zyba-like virus might have found its way into the Rhinolophus intestine via a dietary route rather than infecting bats. Another clade of zyba-like virus sequences includes a yeast-infecting virus of the species Zybavirus bailii, two viruses of plant-infecting fungi (MZ079606, MZ079609), and viruses obtained from water (MW784004) and soil (MN034945) samples. Conidiobolus rugosus virus 1 (MZ600508) was obtained from a soil sample has been associated with the fungus Capillidium rugosum (Conidiobolus rugosus). Chen et al. have associated their Xisha Islands zybavirus (MW784004) from an environmental water sample with putative plant/fungal/bacterial hosts [48]. TC-zyba-like virus 2 from a Teltow Canal water sample matches this correlation.
## TC-Sclerophthora macrospora B-like virus
This virus shows similarity to oomycete-infecting SmBV but also to the Quinjiang sobemo-like viruses 4, 5, 28 (accession nos. PQ055480, PQ055481, and PQ055506) from crabs and a virus sequence from grassland soil (MN033517) (Figs. 4 and5). The availability of only partial sequences, sparse information on host range, and a lack of virus isolates, however, hamper further classification attempts.
Overall, the great divergence of several botourmia-, Ghabrivirales-, mito-, and narna-like viruses from the Teltow Canal and Havel River reflects the fast-growing numbers of new, unassigned mycoviruses and may require the creation of further taxa to integrate them in the present virus taxonomy. Undoubtedly, further sampling will uncover more viruses related to those presented here and contribute to a coherent view of riverine viromes.
viruses are fungal viruses that replicate in mitochondria. Here, we add three new virus sequences from candidate members of the genus Unuamitovirus (Supplementary Fig. S3). Another virus using translation code 4, Havel mito-like virus 1, clusters at the root of the Mitoviridae branch (Supplementary Fig. S3). For this virus, we failed to identify a related viral sequence with significant similarity (data not shown). However, because few mitoviruses lack internal UGA codons, we suggest that this virus has a non-fungal host. While some mito-like viruses that use the standard translation code are proven plant viruses [35,36], a few have been associated with various hosts, e.g., Beihai narna-like virus 26 (source: tunicate [50]), Hubei narna-like virus 24 (source: diptera [50]), Melbourne fly narnavirus 1 (source: drosophila), Halley virus (source: penguin), Dwyer narnalike virus (source: rabbit), Gergich narna-like virus (source: rabbit), and eight mito-like viruses from the Teltow Canal and Havel River. All of these viruses cluster in a single clade that is independent of the four established Mitoviridae genera (Supplementary Fig. S3). Except for sequence data and the source of the sample, no biological information is available. Hence, the significance of this finding is unclear.
## TC-and Havel-narna-like viruses
Recently, Dinan et al. distinguished two clades of narna-like virus, which they named "alphanarnaviruses" and "betanarnaviruses" [37]. The "alphanarnavirus" group is proposed to be comprised of two species, "Narnavirus saccharomajor" and "Narnavirus saccharominor", as well as a number of unclassified viruses, including five viruses from the Havel River and Teltow Canal. As the members of this clade have been detected in fungal and non-fungal hosts, it is unclear whether the narna-like viruses of the Teltow Canal and Havel River are mycoviruses. Interestingly, a few "alphanarnaviruses" have an aligned, uninterrupted complementary reverse-frame ORF (ambigrammatic sequence) [37][38][39][40]. This interesting feature requires further investigation, as it is still unclear whether the encoded protein is really expressed or has a function, but the avoidance of CUA, UUA, and UCA codons, which correspond to translation termination codons in the complementary strand, suggest that the rORF encodes a protein [37,38]. Narna-like viruses with an rORF have been named "mycoambinarnaviruses" [40], but the phylogenetic analyses of Dinan et al. [37], DiRisi et al. [39], and our study (Supplementary Fig. S3) do not indicate that these viruses belong to a single monophyletic clade.
## TC-Zyba-like amalgaviruses
Both TC-zyba-like viruses exhibit the characteristic zybavirus gene layout with fusion of ORFs 1 and 2 by a
## References
1. Sutela, Poimala, Vainio (2019) "Viruses of fungi and oomycetes in the soil environment" *FEMS Microbiol Ecol*
2. Grossart, Van Den Wyngaert, Kagami et al. (2019) "Fungi in aquatic systems" *Nat Rev Microbiol*
3. Ghabrial, Caston, Jiang et al. (2015) "50-plus years of fungal viruses" *Virology*
4. Kondo, Botella, Suzuki (2022) "Mycovirus diversity and evolution revealed/inferred from recent studies" *Annu Rev Phytopathol*
5. Dawe, Kuhn (1016) "Virus-like particles in the aquatic fungus" *Rhizidiomyces. Virology*
6. Ayllón, Vainio (2023) "Mycoviruses as a part of the global virome: Diversity, evolutionary links and lifestyle" *Adv Virus Res*
7. Sato, Suzuki (2023) "Continued mycovirus discovery expanding our understanding of virus lifestyles, symptom expression, and host defense" *Curr Opin Microbiol*
8. Abdoulaye, Foda, Kotta-Loizou (2019) "Viruses infecting the plant pathogenic fungus Rhizoctonia solani" *Viruses*
9. Forgia, Daghino, Chiapello et al. (2024) "New clades of viruses infecting the obligatory biotroph Bremia lactucae representing distinct evolutionary trjectory for viruses infecting oomycetes" *Virus Evol*
10. Andika, Tian, Bian et al. (2023) "Cross-Kingdom interactions between plant and fungal viruses" *Annu Rev Virol*
11. Zell, Groth, Selinka et al. (2022) "Picorna-like viruses of the Havel River" *Germany. Front Microbiol*
13. Zell, Groth, Selinka et al. (2023) "Hepeliviruses in two waterbodies in Berlin" *Germany. Arch Virol*
14. Zell, Groth, Selinka et al. (2023) "Exploring the diversity of plant-associated viruses and related viruses in riverine freshwater samples collected in Berlin" *Germany. Pathogens*
15. Zell, Groth, Selinka et al. (2024) "Diversity of picorna-like viruses in the Teltow Canal" *Viruses*
16. Zell, Groth, Selinka et al. (2024) "Metagenomic analyses of water samples of two urban freshwaters in Berlin, Germany, reveal new highly diverse invertebrate viruses" *Microorganisms*
17. Carducci, Cook, Md et al. (2011) "Surveillance of adenoviruses and noroviruses in European recreational waters" *Water Res*
18. Martin (2011) "Cutadapt removes adapter sequences from highthroughput sequence reads" *EMBnet J*
19. Nurk, Melshko, Korobeynikov et al. (2017) "metaSPAdes: A new versatile metagenomic assembler"
20. Ghabrial, Suzuki (2009) "Viruses of plant pathogenic fungi" *Annu Rev Phytopathol*
21. Parratt, Laine (2016) "The role of hyperparasitism in microbial pathogen ecology and evolution" *ISME J*
22. Liu, Xie, Cheng et al. (2016) "Fungal DNA virus infects a mycophagous insect and utilizes it as a transmission vector" *Proc Natl Acad Sci*
23. Dai, Yang, Pang et al. (2024) "Identification of a negativestrand RNA virus with natural plant and fungal host" *Proc Natl Acad Sci*
24. Jia, Jiang, Xie (2024) "Viruses shuttle between fungi and plants" *Trends Microbiol*
25. Andika, Wei, Cao et al. (2017) "Phytopathogenic fungus hosts a plant virus: a naturally occurring cross-kingdom viral infection" *Proc Natl Acad Sci*
26. Chiapello, Rodríguez-Romero, Ayllón et al. (2020) "Analysis of the virome associated to grapevine downy mildew lesions reveals new mycovirus lineages" *Virus Evol*
27. Depierreux, Vong, Nibert (2016) "Nucleotide sequence of Zygosaccharomyces bailii virus Z: Evidence for +1 programmed ribosomal frameshifting and for assignment to family Amalgaviridae" *Virus Res*
28. Honkura, Shrako, Ehara et al. (1983) "Two types of virus-like particles isolated from downy mildew diseased rice plants" *Ann Phytopathol Soc Jpn*
29. Shirako, Ehara (1985) "Composition of viruses isolated from Sclerophthora macrospora-infected rice plants" *Ann Phytopath Soc Jpn*
30. Heller-Dohmen, Göpfert, Hammerschmidt et al. (2008) "Different pathotypes of the sunflower downy mildew pathogen Plasmopara halstedii all contain isometiric virions" *Mol Plant Pathol*
31. Heller-Dohmen, Göpfert, Pfannstiel et al. (2011) "The nucleotide sequence and genome organization of Plasmopara halstedii virus" *Virology J*
32. (0123)
33. Yokoi, Yamashita, Hibi (2003) "The nucleotide sequence and genome organization of Sclerophthora macrospora virus A" *Virology*
34. Yokoi, Takemoto, Suzuki et al. (1999) "The nucleotide sequence and genome organization of Sclerophthora macrospora virus B" *Virology*
35. Chen, Sadiq, Tian et al. (0118) "RNA viromes from terrestrial sites across China expand environmental viral diversity" *Nat Microbiol*
36. Sadiq, Harvey, Mifsud et al. (2024) "Australian terrestrial environments harbour extensive RNA virus diversity" *Virology*
37. Shi, Lin, Tian et al. (2016) "Redefining the invertebrate RNA viroshere" *Nature*
38. Pyle, Keeling, Nibert (2017) "Amalga-like virus infecting Antonospora locustae, a microsporidian pathogen of grasshoppers, plus related viruses associated with other arthropods" *Virus Res*
39. "Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations"
40. Buchfink, Xie, Huson (2015) "Fast and sensitive protein alignment using DIAMOND" *Nat Methods*
41. Kumar, Stecher, Li et al. (2018) "Molecular evolutionary genetics analysis across computing platforms" *Mol Biol Evol*
42. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: A fast and effective stochastic algorithm for estimating maximum likelihood phylogenies" *Mol Biol Evol*
43. Hoang, Chernomor, Haeseler et al. (2018) "UFBoot2: Improving the ultrafast bootstrap approximation" *Mol Biol Evol*
44. Donaire, Xie, Nerva et al. (2024) "ICTV virus taxonomy profile: Botourmiaviridae 2024" *J Gen Virol*
45. Wolf, Kazlauskas, Iranzo et al. (2018) "Origins and evolution of the global RNA virome" *mBio*
46. Koonin, Dolja, Krupovic et al. (2019) "Proposal 2019.006G.a.v1.Riboviria. Create a metataxonomic framework, filling all principal taxonomic ranks, for realm Riboviria"
47. Naitow, Tang, Canady et al. (2002) "L-A virus at 3.4 Å resolution reveals particle architecture and mRNA decapping mechanism" *Nat Struct Biol*
48. Botella, Manny, Nibert et al. (2021) "Create 100 new species and four new genera (Cryppavirales: Mitoviridae)"
49. Nibert (2017) "Mitovirus UGA(Trp) codon usage parallels that of host mitochondria" *Virology*
50. Nibert, Vong, Fugate et al. (2018) "Evidence for contemporary plant mitoviruses" *Virology*
51. Nerva, Vigani, Silvestre et al. (2018) "Biological and molecular characterization of Chenopodium quinoa mitovirus 1 reveals a distinct small RNA response compared to those of cytoplasmic RNA viruses" *J V I . 0*
52. Dinan, Lukhovitskaya, Olendraite et al. (2020) "A case for a negative-strand coding sequence in a group of positivesense RNA viruses" *Virus Evol*
53. Cook, Chung, Bass et al. (2013) "Novel virus discovery and genome reconstruction from field RNA samples reveals highly divergent viruses in dipteran hosts" *PLoS ONE*
54. Dirisi, Huber, Kistler et al. (2019) "An exploration of ambigrammatic sequences in" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12517474&blobtype=pdf | # Is vaccination a feasible public health strategy against fatal Borna disease virus 1 (BoDV-1) encephalitis? An epidemiological perspective
Kirsten Pörtner, Christina Frank, Hendrik Wilking, Klaus Stark, Christiane Herden, Martin Beer, Dennis Rubbenstroth, Dennis Tappe
## Human Borna disease virus 1 (BoDV-1) encephalitis is characterized by rapid clinical progression, an absence of a causal therapy and an extremely high case fatality rate.
Here, we discuss prevention options through a hypothetical vaccine focusing on epidemiological features.
## Zoonotic Borna disease virus 1 (BoDV-1) causes fatal human encephalitis
In 2018, BoDV-1 was demonstrated to cause severe and mostly fatal encephalitis [1,2], following a long and controversial scientific debate on BoDV-1 pathogenicity in humans [3]. In veterinary medicine, BoDV-1 encephalitis had long been known as Borna disease, affecting mainly horses and sheep [3,4]. Mandatory reporting of direct pathogen detection in human cases was introduced in 2020 in Germany, and active case finding [5], and an increased awareness among clinicians led to the identification of 50 molecularly confirmed (partially retrospective) sporadic human cases as of December 2024 (source: Robert Koch Institute), with a focus on Bavaria (Fig 1 ). A few more confirmed cases (among them at least one retrospective case) have been detected in the current year 2025. Almost all cases (49/50) were fatal. However, ever since the first description of human BoDV-1 encephalitis, epidemiological and medical studies on this severe zoonotic disease have been challenging. This is mainly due to the low incidence with an estimated maximum of 10 incident cases per year (5 confirmed incident cases in 2022, 5 in 2023, 1 in 2024, Source: Robert Koch Institute as of December 31, 2024), case restriction to parts of Germany [6], assumed low level of awareness, and the variable, but often short disease course of 32 (IQR 21-41) days in median after initial hospitalization [7].
Thus, powerful prospective epidemiological, clinical, or therapeutic studies are virtually unfeasible. In recent years, though, larger (retrospective) case series and smaller studies on the epidemiology and phylogeny [4,8], the clinical and paraclinical picture and diagnostics [5,7,[9][10][11] as well as therapeutic attempts [7] helped to increase our knowledge.
## The current epidemiological characteristics limit possible preventive measures
BoDV-1 infections of dead-end hosts such as humans (and other non-reservoir hosts like horses) are the result of spill-over transmission from a reservoir host, the bicolored white-toothed shrew (Crocidura leucodon) shedding the virus in various excretions [3]. BoDV-1-endemic regions have been identified in Germany, Switzerland, Austria, and Liechtenstein, [3,4,12]. However, transmission events or transmission routes remain unknown or at least uncertain: A large case-control study could not identify any potential transmission event or risk factor other than a rural place of residence on the fringe of a settlement close to nature with all cases living in cities or communities with less than 41.000 inhabitants [8]. Consequently, it is challenging to propose any preventive measures. Transmission likely occurs in the peridomestic area, covertly for the infected person, and possibly indirectly from the environment contaminated with shrew excretions containing the virus [4,8].
## Key characteristics of BoDV-1 pathogenesis-Persistent infection, immune-mediated disease, and the lack of an effective therapy
Several studies indicate that inapparent or mild clinical courses of BoDV-1 infection are unlikely; the disease rather presents as fulminant encephalitis with almost always fatal outcomes [7,13,14]. Experimental animal models have shown that BoDV-1 can enter the host via the nasal mucosa or by subcutaneous injection, followed by retrograde intraaxonal transport (for example, along the olfactory route) to the central nervous system (CNS) [15,16].
BoDV-1 establishes persistent, non-cytopathic infection (reviewed in [17,18]), but triggers an inflammatory reaction with mononuclear infiltration of the brain, paralleled by astrocyte and microglia activation with the release of pro-inflammatory cytokines, and a widespread destruction of human brain tissue [19,20]. Consistently, animal models have demonstrated the encephalitis to result from an immunopathogenesis mediated by virus-specific T-lymphocytes [21,22].
Unawareness among clinicians, a facultative and unspecific prodromal phase in which lumbar puncture or imaging is typically not carried out, late seroconversion and the progression to coma within 3 days in median (IQR 2-5) after hospitalization hinders diagnosis early in the disease course [7,10] or even after potential exposure, such as shrew bites, or contact with shrew excretions. Sensitive tests, such as RT-PCR from CSF, demand awareness in the first place and are often ordered far too late. Thus, likely due to late diagnosis and the markedly progressed disease state, sustainable clinical improvement under experimental therapy could not be seen, and treatment remained restricted to individual attempts [7]. Treatment recommendations regarding substance and dosage for BoDV-1 encephalitis therapy (or even post-exposure prophylaxis) in humans are lacking.
## Could the strategy in rabies prevention be a role model for BoDV-1?
For infectious diseases with limited preventive measures, lack of therapy, and high fatality, the prevention by vaccination is an option to wish for. Human rabies, caused by rabies virus, is another zoonotic infectious disease with striking similarities to human bornavirus encephalitis, both viruses belonging to the same order (Mononegavirales), resulting in nearly universal fatality rates after traveling to the CNS and causing severe encephalitis, and both having no curative treatment option.
There are, however, also several important differences regarding prevention: (i) Rabies is widespread in many parts of the world with a large population at risk, whereas BoDV-1 exhibits a small target group for vaccination [4]. (ii) For rabies, potential exposure events (i.e., dog bites in endemic regions) are defined and mostly clearly identifiable, but remain completely unclear for BoDV-1 [8]. Formulating indications for post-(or even pre-) exposure prophylaxis, therefore, seems impossible for BoDV-1, whereas the specific nature of potential exposure to rabies determines the type of post-exposure prophylaxis as well as the indication for pre-exposure vaccination. (iii) A very effective and safe vaccine is available to prevent rabies in humans. Furthermore, vaccinating the reservoir foxes effectively eradicated terrestrial rabies in parts of the world, while eradicating BoDV-1 in shrews-a protected species-would require highly unconventional strategies and would be a complex endeavor. Although the comparison between BoDV-1 and rabies seems obvious, it is misleading when it comes to administering a protective vaccine.
## A future safe and efficacious BoDV-1 vaccine-Likely a challenging endeavor
The idea of developing a (animal) BoDV-1 vaccine already existed in Germany among veterinarians in the early 20 th century after repeated outbreaks in horse and sheep husbandries [23]. A supposedly attenuated live vaccine passaged in rabbit brains was used to vaccinate animals in the endemic regions of mainly East Germany over years. However, potentially insufficient attenuation, questionable efficacy, concerns regarding possible post-vaccination shedding at that time, and the collapse of East Germany led to the suspension of the animal vaccination in 1992 [24].
Several experimental approaches to develop a vaccine against bornaviruses in animals yielded variable results regarding efficacy against persistent infection and (fatal) disease as well as vaccine-triggered exacerbation of the disease and were not pursued any further. Protective immunity was achieved in immunocompetent mice, but not in animals deficient for CD8-positive T-lymphocytes [25]. This is in line with previous studies demonstrating T-lymphocytes to mediate protection against BoDV-1 infection, while humoral immunity appears to play a minor role (reviewed in [21,22]). Given the potential of persistent BoDV-1 infection and the known immunopathogenesis, a prophylactic vaccine must provide a robust (likely T-cell-driven, mucosal) immunity to reliably eliminate the virus immediately at the site of entry or at least before it enters the CNS. In contrast, an immune response providing only incomplete protection against the infection might even result in enhanced immunopathology [22]. It remains to date questionable, if sterile immunity preventing both, the infection and the disease, can be achieved with sufficient certainty. T-cells seem to be responsible for both, protection against the infection as well as immunopathogenesis of the disease (reviewed in [22]).
## The target population-Whom to vaccinate?
In addition to these challenges in vaccine development, there are challenges regarding the target population: Postexposure vaccination of persons known to be exposed is non-practical for BoDV-1, as the exposure event usually remains elusive. For pre-exposure vaccination, taking present census and level of urbanization data into account, as well as the currently known endemic area of BoDV-1, an estimated 5-8 million people in rural localities of parts of Germany, Austria, Switzerland, and Liechtenstein could theoretically be at risk for a BoDV-1 infection. However, given the rarity of BoDV-1 encephalitis, likely only a minor fraction of this population might actually be exposed to the virus on the basis of known or unknown risk or host factors. Real-world effectiveness data (as obtained for SARS-CoV-2 vaccinations) with massive infections will not be obtainable for BoDV-1 for obvious reasons. The entire population living rurally in virus-endemic areas would have to be vaccinated to prevent a small absolute number of infections. The number needed to vaccinate, a widely used number to quantify immunization benefits, would be very large to reduce one human case only. This emphasizes the need for the development of a vaccine with an extremely high safety profile and extensive preclinical and clinical testing, raising also economical and ethical questions.
## Conclusion
Both, humans and animals would certainly benefit from a vaccine against BoDV-1, especially as long as there is no causal therapy and the infection is almost universally fatal. However, the complex immunological mechanisms of BoDV-1 complicate vaccine development. Even if there was a safe and efficacious vaccine, targeted application of either pre-or post-exposure vaccination would seem unrealistic since exposure events remain unknown. Thus, at this point in time, awareness campaigns, the development of novel antivirals and concerted and standardized (theory-based) therapy recommendations although also limited remain the current focus to improve human BoDV-1 control.
## References
1. Korn, Coras, Bobinger et al. (2018) "Fatal encephalitis associated with Borna disease virus 1" *N Engl J Med*
2. Schlottau, Forth, Angstwurm et al. (2018) "Fatal encephalitic Borna disease virus 1 in solid-organ transplant recipients" *N Engl J Med*
3. Rubbenstroth, Schlottau, Schwemmle et al. (2019) "Human bornavirus research: back on track!" *PLoS Pathog*
4. Ebinger, Santos, Pfaff et al. (2024) "Lethal Borna disease virus 1 infections of humans and animalsin-depth molecular epidemiology and phylogeography" *Nat Commun*
5. Eisermann, Rubbenstroth, Cadar et al. (2018) "Active case finding of current bornavirus infections in human encephalitis cases of unknown etiology" *Emerg Infect Dis*
6. (2024) "Humane Infektion mit dem Borna Disease Virus 1-gemeinsames Merkblatt von"
7. Pörtner, Wilking, Frank et al. "Clinical analysis of Bornavirus encephalitis cases demonstrates a small time window for etiological diagnostics and treatment attempts, a large case series from Germany 1996-2022"
8. Pörtner, Wilking, Frank et al. (2023) "Risk factors for Borna disease virus 1 encephalitis in Germany-a case-control study" *Emerg Microbes Infect*
9. Finck, Liesche-Starnecker, Probst et al. (2020) "Bornavirus encephalitis shows a characteristic magnetic resonance phenotype in humans" *Ann Neurol*
10. Allartz, Hotop, Muntau et al. (2024) "Detection of bornavirus-reactive antibodies and BoDV-1 RNA only in encephalitis patients from virus endemic areas: a comparative serological and molecular sensitivity, specificity, predictive value, and disease duration correlation study" *Infection*
11. Neumann, Hierl, Wunderlich et al. (2023) "Cerebrospinal fluid in Borna disease virus 1 (BoDV-1) encephalitis" *J Neurol Sci*
12. Dürrwald, Kolodziejek, Weissenböck et al. (2014) "The bicolored white-toothed shrew Crocidura leucodon (HERMANN 1780) is an indigenous host of mammalian Borna disease virus" *PLoS One*
13. Tappe, Frank, Offergeld et al. (2019) "Low prevalence of Borna disease virus 1 (BoDV-1) IgG antibodies in humans from areas endemic for animal Borna disease of Southern Germany" *Sci Rep*
14. Niller, Angstwurm, Rubbenstroth et al. (2019) "Zoonotic spillover infections with Borna disease virus 1 leading to fatal human encephalitis, 1999-2019: an epidemiological investigation" *Lancet Infect Dis*
15. Carbone, Duchala, Griffin et al. (1987) "Pathogenesis of Borna disease in rats: evidence that intra-axonal spread is the major route for virus dissemination and the determinant for disease incubation" *J Virol*
16. Morales, Herzog, Kompter et al. (1988) "Axonal transport of Borna disease virus along olfactory pathways in spontaneously and experimentally infected rats" *Med Microbiol Immunol*
17. Herden, Lipkin, Richt et al. (2013) "Fields virology"
18. Nobach, Müller, Tappe et al. (2020) "Update on immunopathology of bornavirus infections in humans and animals" *Adv Virus Res*
19. Liesche, Ruf, Zoubaa et al. (2019) "The neuropathology of fatal encephalomyelitis in human Borna virus infection" *Acta Neuropathol*
20. Rauch, Steffen, Muntau et al. (2022) "Human Borna disease virus 1 encephalitis shows marked pro-inflammatory biomarker and tissue immunoactivation during the course of disease" *Emerg Microbes Infect*
21. (2022)
22. Stitz, Bilzer, Planz (2002) "The immunopathogenesis of Borna disease virus infection" *Front Biosci*
23. Rubbenstroth (2022) "Avian bornavirus research-a comprehensive review" *Viruses*
24. Dürrwald, Ludwig (1997) "Borna disease virus (BDV), a (zoonotic?) worldwide pathogen. A review of the history of the disease and the virus infection with comprehensive bibliography" *Zentralbl Veterinarmed B*
25. Dürrwald, Kolodziejek, Oh et al. (2022) "Vaccination against Borna disease: overview, vaccine virus characterization and investigation of live and inactivated vaccines" *Viruses*
26. Hausmann, Baur, Engelhardt et al. (2005) "Vaccine-induced protection against Borna disease in wild-type and perforin-deficient mice" *J Gen Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645941&blobtype=pdf | # Ribosomal protein L35 negatively regulates FMDV replication by recruiting AMFR to promote the ubiquitination and degradation of VP2
Wenhua Shao, Wei Zhang, Yang Yang, Xiaoyi Zhao, Weijun Cao, Chuangwei Chen, Wei Wang, Mengyao Huang, Tingting Zhou, Zixiang Zhu, Fan Yang, Haixue Zheng
## Abstract
The control of foot-and-mouth disease virus (FMDV) primarily relies on vaccine immunization; however, this approach is not always fully effective, underscoring the urgent need for novel antiviral strategies. This study identifies RPL35 as a host antiviral protein that targets FMDV. Further mechanistic investigations demonstrate that RPL35 directly interacts with the FMDV structural protein VP2, mediating its K48-linked polyubiquitination and subsequent degradation. The Lys217 residue of VP2 is critical for RPL35's antiviral activity, as evidenced by the increased viral virulence observed with the rO-VP2K217R mutant virus. Through an unbiased proteomic screen, we revealed that RPL35 recruits the E3 ligase AMFR to ubiquitinate and degrade VP2. Additionally, FMDV induces the degradation of KPNA3, thereby blocking RPL35's nuclear translocation. This study advances our understanding of host-virus interactions and provides new insights into developing antiviral drugs targeting the ubiquitin-proteasome pathway. IMPORTANCE This investigation elucidated the antiviral role of RPL35 in the context of FMDV infection. Our results indicate that RPL35 facilitates the recruitment of AMFR, which, in turn, promotes K48-linked polyubiquitination and subsequent proteasomal degradation of the viral protein VP2. This process thereby mitigates viral infection. Further analysis identified Lys217 of VP2 as a critical ubiquitination site for RPL35, with the inhibitory effect of RPL35 being abolished in the recombinant mutant virus rO-VP2K217R. Additionally, we found that FMDV induces the degradation of KPNA3, which obstructs the nuclear translocation of RPL35. Collectively, these findings suggest that RPL35 functions as a potent antiviral effector in suppressing FMDV infection.
F oot-and-mouth disease virus (FMDV) is a highly contagious pathogen that primarily infects cloven-hoofed animals, such as cattle, pigs, and sheep, causing significant economic losses in the global livestock industry (1,2). Current prevention and control strategies rely heavily on vaccination; however, the virus's high genetic variability and the limitations of existing vaccines underscore the urgent need for novel approaches (3,4). The FMDV capsid, composed of the structural proteins VP1, VP2, VP3, and VP4, plays a critical role in viral assembly, stability, and infection processes (5-7). Among these proteins, VP2 is a key structural component that significantly contributes to the viral lifecycle, making it a focal point for both scientific investigation and applied research (8).
The VP2 protein is crucial for FMDV's ability to infect host cells, maintain viral particle stability, and potentially evade immune responses (9). Investigating VP2 functions offers valuable insights into virus-host interactions and the mechanisms underlying viral pathogenesis. Moreover, VP2 contain key antigenic epitopes capable of eliciting immune responses, thereby providing a foundation for developing more effective and broad-spectrum FMDV vaccines (10). Beyond its role in vaccine development, VP2 also shows promise as a target for antiviral drug discovery. By elucidating its structural and functional roles in viral replication and assembly, researchers may identify potential targets for small-molecule inhibitors or other therapeutic interventions (11). Disrupting VP2 function could effectively suppress viral replication and transmission, presenting a novel strategy for controlling FMDV outbreaks (12,13).
Ribosomal proteins (RPs), traditionally recognized for their roles in protein synthesis, have recently been identified as significant contributors to viral infections (14)(15)(16). RPL35, a component of the ribosomal subunit and a member of the RP family, has been implicated in various stages of the viral life cycle. Ribosomes, essential for cellular protein synthesis, consist of large and small subunits and are frequently exploited by viruses through mechanisms such as internal ribosome entry sites (IRES), which facilitate viral protein synthesis (17,18). Interactions between viral proteins and RPs are well documen ted. For example, RPSA interacts with FMDV VP1 to enhance viral replication (19); RPL4 modulates viral replication as an interaction partner of IBDV VP3 (20), and RPL18 is involved in respiratory syncytial virus (RSV) infection by binding to the nucleocapsid protein (21). Similarly, RPL9 aids in virus particle assembly as a binding partner of MMTV Gag (22), and RPL13 promotes IRES-driven translation of FMDV in a helicase DDX3-dependent manner (23).
Although RPs are often associated with promoting viral infections, recent studies have revealed their potential as antiviral agents. Two primary antiviral mechanisms have been proposed: (i) direct interaction with viral proteins to inhibit viral transcription or translation, as demonstrated by RPL9 binding to the phosphoprotein P of the rabies virus (RABV) (24), and RPS10, 18S rRNA, and tRNAs forming a complex with the HIV-1 Nef protein to reduce viral protein synthesis (25); and (ii) activation of antiviral defense signaling pathways, exemplified by RPS20 modulating Toll-like receptor 3 (TLR3) to inhibit classical swine fever virus (CSFV) replication (26), and RRL13a assembling an interferon-γ-independent antiviral complex to suppress the translation of the respiratory syncytial virus matrix protein M (27).
In this study, we identify RPL35 as a potential inhibitor of FMDV by targeting the VP2 protein. Our findings suggest that RPL35-mediated degradation of VP2 could provide a novel therapeutic strategy for treating FMDV infections. This research not only advances our understanding of FMDV biology but also provides a foundation for developing innovative vaccines, antiviral drugs, and control strategies.
## RESULTS
## RPL35 interacts with the FMDV VP2 protein
It has been reported that the FMDV VP2 protein plays a critical role in viral replication by contributing to virus assembly, stability, host cell recognition, immunogenicity, and release. To identify host proteins that interact with FMDV VP2, immunoprecipitation coupled with mass spectrometry (IP-MS) was performed. Through an unbiased screening approach, the cellular protein RPL35 was identified as a candidate interacting partner. To validate the interaction between VP2 and RPL35, co-immunoprecipitation (Co-IP) assays were conducted, confirming that RPL35 binds to VP2 (Fig. 1A). To further investigate the endogenous interaction between VP2 and RPL35, cellular lysates from FMDV-infec ted cells were subjected to immunoprecipitation. The results demonstrated that VP2 interacts with endogenous RPL35 under viral infection conditions (Fig. 1B). Additionally, confocal microscopy analysis revealed that RPL35 colocalizes with VP2 in cells (Fig. 1C), a finding also confirmed in endogenous experiments (Fig. 1D). To map the functional domain of RPL35 responsible for its interaction with VP2, a series of truncation mutants of RPL35 were generated using PCR-based site-directed mutagenesis (Fig. 1E). Co-IP analysis of these mutants indicated that two or more regions of RPL35 are involved in the interaction with VP2 (Fig. 1F). Collectively, these findings provide compelling evidence that RPL35 interacts with the FMDV VP2 protein.
## RPL35 inhibits FMDV replication
To investigate the impact of RPL35 on FMDV infection via its interaction with the VP2 protein, PK-15 cells were transfected with Myc-RPL35 and subsequently infected with FMDV. Viral protein expression, viral RNA levels, and viral titers were assessed using immunoblot analysis, RT-PCR, and median tissue culture infective dose (TCID 50 ) assays, respectively. Notably, these results revealed that overexpression of RPL35 reduced FMDV protein levels, RNA levels, and viral titers in a dose-dependent manner (Fig. 2A through C). Collectively, these results indicate that RPL35 overexpression significantly suppresses FMDV replication. To further investigate the role of RPL35 in FMDV replication, three RPL35-specific siRNAs (siRNA-1, siRNA-2, and siRNA-3) were designed and synthesized. Their effects on FMDV replication were assessed in RPL35-knockdown PK-15 cells. The silencing efficiencies of the siRNAs were evaluated using immunoblot analysis, which revealed that siRNA-3 exhibited the highest efficiency in reducing RPL35 expression (Fig. 2D andE). Consequently, siRNA-3 was selected for subsequent RPL35 knockdown experiments. PK-15 cells were transfected with either negative control (NC) siRNA or siRNA-3, followed by FMDV infection. Viral protein levels, viral RNA, and viral titers were compared between siRNA-3-transfected and NC siRNA-transfected cells at specified time points postinfection. Immunoblot analysis demonstrated that RPL35 knockdown significantly enhanced FMDV protein expression (Fig. 2F). Similarly, RT-PCR results indicated a marked increase in FMDV genomic RNA levels in RPL35-knockdown cells (Fig. 2G). Additionally, TCID 50 assays revealed higher viral titers in RPL35-knockdown cells compared with NC siRNA-transfected cells (Fig. 2H). Collectively, these findings demonstrate that FMDV replication is significantly elevated in RPL35-knockdown cells, underscoring the role of RPL35 in restricting FMDV replication.
## RPL35 affects FMDV RNA synthesis and virus assembly/release
We next investigated which step of the viral replication cycle is targeted by RPL35. PK-15 cells were transfected with either RPL35 or control vector plasmids, followed by FMDV infection and incubation at 4°C for 1 h. After adsorption, unbound viruses were removed, and the abundance of cell-bound FMDV RNA was measured. No significant reduction in cell-bound FMDV RNA was observed in RPL35-expressing cells (Fig. 3A). To determine whether RPL35 affects FMDV internalization, the cells were inoculated with FMDV at 4°C for 1 h, followed by incubation at 37°C for an additional hour to allow internalization. RT-PCR analysis revealed that RPL35 had no impact on FMDV internalization (Fig. 3B). Based on these findings, we further explored whether RPL35 regulates viral mRNA translation or RNA synthesis. A bicistronic reporter plasmid was employed to assess FMDV IRES activity (Fig. 3C). In this system, translation of the first cistron (Renilla luciferase [Rluc]) is cap-dependent, whereas translation of the second cistron (firefly luciferase [Fluc]) is driven by FMDV IRES activity. Relative IRES activity was calculated as the ratio of Fluc to Rluc expression. The bicistronic reporter plasmid was transfected into RPL35-expressing or control cells, and cell lysates were collected 36 h post-transfection to measure Fluc and Rluc activities. The results showed no significant difference in FMDV IRES activity between RPL35-expressing and control cells (Fig. 3D). Additionally, we assessed the effect of endogenous RPL35 on FMDV translation and observed no impact (Fig. 3G).
During FMDV infection, the viral genomic RNA serves as a template for both translation and RNA replication, tightly coupling these processes. To evaluate the effect of RPL35 on viral RNA (vRNA) synthesis, total RNA and negative-strand RNA were quantified by RT-PCR. As shown in Fig. 3E andF, RPL35-expressing cells exhibited a significant reduction in the synthesis of viral negative-strand RNA compared to control cells. Conversely, knocking down RPL35 significantly promoted FMDV RNA synthesis (Fig. 3H andI). Additionally, RPL35-expressing and control cells were infected with FMDV for 6 h, and the synthesis of viral double-stranded RNA (dsRNA), a marker of FMDV replication, was detected by immunofluorescence. The results showed that overexpression of RPL35 significantly reduced the levels of FMDV dsRNA (Fig. 3J). Furthermore, we examined the effect of RPL35 on intracellular and extracellular viral RNA levels. The ratio of extracellular to intracellular RNA was used to assess the efficiency of virion assembly and release (28). At 8 and 12 h post-infection (hpi), RPL35-expressing cells exhibited reduced intracellular viral RNA levels compared to control cells, along with significantly decreased extracellular viral RNA levels (Fig. 3K). This resulted in a diminished extracellular-to-intracellular viral RNA ratio (Fig. 3L). These results suggest that RPL35 is involved in regulating FMDV virion assembly and release. In summary, these findings demonstrate that RPL35 plays dual roles in inhibiting FMDV replication and modulating virion assembly/release.
## RPL35 induces FMDV VP2 protein degradation in a proteasome-dependent manner
In the experimental process described above, we consistently observed that overexpres sion of RPL35 led to a decrease in VP2 protein levels. This observation prompted us to hypothesize that RPL35 may regulate VP2 protein stability or inhibit its expression. To test this hypothesis, we co-transfected Myc-RPL35 and Flag-VP2 into HEK293T cells. Our results indicated that VP2 protein levels decreased in a dose-dependent manner with RPL35 overexpression (Fig. 4A), whereas VP2 mRNA levels remained unchanged (Fig. 4B). To further investigate the suppressive effect of RPL35 on VP2 expression, we co-transfec ted HEK293T cells with either RPL35 or control vector plasmids alongside VP2 plasmids. The transfected cells were then treated with cycloheximide (CHX), a specific inhibitor of protein synthesis, to evaluate the half-life of VP2. Immunoblot analysis revealed that RPL35 overexpression significantly accelerated VP2 degradation, indicating that RPL35 regulates the half-life of VP2 (Fig. 4C andD).
To investigate the underlying mechanisms of RPL35-induced VP2 reduction, we examined the potential involvement of proteasomal, lysosomal, and caspase-dependent pathways. Myc-RPL35 and Flag-VP2 plasmids were co-transfected into HEK293T cells, which were subsequently treated with the proteasome inhibitor MG132, the lysosome inhibitor NH 4 Cl, or the pan-caspase inhibitor Z-VAD-FMK. As shown in Fig. 4E andF, treatment with MG132 restored VP2 levels in RPL35-overexpressing cells, whereas NH 4 Cl and Z-VAD-FMK had no comparable effect. These results suggest that RPL35 promotes VP2 degradation via the proteasome pathway. with FMDV at an MOI of 1 at the indicated time points. Total (E) or negative-strand (F) viral RNA was quantified using RT-PCR. (G) Effect of endogenous RPL35 on FMDV IRES-driven translation. PK-15 cells were transfected with RPL35 siRNA or a negative control (NC) for 12 h and then transfected with bicistronic FMDV-IRES plasmids. At 36 h post-transfection, the activities of Renilla luciferase (Rluc) and Firefly luciferase (Fluc) were measured. (H and I) Effect of endogenous RPL35 on viral RNA (vRNA) synthesis. PK-15 cells were transfected with RPL35 siRNA or NC, followed by infection with FMDV at an MOI of 1 at the indicated time points.
Total (H) or negative-strand (I) viral RNA was quantified using RT-PCR. (J) Overexpression of RPL35 significantly reduced the levels of FMDV double-stranded RNA (dsRNA). PK-15 cells were transfected with RPL35 or vector plasmids, then infected with FMDV at an MOI of 1 for 6 h. Anti-dsRNA antibodies were used to detect FMDV dsRNA synthesis via immunofluorescence assay (IFA). (K and L) Effect of RPL35 on the efficiency of FMDV virion assembly and release. PK-15 cells were transfected with RPL35 or vector plasmids, then infected with FMDV at an MOI of 0.5 for 8 and 12 h. Quantitative RT-PCR was employed to measure extracellular and intracellular viral RNA levels (K). The ratio of extracellular to intracellular viral RNA was calculated to assess virion assembly and release efficiency (L).*P < 0. To identify the functional domain of RPL35 responsible for VP2 degradation, Flag-VP2 was co-transfected with plasmids expressing various truncated forms of RPL35, and VP2 levels were subsequently assessed. The results demonstrated that the first, second, and third truncated mutants of RPL35 are all essential for VP2 degradation, as the absence of any of these regions abolished the degradation effect (Fig. 4G). In summary, these findings indicate that RPL35 induces the degradation of FMDV VP2 via the proteasome pathway, with the first three regions of RPL35 playing a critical role in this process.
## RPL35 catalyzes the K48-linked polyubiquitination of VP2
Protein ubiquitination plays a crucial role in the proteasome-mediated degradation pathway (29,30). To investigate the polyubiquitination of VP2, we overexpressed RPL35 or control vectors and treated the cells with MG132. As shown in Fig. 5A, MG132 treatment resulted in the accumulation of polyubiquitinated VP2 in cells overexpressing RPL35. To further characterize the types of VP2 polyubiquitination influenced by RPL35, we utilized a panel of ubiquitin (Ub) mutants (K6, K11, K27, K29, K33, K48, and K63). The results revealed that RPL35 specifically enhanced K48-linked polyubiquitination of VP2 (Fig. 5B). Additionally, experiments on endogenous ubiquitination demonstrated that RPL35 promoted both total ubiquitination and K48-linked ubiquitination of endogenous VP2 during viral infection (Fig. 5C). E3 ubiquitin ligases primarily mediate the ubiquitina tion of target proteins by modifying lysine residues (31). FMDV VP2 contains nine lysine residues distributed across its domains. To identify the specific ubiquitination sites on VP2, we generated a series of VP2 mutants (K2R, K3R, K63R, K88R, K159R, K172R, K175R, K198R, and K217R), in which lysine residues were replaced with arginine. We co-transfec ted HEK293T cells with Ub, RPL35, and either wild-type VP2 or its mutants, followed by MG132 treatment. Cell lysates were subjected to immunoprecipitation using anti-Flag antibodies. The results indicated that the K217R mutation significantly reduced the ubiquitination of VP2 mediated by RPL35 (Fig. 5D). In summary, these findings demon strate that RPL35 catalyzes K48-linked polyubiquitination of VP2, with K217 identified as a critical ubiquitination site.
## The VP2 K217 residue is essential for RPL35-mediated restriction of virus replication
The PyMOL illustration indicates that VP2 Lys217 is located on the surface of the virion, suggesting its potential role in binding to RPL35 (Fig. 6A). Subsequently, we compared the amino acid sequences of seven FMDV VP2 subtypes and found that the residues at position 217 (Lys) are conserved (Fig. 6B). Mutant viruses containing the K217R substitution (rVP2-K217R) were generated using reverse genetics technology, with the wild-type virus (WT) serving as a control to investigate the impact of Lys217 on FMDV replication. Suckling mice assays were conducted to compare the pathogenicity of the two viruses in vivo. Both viruses were virulent in suckling mice but exhibited different lethality levels. After 7 days post-inoculation, the WT group had a 25% survival rate, whereas all rVP2-K217R-infected mice died within 4 days (Fig. 6C). Additionally, tissue samples from the heart, liver, spleen, lung, and kidney of suckling mice at 3 days post-infection were analyzed by RT-PCR. The results revealed elevated levels of FMDV mRNA in tissues infected with rVP2-K217R compared with those infected with the wild-type virus (Fig. 6D). Furthermore, histopathological examination of the lungs showed that rVP2-K217R exacerbated lung lesions, such as alveolar shrinkage (Fig. 6E). These findings suggest that disruption of the Lys217 ubiquitination site in VP2 enhances FMDV pathogenesis in vivo.
## RPL35 recruits E3 ligase AMFR to degrade VP2 protein
Although RPL35 modulates the stability of the VP2 protein through post-translational regulation, its intrinsic lack of E3 ubiquitin ligase activity prompted us to investigate potential collaborating ubiquitination enzymes. To elucidate the molecular mechanism underlying RPL35-mediated VP2 degradation, we employed immunoprecipitation-mass spectrometry (IP-MS) analysis and identified AMFR, a known E3 ubiquitin ligase (32), as a novel interacting partner of RPL35 (Fig. 7A). Subsequent functional validation revealed that exogenous overexpression of AMFR significantly enhanced RPL35-mediated VP2 proteolysis (Fig. 7B). To establish the necessity of AMFR in this degradation pathway, we systematically evaluated three distinct AMFR-targeting siRNAs. Quantitative analysis identified siRNA-3 as demonstrating superior knockdown efficiency (Fig. 7C), which was subsequently employed in loss-of-function studies. Depletion of AMFR markedly attenuated RPL35-induced VP2 degradation (Fig. 7D), accompanied by a reduction in VP2 ubiquitination levels (Fig. 7E). These findings collectively establish AMFR as the critical E3 ubiquitin ligase recruited by RPL35 to orchestrate VP2 ubiquitination and subsequent proteasomal degradation.
## The degradation of KPNA3 by FMDV proteins impairs the nuclear transloca tion of RPL35
Normally, RPL35 mRNA is translated into protein in the cytoplasm. The protein then translocates to the nucleus (33), where it assembles with ribosomal RNA to form the 60S ribosomal subunit. This subunit, together with the 40S ribosomal subunit, exits the nucleus and reenters the cytoplasm. In the cytoplasm, the mature 60S and 40S subunits combine to form 80S ribosomes, which subsequently translate mRNA into protein (34)(35)(36)(37). To investigate whether FMDV infection affects the nuclear translocation of RPL35, we performed confocal microscopy analysis. The results indicate that following FMDV infection, RPL35 does not enter the nucleus and remains in the cytoplasm (Fig. 8A). To further elucidate the viral factors that inhibit the nucleocytoplasmic translocation of RPL35, we examined the subcellular localization of RPL35 in cells ectopically expressing individual viral proteins using immunofluorescence. Our observations revealed that the FMDV proteins 2C, L, 2B, 3C, and 3D significantly inhibit the nuclear translocation of RPL35 (Fig. 8B). Nucleocytoplasmic transport is a vital process in eukaryotic cells (38). The molecular mechanisms underlying nuclear transport involve the nuclear transport receptor importin α, also known as karyopherin α (KPNA). KPNA comprises seven subtypes, specifically KPNA1 through KPNA7 (39,40). To identify the primary subtype responsible for transporting RPL35 into the nucleus, we conducted immunoprecipitation experi ments. The results indicate that RPL35 interacts with KPNA3 (Fig. 9A), and this interaction was confirmed by confocal microscopy (Fig. 9B). These findings suggest that KPNA3 is the transporter protein facilitating RPL35's entry into the nucleus. This raises the possibility that FMDV may target nuclear localization signaling receptors for degradation. We examined whether KPNA3 was degraded in FMDV-infected cells. The results demon strated that FMDV induces the degradation of endogenous KPNA3 (Fig. 9C). To identify which viral protein is responsible for KPNA3 degradation, HEK293T cells were cotransfected with KPNA3-expressing plasmids and plasmids expressing various Flagtagged viral proteins. It was observed that the expression of the 3D, 3C, 2B, 2C, and L proteins significantly decreased KPNA3 abundance (Fig. 9D). However, dose-dependent analysis revealed that only the L and 2B proteins degraded KPNA3 in a dose-dependent manner (Fig. 9E). These results indicate that the FMDV proteins 2B and L induce KPNA3 degradation, thereby blocking RPL35 nuclear translocation.
## DISCUSSION
Ribosomal proteins (RPs), together with ribosomal RNA (rRNA), are essential components of ribosomes that facilitate the cellular process of protein biosynthesis, commonly known as "translation" (41,42). Viruses, as small infectious agents with limited genomic capacity, must recruit various host factors, including RPs, to ensure their survival and replication (43). Recent research has increasingly elucidated the functional interplay between RPs and viral infections. Most of these interactions are critical for viral translation and replication, thereby enhancing viral infection and proliferation, whereas a smaller subset activates host cell defense mechanisms by triggering immune pathways against the virus (44). Exploring antiviral strategies based on RPs will guide future research. In this study, we identify the ribosomal protein RPL35 as a crucial host antiviral factor that targets the FMDV VP2 protein for proteasomal degradation. This discovery reveals a novel mecha nism of host-mediated viral restriction and contributes to the development of potential therapeutic strategies against FMDV.
In contrast to established host-mediated ubiquitination pathways, such as those targeting influenza PB1 (e.g., TRIM32-mediated ubiquitination of PB1 [45]) or the rabies matrix protein (e.g., TRIM72-mediated ubiquitination of the matrix protein [46]), the RPL35-driven mechanism represents a unique antiviral strategy. Although both are host proteins that directly bind to viral proteins and induce K48-linked ubiquitination and degradation, RPL35 is not an E3 ligase. Instead, it operates through a distinct mechanism that recruits E3 ligases in host cells to ubiquitinate and degrade VP2. This highlights the diversity of host defense strategies against viral pathogens. Mechanistically, RPL35 overexpression reduces VP2 protein levels in a dose-dependent manner without affecting mRNA stability, whereas proteasome inhibition with MG132 fully restores VP2 accumulation. These findings position RPL35 as a post-translational regulator orchestrat ing viral protein turnover. Importantly, truncation analyses demonstrate that three distinct RPL35 domains cooperatively mediate VP2 degradation, suggesting a multiva lent interaction mode that ensures robust antiviral activity. This contrasts with previous reports of single-domain antiviral ribosomal proteins, revealing an evolutionary refinement in host defense mechanisms.
Functional mapping has identified Lys217 of VP2 as the critical site for ubiquitination. Notably, the conservation of VP2's Lys217 across various FMDV serotypes suggests the potential for developing pan-serotype ubiquitination enhancers, which could help overcome the limitations of current vaccines. The increased virulence observed in the rO-VP2K217R mutant indicates that FMDV may utilize residue-specific modifications to evade RPL35-mediated degradation, a strategy similar to those employed by other viruses to evade immune responses. For example, TRIM7 inhibits enterovirus replication and promotes the emergence of viral variants with increased pathogenicity (47). Additionally, TRIM21 binds to residue R95 of M1 and facilitates K48-linked ubiquitination of M1 at K242, leading to proteasome-dependent degradation and inhibition of H3, H5, and H9 IAV replication (48). Furthermore, the E3 ligase RNF5 interacts with VP1 and targets it for degradation through ubiquitination at Lys200, ultimately inhibiting FMDV replication (11). (D) The effect of AMFR E3 ligase knockdown on RPL35-mediated degradation of VP2 was evaluated. Immunoblot analysis was performed on PK-15 cells transfected with VP2, RPL35, or AMFR siRNA, using anti-Flag, anti-Myc, and anti-AMFR antibodies. (E) The influence of AMFR knockdown on RPL35-mediated ubiquitination of VP2 was analyzed. Immunoblot analysis using anti-HA antibodies was performed on proteins immunoprecipitated with anti-Flag antibodies from lysates of PK-15 cells transfected for 24 h with various combinations. **** P < 0.0001.
Ribosomal proteins are typically involved in protein biosynthesis. However, recent studies have revealed an atypical role for ribosomal proteins in antiviral responses. This antiviral effect may extend beyond their traditional function in translation as ribosomal components. For example, ribosomal protein L9 interacts with the Gag protein of mouse mammary tumor virus (MMTV) to inhibit the assembly of MMTV virus particles (22). RPLP1 restricts HIV-1 transcription by disrupting C/EBPβ binding to the LTR (49). In this study, RPL35 exerts broad-spectrum antiviral effects by suppressing both viral RNA synthesis and virion assembly. As a crucial component of the 60S ribosomal subunit, RPL35 plays a significant role in protein translation and endoplasmic reticulum docking (50). The suppression of viral RNA synthesis mediated by RPL35 may result from infec tion-induced impairment of its nucleocytoplasmic trafficking, which disrupts the production of host proteins necessary for viral RNA replication, thereby indirectly attenuating viral RNA synthesis (33). This dual-phase inhibition-targeting both genomic replication and particle maturation-positions RPL35 as a gatekeeper at multiple stages of the viral lifecycle. Such multi-tiered antiviral activity may account for its potent suppression of viral titers, surpassing the efficacy of single-mechanism inhibitors. RPL35, as a ribosomal subunit component, plays a role independent of its translation function in antiviral responses. The aberrant localization of RPL35 induced by infection may selectively affect the translation of viral and host mRNAs; however, the mechanism underlying this selective effect requires further in-depth investigation. Notably, proteomic screening has identified AMFR, an endoplasmic reticulum-asso ciated E3 ubiquitin ligase, as a key mediator of RPL35-driven VP2 degradation. The recruitment of AMFR by RPL35 for the ubiquitination of VP2 reveals a novel function for AMFR that extends beyond its established role in STING activation (51). Although AMFR typically facilitates K27-linked ubiquitination of STING to enhance antiviral signaling (51), its repurposing here for K48-linked degradation of a viral structural protein broadens the functional repertoire of E3 ligases in antiviral immunity. Our findings align with emerging themes in host-pathogen interactions, where ubiquitination pathways act as double-edged swords-either promoting viral clearance or being exploited by viruses to undermine host defenses.
The dependence of FMDV on the stability of its structural protein VP2 for infectivity making VP2 proteins as a promising therapeutic target. Structural proteins are increas ingly recognized as viable antiviral targets due to their conserved roles in viral assembly and entry. For example, pleconaril targets picornavirus capsids to block uncoating (52), whereas allosteric modulators of the hepatitis B virus core protein (CpAMs) disrupt capsid assembly (53,54). Our research suggests that small molecules mimicking the interaction between RPL35 and VP2, or compounds that enhance AMFR-mediated ubiquitination, could destabilize VP2 and inhibit FMDV replication.
Furthermore, the identification of KPNA3 degradation as a viral countermeasure highlights the potential of stabilizing nuclear transport machinery to enhance host antiviral responses. For example, the African swine fever virus (ASFV) MGF360-12L disrupts KPNA3/4-dependent nuclear transport of NF-κB, thereby suppressing interferon responses (55). Similarly, the FMDV 3C protease degrades KPNA1 to inhibit STAT1/STAT2 nuclear translocation (56). Our study contributes to this paradigm by demonstrating that FMDV counteracts the antiviral function of RPL35 by inducing KPNA3 degradation, which restricts its nuclear trafficking. This reciprocal antagonism exemplifies the evolutionary arms race between host defenses and viral immune evasion.
In conclusion, this study demonstrates that RPL35 negatively regulates FMDV replication by recruiting AMFR, which promotes the ubiquitination and degradation of VP2. Conversely, FMDV induces the degradation of KPNA3, thereby inhibiting RPL35's nuclear translocation (Fig. 10). Our work connects virology fundamentals with drug discovery, offering a roadmap to combat an economically damaging veterinary pathogen.
## MATERIALS AND METHODS
## Cells and viruses
Porcine kidney cell lines (PK-15 and IBRS-2) and baby hamster kidney-21 (BHK-21) cells were cultured in minimum essential medium (MEM, Gibco, USA), whereas human embryonic kidney 293T (HEK293T) cells were maintained in Dulbecco's modified Eagle medium (DMEM, Gibco, USA). All culture media were supplemented with 10% fetal bovine serum (FBS), 1% streptomycin (0.2 mg/mL), and penicillin (200 U/mL). Cells were incubated at 37°C in a humidified atmosphere containing 5% CO 2 . For viral infection experiments, the FMDV O/BY/CHA/2010 strain was used.
## Reagents and antibodies
MG132 was purchased from Merck & Co (Germany), whereas NH 4 Cl and benzyloxycar bonyl (Cbz)-Val-Ala-Asp (OMe)-fluoromethylketone (Z-VAD-FMK) were obtained from Sigma-Aldrich (USA). The following antibodies were used in this study: anti-HA (Biol egend, Cat #901513), anti-Flag (Sigma-Aldrich, Cat #F1804), anti-Myc (Sigma-Aldrich, Cat #M5546), anti-β-actin (Sigma-Aldrich, Cat #A5441), anti-RPL35 (Proteintech, Cat #14826-1-AP), anti-dsRNA (Sigma-Aldrich, Cat #MABE1134), anti-ubiquitin (Cell Signaling Technology (CST), Cat #3933), anti-K48-linkage specific polyubiquitin (CST, Cat #4289), and anti-K63-linkage specific polyubiquitin (D7A11) (CST, Cat #5621). Secondary antibodies included Mouse anti-goat IgG/Alexa Fluor 594 (Invitrogen, Cat #A11005), Rabbit anti-goat IgG/Alexa Fluor 488 (Invitrogen, Cat #A11008), Rabbit anti-goat IgG/ Alexa Fluor 594 (Invitrogen, Cat #A11037), and Mouse anti-goat IgG/Alexa Fluor 488 (Invitrogen, Cat #A28175). Guinea pig anti-FMDV positive serum and mouse anti-VP2 antibody were provided by the Lanzhou Veterinary Research Institute (LVRI).
## Plasmids
The genes encoding VP2, RPL35, RPL35 (deletion of residues 1-30), RPL35 (deletion of residues 31-60), RPL35 (deletion of residues 61-90), RPL35 (deletion of residues 91-123), AMFR, and KPNA1-7 were cloned into the pcDNA3.1/myc-His A or pxj41 vector (Invitro gen, USA) to generate Flag-VP2, Myc-RPL35, Myc-RPL35-Δ1, Myc-RPL35-Δ2, Myc-RPL35-Δ3, Myc-RPL35-Δ4, HA-AMFR, and HA-KPNA1-7, respectively. Additionally, a series of Flag-tagged VP2 mutants (K2R, K3R, K63R, K88R, K159R, K172R, K175R, K198R, and K217R), in which lysine residues were substituted with arginine, were constructed using site-directed mutagenesis PCR. The pCMV-HA-Ub wild-type (WT) and its mutants (K6, K11, K27, K29, K33, K48, and K63) were obtained from BioZY Co., LTD. The primer pairs used for PCR amplification are listed in Table 1. All constructed plasmids were validated by DNA sequencing to ensure accuracy.
## Co-immunoprecipitation assay (Co-IP) and immunoblot analysis
HEK293T cells were cultured in 10-cm dishes until reaching monolayer confluency, followed by co-transfection with the indicated plasmids. After transfection, the cells were lysed in 1 mL of lysis buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 10 µg/mL aprotinin, 10 µg/mL leupeptin, and 1 mM PMSF. The lysates were then incubated with 0.5 mg of the appropriate antibody and 40 µL of protein G-Sepharose (GE Healthcare) in 20% ethanol for 12 h at 4°C. The Sepharose beads were washed three times with 1 mL of lysis buffer supplemented with 500 mM NaCl. The immunoprecipitated proteins were subsequently analyzed by immunoblotting. For immunoblot analysis, the proteins were separated by SDS-PAGE and transferred onto an Immobilon-P membrane (Millipore, USA). The membrane was blocked and probed with the appropriate primary and secondary antibodies. Protein-antibody complexes were visualized using enhanced chemiluminescence (ECL) detection reagents (Thermo, USA).
## Immunofluorescence microscopy
Virus-infected or transfected cells were fixed with 4% paraformaldehyde for 30 min and permeabilized with 0.1% Triton X-100 for 15 min. After blocking with 5% bovine serum albumin (BSA) at 4°C for 4 h, the cells were incubated with the appropriate primary antibody, followed by staining with Alexa Fluor 488-or 594-conjugated secondary antibodies. Cellular images were acquired using a laser-scanning confocal microscope (LSCM, Leica SP8, Solms, Germany).
## Adsorption and internalization assay
In the FMDV adsorption assay, cells overexpressing RPL35 or the empty vector were incubated with FMDV at a multiplicity of infection (MOI) of 10 at 4°C for 1 h. Following adsorption, unbound viruses were removed by washing with ice-cold PBS, and the amount of cell-associated viral RNA was quantified by RT-PCR. For the FMDV internali zation assay, cells were first incubated with FMDV at an MOI of 10 at 4°C for 1 h to allow viral attachment. After removing unbound viruses with ice-cold PBS, the cells were shifted to 37°C for 1 h to promote viral internalization. Non-internalized viruses were then eliminated by treatment with PBS containing proteinase K, and the levels of internalized viral RNA were measured using RT-PCR.
## Primer
Sequence (5' to 3') a Target gene
The italics represent enzymes.
$$VP2-F CGTCTAGCTAGCGATAAGAAAACCGAGGAGAC FMDV VP2 gene VP2-R CGCGGATCCCTCTTTGGAAGGGAACTCAC RPL35-F CGTCTAGCTAGCATGGCCAAGATTAAGGCTCG RPL35 gene RPL35-R CGCGGATCCGGCCTTGACGGCGAACTTCC KPNA1-F CGTCTAGCTAGCATGCTTCCAGAGAAAGACCG KPNA1 KPNA1-R CGCGGATCCAAGCTGGAACCCTTCCATAG KPNA2-F CGTCTAGCTAGCATGTCCACCAATGAGAATGCTA KPNA2 KPNA2-R CGCGGATCCAAAGTTGAAGGTCCCAGTAGC KPNA3-F CGTCTAGCTAGCATGGCCGAGAACCCCGGCTTGGA KPNA3 KPNA3-R CGCGGATCCAAAATTAAATTCTTTCGTTTGAAG KPNA4-F CGTCTAGCTAGCATGGCGGACAGCGAGAAACTGGACA KPNA4 KPNA4-R CGCGGATCCAAACTGGAACCCTTCTGTTGGT KPNA5-F CGTCTAGCTAGCATGAATGCCATGGCTAGTCCAGG KPNA5 KPNA5-R CGCGGATCCAAGGTGAAACCCTTCCATTG KPNA6-F CGTCTAGCTAGCATGTTAGAGACCATGGCGAGCCCAG KPNA6 KPNA6-R CGCGGATCCTAGCTGGAAGCCCTCCATGGGGGCC KPNA7-F CGTCTAGCTAGCATGCCGATTTTAGAAGCTCCCG KPNA7 KPNA7-R CGCGGATCCGGTTTTTGTTAAGGGCGTT a$$
## TCID 50 titration
BHK-21 cells were used to titrate the released infectious virus. The infected cells were harvested at the indicated time post-infection, and the titers were determined in terms of 50% tissue infection dose (TCID 50 )/100 μL by using the Reed-Muench method (57).
## Reverse genetics
The strategy for construction of the plasmid used to produce the recombinant virus was as described previously by our laboratory (58). At 48 h post-transfection, the cell supernatants were harvested by centrifugation at 6,000 × g for 10 min at 4°C and passaged in BHK-21 cells four times. The recovered viruses were named as rVP2-K217R mutant virus and WT virus, which were stored in the National Foot-and-Mouth Disease Reference Laboratory (ABSL-3), Lanzhou Veterinary Research Institute, Chinese Academy of Agricultural Sciences, following the standard protocols and biosafety regulations provided by the Institutional Biosafety Committee. Both viruses were amplified by PCR using the primers used previously, and the PCR products were sequenced using the amplified primers.
## RNA extraction and RT-PCR
Total RNA was isolated utilizing TRIzol Reagent (Invitrogen, USA), followed by cDNA synthesis from the extracted RNA samples employing M-MLV reverse transcriptase (Promega, USA) and random hexamer primers (TaKaRa, Japan). The resulting cDNA served as the template for assessing the expression levels of FMDV RNA and host cellular mRNA. RT-PCR was conducted using the Mx3005P QPCR system (Agilent Technologies, USA) and SYBR Premix ExTaq reagents (TaKaRa, Japan) to quantify RNA levels, with the specific primers detailed in Table 2. The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was utilized as an internal reference control. The relative mRNA expression was determined through the comparative cycle threshold (2 -ΔΔCT ) method.
## RNA interference (RNAi)
The small interfering RNA (siRNA) utilized in the RNA interference (RNAi) experiment was synthetically produced by GenePharma in Shanghai, China. The silencing of endogenous RPL35 was achieved through the transfection of RPL35 siRNA (5′-GAGCUGCUGAAACA ACUGGTT-3′, 5′-GCGUUCUCACCGUCAUCAATT-3′, or 5′-GAGGAGAACCUGAAGACCATT-3′) into PK-15 cells using Polyplus jetPRIME transfection reagent. A non-targeting siRNA (NC siRNA: 5′-UUCUCCGAACGUGUCACGUTT-3′) served as a negative control in the study.
## Cell viability assay
The Cell Counting Kit-8 (CCK-8) from Yeasen was employed to evaluate cell viability in this study. Cells were plated in 96-well plates and incubated for 6-8 h before being treated with 10 μL of CCK-8 solution as per the manufacturer's guidelines. Cell viabil ity was assessed by measuring the absorbance of the cells at OD 450 nm after a 2-h incubation period.
## Mouse infection
For survival analysis, 3-day-old suckling mice were subcutaneously inoculated in the neck with 0.2 mL of PBS, WT, or rVP2-K17R (20LD 50 ) in PBS. The survival rates of all groups (n = 8 per group) were monitored for a period of 7 days. For the histopathological analysis, three mice per group were euthanized on day 3 post-infection to check for lesions in the lungs. For RT-PCR analysis, five mice per group were euthanized on day 3 post-infection to check for virus replication in the muscle, heart, liver, spleen, lung, kidney, and duodenum.
## Histological assessment
Following euthanasia of the mice, their tissues were gathered and promptly preserved in 10% neutral-buffered formalin. The preserved tissues were then embedded in paraffin, sectioned, and subjected to staining with hematoxylin and eosin (H&E) for subsequent histopathological examinations, as conducted by Wuhan Servicebio Technology Co., Ltd.
## Statistical analysis
All data were presented as means ± SD and analyzed using GraphPad Prism software (version 10.1.0). Individual statistical tests are specified within the figure legends. For data with two groups, unpaired Student's t tests were used under the assumption of normality. Data with more than two groups were analyzed by analysis of variance (ANOVA) under the assumption of normality. In general, at least three independent biological replicates (n) were carried out for each experiment. Data were reproduced in independent experiments as indicated in the legends. Significant differences are denoted in the figures as follows: *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; n.s., no statistical significance.
## References
1. Weaver, Domenech, Thiermann et al. (2013) "Foot and mouth disease: a look from the wild side" *J Wildl Dis*
2. Poonsuk, Giménez-Lirola, Zimmerman (2018) "A review of foot-andmouth disease virus (FMDV) testing in livestock with an emphasis on the use of alternative diagnostic specimens" *Anim Health Res Rev*
3. Mahapatra, Parida (2018) "Foot and mouth disease vaccine strain selection: current approaches and future perspectives" *Expert Rev Vaccines*
4. Park (2013) "Requirements for improved vaccines against foot-andmouth disease epidemics" *Clin Exp Vaccine Res*
5. Mason, Grubman, Baxt (2003) "Molecular basis of pathogenesis of FMDV" *Virus Res*
6. Zhang, Yang, Yang et al. (2024) "KIF5B-mediated internalization of FMDV promotes virus infection" *Virol Sin*
7. Ashkani, Rees (2016) "The critical role of VP1 in forming the necessary cavities for receptor-mediated entry Of FMDV to the host cell" *Sci Rep*
8. Salem, El-Kholy, Omar et al. (2019) "Construction, expression and evaluation of recombinant VP2 protein for serotype-independent detection of FMDV seropositive animals in Egypt" *Sci Rep*
9. Ganji, Biswal, Lalzampuia et al. (2018) "Mutation in the VP2 gene of P1-2A capsid protein increases the thermostability of virus-like particles of foot-and-mouth disease virus serotype O" *Appl Microbiol Biotechnol*
11. Hong, Qian, Li et al. (2007) "A recombinant pseudora bies virus co-expressing capsid proteins precursor P1-2A of FMDV and VP2 protein of porcine parvovirus: a trivalent vaccine candidate" *Biotechnol Lett*
12. Zhang, Li, Yang et al. (2025) "RING finger protein 5 is a key anti-FMDV host factor through inhibition of virion assembly" *PLoS Pathog*
13. Carrillo, Tulman, Delhon et al. (2005) "Comparative genomics of foot-and-mouth disease virus" *J Virol*
14. Mushtaq, Shah, Zarlashat et al. (2024) "Cell culture adaptive amino acid substitutions in FMDV structural proteins: a key mechanism for altered receptor tropism" *Viruses*
15. Dong, Zhang, Kuang et al. (2021) "Selective regulation in ribosome biogenesis and protein production for efficient viral translation" *Arch Microbiol*
16. Akhter, Hossain, Kitab et al. (2025) "Common host factors for internal ribosomal entry site-mediated translation of viral genomic RNA: an investigation in foot-and-mouth disease and classical swine fever viruses" *Virus Res*
17. Cardinali, Fiore, Campioni et al. (1999) "Resistance of ribosomal protein mRNA translation to protein synthesis shutoff induced by poliovirus" *J Virol*
18. Malygin, Kossinova, Shatsky et al. (2013) "HCV IRES interacts with the 18S rRNA to activate the 40S ribosome for subsequent steps of translation initiation" *Nucleic Acids Res*
19. Byrd, Zamora, Lloyd (2005) "Translation of eukaryotic translation initiation factor 4GI (eIF4GI) proceeds from multiple mRNAs containing a novel cap-dependent internal ribosome entry site (IRES) that is active during poliovirus infection" *J Biol Chem*
20. Zhu, Li, Zhang et al. (2020) "Foot-and-mouth disease virus capsid protein VP1 interacts with host ribosomal protein SA to maintain activation of the MAPK signal pathway and promote virus replication" *J Virol*
21. Chen, Lu, Zhang et al. (2016) "Ribosomal protein L4 interacts with viral protein VP3 and regulates the replication of infectious bursal disease virus" *Virus Res*
22. Li, Li, Zhou (2018) "Ribosomal protein L18 is an essential factor that promote rice stripe virus accumulation in small brown planthopper" *Virus Res*
23. Beyer, Bann, Rice et al. (2013) "Nucleolar trafficking of the mouse mammary tumor virus gag protein induced by interaction with ribosomal protein L9" *J Virol*
24. Han, Sun, Li et al. (2020) "Ribosomal protein L13 promotes IRES-driven translation of footand-mouth disease virus in a helicase DDX3-dependent manner" *J Virol*
25. Li, Dong, Shi et al. (2016) "Rabies virus phosphoprotein interacts with ribosomal protein L9 and affects rabies virus replication" *Virology (Auckl)*
26. Abbas, Dichamp, Herbein (2012) "The HIV-1 Nef protein interacts with two components of the 40S small ribosomal subunit, the RPS10 protein and the 18S rRNA" *Virol J*
27. Cao, Yang, Zheng et al. (2019) "Classical swine fever virus non-structural proteins modulate Toll-like receptor signaling pathways in porcine monocyte-derived macrophages" *Vet Microbiol*
28. Mazumder, Poddar, Basu et al. (2014) "Extraribosomal l13a is a specific innate immune factor for antiviral defense" *J Virol*
29. Zhang, Xie, Xia et al. (2019) "Zika virus NS2A-mediated virion assembly" *mBio*
30. Matsuda, Nakano (1998) "RMA1 an Arabidopsis thaliana gene whose cDNA suppresses the yeast secl5 mutation, encodes a novel protein with a RING finger motif and a membrane anchor" *Plant and Cell Physiology*
31. Jiang, Chen (2011) "The role of ubiquitylation in immune defence and pathogen evasion" *Nat Rev Immunol*
32. Ardley, Robinson (2005) "E3 ubiquitin ligases" *Essays Biochem*
33. Wang, Liu, Cui et al. (2014) "The E3 ubiquitin ligase AMFR and INSIG1 bridge the activation of TBK1 kinase by modifying the adaptor STING" *Immunity*
34. Chen, Wang, Tai (2008) "The role of expansion segment of human ribosomal protein L35 in nuclear entry, translation activity, and endoplasmic reticulum docking" *Biochem Cell Biol*
35. Peña, Hurt, Panse (2017) "Eukaryotic ribosome assembly, transport and quality control" *Nat Struct Mol Biol*
36. Baßler, Hurt (2019) "Eukaryotic ribosome assembly" *Annu Rev Biochem*
37. Greber (2016) "Mechanistic insight into eukaryotic 60S ribosomal subunit biogenesis by cryo-electron microscopy" *RNA*
38. Broeck, Klinge (2024) "Eukaryotic ribosome assembly" *Annu Rev Biochem*
39. Yang, Guo, Chen et al. (2023) "Nuclear transport proteins: structure, function and disease relevance" *Sig Transduct Target Ther*
40. Oka, Yoneda (2018) "Importin α: functions as a nuclear transport factor and beyond" *Proc Jpn Acad Ser B Phys Biol Sci*
41. Miyamoto, Yamada, Yoneda (2016) "Importin α: a key molecule in nuclear transport and non-transport functions" *J Biochem*
42. Lindström (2009) "Emerging functions of ribosomal proteins in genespecific transcription and translation" *Biochem Biophys Res Commun*
43. Murakami, Singh, Morris et al. (2018) "The interaction between the ribosomal stalk proteins and translation initiation factor 5B promotes translation initiation" *Mol Cell Biol*
44. Lee, Wu, Wu et al. (2020) "The RNA-dependent RNA polymerase of enterovirus A71 associates with ribosomal proteins and positively regulates protein translation" *RNA Biol*
45. Li (2019) "Regulation of ribosomal proteins on viral infection" *Cells*
46. Fu, Wang, Ding et al. (2015) "TRIM32 senses and restricts influenza A virus by ubiquitination of PB1 polymerase" *PLoS Pathog*
47. Sui, Zheng, Fu et al. (2024) "TRIM72 restricts lyssavirus infection by inducing K48-linked ubiquitination and proteasome degradation of the matrix protein" *PLoS Pathog*
48. Fan, Mar, Sari et al. (2021) "TRIM7 inhibits enterovirus replication and promotes emergence of a viral variant with increased pathogenicity" *Cell*
49. Lin, Wang, Chen et al. (2023) "TRIM21 restricts influenza A virus replication by ubiquitinationdependent degradation of M1" *PLoS Pathog*
50. Yang, Wang, Li et al. (2024) "RPLP1 restricts HIV-1 transcription by disrupting C/EBPβ binding to the LTR" *Nat Commun*
51. Jiang, Hu, Liu et al. (2015) "60S ribosomal protein L35 regulates β-casein translational elongation and secretion in bovine mammary epithelial cells" *Arch Biochem Biophys*
52. Thomsen, Skouboe, Møhlenberg et al. (2025) "Length Text Journal of Virology November"
53. Me, Zhang, Mikkelsen et al. (2024) "Impaired STING activation due to a variant in the E3 ubiquitin ligase AMFR in a patient with severe VZV infection and hemophagocytic lymphohistiocytosis" *J Clin Immunol*
54. Abdelnabi, Geraets, Ma et al. (2019) "A novel druggable interprotomer pocket in the capsid of rhino-and enteroviruses" *PLoS Biol*
55. Schlicksup, Zlotnick (2020) "Viral structural proteins as targets for antivirals" *Curr Opin Virol*
56. Feng, Gao, Han et al. (2018) "Discovery of small molecule therapeutics for treatment of chronic HBV infection" *ACS Infect Dis*
57. Zhuo, Guo, Ba et al. (2021) "African swine fever virus MGF360-12L inhibits type i interferon production by blocking the interaction of importin α and NF-κB signaling pathway" *Virol Sin*
58. Du, Bi, Liu et al. (2014) "3C pro of foot-and-mouth disease virus antagonizes the interferon signaling pathway by blocking STAT1/STAT2 nuclear translocation" *J Virol*
59. Zamoiskii Ea (1956) "Evaluation of Reed-Muench method in determina tion of activity of biological preparations" *Zh Mikrobiol Epidemiol Immunobiol*
60. Yang, Zhu, Cao et al. (2020) "Genetic determinants of altered virulence of type O foot-and-mouth disease virus" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12643176&blobtype=pdf | # Immunogenicity of JN.1 and KP.2 COVID-19 mRNA vaccines against emerging SARS-CoV-2 variants
Ninaad Lasrado, Annika Rössler, Isabella Mcconnell, Katherine Molloy, Ritobhas Bhowmik, Christine Happle, Ruoran Guan, Katherine Mcmahan, Juliana Pereira, Jinyan Liu, Erica Borducchi, Liping Wang, Krishna Shah, Bridget Wixted, Metodi Stankov, Alexandra Dopfer-Jablonka, Ai-Ris Collier, Georg Behrens, Dan Barouch
## Abstract
Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) continues to evolve five years after the initial outbreak. Although mRNA vaccines encoding the JN.1 and KP.2 Spike proteins were authorized in fall 2024, it remains unclear whether vaccine updates will be necessary for variants containing antigenically closely related Spike proteins. In this study, we evaluated the immunogenicity of JN.1 and KP.2 mRNA boosters in participants from Germany and the United States, respectively. Both vaccines induced robust and similar neutralizing antibody (NAb)
## Introduction
In fall 2024, the European Medicines Agency (EMA) and US Food and Drug Administration (FDA) approved COVID-19 mRNA vaccines expressing SARS-CoV-2 Spike from the JN.1 and KP.2 variants, respectively. Initial reports indicate that these vaccines successfully raise neutralizing antibody (NAb) titers against various JN.1 sublineages and provide protection against severe disease [1][2][3][4][5] . However, a debate has emerged regarding the comparative immunogenicity of antigenically related COVID-19 vaccines and the potential utility of updating COVID-19 vaccines for closely related variants, such as current circulating SARS-CoV-2 variants (Fig. S1). In this study, we evaluated the immunogenicity of the JN.1 and KP.2 COVID-19 mRNA boosters.
## Results and Discussion
We assessed immune responses in 40 individuals who received the JN.1 mRNA booster in Hannover, Germany and in 63 individuals who received the KP.2 mRNA booster in Boston, USA in fall 2024 (Tables S1,S2). Participants had a median of 5 COVID-19 vaccine doses prior to the JN.1 or KP.2 mRNA booster, and at least 90% of participants in the JN.1 cohort and 69% in the KP.2 cohort had at least one documented SARS-CoV-2 infection, although we expect the true rate of natural infection to be substantially higher 6 .
We evaluated NAb responses in participants who received the KP.2 mRNA boost against historical SARS-CoV-2 variants as well as JN.1, KP.2, and recently circulating subvariants LB.1, XEC, LP.8.1.1, and NB.1.8.1 (Fig. 1A). NAb responses increased at 3 weeks after the boost against all tested variants (Fig. 1A), with a 2.6-and 4.5-fold increase against JN.1 and KP.2, respectively, and a 20.1-and 23.8-fold increase in NAb titers against LP.8.1.1 and NB.1.8.1, respectively (Fig. 1B). In participants who received the KP.2 mRNA boost, NAb titers similarly increased against all tested variants (Fig. 1C), with a 15.3-and 11.7-fold increase against JN.1 and KP.2, respectively, and a 6.0-and 13.7-fold increase in NAb titers against LP.8.1.1 and NB.1.8.1, respectively (Fig. 1D). These data demonstrate that both the JN.1 and KP.2 mRNA vaccines boosted peripheral NAb responses against all current circulating variants tested. The different fold increases in the two cohorts likely reflected the higher baseline NAb titers against JN.1 and KP.2 in the JN.1 booster cohort, presumably reflecting differences in natural infection history.
We next assessed NAb responses to other sarbecoviruses, including SARS-1, WIV-1, Rs4874, BANAL-20-103, PCoV-GD1-2019, and hCoV-NL63. Both cohorts showed antibodies against these diverse sarbecoviruses at baseline, suggesting acquired natural immunity, but these responses were not substantially increased by mRNA boosting (Fig. S2). These data suggest that a broad pan-sarbecovirus vaccine will likely require a different antigen strategy than sequential SARS-CoV-2 boosters. We also evaluated binding IgG responses against SARS-CoV-2 WA1/2020, B.1.617.2, BA.5, XBB.1.5, JN.1 and KP.2 by ELISA and ECLA assays, which showed substantial increases to all variants following both JN.1 and KP.2 mRNA boosting (Figs. S3,S4). In contrast, we observed no increases in WA1/2020, BA.5, XBB.1.5, or JN.1 Spike-specific cellular immune responses following KP.2 mRNA boosting (Fig. S5), consistent with prior reports 7,8 .
As mucosal immunity is likely important for preventing SARS-CoV-2 acquisition, we evaluated mucosal NAb responses in nasal swabs following JN.1 or KP.2 mRNA boosting and observed minimal to no increases in nasal NAb titers (Fig. 2), similar to previous findings 8 . We observed a modest increase in mucosal IgG but no increase in mucosal IgA (Fig. S6), indicating that current intramuscular COVID-19 vaccines do not induce robust mucosal antibody responses 8 .
Finally, we evaluated the durability of humoral immune responses at 6 months following KP.2 mRNA boosting. NAb responses waned over this period but remained higher in individuals who received the boost compared with individuals who did not receive a boost, including against currently circulating variants (Fig. 3). These data suggest that it may not be necessary to time boosting precisely in populations with high levels of baseline immunity.
Our findings indicate that both the JN.1 and KP.2 mRNA vaccines substantially increased peripheral, but not mucosal, antibody responses against current circulating SARS-CoV-2 variants, including LP.8.1.1 and NB.1.8.1. The similar profile of NAb responses induced by these two antigenically related mRNA vaccines suggests that minor differences in the vaccine antigen sequence, especially for variants within the same antigenic cluster as defined by antigenic cartography [9][10][11] , may not have a major impact on immunogenicity. These data suggest that vaccine strain updates may not be necessary for closely related circulating variants in the absence of a major antigenic shift. The FDA and EMA suggested manufacturers produce any JN.1 lineage vaccines for fall 2025 and consider LP.8.1 12,13 . Our data support these recommendations, and the immunologic cross-reactivity between JN.1, KP.2, LP.8.1, and JN.1-related variants suggest that minor strain updates for variants within the same antigenic cluster may provide only modest additional benefits in the context of widespread hybrid immunity.
## References
1. Happle, Hoffmann, Kempf (2024) "Humoral immunity after mRNA SARS-CoV-2 omicron JN.1 vaccination" *The Lancet Infectious Diseases*
2. Wang, Mellis, Wu (2025) "KP.2-based monovalent mRNA vaccines robustly boost antibody responses to SARS-CoV-2. The Lancet Infectious Diseases"
3. Suthar, Manning, Ellis (2025) "The KP.2-adapted COVID-19 vaccine improves neutralising activity against the XEC variant" *The Lancet Infectious Diseases*
4. Zhang, Kempf, Nehlmeier (2025) "Host cell entry and neutralisation sensitivity of the emerging SARS-CoV-2 variant LP.8.1. The Lancet Infectious Diseases"
5. Hansen, Lassaunière, Rasmussen et al. (2025) "Effectiveness of the BNT162b2 and mRNA-1273 JN.1-adapted vaccines against COVID-19-associated hospitalisation and death: a Danish, nationwide, register-based, cohort study. The Lancet Infectious Diseases"
6. Nkolola, Liu, Collier (2024) "High Frequency of Prior Severe Acute Respiratory Syndrome Coronavirus 2 Infection by Sensitive Nucleocapsid Assays" *J Infect Dis*
7. Lasrado, Collier, Miller (2024) "Waning immunity and IgG4 responses following bivalent mRNA boosting" *Science Advances*
8. Lasrado, Rowe, Mcmahan (2024) "SARS-CoV-2 XBB.1.5 mRNA booster vaccination elicits limited mucosal immunity" *Science Translational Medicine*
9. Rössler, Netzl, Lasrado (2025) "Nonhuman primate antigenic cartography of SARS-CoV-2" *Cell Rep*
10. Mellis, Wu, Hong (2025) "Antibody evasion and receptor binding of SARS-CoV-2 LP.8.1.1, NB.1.8.1, XFG, and related subvariants" *Cell Reports*
11. Guo, Yu, Liu (2025) "Antigenic and virological characteristics of SARS-CoV-2 variants BA.3.2, XFG, and NB.1.8.1. The Lancet Infectious Diseases"
12. Fda (2025) "COVID-19 Vaccines for 2025-2026 Formula for Use in the United States Beginning in Fall"
13. Ema (2025) "EMA recommendation to update the antigenic composition of authorised COVID-19 vaccines for" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12259239&blobtype=pdf | for just about anything from solo traveling and food history to language learning and sustainable gardening. Biomedical researchers consume a substantial part of this virtual space, publishing a plethora of podcasts ranging from microbiology to clinical psychology. Some podcasts are even purely dedicated to discussing the daily trials and tribulations in the lab. Everyone or what questions to ask, let alone who sat at the table in each field. However, podcasts allowed me to sit at various tables with full anonymity and unlimited curiosity. No matter where I was and what I was doing, I could listen to scientists in diverse fields talk about their research and careers. All I needed was internet access. This gave me a full picture of what it was like to be a research scientist. Specifically, I learned how past research inspired current investigations; I virtually "met" and heard the career stories of scientists from across the world; and I gained an invaluable understanding of all the hats researchers wear inside and outside the lab. I also learned to not just absorb information, but also critically evaluate it, which is something I was never taught in the classroom. This was the immersive learning experience I needed to determine if research was right for me, an experience undoubtedly shared with many other young scientists. Thus, after a few intriguing episodes about the 2013 West Africa Ebola virus epidemic, virology claimed my interest, and I started the journey I'm on today.
Yet podcasts are so much more than highly accessible, geographically unrestricted outlets for aspiring scientists. They're a multifaceted tool for scientists of all stages, backgrounds, and resources. First, constant access to diverse thoughts on current topics may very well help you troubleshoot that assay or write that grant you've been working on for months, because you can think outside the box figuratively and literally. And that's just because you decided to listen to something on your morning commute. Subsequently, your comfort with diving into different fields can foster much-needed interdisciplinary interactions to answer bigger questions that have been unanswered for decades, like chronic disease etiology. In the end, regardless of whether the interactions are in-person or not, this results in you having a more highly connected research community that may span the globe. Engaging with research via dialogue becomes increasingly important as we navigate an era inundated with scientific (mis)information: Only direct discussion with the original sources and their accompanying experts can truly help us navigate this crisis. We might not agree with the perspectives we hear on podcasts about this, but they offer us opportunities to practice communicating more than just what we printed in our Cell article. We ultimately walk away from each listening session with better critical thinking, networking, and communication skills.
But most of all, podcasts are a way to connect and support researchers on a personal, not just educational and professional, level. Being a researcher is tough. We don't realize until we're kneedeep that failure and struggle are routine, predictability is rare, and external factors can work against us. We can easily feel isolated, since emphasis is put on our scientific contributions rather than our well-being. No one seems to talk about it much either. But hearing podcast hosts discuss their stress during a period of continually failed experiments and guest speakers describe how continual manuscript rejection triggered their depression opens the conversation to how we can improve our research by starting with both the individual and how and what we communicate with each other. After all, we can only do good research when we're in the right head space to think through challenges and celebrate successes together. Having a faceless environment to start this conversation is vital for us to sustain a strong, inclusive research community.
Therefore, podcasts serve as tools to teach, inspire, connect, and help us grow as an integrated community that will continue to thrive as creative thinkers, excellent communicators, and supportive colleagues. So, is all this chatter worth it? Absolutely. I hope you continue listening.
Are you listening? |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12797530&blobtype=pdf | # N-terminal insertion peptides produce varied effects over the cleavage and assembly of norovirus major capsid protein VP1
Yuqi Huo, Jie Ma, Jinjin Liu
## Abstract
of expression systems [1][2][3]. These VLPs are structurally and antigenically similar to native virions, making them valuable tools for vaccine development, diagnostic assays, and structural studies. Structurally, VP1 is divided into two distinct domains: the N-terminal shell (S) domain and the C-terminal protruding (P) domain [4]. Interestingly, both the S and P domains are capable of independently forming smaller particles, known as subviral particles (SVPs) [5,6]. Among these, the P domainformed particles, often referred to as P particles, exhibit superior stability compared to those formed by the S domain. This is evidenced by the abundance of P particles observed in fecal samples [7,8]. The P domain is particularly significant as it harbors critical antigenic epitopes. When antibodies bind to these epitopes, they block the interaction between the P domain and histo-blood group
## Introduction
Norovirus (NoV) is a leading cause of acute gastroenteritis worldwide, responsible for both sporadic cases and outbreaks in various settings. The virus is highly contagious and poses a significant public health burden. The major capsid protein VP1 encoded by the second open reading frame (ORF2) of the viral genome, has the remarkable ability to self-assemble into virus-like particles (VLPs) when expressed in vitro using a variety antigens (HBGAs), which are presumed to function as co-receptors for NoV [6,9].
A common observation during the expression of VP1 is its association with N-terminal cleavage, which is often followed by assembly into SVPs [10,11]. This phenomenon has been consistently reported across various expression systems, suggesting the involvement of a conserved pathway. However, no specific protease has been definitively linked to this cleavage. Typically, the N-terminal 44 and 38 amino acids are lost in GI and GII VP1 proteins, respectively [11,12]. In the context of expressed VP1 proteins undergoing N-terminal cleavage, two types of particles are observed: VLPs and SVPs. Both cleaved and uncleaved forms of VP1 participate in the assembly of these VLPs and SVPs, indicating that the final morphology of the assembled products is determined by interactions between dimers or higher-order oligomers [13].
The assembly process of NoV VP1 is hypothesized to involve intermediates mediated by dimers and pentamers of dimers [4]. Time-resolved small-angle X-ray scattering studies have revealed the presence of only two intermediates during assembly: dimers and species composed of eleven dimers [14]. While it is likely that the cleavage of expressed VP1 occurs prior to the formation of mature VLPs, the precise mechanisms by which these intermediates become susceptible to cleavage remain elusive. As N-terminal sequence of VP1 plays a regulatory role in VLP assembly, the addition of extra short peptide sequences to the N-terminus followed with analysis of cleavage patterns and assembly products might provide insights into the mechanisms of cleavage and assembly. To this end, we expressed and characterized a series of recombinant VP1 proteins and our results indicate that the cleavage is dependent on N-terminal insertion sequences.
## Materials and methods
## Generation of recombinant baculoviruses
To generate recombinant baculoviruses, 75-amino acid peptide sequences derived from the VP1 of previously reported GI.7 (GenBank accession number, JN899243) and GII.3 (GenBank accession number, KY767664) strains, as well as the VP2 of a GII.4 strain (GenBank accession number, KF306214), were fused to the N-terminus of the VP1 protein (designated as GII.6-AB) from a GII.6 strain (GenBank accession number, AB818400). These sequences were codon-optimized, synthesized, and ligated into the multi-cloning site downstream of the polyhedrin promoter in the pFastBac-Dual vector. To investigate whether cleavage occurs at specific sites, a 74-amino acid peptide sequence containing an N-terminal FLAG tag was similarly codon-optimized, synthesized, and inserted into the pFastBac-Dual vector. These constructs were used to generate donor plasmids, which were then transformed into competent E.coli DH10Βac cells to produce bacmids. The bacmids were purified and subsequently used to transfect Sf9 cells. Approximately 3-5 days post-transfection, the P1 baculoviruses were harvested and used to infect fresh Sf9 cells to produce P2 baculoviruses.
## Infection of Sf9 cells and purification of VLPs
Sf9 cells were infected with the recombinant baculoviruses, and the expressed proteins were purified as previously described with minor modifications [15]. Briefly, Sf9 cells were cultured in SF-SFM culture medium (Womei Biology, China) at 27 °C. For infection, cells were infected with recombinant baculoviruses at a multiplicity of infection (MOI) of 5. After 5-7 days post-infection, the medium was harvested by centrifuging at 10,000 g for 30 min at 4 °C to remove cell debris. The VLPs in the supernatant were first precipitated by centrifuging at 137,800 g in a P32ST-0159 rotor (CP-80NX, Hitachi, Japan) for 3 h at 8 °C and then further separated using cesium chloride density gradient centrifugation at 284,000 g in a P40ST-2588 rotor (CP-80NX, Hitachi, Japan) for 18 h at 8 °C. The protein bands were collected and further precipitated by centrifuging at 137,800 g in a P32ST-0159 rotor for 3 h at 8 °C. Protein pellet was resuspended in 0.01 M phosphate buffered saline (PBS, pH7.2). The integrity and assembly of the purified proteins were assessed using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transmission electron microscopy (TEM) (JEM-1400, JEOL, Japan) following negative staining using phosphotungstic acid at a concentration of 0.25 mg/mL.
## SDS-PAGE and Western blot (WB) analysis
For SDS-PAGE analysis, purified VLPs (2 µg) or appropriate amount of cell lysates were boiled and loaded onto a discontinuous 8-16% precast gel. After electrophoresis, the proteins were visualized using Coomassie blue staining. For WB analysis, the separated proteins were transferred to a nitrocellulose membrane and sequentially incubated with (a) rabbit anti-GII.6 VLP hyperimmune serum (1:2,000), or mouse anti-FLAG mAb (1:1,000), followed by (b) horseradish peroxidase (HRP)-conjugated goat anti-rabbit or anti-mouse IgG antibodies. Protein bands were detected using 3,3'-diaminobenzidine (DAB) as the color-developing agent. Recombinant baculoviruses produced using empty donor plasmid were used as wild type viruses. Sf9 cell was infected with wild type viruses and lysed as a negative control.
## Peptide mass fingerprinting (PMF) analysis
The proteins were reduced by 10 mM DTT at 56 °C for 30 min, alkylated by 50 mM iodoacetamide (IAA) at room temperature in the dark for 15 min, sequentially digested by trypsin, chymotrypsin, and endoproteinase Glu-C, and desalted. The peptides were first separated using an Easy-nLC 1200 system with a Trap column packed with PepMap100 C18 (3 μm, 75 μm x 2 cm, Ther-moFisher Scientific, USA). Mass analysis was performed using an Obitrap Fusion Lumos Mass Spectrometer (ThermoFisher Scientific, USA). The raw MS files were analyzed and searched against target protein database based on the species of the samples using PEAKS Studio.
## Results
## Selection of peptide sequences for Recombinant protein expression
In this study, peptide sequences derived from the VP1 of previously reported GI.7 and GII.3 strains, as well as the VP2 of the GII.4 strain, were selected for recombinant protein expression [16][17][18]. This selection was primarily motivated by the availability of a panel of GI-and GIIspecific monoclonal antibodies, and these recombinant proteins could be used to identify their binding epitopes (linear B-cell epitopes) [19]. Secondly, sequences derived from different regions of VP1 might have varied functions (initiation of assembly or stabilization of assembled VLPs) in VLP assembly, we were thus interested in how short homologous sequence insertion from a different strain to the N-terminus could affect the cleavage and assembly of the GII.6-AB proteins. Additionally, we encountered difficulties in the expression of VP2 proteins with either prokaryotic or eukaryotic expression systems. Therefore, recombinant proteins containing VP2 sequences were intended for the production of polyclonal antibodies, which could subsequently be employed to assess the expression status of VP2 (Fig. 1). The chosen peptide length of 75 amino acids was based on the observation that the GII.6-AB can accommodate this length without compromising its expression level [18].
## Effects of GI.7-derived sequences on cleavage and assembly of Recombinant GII.6-AB proteins
A total of eight recombinant proteins (designated as GII.6-AB/P1 through GII.6-AB/P8) were designed, each containing sequences derived from the VP1 of the GI.7 strain. These insertion peptides spanned the entire region of the GI.7 VP1, with each adjacent peptide sharing a five-amino acid overlap, except for GII.6-AB/P1 Fig. 1 Schematic representation of constructs used in this study and GII.6-AB/P2, which had a 25-amino acid overlap. All eight recombinant proteins were purified to high purity. SDS-PAGE and WB analysis revealed extensive cleavage for GII.6-AB/P1, GII.6-AB/P4, GII.6-AB/P5, GII.6-AB/ P6, and GII.6-AB/P7, while GII.6-AB/P2, GII.6-AB/P3, and GII.6-AB/P8 exhibited either reduced or no cleavage (Fig. 2). The cleavage could be summarized as a fourband pattern (based on SDS-PAGE, B1 to B4) with some recombinant proteins missing one, two or three bands. However, two primary protein bands corresponding to B3 and B4 were generally observed for most recombinant proteins. The lower migrating bands between 15 and 35 kDa (e.g., GII.6-AB/P6 and GII.6-AB/P7) were most possibly derived from the cleaved products in the predicted surface-exposed loop region as reported previously. The migrating band with MW close to 55 kDa requires further determination.
TEM analysis confirmed the successful assembly of all recombinant proteins into VLPs or SVPs, though significant morphological differences were observed (Fig. 3 and Supplementary Fig. 1). GII.6-AB/P1, GII.6-AB/P5, GII.6-AB/P6, GII.6-AB/P10, and GII.6-AB/P11 formed a mixture of VLPs and SVPs, whereas GII.6-AB/P3 and GII.6-AB/P8 assembled into homogeneous VLPs or SVPs that tend to aggregate. Notably, GII.6-AB/P2 assembled into homogeneous VLPs, but accompanied with evenly distributed protein complexes, likely resulting from disassembled VLPs or SVPs. GII.6-AB/P3 formed VLPs resistant to staining reagents (metal salts unable to enter into formed VLPs), while GII.6-AB/P8 assembled into VLPs resembling protein subunit aggregates.
## Cleavage patterns and assembly products of Recombinant proteins with GII.4 VP2 and GII.3 VP1 sequences
In a previous study, we expressed a recombinant GII.6-AB protein with an N-terminal insertion peptide derived from the GII.4 VP2 [20]. Here, we extended this approach by using additional regions of the GII.4 VP2 for recombinant protein expression (designated as GII.6-AB/P10, GII.6-AB/P11, and GII.6-AB/P12). Additionally, to characterize the predicted surface-exposed loop region of the GII.3 VP1, we expressed a recombinant protein containing this region (designated as GII.6-AB/P9). All four recombinant proteins were successfully purified using CsCl density gradient centrifugation. SDS-PAGE and WB analysis revealed varied cleavage patterns: GII.6-AB/P12 displayed two primary bands, while GII.6-AB/P9 exhibited a pattern similar to that of GII.6-AB/P3. In contrast, GII.6-AB/P10 and GII.6-AB/P11 showed a cleavage pattern akin to GII.6-AB/P1, GII.6-AB/P4, GII.6-AB/P5, GII.6-AB/P6, and GII.6-AB/P7 (Fig. 2).
TEM analysis revealed significant morphological differences among the four recombinant proteins. GII.6-AB/P9 and GII.6-AB/P12 assembled into VLPs with similar morphology, though GII.6-AB/P12 contained a larger proportion of full VLPs. GII.6-AB/P10 formed VLPs of varied shapes that tend to aggregate, a character similar to that observed for GII.6-AB/P3 and GII.6-AB/ P8. GII.6-AB/P11 assembled into VLPs exhibiting typical characteristics of cleaved VP1 proteins, reflected as a mixture of VLPs and SVPs (Fig. 3).
Fig. 2 Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot analysis. CsCl-density gradient-purified VLPs were loaded on precast gels for SDS-PAGE (A) and WB analysis (B). Rabbit anti-GII.6-AB hyperimmune serum (1:1,000) were used. Supernatant from lysed sf9 cells infected with recombinant baculoviruses produced by empty donor vector was used as negative control. Primary cleaved products are indicated with arrows. A typical four-band pattern (e.g., GII.6-AB/P1, bands B1-B4 with approximate molecular weights of 67, 62, 59, and 56 kDa, respectively) was observed, with some recombinant proteins missing one or more of these bands
## PMF analysis demonstrated conserved cleavage site at the N-terminus
To determine the exact residues of both N-and C-terminus of cleaved VP1 proteins, PMF analysis was performed for GII.6-AB/P1 proteins as a representative. The two protein bands corresponding to the full-length GII.6-AB (B-3, as indicated in Fig. 2A) and cleaved product (B-4, as indicated in Fig. 2A) were selected. Based on the results, the N-terminal residue started at residue 3 (A) and ended at residue 547 (Q) for B-3 when GII.6-AB was selected as reference, and the N-terminal residue started at residue 35 (A) and ended at residue 547 (Q) for B-4 (Fig. 4A-D).
## N-terminal FLAG-tagged fragment was not detectable
To investigate whether N-terminal cleavage is mediated by a specific protease, we introduced a FLAG tag at the N-terminus of the GII.6-AB/P1 protein (designated as GII.6-AB/P13). This approach was based on the identification of a proteolytic fragment characteristic of typical VP1 processing. If cleavage was mediated by a specific protease, a fragment of 11.6 kDa (corresponding to the N-terminal 113 amino acids) should be detectable using anti-FLAG antibodies. Sf9 cells infected with the recombinant baculovirus encoding GII.6-AB/P13 were harvested three days post-infection and lysed for WB analysis. While a distinct band corresponding to the fulllength GII.6-AB/P13 protein was consistently detected on the WB membranes, no specific bands corresponding to the predicted 11.6 kDa fragment were observed, regardless of whether rabbit anti-GII.6 VLP hyperimmune serum or mouse anti-FLAG antibody was used for detection (Fig. 5). No such band was detected in wildtype baculovirus-infected Sf9 cells, confirming the specificity of the results.
## Discussion
The N-terminal cleavage of NoV major capsid protein VP1 during in vitro expression is a well-documented phenomenon, yet its underlying mechanisms and functional implications remain poorly understood. In this study, we systematically investigated sequence variations associated with this cleavage by expressing a series of recombinant GII.6 VP1 proteins fused with peptide sequences derived from the VP1 of GI.7 and GII.3 strains, as well as the VP2 of GII.4 strain. Our previous study focuses on the extra sequence length tolerance of GII.6 VP1 N-terminus, this work further explores this phenomenon with varied sequences [18]. Our findings provide new insights into the cleavage patterns and its impact on the morphology of VLPs.
One of the key findings of this study is the nature of irregularity of the N-terminal cleavage observed in recombinant VP1 proteins. The varied cleavage patterns observed among the recombinant proteins, as evidenced by SDS-PAGE and WB analysis, provides no evidence that the cleavage was mediated by a single specific protease. This conclusion is further supported by the failure to detect an N-terminal FLAG-tagged fragment in the GII.6-AB/P13 construct, which was designed to test the hypothesis of protease-specific cleavage. The absence of the predicted 11.6 kDa fragment, despite the presence of the full-length protein, indicates that the cleavage likely occurs through sequential order, possibly involving multiple host cell proteases or other environmental factors.
Another finding is that there is a clear correlation between the extent of N-terminal cleavage and the morphology of the assembled VLPs, as assessed by SDS-PAGE and TEM. Proteins exhibiting extensive cleavage (GII.6-AB/P1, GII.6-AB/P4, GII.6-AB/P5, GII.6-AB/P6, GII.6-AB/P7, GII.6-AB/P10, and GII.6-AB/P11) formed a mixture of VLPs and SVPs, indicating that cleavage interferes with the proper assembly of VP1 into well-defined VLPs. In contrast, proteins with minimal or no cleavage (GII.6-AB/P2, GII.6-AB/P8, GII.6-AB/P9, and GII.6-AB/P12) assembled into homogeneous VLPs or SVPs, suggesting that the integrity of the N-terminal region is critical for the formation of stable and uniform VLPs. The N-terminus of VP1 may play a role in stabilizing assembly intermediates, such as dimers or pentamers of dimers, which are essential intermediates for the formation of mature VLPs [5]. Cleavage at the N-terminus could possibly disrupt these interactions, leading to the formation of defective VLPs or SVPs.
The inclusion of insertion peptides derived from GI.7, GII.3, and GII.4 strains provided an opportunity to explore the role of specific sequences in modulating cleavage and assembly of VP1 proteins. Interestingly, the effects of these peptides varied depending on their origin and position within the VP1 and VP2 protein. For example, recombinant proteins containing sequences from the GI.7 VP1 (e.g., GII.6-AB/P1 to GII.6-AB/P8) exhibited a Fig. 5 Western blot analysis. Cell lysates and CsCl-density gradient-purified VLPs were loaded on precast gels for WB analysis. Rabbit anti-GII.6-AB hyperimmune serum and mouse anti-FLAG mAb were separately used as detection antibodies. Supernatant from lysed sf9 cells infected with recombinant baculoviruses produced by empty donor vector was used as negative control wide range of cleavage patterns and assembly outcomes, suggesting that the structural context of the insertion peptides plays a critical role in determining their susceptibility to cleavage and their impact on VLP assembly. Similarly, recombinant proteins containing sequences from the GII.4 VP2 and GII.3 VP1 (e.g., GII.6-AB/P9 to GII.6-AB/P12) also showed varied cleavage patterns and assembly products. Furthermore, we also observed that recombinant proteins with limited cleavage (e.g., GII.6-AB/P2 and GII.6-AB/P8) showed poor yields in VLPs (data not provided), and how this is related with N-terminal cleavage and assembly efficiency requires further investigation.
PMF results indicated that the N-terminal of recombinant GII.6-AB proteins were susceptible to cleavage with a similar pattern to that of wild type GII.6-AB proteins, with generally 3 or 35-38 amino acids lost from the N-terminus [15]. As the N-terminal FLAG-tagged fragment was not detected, this cleavage at residues 35-38 might not be specific, but likely is more prone to structural restrictions.
There are several limitations in our study. Firstly, there is a lack of controls to support the involvement or absence of a specific protease in the N-terminal FLAGtagged experiment. Secondly, all experiments were performed in Sf9 cells using baculoviruse expression system. The possibility remains that the observed cleavage patterns reflect host-specific protease activity or other system-specific effects. Thirdly, functional evaluation of those constructs were not performed and will be addressed in our future work. Fourthly, it should be noted that only one representative recombinant protein was analyzed, the actual cleavage pattern of other proteins should be further investigated.
In summary, this study provides valuable insights into the effects of incorporating peptide sequences from different NoV strains into the GII.6-AB protein. The findings highlight the importance of sequence variations in influencing the cleavage patterns, assembly behaviors, and morphological characteristics of recombinant proteins. While sequence-dependent effects were observed, the precise biochemical or structural determinants remain to be determined. These insights have important implications for the development of NoV vaccines and diagnostic tools, as well as for our understanding of the structural and functional properties of NoV capsid proteins. Further studies are needed to explore the mechanisms underlying the cleavage and assembly of these recombinant proteins and to determine their potential applications in virology and immunology.
## References
1. Jiang, Wang, Graham et al. (1992) "Expression, self-assembly, and antigenicity of the Norwalk virus capsid protein" *J Virol*
2. Huo, Ling, Wu et al. (2018) "Expression and purification of Norovirus virus like particles in Escherichia coli and their immunogenicity in mice" *Mol Immunol*
3. Diamos, Mason (2018) "High-level expression and enrichment of Norovirus virus-like particles in plants using modified geminiviral vectors" *Protein Expr Purif*
4. Prasad, Hardy, Dokland et al. (1999) "X-ray crystallographic structure of the Norwalk virus capsid" *Science*
5. Bertolotti-Ciarlet, White, Chen et al. (2002) "Structural requirements for the assembly of Norwalk virus-like particles" *J Virol*
6. Tan, Hegde, Jiang (2004) "The P domain of Norovirus capsid protein forms dimer and binds to histo-blood group antigen receptors" *J Virol*
7. Greenberg, Valdesuso, Kalica et al. (1981) "Proteins of Norwalk virus" *J Virol*
8. Hardy, White, Ball et al. (1995) "Specific proteolytic cleavage of recombinant Norwalk virus capsid protein" *J Virol*
9. Lochridge, Jutila, Graff et al. (2005) "Epitopes in the P2 domain of Norovirus VP1 recognized by monoclonal antibodies that block cell interactions" *J Gen Virol*
10. Pogan, Weiss, Bond et al. (2020) "N-terminal VP1 truncations favor T = 1 norovirus-like particles" *Vaccines (Basel)*
11. Huo, Wang, Meng et al. (2015) "Production of Norovirus VLPs to size homogeneity" *Virus Res*
12. Huo, Ling, Shen (2016) "Biological and immunological characterization of Norovirus major capsid proteins from three different genotypes" *Microb Pathog*
13. White, Hardy, Estes (1997) "Biochemical characterization of a smaller form of recombinant Norwalk virus capsids assembled in insect cells" *J Virol*
14. Tresset, Coeur, Bryche et al. (2013) "Norovirus capsid proteins self-assemble through biphasic kinetics via longlived stave-like intermediates" *J Am Chem Soc*
15. Ma, Liu, Fu et al. (2024) "GII.6 norovirus major capsid protein VP1 derived from distinct clusters induce cross-blocking effects" *Infect Genet Evol*
16. Zheng, Wang, Liu et al. (2018) "Characterization of a Norovirus-specific monoclonal antibody that exhibits wide spectrum binding activities" *J Med Virol*
17. Huo, Wang, Zheng et al. (2017) "Enzymatic cleavage promotes disassembly of GII.3 Norovirus virus like particles and its binding to salivary histo-blood group antigens" *Virus Res*
18. Ma, Liu, Zheng et al. (2022) "Sequence addition to the N-or C-terminus of the major capsid protein VP1 of norovirus affects its cleavage and assembly into virus-like particles" *Microb Pathog*
19. Ma, Liu, Liu et al. (2025) "Effect of trypsin digestion on the integrity and antigenic epitopes of GII.6 norovirus virus-like particles" *Arch Virol*
20. Ma, Liu, Huo (2024) "Characterization of cleavage patterns and assembly of N-terminally modified GII.6 norovirus VP1 proteins" *Arch Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12592196&blobtype=pdf | # Development of siRNA therapeutics to combat microbial infections: a bibliometric analysis
Yi He, Yulin Yuan, Meihua Zhou, Miao Li, Lingjin Li, Chunhong Li, Xiaocui Liang, Panyan Liu, Wei Wang, Zhenfeng Deng, Yanling Hu, Jingcheng Yang, Achimugu Dickson, Musa
## Abstract
Background: The rise of antimicrobial resistance and the COVID-19 pandemic highlight the limitations of traditional therapies. Small interfering RNA (siRNA) therapeutics, which utilize RNA interference for targeted gene silencing, present a promising approach to combating microbial infections. However, research in this area remains fragmented. This study employs a comprehensive bibliometric analysis to chart research trends and inform future directions. Methods: A total of 8,426 publications from the Web of Science Core Collection were analyzed using CiteSpace and VOSviewer software to examine annual publication trends, geographic and institutional contributions, author networks, journal impacts, and keyword evolution. Data extraction focused on English-language articles. Results: The publication trends for siRNA therapeutics in microbial infections have evolved in three phases: rapid growth, stabilization at a peak, and subsequent cyclical fluctuations. Research contributions spanned 99 countries and regions, with 5,564 institutions and 1,234 journals involved. China (2,849 publications) and the United States (2,820 publications) led in publication volume. While the United States maintained dominance in academic influence and collaboration, China has steadily increased its research output in this area. The Journal of Virology emerged as the leading journal in terms of both productivity and citation impact. Key research areas include delivery systems, target selection, manufacturing technologies, antiviral therapeutics, and combination therapies. The field has shifted from basic mechanistic studies to clinical applications, with future research poised to focus on organ-specific delivery beyond the liver, exploration of diverse administration routes, integration of artificial intelligence-driven strategies, and enhanced global collaboration.Frontiers in Cellular and Infection Microbiology frontiersin.org 01
## 1 Introduction
The emergence of antimicrobial resistance and the COVID-19 pandemic has exposed significant vulnerabilities in conventional therapeutic strategies. Traditional antibiotics and antivirals, which target conserved microbial structures or enzymes, are increasingly ineffective due to pathogen evolution and mutation, and exhibit limited efficacy against viruses (Cook and Wright, 2022;Karim et al., 2023). Rapidly mutating viruses such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), persistent viruses like hepatitis B virus (HBV), and pathogens such as methicillin-resistant Staphylococcus aureus (MRSA) exemplify the constraints of existing therapies. These challenges demand innovative approaches that adapt to genetic variability and overcome resistance mechanisms. In this context, small interfering RNA (siRNA) has emerged as a transformative modality, offering a promising strategy to combat microbial infections through sequence-specific gene silencing that disrupts microbial replication, virulence, or host dependency factors (Ebenezer et al., 2025). Unlike small-molecule drugs, siRNA enables precise targeting of pathogen genomes or host pathways, minimizing off-target effects and mitigating the risk of drug resistance. Early successes, such as siRNA-mediated suppression of HBV surface antigen (HBsAg) in clinical trials (Yuen et al., 2020), demonstrate its potential to redefine antimicrobial therapeutics.
SiRNA therapeutics operate through RNA interference (RNAi), a conserved mechanism that facilitates the degradation of complementary messenger RNA (mRNA). The process begins when siRNA duplexes are incorporated into the RNA-induced silencing complex (RISC), where the guide strand directs target recognition and cleavage (Carthew and Sontheimer, 2009). The specificity of siRNA is derived from its ability to bind to unique genomic sequences, making it particularly effective against pathogens with high mutation rates. For instance, siRNA targeting ultra-conserved regions of SARS-CoV-2 demonstrated broad-spectrum activity across variants, highlighting its adaptability (Xu et al., 2022;Rabdano et al., 2024). Moreover, chemical modifications and advanced delivery systems, such as lipid nanoparticles (LNPs) and triantennary N-acetylgalactosamine (tri-GalNAc) conjugates, have improved siRNA stability, biodistribution, and cellular uptake (Hu et al., 2020). These innovations address early challenges of immunogenicity and inefficient delivery, facilitating clinical applications for infections like HBV and human immunodeficiency virus (HIV).
Despite these advancements, significant challenges remain in the application of siRNA for microbial infections. Bacterial and fungal targets are underrepresented due to the absence of RNAi machinery in prokaryotes and delivery barriers, while viral applications dominate the field (Ebenezer et al., 2025). Additionally, clinical heterogeneity in outcomes, stemming from variability in delivery platforms and dosing regimens, highlights the need for standardized methodologies. Bibliometric analyses reveal fragmented research efforts, with notable disparities in geographic contributions and thematic focus. A comprehensive evaluation of the field's intellectual structure and emerging trends is therefore crucial to prioritize high-impact research, foster global collaboration, and address translational gaps. This study aims to construct a visualization model to assess current research trends, track the evolution of the field, and forecast future directions. By providing an extensive visual knowledge map, it offers critical insights for researchers and guides future initiatives in siRNA therapeutics for microbial infections.
## 2 Methods
## 2.1 Database and search strategy
All publications in this study were sourced from the Web of Science Core Collection (WOSCC), a database indexing over 12,000 high-impact academic journals. The retrieval strategy employed the following search terms: TS = ("small interfering RNA" OR "siRNA") AND TS = (infection OR infectious diseases OR pathogens OR viruses OR bacteria OR fungi OR parasites) NOT TS = (plant OR agriculture OR agronomy OR environment OR ecology OR food OR veterinary OR entomology). To focus on human infections, the search excluded clearly irrelevant fields such as agriculture, environmental sciences, food science, and veterinary medicine using the "NOT" operator. The search period covered the database's inception through June 30, 2025, with the language limited to English. To ensure that the included literature is original research that has undergone strict peer review and to maintain the comparability of influence among different studies, the document types for this study is limited to articles only, avoiding other types of literature such as reviews and conference abstracts that may interfere with the analysis results. The literature screening process for this study is illustrated in Figure 1.
## 2.2 Data analysis
## 3.2 Distribution of countries/regions and institutions
The analysis revealed contributions from 5,564 institutions across 99 countries/regions to the field of siRNA therapeutics for microbial infections. Figure 2B shows the geographical distribution of these countries/regions in global publications. Table 1 lists the top 10 most productive countries, including their total citations (TC), average citations, and total link strength (TLS). China (2,849) and the United States (2,820) lead in publication volume, ranking first and second, respectively, followed by Japan (690), Germany (582), and South Korea (431). However, the USA leads in TC, significantly surpassing China, Japan, Germany, and Canada, which occupy the second to fifth positions. The TLS, which reflects the intensity of research collaboration, shows that the USA and several European countries, particularly Germany, France, and the UK, lead in academic cooperation networks, occupying four of the top five TLS positions, with China ranking second. Figure 2C illustrates the co-occurrence map among countries/regions involved in this research topic, while Figure 2D presents the proportional trends in publication numbers for the top 10 countries over different years, shown through a stacked area graph.
Table 2 lists the top 10 most productive institutions, with 7 from China, 2 from Japan, and 1 from the USA. The Chinese Academy of Sciences leads with 141 publications, while Harvard University ranks first in TC with 13,816 citations. Both institutions tie for first place in TLS. Figure 3A shows a cluster analysis of institutions (with a minimum of 30 publications), revealing five closely collaborating clusters, each distinguished by different colors. This clustering highlights the collaborative networks among institutions, providing insights for future inter-institutional cooperation. Figure 3B further illustrates the evolution of these institutions over time. Additionally, the top 10 institutions in terms of TC are listed in Supplementary Table S1, with Harvard University, the University of Pennsylvania, and the Massachusetts Institute of Technology (MIT) ranking the top three, all from the USA, reflecting their strong academic influence and significant contributions to this field of research.
## 3.3 Analysis of authors and co-cited authors
Representative scholars and research teams in the field of siRNA therapeutics for microbial infections can be identified through author and co-cited author analysis. Table 3 lists the top 10 most prolific authors, each having published over 20 articles. Wakita, Takaji and Li, Chenghua tied for first place with 25 articles. Figure 4A illustrates the collaborative network of authors with 10 or more publications, highlighting the active scholars with extensive collaboration in this field. Supplementary Table S2 and Figure 4B provide a statistical and clustering analysis of co-cited authors. Among the top 10 co-cited authors, Elbashir, SM, Fire, A, and Livak, KJ ranked the highest in co-citation, while Elbashir, SM, Fire, A, and Brummelkamp, TR led in TLS, indicating their pivotal roles as pioneers and significant contributors to the field.
## 3.4 Analysis of journal and co-cited journal
Since 2001, research on siRNA therapeutics for microbial infections has been published in 1,234 academic journals. Table 4 and Supplementary Table S3 list the top 10 most productive and cocited journals. Journals such as Journal of Virology, Journal of Biological Chemistry, Journal of Immunology, Proceedings of the National Academy of Sciences of the United States of America, PLOS One, and Virology appear in the top 10 of both lists, signifying their central role in this research area. The Journal of Virology stands out as the most productive and influential journal. The collaborative network of journals is presented in Figure 5A, while Figure 5B shows the co-citation network of journals.
## 3.5 Cluster and burst analysis of keywords
Keyword clustering is an effective method for identifying core topics and emerging trends in the field of siRNA therapeutics for microbial infections by analyzing co-occurrence relationships among keywords in the literature, thereby reflecting the dynamics of disciplinary development. Through VOSviewer analysis, 97 keywords with more than 100 occurrences were identified, forming four distinct clusters, as shown in Figure 6A. Cluster 1 (red) consists of 30 terms such as "siRNA," "replication," "gene expression," "inhibition," "RNA interference," "in-vitro," "doublestranded RNA," and "mammalian cells," focusing on the fundamental mechanisms of siRNA through RNAi. Cluster 2 (green) includes 26 terms like "expression," "apoptosis," "gene," "pathway," "cancer," "proliferation," "growth," "autophagy," and "oxidative stress," highlighting research directions related to disease mechanisms and the development of targeted therapies. Cluster 3 (blue) encompasses 21 terms, such as "infection," "protein," "cells," "identification," "virus," "receptor," and "binding," addressing antiviral strategies and host-virus interaction mechanisms. Cluster 4 (yellow) contains 20 terms, including "activation," "NF-kappa-B," "inflammation," "induction," and "innate immunity," emphasizing inflammatory signaling pathways related to siRNA and their interaction with the immune system. Figure 6B provides further insights into the temporal evolution of these keywords.
The phenomenon of keyword citation bursts often signals the emergence of new research hotspots. Figure 6C displays the top 20 keywords with the most significant citation bursts identified by CiteSpace analysis, with red lines indicating the burst duration. Research hotspots can be categorized into three phases: the early phase (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) focused on exploring fundamental mechanisms, with burst keywords such as "RNA interference," "small interfering RNA," "gene expression," "short hairpin RNA," and "lentiviral vector." The middle phase (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018) witnessed a peak in publication output, accompanied by burst keywords like "siRNA delivery," "nasopharyngeal carcinoma," "hepatocellular carcinoma," "Helicobacter pylori," "influenza virus," "viral replication," "innate immunity," "cervical cancer," and "IL-1b."
The latest phase (2019-present) includes keywords like "NFkappa B," "virus replication," "oxidative stress," "drug delivery," and "lipid nanoparticles." Keywords such as RNAi, small interfering RNA, siRNA delivery, hepatocellular carcinoma, Helicobacter pylori, viral replication, cervical cancer, and oxidative stress have exhibited burst durations exceeding five years, indicating sustained research attention.
## 3.6 Citation and burst analysis of studies
Through citation analysis of the studies, representative articles in the field of siRNA therapeutics for microbial infections were identified. Table 5 lists the top 10 most highly cited studies. The highest-ranked work is Taro Kawai's 2005 study in Nature Immunology, titled "IPS-1, an adaptor triggering RIG-I-and Mda5-mediated type I interferon induction," which demonstrated that siRNA-mediated "knockdown" of interferon-beta promoter stimulator 1 (IPS-1) could block virus-induced interferon activation. This study directly validated siRNA-guided interference effects and emphasized its potential as a genesilencing tool. The second most cited study is Gunter Meister's 2004 Nature paper, "Mechanisms of gene silencing by doublestranded RNA," which elucidated the RNAi mechanism by revealing how double-stranded RNA (dsRNA) is processed into short RNA duplexes with distinct size and structure, while establishing their unique gene-silencing capabilities. Ranking third is Thimmaiah P. Chendrimada's 2005 Nature study, "TRBP recruits the Dicer complex to AGO2 for microRNA processing and g e n e s i l e n c i n g , " w h i c h u n c o v e r e d t h a t t h e h u m a n immunodeficiency virus transactivating response RNA-binding protein (TRBP) regulates RISC assembly through interactions with Dicer and Argonaute 2. siRNA experiments confirmed its essential role in miRNA biogenesis and RNAi functionality. Figure 7A presents the collaboration networks of these highly cited studies.
CiteSpace analysis identified the top 20 references with the strongest citation bursts, visualized in Figure 7B, with burst durations marked in red. The first significant citation burst emerged in 2002, triggered by Elbashir SM's landmark 2001 Nature study, "Duplexes of 21-nucleotide RNAs mediate RNAi in cultured mammalian cells," which maintained the burst condition until 2006. This groundbreaking work experimentally demonstrated that 21-nucleotide siRNA duplexes could specifically inhibit endogenous and exogenous gene expression across multiple mammalian cell lines, highlighting their potential as functional genomics tools and gene-specific therapeutics. Notably, two recent burst studies include Bo Hu's 2020 Signal Transduction and Targeted Therapy review, "Therapeutic siRNA: state of the art," which synthesizes advancements in siRNA chemical modifications and delivery platforms, discussing both milestones and ongoing challenges in specificity and delivery. Another is Markus Hoffmann's 2020 Cell paper, "SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor." While not directly related to siRNA, its revelation of the ACE2/TMPRSS2 signaling pathway as critical for SARS-CoV-2 entry has been extensively cited in siRNAbased antiviral studies targeting this pathway.
## 4 Discussion
## 4.1 Global research status and trends
Over the past two decades, the field of siRNA therapeutics for microbial infections has experienced transformative growth, marked by evolving publication trends, shifting geographic contributions, and changing research priorities. A comprehensive bibliometric analysis of 8,426 publications in this study reveals a three-phase developmental trajectory: the early phase (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) was characterized by explosive growth, with a 110.53% annual increase in publication volume, focusing on RNAi mechanisms and validation; the middle phase (2010-2018) saw a period of dualpeak stabilization with growth rates below 1%, emphasizing drug delivery systems and clinical applications; the latest phase (2019present) exhibits cyclical fluctuations and moderate declines in publication numbers, with a focus on COVID-19 pandemicresponsive innovations and novel strategies for siRNA therapeutics in specific infectious diseases. This progress has been driven by continuous technological advancements, including the elucidation of RNAi mechanisms, innovations in drug manufacturing, the development of novel delivery systems, and the emergence of combination therapeutic strategies, coupled with the urgency of global health challenges, particularly during the COVID-19 pandemic, which significantly accelerated the clinical translation of siRNA therapies.
Geographically, research output is primarily led by China and the USA, contributing 2,849 and 2,820 publications, respectively, accounting for 33.8% and 33.5% of the total publications. While China leads in publication volume, the USA maintains unparalleled academic influence, as evidenced by its higher total and average citation counts and its dominance in international collaboration networks. The steady increase in China's publication share reflects its increased research investment and growing application scale. Notably, while China's institutional output is concentrated within the Chinese Academy of Sciences, top institutions in the USA such as Harvard University, the University of Pennsylvania, and MIT lead high-impact studies, underscoring their pivotal role in pioneering clinical applications and innovations. There are two main reasons why China leads in publications but a lower total citations compared to the USA. First, the USA is the pioneer in this field and holds significant academic influence. Second, China's publications is predominantly concentrated in the latest phase, resulting in a citation time-lag bias where recent studies tend to be cited less frequently. Additionally, countries like Japan, South Korea, Canada, and European nations, particularly Germany, France, and the UK, show lower publication volumes in comparison to China and the USA, likely due to disparities in population size and research investment. However, these nations have also made significant contributions, collectively establishing a complementary global framework for siRNA therapeutics alongside China and the USA. At the author level, Wakita, Takaji (Japan Institute for Health Security) and Li, Chenghua (Ningbo University) are the most productive authors, each with 25 publications. Their main research areas focus on siRNA therapeutic targets for hepatitis viruses and innate immune signaling pathways, respectively. However, foundational influence lies with co-cited pioneers such as Fire A (Nobel laureate for RNAi discovery) (Fire et al., 1998) and Elbashir SM (seminal 2001 Nature paper on siRNA-mediated RNAi in mammalian cells) (Elbashir et al., 2001a), whose work forms the theoretical basis of siRNA therapeutics. At the journal level, Journal of Virology is the most productive journal, while Nature holds the highest citation count, serving as key platforms for disseminating research in this field.
## 4.2 Early phase (2001-2009): mechanism of RNA interference
The early phase laid the conceptual and technological foundation for siRNA therapeutics against microbial infections, marked by explosive growth in publications and the discovery of RNAi's foundational mechanisms. RNAi is a natural cellular process that silences gene expression by targeting and degrading mRNA. The seminal discovery of RNAi by Fire A and Mello C in 1998 revolutionized our understanding of gene regulation in cells (Fire et al., 1998), paving the way for the development of RNAbased therapeutics. However, initial applications of siRNA encountered challenges due to interferon responses triggered by dsRNA of varying lengths, which resulted in nonspecific mRNA degradation. To overcome this, researchers identified that dsRNA fragments with two-nucleotide 3′ overhangs could effectively degrade sequence-specific mRNA without eliciting interferon production (Elbashir et al., 2001b;Schubert et al., 2005). These siRNA molecules, typically 21-23 nucleotides in length, are designed to complement target mRNA sequences, enabling precise gene silencing in mammalian cells (Zamore et al., 2000;Elbashir et al., 2001a;Sørensen et al., 2003). The siRNA molecules bind to mRNA, either blocking its translation into proteins or inducing its degradation, thereby modulating gene expression (Lipardi et al., 2001;Shyu et al., 2008;Prevost et al., 2011). Moreover, researchers developed methods to achieve targeted gene silencing through synthetic siRNA delivery. These siRNA molecules, designed to specifically silence pathogenic genes, enter cells and trigger enzymatic cascades that form the RISC (Rand et al., 2004;Schwarz et al., 2004). The Dicer enzyme processes dsRNA into siRNA duplexes, which consist of a passenger (sense) strand and a guide (antisense) strand. The siRNA is then loaded into the RISC complex. Within RISC, strand separation occurs: the guide strand, with a more stable 5′ end, is retained, while the passenger strand is cleaved by AGO2 nuclease and degraded (Chendrimada et al., 2005). The guide strand-RISC complex then scans cellular mRNA for complementary sequences, and upon target recognition, RISC induces site-specific mRNA cleavage, silencing gene expression (Valencia-Sanchez et al., 2006;Ameres et al., 2007). Notably, dsRNA exceeding 30 nucleotides can trigger innate immune responses in mammalian cells, inducing interferon production as a defense mechanism against viral dsRNA generated during replication. Consequently, therapeutic siRNA design must optimize length to minimize unintended immunogenicity and determine the minimal effective concentrations for gene silencing (Judge and MacLachlan, 2008). RNAi plays a pivotal role in gene regulation and innate antiviral immunity. As emerging therapeutic agents in molecular biology and gene regulation, siRNA-based drugs show significant promise. Moreover, siRNA has become a pivotal tool in functional genomics research, marking the onset of this phase (Dorsett and Tuschl, 2004;Van De Water et al., 2006).
## 4.3 Middle phase (2010-2018): drug delivery and clinical application
The middle phase marked the transition of siRNA therapeutics from mechanistic exploration to clinical application, characterized by a peak in publications and significant breakthroughs in drug delivery systems. SiRNA therapeutics show immense promise in treating microbial infections; however, their clinical translation hinges on the safety and efficiency of siRNA delivery systems, which must overcome challenges such as instability, immunogenicity, limited tissue penetration, and potential offtarget effects (Cullis and Felgner, 2024). To effectively target pathogens or specific cells, siRNA must reach the infection site without being cleared or degraded prematurely. The safety and efficiency of delivery systems play a critical role in determining whether these therapeutics can engage their intended targets and maximize therapeutic efficacy (Vicentini et al., 2013). During this phase, substantial advancements in delivery technologies, including viral vectors, LNPs, tri-GalNAc conjugates, polymer carriers, and chemical modifications, provided new therapeutic options for microbial infections.
Viral vectors, capable of mediating efficient gene transduction and enabling long-term expression, were extensively explored for delivering genes encoding short hairpin RNA (Moffat et al., 2006;Meerbrey et al., 2011;Börner et al., 2013;Osoŕio et al., 2014). These vectors have the potential for a single administration to achieve continuous, long-term siRNA production in vivo, a feature highly attractive for chronic infections requiring prolonged treatment, such as HIV (Chung et al., 2014;Spanevello et al., 2016;Swamy et al., 2016) and HBV (Li et al., 2009(Li et al., , 2016;;Xia et al., 2013). However, their clinical application has been hindered by immunogenicity concerns, risks of insertional mutagenesis, and limited nucleic acid payload capacity (Zhang et al., 2023;Zhou et al., 2009), leading to the shift toward non-viral vectors as the dominant delivery strategy. Among these, LNPs have emerged as the most successfully developed non-viral platform. LNPs use ionizable cationic lipids to encapsulate siRNA efficiently under acidic conditions, forming stable nanoparticles. Upon intravenous injection, their nearneutral surface charge at physiological pH minimizes nonspecific interactions and toxicity (Patil and Panyam, 2009;Buyens et al., 2012). Once inside the cells, LNPs undergo protonation in the acidic environments of endosomes or lysosomes, which triggers a charge reversal, facilitating the dissociation of siRNA from the LNPs. This enables efficient "endosomal escape" and the release of functional siRNA into the cytoplasm (Kolli et al., 2013). An additional advantage of LNPs is their natural liver-targeting property. After intravenous injection, LNPs are coated by apolipoproteins and selectively internalized by hepatocytes via low-density lipoprotein receptors (Akinc et al., 2010). This liver targeting feature makes LNPs particularly suitable for treating liver-enriched viral infections such as HBV and hepatitis C virus (HCV) (Cho et al., 2009;Mevel et al., 2010;Wooddell et al., 2013;Sato et al., 2017). Currently, ARB-1740, an LNPs-delivered siRNA therapeutic targeting HBV genes, has entered clinical trials and demonstrated sustained reductions in HBsAg levels (Thi et al., 2019). For liver-tropic pathogens, tri-GalNAc conjugation has revolutionized targeted delivery by exploiting hepatocyte-specific asialoglycoprotein receptors (Nair et al., 2014(Nair et al., , 2017)), making it one of the most promising delivery systems in clinical applications. Subcutaneously administered GalNAc-siRNA conjugates enable efficient hepatic uptake without the need for complex carriers. This strategy simplifies manufacturing, supports patient-friendly dosing, and underpins HBV therapeutics (Zimmermann et al., 2017;Foster et al., 2018). Several GalNAc-siRNA therapies, such as ARC-520 (Yuen et al., 2020(Yuen et al., , 2022b)), JNJ-3989 (Yuen et al., 2022a), bersacapavir (Yuen et al., 2023), and VIR-2218 (Yuen et al., 2024), have entered clinical trials, demonstrating durable suppression of all HBV antigens and offering hope for the functional cure of hepatitis B. Due to the livertargeting properties of the two primary delivery vehicles, LNPs and tri-GalNAc conjugation, research on siRNA therapeutics for liver viral infections has progressed significantly ahead of studies targeting other organs or tissues. While LNPs and tri-GalNAc conjugates demonstrated clinical promise, their translational bottlenecks, including liver-centric biodistribution limiting broader organ applications and manufacturing scalability challenges, as well as uncertain long-term safety profile and high production costs, remained critical barriers. Polymeric siRNA delivery systems also hold promise. Polymeric carriers form stable nanocomplexes by encapsulating siRNA, effectively protecting it and prolonging its circulation time in vivo. Through chemical modifications or incorporation of targeted ligands, these polymers exhibit enhanced biocompatibility and specificity (Takemoto and Nishiyama, 2017). Key advantages include stability and the ability to facilitate endosomal escape. Despite ongoing challenges related to immunogenicity and largescale manufacturing, these systems have shown promising antiviral efficacy in infection models, including HIV and HBV (Weber et al., 2008(Weber et al., , 2012;;Wooddell et al., 2013). As a new generation of carriers, extracellular vesicles (EVs) are generating significant interest. EVs can carry various nucleic acid molecules, including siRNA, and offer advantages such as low immunogenicity, ease of body barrier penetration, and the potential for target modification through engineering (Jiang et al., 2017;O'Loughlin et al., 2017;Guo et al., 2025). However, their application in treating microbial infections still requires extensive research and optimization. Additionally, while chemical modification is not a delivery carrier itself, it plays a pivotal role in enhancing the inherent characteristics of siRNA. Appropriate chemical modifications can significantly improve siRNA stability, immune evasion, and assembly efficiency with RISC. Existing studies have demonstrated that chemical modifications can enhance the antiviral efficacy of siRNA therapeutics (Marimani et al., 2013(Marimani et al., , 2015;;Kalke et al., 2022).
## 4.4 Latest phase (2019-present): COVID-19 pandemic-driven innovation
The latest phase marked by the COVID-19 pandemic has driven innovation in siRNA therapeutics, characterized by a modest decline and cyclical fluctuations in publication numbers, as well as the integration of multiple therapeutic strategies. The pandemic has fundamentally transformed the trajectory of siRNA therapeutics, accelerating their clinical translation through unprecedented global collaboration and resource mobilization. To address SARS-CoV-2's high mutability and respiratory tract tropism, siRNA therapeutics have advanced significantly across three areas. Firstly, in terms of delivery systems, researchers have developed novel LNPs for delivering anti-SARS-CoV-2 siRNA. One such innovation is stealth LNPs, which protect siRNA, enabling stable circulation in the serum while effectively targeting the lungs (Idris et al., 2021). Another breakthrough involves the DOTAP +MC3 LNPs, which reduce the proportion of cationic lipid 1,2dioleoyl-3-trimethylammonium-propane (DOTAP) and introduce the ionizable lipid DLin-MC3-DMA (MC3). This modification reduces the positive charge of the delivery vector, minimizing immunogenicity and enabling targeted lung delivery after intravenous injection (Cheng et al., 2020;Idris et al., 2021). These advancements in delivery technologies have made LNPs a prominent research hotspot once again, highlighting their potential to target sites beyond the liver and lungs. Additionally, the development of new nanoscale polymers (Athaydes Seabra Ferreira et al., 2025) and EVs (Fu and Xiong, 2021) for SARS-CoV-2 has shown phased progress. Secondly, regarding target design, siRNA sequences have been tailored to target highly conserved regions of the SARS-CoV-2 genome, such as the RNA- dependent RNA polymerase (Xu et al., 2022) nucleocapsid protein (N protein)-coding regions (Rabdano et al., 2024;Xue et al., 2024), with the aim of inhibiting multiple variants from Alpha to Omicron. An exciting development in this phase is the application of artificial intelligence (AI) to optimize target selection. A recent study leveraged generative AI models, DeepFrag and EMPIRE, to target the N protein and optimize the structure of phenanthridine SARS-CoV-2 inhibitors. These optimized compounds, with strong binding affinity, were subsequently validated for their antiviral activity in vitro (Xiang et al., 2024), demonstrating the promising role of AI in accelerating antiviral drug development and addressing viral mutations. Additionally, AI is being used to optimize siRNA delivery systems, including the prediction of LNP characteristics and formulation optimization (Amoako et al., 2025), as well as in the broader fight against infectious diseases (Wong et al., 2023). Thirdly, regarding administration routes, the first inhaled siRNA drug, SNS812, for treating SARS-CoV-2, has been developed and entered clinical trials (Chang et al., 2022(Chang et al., , 2025)). Inhaled drug delivery represents a major innovation in achieving targeted lung delivery, directly delivering siRNA to respiratory tract infections. This method increases drug concentration in the lungs while reducing systemic exposure, thereby minimizing side effects. This advancement marks a significant step in the development of siRNA therapeutics for respiratory infections. Furthermore, preliminary breakthroughs have been achieved in the delivery to major organs or targets such as brain, tumor, muscle, spleen, and kidneys (Khare et al., 2023;Song et al., 2024;Vaidya et al., 2024;Quijano et al., 2025), offering new potential for extrahepatic targeted therapies.
Currently, siRNA therapeutics have achieved the greatest success in antiviral applications within microbial infections, largely due to the RNAi mechanism's ability to effectively suppress viral pathogenesis. Beyond its relatively established applications in challenging viral infections such as HBV, HIV, and SARS-CoV-2, siRNA-based approaches have also been explored to varying extents for treating infections caused by respiratory syncytial virus, influenza, herpesviruses, Ebola, Zika, enteroviruses, dengue virus, and rabies virus (Liang et al., 2015;Martin et al., 2018;Martıń-Vicente et al., 2019;Scott and Nel, 2021;Kalke et al., 2022;Zhang et al., 2022;Haas et al., 2023;Bie et al., 2024). In contrast, siRNA therapeutics have encountered significant obstacles in antibacterial applications. Bacteria lack the core RNAi machinery components, specifically the RISC, rendering siRNA ineffective for directly regulating bacterial genes. Furthermore, the historical absence of efficient delivery systems for bacterial cells has hindered progress. However, the emergence of EVs, particularly exosomes, as a novel biological delivery system, has begun to address this limitation. A groundbreaking study demonstrated the use of exosomes to deliver siRNA-AGO2 complexes, successfully suppressing the expression of drug resistance genes in MRSA. This innovative technology significantly reduced resistance protein levels, restoring MRSA's sensitivity to antibiotics and providing a transformative strategy against global superbug infections. Additionally, siRNA has been applied to silence genes associated with Helicobacter pylori cytotoxin production and urease enzyme activity, effectively reducing inflammatory responses and colonization rates, which aids in the treatment of gastric cancer (Motamedi et al., 2023;Reza et al., 2023). For fungal pathogens, which possess primitive RNAi mechanisms, siRNA could theoretically be designed to silence drug resistance or virulence genes directly. Studies have confirmed siRNA-mediated silencing of
## Rank
First author Title Journals Year DOI TC critical genes in pathogenic fungi such as Aspergillus fumigatus and Mucor species, although these are confined to in vitro and animal models (Guo et al., 2012;Nicolaś et al., 2015;Yu et al., 2025). Practical challenges remain, as exogenous siRNA exhibits lower silencing efficiency in fungal cells compared to mammalian cells (Sahay et al., 2013;Dong et al., 2019;Mosquera et al., 2025). Consequently, siRNA-based antifungal therapies are still in a much earlier exploratory phase compared to viral and bacterial applications.
## 4.5 Limitations
Although bibliometric analysis provides a comprehensive and objective overview, several limitations must be acknowledged. Firstly, this study relied solely on publications from the WOSCC database, which may not encompass all relevant research on siRNA therapeutics. To ensure the integrity of citation data, maintain data consistency, and literature quality, joint analyses utilizing multisource databases such as PubMed and SCOPUS were excluded. Secondly, our exclusive focus on English-language literature may result in the exclusion of studies published in other languages, potentially introducing language bias. In addition, there is a timelag bias in citation analysis. This means that compared with earlier studies, recent studies may receive fewer citations due to being newly published. Thirdly, keyword-based retrieval strategies, such as "small interfering RNA" or "siRNA", may inadvertently exclude pioneering terminology from early studies or alternative terminology of emerging fields. This inherent limitation of bibliometric methods underscores the necessity for iterative keyword refinement in future analyses. Lastly, the quality of the literature was not evaluated, meaning that high-quality and lowquality studies were treated equally. These inherent limitations should be considered when interpreting the results.
## 5 Conclusion
In conclusion, siRNA therapeutics have emerged a promising approach for treating a wide range of diseases, with clinical applications spanning cancer, genetic disorders, microbial infections, and inflammatory diseases. This study examines the evolving landscape of siRNA therapeutics for microbial infections over the past two decades through bibliometric analysis. The field has progressed from foundational mechanistic studies to clinical applications, fueled by technological innovations and the urgency created by the COVID-19 pandemic. Current research hotspots include delivery systems, target selection, manufacturing technologies, antiviral therapeutics, and combination strategies. Future research should focus on organ-specific delivery beyond the liver, explore diverse administration routes, integrate AI with drug design, and foster global collaborations. By addressing existing challenges, siRNA-based therapies hold significant potential for advancing the treatment of microbial infections and offering transformative solutions for the future.
## References
1. Kawai (2005) "Taro IPS-1, an adaptor triggering RIG-I-and Mda5-mediated type I interferon induction Nature immunology"
2. Meister (2004) "Gunter Mechanisms of gene silencing by double-stranded RNA Nature"
3. Chendrimada, Thimmaiah (2005) "TRBP recruits the Dicer complex to Ago2 for microRNA processing and gene silencing Nature"
4. Stewart (2001) "Lentivirus-delivered stable gene silencing by RNAi in primary cells RNA"
5. Brass (2008) "Identification of host proteins required for HIV infection through a functional genomic screen" *Science*
6. (2021) "Melanie Noncoding RNA therapeutics -challenges and potential solutions Nature reviews Drug discovery"
7. Kumar (2007) "Priti Transvascular delivery of small interfering RNA to the central nervous system Nature"
8. Morrissey (2005) "Potent and persistent in vivo anti-HBV activity of chemically modified siRNAs Nature biotechnology"
9. (2005) "Erwei Antibody mediated in vivo delivery of small interfering RNAs via cell-surface receptors Nature biotechnology"
10. Doi ; Akinc, Querbes, De et al. (2010) "Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms" *Mol. Ther*
11. Ameres, Martinez, Schroeder (2007) "Molecular basis for target RNA recognition and cleavage by human RISC" *Cell*
12. Amoako, Mokhammad, Malik et al. (2025) "Enhancing nucleic acid delivery by the integration of artificial intelligence into lipid nanoparticle formulation" *Front. Med. Technol*
13. Seabra Ferreira, Cordeiro Guimarães, Carvalho Costa et al. (2025) "Targeting CCL2 silencing using polymer-lipid hybrid nanoparticles to reduce cytokine storm and inflammation in severe COVID-19" *Int. J. Pharm*
14. Bie, Zhang, Chen et al. (2024) "Design of antiviral AGO2-dependent short hairpin RNAs" *Virol. Sin*
15. Börner, Niopek, Cotugno et al. (2013) "Robust RNAi enhancement via human argonaute-2 overexpression from plasmids, viral vectors and cell lines" *Nucleic Acids Res*
16. Buyens, De Smedt, Braeckmans et al.
17. Grunsven (2012) "Liposome based systems for systemic siRNA delivery: stability in blood sets the requirements for optimal carrier design" *J. Controlled Release*
18. Carthew, Sontheimer (2009) "Origins and mechanisms of miRNAs and siRNAs" *Cell*
19. Chang, Chen, Yang et al. (2025) "Pharmacokinetics and safety profile of <span style="
20. ">SNS812, a first in human fully modified SIRNA targeting wide-spectrum SARS -COV -2, in healthy subjects" *Clin. Transl. Sci*
21. Chang, Yang, Chen et al. (2022) "A siRNA targets and inhibits a broad range of SARS-CoV-2 infections including delta variant" *EMBO Mol. Med*
22. Chendrimada, Gregory, Kumaraswamy et al. (2005) "TRBP recruits the dicer complex to Ago2 for microRNA processing and gene silencing" *Nature*
23. Cheng, Wei, Farbiak et al. (2020) "Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR-cas gene editing" *Nat. Nanotechnol*
24. Cho, Park, Kim et al. (2009) "Suppression of hepatitis B virus-derived human hepatocellular carcinoma by NF-kBinducing kinase-specific siRNA using liver-targeting liposomes" *Arch. Pharm. Res*
25. Cook, Wright (2022) "The past, present, and future of antibiotics" *Sci. Transl. Med*
26. Cullis, Felgner, Dong et al. (2019) "The 60-year evolution of lipid nanoparticles for nucleic acid delivery" *Nat. Rev. Drug Discov*
27. Dorsett, Tuschl (2004) "siRNAs: applications in functional genomics and potential as therapeutics" *Nat. Rev. Drug Discov*
28. Ebenezer, Oyebamiji, Olanlokun et al.
29. "Recent update on siRNA therapeutics" *Int. J. Mol. Sci*
30. Elbashir, Harborth, Lendeckel et al. (2001) "Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells" *Nature*
31. Elbashir, Martinez, Patkaniowska et al. (2001) "Functional anatomy of siRNAs for mediating efficient RNAi in drosophila melanogaster embryo lysate" *EMBO J*
32. Fire, Xu, Montgomery et al. (1998) "Potent and specific genetic interference by double-stranded RNA in caenorhabditis elegans" *Nature*
33. Foster, Brown, Shaikh et al. (2018) "Advanced siRNA designs further improve in vivo performance of GalNAc-siRNA conjugates" *Mol. Ther*
34. Fu, Xiong (2021) "Tagged extracellular vesicles with the RBD of the viral spike protein for delivery of antiviral agents against SARS-COV-2 infection" *J. Controlled Release*
35. Guo, Gao, Wu (2012) "Toll-like receptor 2 siRNA suppresses corneal inflammation and attenuates aspergillus fumigatus keratitis in rats" *Immunol. Cell Biol*
36. Guo, Lou, Yang et al. (2025) "Exosomal miRNAs in inflammatory diseases" *Icell*
37. Haas, Mcgregor, Bouhaddou et al.
38. (2020) "Proteomic and genetic analyses of influenza a viruses identify pan-viral host targets" *Signal Transduction Targeting Ther*
39. Sars- "CoV-2 targeted siRNA-nanoparticle therapy for COVID-19" *Mol. Ther*
40. Jiang, Vader, Schiffelers (2017) "Extracellular vesicles for nucleic acid delivery: progress and prospects for safe RNA-based gene therapy" *Gene Ther*
41. Judge, Maclachlan (2008) "Overcoming the innate immune response to small interfering RNA" *Hum. Gene Ther*
42. Kalke, Lund, Nyman et al. (2022) "Swarms of chemically modified antiviral siRNA targeting herpes simplex virus infection in human corneal epithelial cells" *PloS Pathog*
43. Karim, Lo, Einav (2023) "Preparing for the next viral threat with broad-spectrum antivirals" *J. Clin. Invest*
44. Khare, Edgecomb, Hamadani et al. (2023) "Lipid nanoparticle-mediated drug delivery to the brain" *Adv. Drug Deliv. Rev*
45. Kolli, Wong, Harbottle et al. (2013) "pH-triggered nanoparticle mediated delivery of siRNA to liver cells in vitro and in vivo" *Bioconjug. Chem*
46. Li, He, Yao et al. (2009) "Inhibition of HBV replication and gene expression in vitro and in vivo with a single AAV vector delivering two shRNA molecules" *BMB Rep*
47. Li, Sun, Li et al. (2016) "Suppression of hepatitis B virus antigen production and replication by wild-type HBV dependently replicating HBV shRNA vectors in vitro and in vivo" *Antiviral Res*
48. Liang, Chow, Lau et al. (2015) "Inhalable dry powder formulations of siRNA and pH-responsive peptides with antiviral activity against H1N1 influenza virus" *Mol. Pharm*
49. Lipardi, Wei, Paterson (2001) "RNAi as random degradative PCR: siRNA primers convert mRNA into dsRNAs that are degraded to generate new siRNAs" *Cell*
50. Marimani, Ely, Buff et al. (2013) "Inhibition of hepatitis B virus replication in cultured cells and in vivo using 2′-O-guanidinopropyl modified siRNAs" *Bioorg. Med. Chem*
51. Marimani, Ely, Buff et al. (2015) "Inhibition of replication of hepatitis B virus in transgenic mice following administration of hepatotropic lipoplexes containing guanidinopropyl-modified siRNAs" *J. Controlled Release*
52. Martin, Chiramel, Schmidt et al.
53. Martıń-Vicente, Resino, Martıńez (2019) "siRNA-mediated simultaneous regulation of the cellular innate immune response and human respiratory syncytial virus replication" *Biomolecules*
54. Meerbrey, Hu, Kessler et al. (2011) "The pINDUCER lentiviral toolkit for inducible RNA interference in vitro and in vivo" *Proc. Natl. Acad. Sci*
55. Mevel, Kamaly, Carmona et al. (2010) "DODAG; a versatile new cationic lipid that mediates efficient delivery of pDNA and siRNA" *J. Controlled Release*
56. Moffat, Grueneberg, Yang et al.
57. "A lentiviral RNAi library for human and mouse genes applied to an arrayed viral high-content screen" *Cell*
58. Mosquera, Gineśy, Bocos-Asenjo et al.
59. (2025) "Spray-induced gene silencing to control plant pathogenic fungi: a stepby-step guide" *J. Integr. Plant Biol*
60. Motamedi, Abiri, Salari et al. (2017) "Impact of enhanced metabolic stability on pharmacokinetics and pharmacodynamics of GalNAc-siRNA conjugates" *Nucleic Acids Res*
61. Nair, Willoughby, Chan et al. (2014) "Multivalent N -acetylgalactosamine-conjugated siRNA localizes in hepatocytes and elicits robust RNAi-mediated gene silencing" *J. Am. Chem. Soc*
62. Nicolaś, Vila, Moxon et al. (2015) "The RNAi machinery controls distinct responses to environmental signals in the basal fungus mucor circinelloides" *BMC Genomics*
63. O'loughlin, Mäger, De et al. (2017) "Functional delivery of lipid-conjugated siRNA by extracellular vesicles" *Mol. Ther*
64. Osoŕio, Gijsbers, Oliveras-Salva et al. (2014) "Viral vectors expressing a single microRNA-based short-hairpin RNA result in potent gene silencing in vitro and in vivo" *J. Biotechnol*
65. Patil, Panyam (2009) "Polymeric nanoparticles for siRNA delivery and gene silencing" *Int. J. Pharm*
66. Prevost, Desnoyers, Jacques et al. (2011) "Small RNA-induced mRNA degradation achieved through both translation block and activated cleavage" *Genes Dev*
67. Quijano, Martinez-Saucedo, Ianniello et al. (2025) "Systemic administration of an RNA binding and cellpenetrating antibody targets therapeutic RNA to multiple mouse models of cancer" *Sci. Transl. Med*
68. Rabdano, Ruzanova, Vertyachikh et al. (2024) "N-protein vaccine is effective against COVID-19: phase 3, randomized, double-blind, placebo-controlled clinical trial" *J. Infect*
69. Rand, Ginalski, Grishin et al. (2004) "Biochemical identification of argonaute 2 as the sole protein required for RNA-induced silencing complex activity" *Proc. Natl. Acad. Sci*
70. Reza, Mahmud, Ferdous et al. (2023) "Gene silencing of helicobacter pylori through newly designed <span style="font-variant:Small-caps;">siRNA convenes the treatment of gastric cancer" *Cancer Med*
71. Sahay, Querbes, Alabi et al. (2013) "Efficiency of siRNA delivery by lipid nanoparticles is limited by endocytic recycling" *Nat. Biotechnol*
72. Sato, Matsui, Yamamoto et al. (2017) "Highly specific delivery of siRNA to hepatocytes circumvents endothelial cell-mediated lipid nanoparticle-associated toxicity leading to the safe and efficacious decrease in the hepatitis B virus" *J. Controlled Release*
73. Schubert, Grünweller, Erdmann et al. (2005) "Local RNA target structure influences siRNA efficacy: systematic analysis of intentionally designed binding regions" *J. Mol. Biol*
74. Schwarz, Tomari, Zamore (2004) "The RNA-induced silencing complex is a Mg2+-dependent endonuclease" *Curr. Biol*
75. Scott, Nel (2021) "Rabies prophylactic and treatment options: an in vitro study of siRNA-and aptamer-based therapeutics" *Viruses*
76. Shyu, Wilkinson, Van Hoof (2008) "Messenger RNA regulation: to translate or to degrade" *EMBO J*
77. Song, Zhao, Wang et al. (2024) "Tuning lipid nanoparticles for RNA delivery to extrahepatic organs" *Adv. Mater*
78. Sørensen, Leirdal, Sioud (2003) "Gene silencing by systemic delivery of synthetic siRNAs in adult mice" *J. Mol. Biol*
79. Spanevello, Calistri, Del Vecchio et al. (2016) "Development of lentiviral vectors simultaneously expressing multiple siRNAs against CCR5, vif and tat/rev genes for an HIV-1 gene therapy approach" *Mol. Ther.-Nucleic Acids*
80. Swamy, Wu, Shankar (2016) "Recent advances in RNAi-based strategies for therapy and prevention of HIV-1/AIDS" *Adv. Drug Deliv. Rev*
81. Takemoto, Nishiyama (2017) "Functional polymer-based siRNA delivery carrier that recognizes site-specific biosignals" *J. Controlled Release*
82. Thi, Dhillon, Ardzinski et al. (2019) "ARB-1740, a RNA interference therapeutic for chronic hepatitis B infection" *ACS Infect. Dis*
83. Vaidya, Moore, Chatterjee et al. (2024)
84. "Expanding RNAi to kidneys, lungs, and spleen via selective ORgan targeting (SORT) siRNA lipid nanoparticles" *Adv. Mater*
85. Valencia-Sanchez, Liu, Hannon et al. (2006) "Control of translation and mRNA degradation by miRNAs and siRNAs" *Genes Dev*
86. Van De Water, Boerman, Wouterse et al.
87. Masereeuw (2006) "Intravenously administered short interfering rna accumulates in the kidney and selectively suppresses gene function in renal proximal tubules" *Drug Metab. Dispos*
88. Vicentini, Borgheti-Cardoso, Depieri et al. (2013) "Delivery systems and local administration routes for therapeutic siRNA" *Pharm. Res*
89. Weber, Merkel, Kissel et al. (2012) "PEGylated poly(ethylene imine) copolymer-delivered siRNA inhibits HIV replication in vitro" *J. Controlled Release*
90. Weber, Ortega, Clemente et al.
91. (2008) "Characterization of carbosilane dendrimers as effective carriers of siRNA to HIV-infected lymphocytes" *J. Controlled Release*
92. Wong, De La Fuente-Nunez, Collins (2023) "Leveraging artificial intelligence in the fight against infectious diseases" *Science*
93. Wooddell, Rozema, Hossbach et al. (2013) "Hepatocyte-targeted RNAi therapeutics for the treatment of chronic hepatitis B virus infection" *Mol. Ther*
94. Xia, Chen, Gong et al. (2013) "Inhibition of hepatitis B virus gene expression and replication by ribonuclease P" *Mol. Ther*
95. Xiang, Fan, Ren et al. (2024) "Utilizing artificial intelligence for precision exploration of N protein targeting phenanthridine sars-cov-2 inhibitors: a novel approach" *Eur. J. Med. Chem*
96. Xu, Chen, Lu et al. (2022) "An update on inhibitors targeting RNA-dependent RNA polymerase for COVID-19 treatment: promises and challenges" *Biochem. Pharmacol*
97. Xue, Chu, Wang et al. (2024) "Coronavirus nucleocapsid protein enhances the binding of p-PKCa to RACK1: implications for inhibition of nucleocytoplasmic trafficking and suppression of the innate immune response" *PloS Pathog*
98. Yu, Vogel, Hirsch et al. (2025) "Enhanced antifungal activity of siRNA-loaded anionic liposomes against the human pathogenic fungus aspergillus fumigatus" *Nanoscale*
99. Yuen, Asselah, Jacobson et al. (2023) "Efficacy and safety of the siRNA JNJ-73763989 and the capsid assembly modulator JNJ-56136379 (bersacapavir) with nucleos(t)ide analogues for the treatment of chronic hepatitis B virus infection (REEF-1): a multicentre, double-blind, activecontrolled, randomised, phase 2b trial" *Lancet Gastroenterol. Hepatol*
100. Yuen, Lim, Yoon et al.
101. "VIR-2218 (elebsiran) plus pegylated interferon-alfa-2a in participants with chronic hepatitis B virus infection: a phase 2 study" *Lancet Gastroenterol. Hepatol*
102. Yuen, Locarnini, Lim et al. (2022) "Combination treatments including the small-interfering RNA JNJ-3989 induce rapid and sometimes prolonged viral responses in patients with CHB" *J. Hepatol*
103. Yuen, Schiefke, Yoon et al. (2020) "RNA interference therapy with ARC-520 results in prolonged hepatitis B surface antigen response in patients with chronic hepatitis B infection" *Hepatology*
104. Yuen, Wong, Schluep et al. (2022) "Long-term serological, virological and histological responses to RNA inhibition by ARC-520 in chinese chronic hepatitis B patients on entecavir treatment" *Gut*
105. Zamore, Tuschl, Sharp et al. (2000) "RNAi: doublestranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals" *Cell*
106. Zhang, Chen, Gan et al. (2023) "A comprehensive review of small interfering RNAs (siRNAs): mechanism, therapeutic targets, and delivery strategies for cancer therapy" *Int. J. Nanomedicine*
107. Zhang, Fu, Cheng et al. (2022) "sEVsRVG selectively delivers antiviral siRNA to fetus brain, inhibits ZIKV infection and mitigates ZIKV-induced microcephaly in mouse model" *Mol. Ther*
108. Zhou, Zhang, Wang et al. (2009) "A method for detecting and preventing negative RNA interference in preparation of lentiviral vectors for siRNA delivery" *RNA*
109. Zimmermann, Karsten, Chan et al. (2017) "Clinical proof of concept for a novel hepatocyte-targeting GalNAc-siRNA conjugate" *Mol. Ther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671125&blobtype=pdf | # Insights into the viral landscape of the western honey bee and native bees in Bangladesh
Islam Hamim, Lena Wilfert
## Abstract
Bees are important pollinators that are increasingly threatened by viruses. In this study, we investigated the viruses in honey bees in Bangladesh, focusing on western (Apis mellifera) and native bee species (A. cerana, A. dorsata, A. florea, and Trigona sp.). Using high-throughput poly(A)-selected RNA sequencing, we observed that viruses of the order Picornavirales are frequently detected in both western and native bees. However, this pattern may reflect both true biological abundance and methodological bias, as this approach inherently enriches for polyadenylated RNA viruses. Deformed wing virus (DWV), black queen cell virus (BQCV), and sacbrood virus (SBV) were commonly found in western bees, while native bees exhibited a high diversity of viral communities rather than dominance of specific viruses. The common bee viruses also showed high read abundances in western bees. Notably, the study identified unreported viruses in bees belonging to the Iflaviridae and Dicistroviridae families, expanding the known diversity of honey bee pathogens. In addition, plant-associated viruses were identified, suggesting a potential role for honey bees as vectors of plant viruses and highlighting the interactions between bees, plants, and their pathogens. The results of the diversity analysis demonstrated significant differences in the composition of virus populations between western and native bees in our studied samples. These results reveal the occurrence of bee viruses in Bangladesh and highlight the potential interspe cific transmission of viruses, which may pose a significant threat to local bee populations. Our study emphasizes the importance of monitoring known viruses and novel viruses, as well as plant pathogens, and implementing sustainable management practices to reduce the spread of pathogens and protect both native and western bees. IMPORTANCE Pollinators face increasing threats from viral pathogens, yet data on their viromes remain limited in many parts of the world, including South Asia. This study provides insights into the viral communities of both native and non-native bee species in Bangladesh using RNA sequencing. While Apis mellifera showed higher viral loads of known honey bee viruses, native bee species exhibited a broader diversity of viral sequences, including several uncharacterized viruses. Although based on a limited sample set, these findings contribute to a growing understanding of viral diversity in pollinators and underscore the value of continued surveillance to better understand virus-host associations and potential cross-species transmission in regions undergoing rapid apicultural expansion. KEYWORDS virology, environmental microbiology, plant pathology, bee pathology B ees are essential pollinators of agricultural crops and flowering plants, contributing to the conservation of biodiversity and agro-ecosystems. Furthermore, both the western honey bee (Apis mellifera), native to Africa and Europe, and the eastern honey bee (A. cerana), endemic to Asia, produce honey and other substances important for human consumption. Bee populations have declined significantly worldwide, often
resulting in reduced agricultural production due to inadequate pollinator visits (1,2). This has implications for our food supply, economic stability, biodiversity, and ecosystem functioning. The main causes of decline in pollinator and bee populations are habitat loss, agrochemical use, climate change, pests, and pathogens (3,4). In managed bees, particularly in A. mellifera, emerging and widespread diseases are important threats to the health of bee populations and the sustainability of beekeeping (5). Despite the importance of bee pathogens, research is lacking in large parts of the world, including in South and Southeast Asia.
In western populations of the honey bee A. mellifera, viruses have been intensively studied because of their economic importance. Viruses infecting honey bees are linked to a variety of symptoms, ranging from asymptomatic infections to severe mortality and colony collapse. Although most bee viruses cause latent infections, some viruses, such as sacbrood virus (SBV), black queen cell virus (BQCV), and deformed wing virus (DWV), cause apparent symptoms and death of individual bees (6). Members of the DWV and acute bee paralysis virus (ABPV) clades are also associated with the collapse of entire colonies of A. mellifera, particularly in the presence of the ectoparasitic mite Varroa. A. cerana, which is of comparable apicultural importance to A. mellifera in Asia, typically shows lower levels of virus infection, likely due to reduced Varroa destructor infestation (7)(8)(9)(10). This species has co-evolved with V. destructor and employs a range of behavioral and physiological strategies to limit mite reproduction (11). In managed A. cerana, SBV is the main pathogen of concern, with outbreaks across Asia causing high mortality (9). For example, SBV infections in A. cerana led to the collapse of approximately 80% of colonies and a significant reduction in honey production in Bangladesh in 1990 (12).
Asia is the biodiversity hotspot for honey bees, with several wild honey bee species including the cave-dwelling A. cerana. This species occurs in wild populations but has also been kept by beekeepers for centuries. Additionally, managed apiaries of non-native A. mellifera are also common across Asia. In addition, many wild solitary and social bee species, from stingless bees to bumble bees, occur in Asia. Even though the eastern honey bee is generally less affected by infectious diseases than its western counterpart, Asia has been the origin of two recently emerged parasites: both the virus-vectoring V. destructor and the microsporidian Nosema ceranae jumped from A. cerana to non-native western honey bees in the last century and have since been anthropogenically spread around most of the globe (13,14). With the recent advances in meta-transcriptomics, it has also become clear that honey bees harbor a high number of potential viral patho gens, with over 80 viral genomes now identified in honey bees (15), with more being constantly added (15,16). The high risk of disease emergence and the lack of studies in South and Southeast Asia highlight the need for de novo meta-transcriptome-based studies in these regions.
Bangladesh is a country that exemplifies this need for meta-transcriptomic studies. It is home to several honey bee species, including wild bees, such as A. dorsata (giant honey bee) and A. florea (dwarf honey bee). The native species A. cerana (eastern honey bee) has traditionally been used for honey production under partially managed conditions, due to its adaptability to local conditions (12). However, in the early 1990s, A. mellifera was introduced to Bangladesh, following the collapse of A. cerana populations due to an SBV outbreak, and it has since become the dominant species for commercial beekeeping due to its higher honey yields (12). Honey bees in Bangladesh forage on a variety of crops, such as mustard (Brassica campestris), litchi (Litchi chinensis), black cumin (Nigella sativa), coriander (Coriandrum sativum), niger (Guizotia abyssinica), and sesame (Sesamum indicum), as well as wild plants in Sundarbans and the hilly regions of Chittagong (12). Honey is produced mainly in winter, using A. mellifera and A. cerana, while only a limited amount of honey is harvested from wild bees (17). The country has yet to commercialize pollination services as a distinct industry. However, the beekeep ing sector, which is primarily focused on honey production for domestic consumption, plays an important role in increasing agricultural productivity by increasing crop yields through the provision of pollination services (12) and can provide significant benefits for poverty alleviation and rural development through low-investment, high-return opportunities (17). However, honey bee health in much of Asia, including in Bangladesh, has received limited research attention, particularly with respect to virome structure and virus transmission.
The high diversity of honey bees and stingless bees in Southeast Asia and Bangla desh, and the overlap with the non-native western honey bee A. mellifera, may facilitate the interspecific transmission of pathogens and parasites (8). The spread of the virusvectoring Varroa mite in A. mellifera has led to the spillover of DWV into wild bee species (18,19). For example, research in China has shown that A. mellifera is a current source of DWV for A. cerana (10), and the spillover of pathogens from managed A. mellifera to wild species can cause harm, such as in stingless bees (20). Here, we used meta-transcrip tomics to explore the occurrence and diversity of both known and novel viruses in bee species from Bangladesh. This approach allows us to assess not only the viruses that directly affect bees, but also those they may potentially transmit from plants or other environments. The results of this study underscore the importance of applying caution in beekeeping practices with A. mellifera, as this species appears to harbor a higher abundance of specific groups of potentially pathogenic viruses compared to native bee species such as A. cerana.
## RESULTS
## Viral load distribution in bee species
In this study, we used high-throughput, poly(A)-selected RNA sequencing (RNA-Seq) on the DNBSEQ platform (BGI, China) to investigate the viral landscape across different honey bee species in Bangladesh. Additionally, we were able to include one library each of stingless bees (Trigona sp.) and the small hive beetle Aethina tumida, an inquiline of the western honey bee A. mellifera. A total of 15 libraries were examined that include A. mellifera (eight libraries), A. cerana (three libraries), A. dorsata (one library), A. florea (one library), Trigona sp. (one library), and A. tumida (one library) (Fig. 1; Table 1). These included two libraries available from GenBank: one of A. mellifera and the other of A. cerana in Tangail, Bangladesh (NCBI BioProject PRJNA1090432). On average, 8.1% of the total reads were of viral origin (Table 1).
Western honey bees showed a range of viral read percentages from 0.067% to 69.777% (mean = 11.686%, median = 2.774, standard deviation = ±23.785) (Table 1), indicating high variability in viral loads across libraries. Native bees showed low viral load percentages ranging from 0.0007% to 0.1859% (mean = 0.0396%, median = 0.0081 and standard deviation = ±0.073), indicating an overall low viral load with less variation across the samples. The percentage of viral reads differed significantly among insect groups, defined here as combining bee and non-bee insects in this study (χ² = 10.334, d.f. = 2, P = 0.006), with higher viral read percentages in western honey bees. Dunn's test also confirmed a statistically significant difference (P = 0.003) in the percentage of viral reads between native and western bees.
## Bee virome composition
The transcriptome analysis of bee viromes led to the identification and assembly of 50 viral operational taxonomic units (OTUs), including 10 putative plant viruses, through sequence homology. Using sequence homology and phylogenetic analysis, we classified the viruses into insect and plant virus families (Fig. 2; Table S1). Of these, viral families of the order Picornavirales were the most abundant across all bee viromes. This pattern may reflect both true biological abundance and methodological bias, as we used poly(A)selected RNA sequencing approach that inherently enriches for polyadenylated RNA viruses and can exclude other viral groups. However, Iflaviridae was found as the most abundant family, with 95,098.55 reads per million (RPM), followed by Dicistroviridae (7,718.257 RPM) (Fig. 2; Table S1). Dicistroviridae and Iflaviridae were detected in all western honey bee libraries and in the A. florea library. Additionally, the Dicistroviridae were detected in two out of three A. cerana libraries and in A. dorsata, while Iflaviridae were detected in one of three A. cerana libraries. Neither was detected in the Meliponini library. Sinhaliviridae, with 1,062.95 RPM, was detected in five of eight A. mellifera libraries, indicating that its presence is restricted to western honey bees (Fig. 2; Table S1). Besides these, unclassified insect viruses were found in six of the libraries, including four native bee libraries, with a relatively low abundance (43.95 RPM). The family Secoviridae, typically associated with plant viruses, was found in three of the studied libraries, including the Meliponini (Fig. 2). Virus families, such as Alphaflexiviridae, Marnaviridae, and Fusaviridae were the rarest, each appearing in only one library.
## Identification of the phylogenetic positions of insect viruses
We identified the phylogenetic positions of the viruses by comparing the positions of virus sequences in the constructed evolutionary trees relative to known virus species and strains. The findings indicated well-defined monophyletic clades within the Iflaviridae family, including DWV, SBV, and Varroa destructor virus 2 (VDV2) (Fig. 3). These clades have strong bootstrap support and represent established lineages. On the Iflaviridae tree (Fig. 3A), DWV-A, DWV-B, DWV-C, and DWV-D form a monophyletic group, with both DWV-A and DWV-B present in Bangladesh, but no evidence of strains C or D (Fig. 3A). The SBV BD (Bangladesh) isolate clustered with an SBV isolate from Australia, while the VDV 2 BD isolate clustered with an isolate from China (Fig. 3A). In addition to these well-studied clades, the study identified unknown isolates of Iflaviridae collected in Bangladesh (Fig. 3A). These unknown isolates could be new virus species or strains, showing new lineages within the family. The unknown isolates Bee Iflavirus BD 9 and Bee Iflavirus BD 7 were found as a potential outgroup to DWV (Fig. 3A; Fig. S1). Bee Iflavirus BD 7 and Bee Iflavirus BD 9 share 44% sequence similarity at the nucleotide level, while their similarity to DWV ranges from 58% to 68% and 46% to 58%, respectively (Supplemental material S1).
The phylogenetic analysis also revealed strong genetic relationships among various Dicistroviridae viruses (Fig. 3B). Here, BQCV BD clusters most closely with BQCV isolates from Pakistan. For the Sinhaliviridae, we found LSV SA2, LSV 3, and LSV 4 in Bangladesh (Fig. 3C), while the strains LSV 1 and LSV 2 were not found in our samples. In the LSV 4 group, LSV 4 BD is closely related to the previously reported isolates from Bangladesh and Pakistan, indicating a potential regional connection to South Asia. Within the LSV SA2 lineage, LSV SA2 BD is genetically distinct, forming a separate branch. Similarly, LSV 3 BD forms a branch of the LSV3 lineage, which is associated with Chinese and global isolates.
## Distribution of insect viruses
Among the insect viruses identified, the most prevalent virus species in bee samples studied were DWV-A, BQCV, bee-associated cripavirus 1, SBV, and bee dicistrovirus 1 (Dicistroviridae sp.). While DWV-A was absent in all native bee libraries, it was present in all eight western honey bee libraries (mean = 2,136.298 RPM, median = 563.967 RPM, standard deviation = ±3,158.03 RPM), four of which also contained DWV-B (mean = 31.876 RPM, median = 18.36 RPM, standard deviation = ±32.56 RPM) (Fig. 4). While LSV variants were only found in western honey bees, BQCV, SBV, bee-associated cripavirus 1, bee dicistrovirus BD 1 and 3, and Planococcus ficus-associated dicistrovirus 1 occurred in several western and native bee libraries. In contrast, some unknown virus sequences identified in the study exhibited low prevalence in the bee populations or were found to occur in localized instances. For example, bee dicistrovirus BD 4 was detected solely in A. dorsata, with a viral count of 21.64 RPM (Table S2). Other rare viruses, such as bee dicistrovirus BD 8 and bee dicistrovirus BD 9, were detected in A. dorsata and A. florea, respectively. Figure 4 reveals several interesting patterns of co-occurrence. Lake Sinai virus (LSV) occurrence was evident solely in A. mellifera libraries in the presence of BQCV (from Mymensingh-Shutiakhali, Dhaka, Shatkhira, and Tangail). There is a statistically significant dependence between the presence of BQCV and the occurrence of LSV (Fisher's exact test, P-value = 0.02778). Bee-associated cripavirus 1, bee-associated dicistrovirus 1, and Planococcus ficus-associated dicistrovirus 1 appeared to co-occur in the bee libraries, particularly in cases where BQCV was absent or present in minimal amounts (Fig. 4). A single overall Fisher's exact test indicated a statistically significant negative association between the presence of BQCV and the detection of bee-associated cripavirus 1, bee-associated dicistrovirus 1, and Planococcus ficus-associated dicistrovirus 1 (P = 0.02857; odds ratio = 0; 95% CI: 0.000-1.080). This suggests that these viruses are unlikely to co-occur with BQCV, potentially due to competitive exclusion. However, because the confidence interval includes 1, this relationship should be interpreted with caution, and additional data would help confirm the trend. As this was a single test rather than multiple comparisons, statistical correction for multiple testing was not required.
## Diversity of insect viruses
The viral richness investigation showed that both native and western bees generally harbor several viral species, ranging from three to nine species in the native species, compared to two to eight in western bees (A. mellifera) (Fig. 5A; Table S3), with no difference in species richness (W = 10, P-value = 0.8726). The populations also showed a range of alpha diversities, with significantly higher diversity in native species compared to A. mellifera, as indicated by both the Shannon index (W = 28, P-value = 0.048) and the Simpson index (W = 28, P-value = 0.048) (Fig. 5B andC; Table S3). A PERMANOVA analysis of Bray-Curtis distances, with host species as a factor, reveals a significant difference in the virome composition between western honey bees and native honey bees (F = 1.7914, P = 0.019, d.f. = 1). This clear distinction between the viromes of western honey bees and native honey bees is also found in a principal coordinates analysis (PCoA) of the Bray-Curtis distances (Fig. 5D), with native bees closely clustering together, while the western bees are widely spaced.
## Plant viruses
Ten potential viruses related to plants or fungi were identified from bee libraries through sequence analyses followed by phylogenetic analysis (Fig. 6A). Each virus appeared in only one library (Fig. 6B). Secoviruses, such as tomato black ring virus and rehmannia torradovirus, were identified in A. mellifera, and a novel virus sequence, bee-associated nepovirus BD1, was detected in Trigona sp. They clustered with secoviruses previously reported in different plant species from other countries in the phylogenetic tree (Fig. 6A). Similarly, potexviruses such as bee-associated potexvirus BD 1 and bee-associated potexvirus BD 2, potyviruses such as chili veinal mottle virus BD, novel bee-associated plant RNA virus 1, and marnaviruses such as Lactuca sativa marnavirus and Marnaviridae sp., identified in A. mellifera from different locations in Bangladesh, cluster with other virus sequences previously reported in plant species. We also detected a mycovirus sequence, tentatively named bee-associated Fusaviridae sp., from an A. mellifera library in Tangail. Interestingly, none of the tested native Apis species were positive for plant or fungal viruses.
## DISCUSSION
Using poly(A)-selected high-throughput RNA sequencing (RNA-seq), we explored the viromes of the non-native, managed western honey bee (Apis mellifera) and native bee species in Bangladesh, including A. cerana, A. dorsata, A. florea, stingless bees (Trigona spp.), and the inquiline species of the western honey bee, Aethina tumida. Although our data set comprises a limited number of samples (n = 15) collected from distinct geographic regions and diverse hosts, we employed rigorous quality control measures and conservative assembly strategies to address these challenges. This study provides valuable insights into virus occurrence and diversity across multiple bee taxa and lays the groundwork for future investigations into interspecific viral transmission and potential spillover risks in a biodiversity-rich region. We identified 50 distinct viral OTUs, including known insect viruses, novel insect viruses, and plant viruses. The western honey bee populations showed a distinct virome, with a higher proportion of viral reads but less alpha diversity than the native bees, which also showed more variation in beta diver sity (Table 1; Fig. 5). This shows that Apis mellifera is a potential source for viruses in Bangladesh, but that native bees also carry diverse viromes. This work confirms that, while diverse virus families infect honey bees (21), the Iflaviridae and Dicistroviridae are the most common and dominant viral families (15,21,22). However, as we used poly(A)-selected RNA sequencing, this pattern may reflect not only true biological abundance but also methodological bias, since the approach enriches for polyadenylated viruses while underrepresenting others. The family Iflaviridae includes viruses with major threats to global bee populations, such as DWV and SBV. DWV, transmitted by the Varroa destructor mite, causes wing deformities and, in conjunction with Varroa, causes high overwinter colony mortality (23). In our study, DWV-A was detected in all examined libraries of A. melifera, where four libraries also showed the presence of DWV-B (Fig. 4). DWV-B has been shown to have higher virulence at the colony level in A. mellifera in Europe (23). A shift from DWV-A to DWV-B was first observed in A. melifera in Europe, with more recent detections in Asia (24)(25)(26). While DWV-B appears to be spreading in South Asia, these results are in line with DWV-A still remaining more prevalent in Asia (10). While SBV, on the other hand, is comparatively rare and of little concern in western honey bees, it has led to severe colony losses in outbreaks of A. cerana across Asia (9). It can cause pupation failure and death in larvae FIG 4 Prevalence and abundance of insect viruses in bee species in Bangladesh. The x-axis of the bubble plot shows the libraries, while the y-axis shows the insect virus species. The size of each bubble represents the normalized abundance of the virus family in each library, while the color gradient indicates the prevalence across libraries. Larger bubbles indicate higher abundance, and darker blue colors indicate higher prevalence. and adults (27). This globally distributed virus was found in A. mellifera, A. cerana, and A. florea in Bangladesh (Fig. 4). Variation in virus genotypes in different bee hosts has been reported (15,(28)(29)(30), and the high virulence in A. cerana raises concerns about possible spillovers and recombination of viral variants, as seen in DWV. These findings emphasize the need for continuous surveillance and research on the impact of both DWV and SBV on honey bee populations in Asia, considering their prevalence, genetic diversity, and spread.
The Dicistroviridae include important viruses, such as acute bee paralysis virus (ABPV), BQCV, Kashmir bee virus (KBV), and Israeli acute paralysis virus (IAPV) (15,22). Among these, IAPV and BQCV were reported as commonly occurring in honey bee colonies in different countries (22). While we did not detect ABPV, KBV, and IAPV in our studied samples from Bangladesh (Fig. 4), BQCV was detected in seven A. mellifera libraries and one A. cerana library in our study. Other dicistroviruses, i.e., bee-associated cripavirus 1, bee-associated dicistrovirus 1, and Planococcus ficus-associated dicistrovirus 1, were found to co-occur in the A. mellifera and A. dorsata libraries, particularly in cases where BQCV was absent or present in lower amounts. A reminiscent pattern, with the presence of BQCV seemingly inversely related to the presence of other Dicistroviridae such as ABPV, IAPV, KBV, and SBPV, has also been reported in other studies of honey bees (31,32). Here, we find a further potential pattern of co-occurrence with BQCV: LSV. This co-occurrence was only evident in A. mellifera libraries (Fig. 4). LSV could infect ants, solitary bees, bumble bees, and hornets besides honey bees, indicating possible cross-species transmission (33,34). Such potential patterns of exclusion or co-infection warrant further monitoring across populations and potential host species, as well as experimental verification.
In addition to these well-known viruses, several novel Dicistroviridae (bee dicistrovi rus BD 1 to 10), Iflaviridae (bee iflavirus BD 1 to 9), and unclassified viral OTUs were found in A. mellifera, A.cerana, A. dorsata, and A. florea. As the virome of more populations and species of bees is investigated by de novo sequencing methods, the number of viral OTUs has been increasing (21). While these viruses may have a limited geographic or host distribution and their degree of virulence is unknown, it is important to obtain a fuller picture of the virome of honey bees, particularly given the potential increasing need for pollinators to maintain global food security and the linked increase in the potential for transmission across species and geographic regions. For example, a novel dicistrovirus identified in A. mellifera from the Netherlands was later identified in A. florea from India (22,35). Similarly, rare viruses may be rare precisely because they have high virulence. Understanding these less common viruses and their interactions with other honey bee viruses, potential vectors, and biotic or abiotic stressors is thus crucial for future research into pathogens of these beneficial insects.
This study revealed stark differences in the composition of viromes of A. mellifera (western honey bees) and native bees in Bangladesh. The well-known and globally distributed viruses DWV-A and DWV-B, SBV, LSV, and BQCV were not detected or showed very low viral reads in bees native to Bangladesh in our study compared to A. mellifera (Fig. 4). Generally, the virome of A. mellifera has been found to be quite distinct from more distantly related Apidae (e.g., bumble bees and stingless bees, solitary bees) (9,21,36,37). However, an RNAseq study including global A. mellifera and A. cerana, A. dorsata, and A. florea from India did not detect such clear differentiation, with all Asian honey bee species also positive for DWV (22). Additionally, targeted qPCR-based surveys of A. mellifera and A. cerana in China have shown the presence of typical honey bee viruses (DWV, BQCV, IAPV, SBV), albeit to some extent at lower prevalence or titer in A. cerana, except for SBV, which showed the opposite pattern. However, it should be noted that we applied stringent thresholds, including a 15% minimum genome coverage threshold for the reference viral sequence, to minimize false positives in viral read estimation in this study. Here, we found lower alpha diversity in western honey bees compared to native honey bees, while viral richness did not differ between these groups (Fig. 5). This can be explained by the viromes of western honey bees being dominated by few viral strains, particularly DWV-A and DWV-B, in the presence of the Varroa mite. Native honeybees, in contrast, show similarly rich, but more evenly distributed, sets of viruses. However, it should be noted that, with the exception of A. cerana from Faridpur and Tangail (SRR2840858), all of the other native bees were collected in Chittagong division, from which we could not obtain A. mellifera. Thus, the difference between western honey bees and native honey bees, as well as the difference to studies from India (22) and China ( 9), could at least partially be due to environmental factors, such as climate or agricultural practices. These results emphasize the importance of understanding the potential impact of viruses on honey bees as well as the potential for cross-species transmission in different agricultural and climatic environments in South Asia, where commercial pollination and apiculture are developing at a fast pace and are crucial for food security and poverty alleviation.
In addition to diseases that can harm pollinators, bee species can also carry plant pathogens (22,37,38). Indeed, we also detected several plant viruses (e.g., the Secovir idae tomato black ring virus, bee-associated nepovirus, and Rehmannia torradovirus), including novel OTUs (bee-associated plant RNA virus 1, bee-associated nepovirus, and bee-associated potexvirus BD 1) (see Fig. 6). In contrast to the typical bee viruses, whose distribution was found to be determined by phylogenetic relationships between honey bees and bumble bees in a European study, the plant virome carried by pollinators was found to be structured by season and pollinator niche overlap (21). Similarly, Kadlečková et al. (39) found that plant viruses, but not bee viruses, showed small-scale geographic clustering. This shows that the plant virome carried by bees is strongly affected by the plants from which bees forage at any given time and place. It is clear that bees can vector plant diseases via pollination (40,41). Whether bees can also be active vectors of plant viruses remains unclear, although replication of tobacco ringspot virus in honey bees has been reported (38). At the same time, bee pollination can also reduce the vertical spread of diseases in self-pollinated plants (42). From an applied point of view, sequencing pollen collected by bees can be used to monitor the prevalence and spread of plant diseases (43). The presence of plant pathogens in pollinators such as honey bees can also give fundamental insights into plant-pollinator networks, particularly in difficult-to-observe tropical ecosystems such as those in Southeast Asia, where insect pollinators frequently feed in the canopy of flowering trees. Studying the association between plant viruses and bees has great potential in Asia, with the opportunities to better understand insect pollination in tropical agriculture and natural environments, as well as enabling monitoring for plant diseases over large areas. Furthermore, the potential for plant viruses to replicate in bees warrants further research. While cross-king dom replication is evident in plant viruses such as rice stripe virus, vectored by small brown planthopper (Laodelphax striatellus) (44), the potential for such active vectoring remains to be tested in bees, particularly in tropical ecosystems.
Southeast Asia is the center of honey bee biodiversity, but has also given rise to several global emerging bee parasites and pathogens (13,45). Here, we have shown that non-native western honey bees have a distinct virome from native honey bees in Bangladesh, characterized by high viral reads and viruses pathogenic to A. mellifera and A. cerana. This study thus highlights the need for comprehensive virus surveillance programs in Bangladesh and throughout Southeast Asia, including both known and novel viruses carried by bees. This research also calls for future studies on how these viruses impact wild and managed bee species. Moreover, monitoring the viruses across different wild bee populations in agricultural and wild ecosystems, and regions will be crucial for understanding the fundamental ecology of pathogens infecting wild bees, preventing spillovers from wild to managed bees, and reducing the potential for new emerging bee diseases.
## MATERIALS AND METHODS
## Sample collection and processing
The study analyzed samples of diverse foraging bee species collected opportunistically across multiple regions in Bangladesh, providing an overview of occurrence of viruses in different bee populations (Fig. 1; Table 1). Sampling was uneven across species and sites, with collections focused on single foraging locations per region due to field and logistical constraints. Therefore, this study is exploratory in nature. Samples of Apis mellifera (western honey bee) were collected from single foraging sites in Pabna, Faridpur, Dhaka, Satkhira, and Tangail and from two different sites in Mymensingh. A. cerana (Asian honey bee) was sampled from single sites in Faridpur and Chittagong. In Chittagong, we were also able to sample A. dorsata (giant honey bee), A. florea, and stingless bees identified as Trigona sp., which may include multiple species due to taxonomic uncertainty. Additionally, the small hive beetle (Aethina tumida), a known parasite of A. mellifera, was collected from a hive in Dhaka. After collection, each sample was immediately preserved in RNAlater solution to ensure RNA integrity, followed by RNA extraction and sequencing for virome analysis (46). Total RNA was extracted from tissue of laterally single-bisected bee (half-bee) following previously described methods (18,19). The tissue was homogenized at 5 m/s for 25 sec in three cycles with 20-sec ond pauses using a FisherBrand Bead Mill 24. RNA extraction was performed using 1.3 mL TRI Reagent and 0.1 mL 1-bromo-3-chloropropane (both from Sigma-Aldrich). RNA was eluted in 80 µL of nuclease-free water. RNA concentrations were quantified using the QuantiFluor RNA System (Promega). RNA samples were then pooled by species and location, with each pool containing RNA from five individuals of the same species, generating 13 RNA libraries for polyA-selected RNASeq on the DNBSEQ platform (BGI Bioinformatics, China) with 150 bp paired-end reads, yielding sufficient data for viral analysis. Additionally, two publicly available RNA-seq data sets from A. mellifera (SRR28408579) and A. cerana (SRR2840858) from Tangail, Bangladesh, were included, bringing the total to 15 data sets used in the viral detection and diversity analysis.
## Bioinformatic processing of RNA-Seq data
Raw sequencing data were processed using SOAPnuke software to ensure high-quality data for downstream analysis (47). Reads with ≥25% adapter sequence match (allow ing up to two base mismatches) were removed to eliminate adapter contamination. Additionally, reads shorter than 150 base pairs, with ≥0.1% unknown bases (N), and reads containing polyX stretches longer than 50 base pairs were also removed. Reads were additionally quality filtered by removing those where more than 40% of bases had a Phred quality score <20, and the Phred + 33 quality score system was used to assess the accuracy of base calls. These rigorous quality control steps ensured the retention of high--quality data for further analysis. The sequencing data quality was further assessed using FastQC, which identified concerns such as low per-base quality, GC content anomalies, and adapter contamination (https://www.bioinformatics.babra ham.ac.uk/projects/fastqc/). Low-quality reads and adapter sequences (if present) were removed using Sickle. To remove host genetic material, quality-filtered reads were aligned using Bowtie2 to the respective host genomes: A. mellifera, A. cerana, A. florea, A. dorsata, and Aethina tumida (48). This step reduced the volume of non-viral RNA-seq data to decrease data complexity and improve virus detection during de novo assembly (Trinity), particularly by facilitating contig reconstruction for subsequent blast analysis. However, for downstream analyses such as viral read counts (used as a proxy for viral load), we used the entire set of quality-filtered RNA-seq reads and mapped them directly to a reference virus database. For Trigona sp., RNA-seq data were processed without host read removal, using the quality-filtered reads directly for downstream analysis.
## Identification of viral OTUs
Non-host reads were assembled using Trinity (49), with a minimum length of 500 nucleotides. However, for Trigona sp., all quality-filtered reads were used for assembly, which may have included host-derived sequences. After de novo assembly, BLASTX was used to identify potential viral sequences by comparing the assembled contigs to the NCBI virus protein database (https://www.ncbi.nlm.nih.gov/labs/virus/vssi/#/) (49). Viral sequences were classified taxonomically according to their top BLASTx hits, identifying viral families and lineages. To ensure the accuracy of these identifications, we manually verified the closest relatives against the GenBank database, cross-referencing the results to resolve potential misclassifications or sequence match discrepancies. We focused on sequences containing the RNA-dependent RNA polymerase (RdRp) domain for all downstream analyses, including virus detection, diversity assessment, and phylogenetic reconstruction, which serves as a key diagnostic for viral sequences (50), as it is essential for RNA virus genome replication and is therefore highly conserved. The virus sequen ces were studied by identifying open reading frames (ORFs) using NCBI's ORF Finder and translating them into amino acid sequences. The presence of RdRp domains was confirmed by searching NCBI's Conserved Domain Database and Pfam for specific RdRp-related domains to validate domain identity and function. For our study, RdRpcontaining viral sequences were classified into operational taxonomic units (OTUs) based on a 90% amino acid sequence identity threshold. This classification provided informa tion about the composition of the viral community and enabled the identification of novel variants within different viral populations.
Phylogenetic relationships were investigated using maximum likelihood phyloge netic trees. These trees were constructed with 1,000 bootstrap iterations using RaxML (Randomized Axelerated Maximum Likelihood) (51) using the GTR + G (General Time Reversible with Gamma Distribution) sequence evolution model. Sequence alignments were constructed using MAFFT (52). The obtained maximum likelihood trees were visualized using FigTree.
## Estimation of viral reads and normalization by RPM and genome length
To estimate viral reads, clean reads from RNA sequencing were aligned to assembled viral contigs using Bowtie2 with the alignment mode set to highly sensitive (48). The amount of reads mapping to each viral OTU was normalized by reads per million (RPM), giving the number of viral reads scaled to a million total reads in the cleaned library. Further more, the genome length of each virus was included in the normalization procedure, ensuring that viral read estimates were adjusted not only for total sequencing depth and library size, but also for viral genome size. For distribution and diversity analyses, we set a 15% minimum genome coverage threshold for the reference viral sequence to minimize false positives in viral read estimation, ensuring that only reliable viral reads with adequate coverage are included in the analysis. This helped eliminate low-abun dance contaminants and potential errors that could lead to false matches. However, this stringent cut-off led to no viral reads being recorded for two libraries: Trigona sp. and A. cerana from Tangail (SRR2840858). This outcome may reflect true biological absence of detectable viruses at the time of sampling or viral loads below the detection threshold imposed by our 15% genome coverage filter. Moreover, given our use of poly-A selected RNA, we may miss viral genomes that lack polyadenylation.
## Analysis of virome composition and diversity
All analyses were performed in R (https://www.r-project.org/), using the ggplot2 package for plotting (53), dplyr for data handling (https://github.com/tidyverse/dplyr), and scales for scaling log-transformed data (https://github.com/r-lib/scales), for example, for prevalence and abundance based on viral reads. We used species richness as well as the Shannon and Simpson diversity indices to assess virome alpha diversity using the vegan package in R (https://CRAN.R-project.org/package=vegan). Beta diversity was investiga ted via Bray-Curtis distances calculated using the vegan package. PERMANOVA was then applied to look for significant changes across bee groups over 999 permutations using the adonis() function (54). For visualization, the two first coordinates of a Principal Coordinate Analysis (PCoA) were plotted. The two libraries, Trigona sp. and A. cerana from Tangail (SRR2840858), which did not contain viral reads based on our conservative cut-off (see above), were excluded from these analyses.
## References
1. Turo, Reilly, Fijen et al. (2024) "Insufficient pollinator visitation often limits yield in crop systems worldwide" *Nat Ecol Evol*
2. Zattara, Aizen (2021) "Worldwide occurrence records suggest a global decline in bee species richness" *One Earth*
3. Brunet, Fragoso (2024) "What are the main reasons for the worldwide decline in pollinator populations?" *CABI Reviews*
4. Goulson, Nicholls, Botías et al. (2015) "Bee declines driven by combined stress from parasites, pesticides, and lack of flowers" *Science*
5. Smith, Loh, Rostal et al. (2013) "Pathogens, pests, and economics: drivers of honeybee colony declines and losses" *Ecohealth*
6. Mcmenamin, Genersch (2015) "Honeybee colony losses and associated viruses" *Curr Opin Insect Sci*
7. Ai, Yan, Han (2012) "Occurrence and prevalence of seven bee viruses in Apis mellifera and Apis cerana apiaries in China" *J Invertebr Pathol*
8. Chantawannakul, De Guzman, Li et al. (2016) "Parasites, pathogens, and pests of honeybees in Asia" *Apidologie (Celle)*
9. Chen, Wu, Deng et al. (2021) "Seasonal variation of viral infections between the eastern honeybee (Apis cerana) and the western honeybee (Apis mellifera)"
10. Diao, Yang, Zhao et al. (2019) "Prevalence and population genetics of the emerging honeybee pathogen DWV in Chinese apiculture" *Sci Rep*
11. Lin, Page, Li et al. (2016) "Go east for better honeybee health: Apis cerana is faster at hygienic behavior than A. mellifera" *PLoS One*
12. Hossain, Pal (2019) "Beekeeping in Bangladesh-status and prospect" *Bangladesh J Sugarcane*
13. Klee, Besana, Genersch et al. (2007) "Widespread dispersal of the microsporidian Nosema ceranae, an emergent pathogen of the western honeybee, Apis mellifera" *J Invertebr Pathol*
14. Wilfert, Long, Leggett et al. (2016) "Deformed wing virus is a recent global epidemic in honeybees driven by Varroa mites" *Science*
15. Beaurepaire, Piot, Doublet et al. (2020) "Diversity and global distribution of viruses of the western honeybee" *Apis mellifera. Insects*
16. Deboutte, Smet, Brunain et al. (2024) "Known and novel viruses in Belgian honeybees: yearly differences, spatial clustering, and associations with overwintering loss" *Microbiol Spectr*
17. Moniruzzaman, Rahman (2009) "Prospects of beekeeping in Bangladesh" *J Bangladesh Agril Univ*
18. Dobelmann, Manley, Wilfert (2024) "Caught in the act: the invasion of a viral vector changes viral prevalence and titre in native honeybees and bumblebees" *Biol Lett*
19. Manley, Temperton, Doyle et al. (2019) "Knock-on community impacts of a novel vector: spillover of emerging DWV-B from Varroa-infested honeybees to wild bumblebees" *Ecol Lett*
20. Fleites-Ayil, Medina-Medina, Euán et al. (2023) "Trouble in the tropics: pathogen spillover is a threat for native stingless bees" *Biol Conserv*
21. Doublet, Doyle, Carvell et al. (2025) "Host ecology and phylogeny shape the temporal dynamics of social bee viromes" *Nat Commun*
22. Galbraith, Fuller, Ray et al. (2018) "Investigating the viral ecology of global bee communities with highthroughput metagenomics" *Sci Rep*
23. Natsopoulou, Mcmahon, Doublet et al. (2017) "The virulent, emerging genotype B of deformed wing virus is closely linked to overwinter honeybee worker loss" *Sci Rep*
24. Chantaphanwattana, Shafiey, Phokasem et al. (2023) "The presence of identical deformed wing virus sequence variants in co-occurring Apis species in Northern Thailand may represent a potential epidemiological threat to native honeybees of Southeast Asia" *J Invertebr Pathol*
25. Doublet, Oddie, Mondet et al. (2024) "Shift in virus composition in honeybees (Apis mellifera) following worldwide invasion by the parasitic mite and virus vector Varroa destructor" *R Soc Open Sci*
26. Paxton, Schäfer, Nazzi et al. (2022) "Epidemiology of a major honeybee pathogen, deformed wing virus: potential worldwide replacement of genotype A by genotype B" *Int J Parasitol Parasites Wildl*
27. Procházková, Füzik, Škubník et al. (2018) "Virion structure and genome delivery mechanism of sacbrood honeybee virus" *Proc Natl Acad Sci*
28. Chang, Chang, Ko et al. (2020) "Genomic sequencing and comparison of sacbrood viruses from Apis cerana and Apis mellifera in Taiwan" *Pathogens*
29. Choe, Nguyen, Hyun et al. (2012) "Genetic and phylogenetic analysis of South Korean sacbrood virus isolates from infected honeybees (Apis cerana)" *Vet Microbiol*
30. Li, Fei, Sun et al. (2019) "Genetic and phylogenetic analysis of Chinese sacbrood virus isolates from Apis mellifera" *PeerJ*
31. Kwon, Jung, Kil (2023) "Metagenomic analysis of viromes in honeybee colonies (Apis mellifera; Hymenoptera: Apidae) after mass disappearance in Korea" *Front Cell Infect Microbiol*
32. Šimenc, Knific, Toplak (2021) "The comparison of honeybee viral loads for six honeybee viruses (ABPV, BQCV, CBPV, DWV, LSV3 and SBV) in healthy and clinically affected honeybees with taqman quantitative real-time RT-PCR assays" *Viruses*
33. Hou, Liang, Chen et al. (2023) "Lake Sinai virus is a diverse, globally distributed but not emerging multi-strain honeybee virus" *Mol Ecol*
34. Rodríguez-Flores, Lopes, Diéguez-Antón et al. (2024) "Honeybee viruses in the yellow-legged hornet Vespa velutina (Lepelieter 1836): prevalence, loads, and detection of replicative DWV and LSV forms" *J Invertebr Pathol*
35. Remnant, Shi, Buchmann et al. (2017) "A diverse range of novel RNA viruses in geographically distinct honeybee populations" *J Virol*
36. Dolezal, Hendrix, Scavo et al. (2016) "Honeybee viruses in wild bees: viral prevalence, loads, and experimental inoculation" *PLoS One*
37. Schoonvaere, Smet, Smagghe et al. (2016) "Unbiased RNA shotgun metagenomics in social and solitary wild bees detects associations with eukaryote parasites and new viruses" *PLoS One*
38. Li, Cornman, Evans et al. (2014) "Systemic spread and propagation of a plant-pathogenic virus in European honeybees" *Apis mellifera. mBio*
39. Kadlečková, Tachezy, Erban et al. (2022) "The virome of healthy honeybee colonies: ubiquitous occurrence of known and new viruses in bee populations"
40. Darzi, Smith, Shargil et al. (2018) "The honeybee Apis mellifera contributes to cucumber green mottle mosaic virus spread via pollination" *Plant Pathol*
41. Fetters, Ashman (2023) "The pollen virome: a review of pollenassociated viruses and consequences for plants and their interactions with pollinators" *Am J Bot*
42. Mhlanga, Pate, Arinaitwe et al. (2024) "Reduction in vertical transmission rate of bean common mosaic virus in beepollinated common bean plants" *Virol J*
43. Roberts, Ireland, Tay et al. (2018) "Honeybee-assisted surveillance for early plant virus detection" *Ann Appl Biol*
44. Zhang, Yu, Xu et al. (2024) "Rice stripe virus infection facilitates the reproductive potential of Laodelphax striatellus" *Agronomy*
45. Hasegawa, Techer, Adjlane et al. (2023) "Evolutionarily diverse origins of deformed wing viruses in western honeybees" *Proc Natl Acad Sci*
46. Hamim, Suzuki, Borth et al. (2022) "Preserving plant samples from remote locations for detection of RNA and DNA viruses" *Front Microbiol*
47. Chen, Chen, Shi et al. (2018) "SOAPnuke: a MapReduce acceleration-supported software for integrated quality control and preprocessing of high-throughput sequencing data" *Gigascience*
48. Langmead, Salzberg (2012) "Fast gapped-read alignment with Bowtie 2" *Nat Methods*
49. Grabherr, Haas, Yassour et al. (2011) "Full-length transcriptome assembly from RNA-Seq data without a reference genome" *Nat Biotechnol*
50. Koonin, Wolf, Nagasaki et al. (2009) "The complexity of the virus world" *Nat Rev Microbiol*
51. Stamatakis (2014) "RAxML version 8: a tool for phylogenetic analysis and post-analysis of large phylogenies" *Bioinformatics*
52. Katoh, Misawa, Kuma et al. (2002) "MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform" *Nucleic Acids Res*
53. Wickham (2016)
54. Anderson (2001) "A new method for non-parametric multivariate analysis of variance" *Austral Ecol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12090734&blobtype=pdf | # Correction for Yan et al., "Molecular Determinants of Hepatitis B and D Virus Entry Restriction in Mouse Sodium Taurocholate Cotransporting Polypeptide"
Huan Yan, Bo Peng, Wenhui He, Guocai Zhong, Yonghe Qi, Bijie Ren, Zhenchao Gao, Zhiyi Jing, Mei Song, Guangwei Xu, Jianhua Sui, Wenhui Li
## Abstract
In the original article, the two images on the far right (hNTCP-m84-87 and mNTCP-h84-87-mk157-165) were duplicates. This unfortunate duplication error probably occurred while preparing the two images for the figure and remained overlooked at later stages of the publication process. Both images show no binding of FITC-pre-S1 to the two mutated NTCP receptors. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12828704&blobtype=pdf | # Unveiling the RNA virus landscape of Cotesia chilonis: geographic distribution, diversity , and potential roles in parasitoid-host interactions
Zhichao Cao, Cheng Ue, Fei Wan, Qi Fang, Qisheng Song, Gongyin Ye
## Abstract
Cotesia chilonis, a key parasitoid wasp of the important rice pest Chilo suppressalis, plays a critical role in the biological control of its host larvae. While previous studies have predominantly focused on polydnaviruses, associated with this species, the RNA virome of C. chilonis remains largely uncharacterized. To address this gap, we conducted a comprehensive viromic analysis across 17 geographically distinct populations of C. chilonis, identifying nine nove l RNA viruses phylogenetically affiliated with eight families: Xinmoviridae, Artoviridae, Rhabdoviridae, Qinviridae, Orthomyxoviridae, Phenuiviridae, Narnaviridae, and Virgaviridae. These viruses include seven negativesense single-stranded RNA (-ssRNA) viruses and two positive-sense single-stranded RNA (+ssRNA) viruses. Analysis of their geographic distribution revealed significant distribution patterns, with two -ssRNA viruses (C. chilonis Xinmo-like virus, CcXLV, and C. chilonis Artolike virus, CcALV) demonstrating broad prevalence and stable spatiotemporal persistence. This study not only expands the diversity of RNA viruses in parasitoid wasps but also provides foundational insights into potential virus-host interactions, offering potential avenues for optimizing parasitoid-based pest management strategies.
## Introduction
Insects represent the most abundant animal group on Earth. Although 1 million insect species have been described, recent research suggests that ∼80% remain undiscovered (Stork 2018). Viruses, characterized by their small size, diversity, and simplicity, are also widely distributed and abundant in association with insects (Andrade et al. 2018). Owing to their strong environmental adaptability and fecundity, insects are important vectors for viral transmission, posing a great thr eat to agricultural and livestock production and human health (Paixao et al. 2018). More than 70% of known plant viruses depend on insect vectors for survival and transmission (Hogenhout et al. 2008). Insects can transmit both DNA and RNA viruses. However, compared to DNA viruses, RNA viruses have received less research attention, partly due to challenges associated with their instability (Lacey et al. 2015, Liao et al. 2022). The advent of next-generation sequencing (NGS) and third-generation sequencing (TGS) technologies has gradually facilitated the discovery of insect RNA viruses over the past decade (Käfer et al. 2019, Wu et al. 2020). The diversity of negativesense single-stranded RNA (-ssRNA) viruses in arthropods is critical for understanding the origin and evolution of RNA viruses (Li et al. 2015).
Parasitoid wasps are a group of parasitic insects that act as natural enemies of many agricultural pests , including Helicoverpa armigera, Ostrinia furnacalis, and Sitophilus zeamais (Mamoon-ur-Rashid et al. 2025). Parasitoid wasps harbour a vast array of viruses, with which t hey share intricate relationships (Wang et al. 2021). For instance, some parasitoid wasps regulate the host immune response b y injecting polydnaviruses (PDVs) (Zhang et al. 2021). Recent research on RNA viruses unveils that many nonpathogenic viruses are beneficial for parasitoid wasp reproduction and pest suppression. For example, Diachasmimorpha longicaudata Narna-like virus (DlNaLV) can selectively infect female D. longicaudata and induce male-specific lethality (Zhang et al. 2023). Pteromalus puparum negative-strand RNA virus-1 (PpNSRV-1) can regulate the offspring sex ratio of its parasitoid host (Wang et al. 2017). Rondani's wasp virus 1 (RoWV-1), vectored by Pachycrepoideus vindemmiae, enhances the oviposition capacity of its host Drosophila melano gaster and prolongs the developmental period upon infection (Zhang et al. 2021). Although many studies have shown that RNA viruses facilitate parasitoid parasitism and reproduction, most documented work has been limited to virus species identification (Dheilly et al. 2015, Della et al. 2020, Wang et al. 2021, Guinet et al. 2024). The true abundance of RNA viruses in parasitoids is still far beyond our current understanding (Qi et al. 2023). While researchers have created a comprehensive dataset on the spatiotemporal and geographic distribution of mosquitoes and associated viruses in various regions in China (Atoni et al. 2020), similar efforts for parasitoid wasps are lacking. Parasitoid wasps exhibit incredible host richness and diversity of parasitic modes , which may often be facilitated by viruses that aid in host exploitation (Polaszek and Vilhemsen 2023). Thus, it can be inferred that a vast number of RNA viruses exist in Hymenoptera, especially in parasitoid wasps associated with host pests (Qi et al. 2023).
To enrich the RNA virus database for the parasitoid wasps, we performed transcriptome sequencing of the endoparasitoid wasp Cotesia chilonis (Hymenoptera: Braconidae), an effective natural enemy of the rice striped stem borer, Chilo suppressalis (Lepidoptera: Crambidae). Nine novel RNA viruses, comprising seven -ssRNA viruses and two positive-sense single-stranded RNA(+ssRNA) viruses, were identified from samples collected at 17 major rice-producing areas in China. The complete genome sequences of six viruses were validated using rapid amplification of complementary DN A (cDNA) ends (RACE). We also used RT-PCR to analyse the infection rate and viral load of two widely distributed -ssRNA viruses in C. chilonis. Our study provides valuable resources for further research on the interaction between RNA viruses and their host C. chilonis.
## Materials and methods
## Insect rearing
Cotesia chilonis and its host C. suppressalis were collected in the rice fields from 17 locations covering six provinces, autonomous regions, or municipalities directly under the Central Government in China from 2022 to 2023. Both insects were maintained in an incubator at 27 ± 1 • C and 70 ± 5% relative humidity (RH) with a 16 h:8 h light/dark cycle, fed with artificial diet and 20% honey solution, respectively (Wen et al. 2021). The third to fourth instar C. suppressalis larvae were used as hosts to propagate C. chilonis.
## RNA extraction and sequencing
For each of the 17 geographic populations of C. chilonis, RNA libraries were constructed using eight adult males and eight adult females, respectively, with three biological replicates per sex. Wasps were immobilized on ice and surface-sterilized by washing with 1× phosphate-buffered saline (PBS, 10 mM, pH = 7.4 ± 0.1). Extraction of total RNA was performed using the TRIzol reagent (Thermo Fisher Scientific, USA) following the instructions. Quality and concentration of RNA were assessed by 1% agarose gel electrophoresis (Tsingke Biotech, Beijing, China) and NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). Transcriptome sequencing libraries were prepared with the TruSeq RNA Library Preparation Kit (Illumina, USA), and index codes were added to the sequence of each sample. The indexed libraries were clustered on the cBot Cluster Generation System using TruSeq PE Cluster Kit v3-cBot-HS (Illumina). Finally, paired-end (150 bp) sequencing of each library was performed on the Illumina Novaseq sequencing platform.
## Assembly of vir al genome
Adapter and low-quality sequences were removed from the raw sequencing data using F astp software (v0.20.1) to obtain clean reads (Chen et al. 2018). De novo assembly was performed using Trinity (v2.15.1) with default parameters (Haas et al. 2013).
To identify viral sequences, a two-step BLAST approach was employed. First, all assembled contigs were compared against the National Center for Biotechnology Information (NCBI) Viral RefSeq database (taxid:10239) using Diamond BLASTx (v2.0.15) with an E-value cutoff of 1 × 10 -5 (Pan et al. 2024). Second, the resulting contigs were subjected to a similarity search against the entire NCBI nonredundant (nr) database (2023) (Wang et al. 2024). Clean reads were mapped to the potential RNA virus sequences using Bowtie2 version 2.5.1 (Langmead and Salzberg 2012) to obtain the reads of different viruses. The viral abundance was quantified using the salmon software version 1.10.2. Virus abundance was plotted using the R package ggplot2 of R version 4.4.2 according to different populations' library reads.
## PCR-based validation of vir al genomes
Primers were designed using Primer Premier 5.0 based o n the sequencing results (Supplementary Table S1). cDNA was synthesized from total RNA using the TransScript ® One-Step gDNA Removal and cDNA Synthesis SuperMix Kit (TransGen Biotech, Beijing, China). The viral genome was amplified via overlaps R T-PCR (100 bp), and the genome termini were performed using the SMARTER™ RACE 5 /3 Kit (TaKaRa, Dalian, China). The amplicons were purified and sequenced after 1% agarose gel electrophoresis. Finally, the full-length viral genomes were obtained by splicing the verified sequences using SeqMan.
## Phylogenetic analyses of viruses
Potential ORFs in the viral genomes were predicted using ORFfinder version 2023 (https://www.ncbi.nlm.nih.gov/orffinder/). Conserved domains in the viral proteins were identified using Conserved Domain Search version 2023 (https://www.ncbi.nlm. nih.gov/Structure/cdd/wrpsb.cgi). Viral genome structures were illustrated using the Illustrator for Biological Sequences version 2.0 (IBS). Protein properties, including molecular weight and theor etical pI, were predicted using the ProtParam tool version 2023 (https://web.expasy.org/protparam/). Signal peptides were predicted using SignalP-6.0 Server. TMHMM and MemBrain were used for transmembrane domain analysis. Potential glycosylation sites of viral proteins were predicted by NetOGlyc 4.0 server and NetNGlyc 1.0 server, and the phosphorylation sites were predicted and identified by NetPhos 3.1 server. RdRp sequences of the viruses were screened by Multiple Alignment using Fast Fourier Transform (MAFFT) alignment with the default parameters. Poorly aligned regions were trimmed from the multiple sequence alignment using TrimAl. The best-fit substitution model was selected by ModelFinder, and a maximum likelihood (ML) phylogenetic tree was constructed with IQ-TREE, with branch support assessed by 1000 bootstrap replicates.
## Detection of RNA virus infection and viral load of CcXLV and CcALV
To assess viral infection across development and tissues, we collected larvae, cocoons, and adults of C. chilonis. Larval samples (n = 20) were collected on Days 1, 7, and 9 of the larval stage; cocoon samples (n = 20) on Days 1 and 5 after cocooning; and adult samples (n = 10) on day 1 after eclosion. Additionally, heads, thoraxes, and abdomens from Day-1 adults (n = 20); guts, remnants, ovaries, and testes from Day-1 female and male adults (n = 50); and whole female and male adults on Days 1, 2, and 3 after eclosion (n = 10) were collected. Three biological replicates were prepared for each sample type. RNA extraction procedures were described in Section 2.2. The viral loads of CcXLV and CcALV were measured by absolute quantitative RT-PCR (RT-qPCR). Primers for RT-qPCR were designed using Primer 3 (https://primer3.ut. ee/) (Supplementary Table S2). RT-qPCR was performed using ChamQ SYBR qPCR Master Mix (Vazyme, Nanjing, China) on a Bio-Rad CFX 96 Real-Time Detection System (Bio-Rad, Hercules, CA, USA), with three biological replicates per sample. To generate standard curves, the target fragments of CcXLV and CcALV were cloned into the pCE3 Blunt Vector. The plasmid concentration was measured by Nanodrop 2000, and the copy number was calculated using the formula: (6.02 × 10 23 × plasmid concentration in ng/μl × 10 -9 )/(plasmid length × 660) = copies/μl. Serial 10-fold dilutions of the plasmids, ranging from 10 3 to 10 8 copies/μl, were used as templates for qPCR to generate standard curves. The standard curve for CcXLV was y = -3.5737x + 41.555 (R 2 = 0.9976), and for CcALV was y = -3.0326x + 39.079 (R 2 = 0.9967), where y represents the Cq value and x is the log 10 (plasmid copy number).
## Statistical anal ysis
All experimental values were expressed as mean ± standard deviation (SD). The qPCR data were tested for homogeneity of variance, followed by analysis of variance (ANOVA) and post hoc multiple comparison tests using Tukey's test (P < .05). Student's t-test was used to analyse the viral load data of different generations (ns P > .05; * P < .05; * * P < .01; * * * P < .001; * * * * P < .0001). Statistical analyses of all data were performed using R v.4.2. Images were drawn using GraphPad Prism 9.0 (GraphPad, San Diego, CA, United States).
## Results
## Metaviromic analysis revealing nine novel RNA viruses with minimal geographic variation
In total, nine complete or nearly complete viral genomes were assembled from 167 contigs. Phylogenetic alignment identified nine viral-like genomes, comprising seven ssRNA viruses (affiliated with families Xinmoviridae, Artoviridae, Rhabdoviridae, Qinviridae, Orthomyxoviridae, and Phenuiviridae) and two +ssRNA viruses (families Virgaviridae and Narnaviridae). These viruses were named C. chilonis Xinmo-like virus (CcXLV), C. chilonis Artolike virus (CcALV), C. chilonis Rhabdo-like virus (CcRLV), C. chilonis Qin-like virus (CcQLV), C. chilonis Orthomyxo-like virus (CcOLV), C. chilonis Phenui-lik e virus (CcPLV), C. chilonis Bunya-like virus (CcBLV), C. chilonis Virga-like virus (CcVLV), and C. chilonis Narnalike virus (CcNLV), respectively. Notably, the full-length genomes of CcXLV, CcALV, CcRLV, CcQLV, CcNLV, and CcVLV were obtained using RACE.
We conducted an in-depth analysis of the virome of C. chilonis using 51 samples collected from 17 sampled locations in China (Supplementary Table S 3). After extracting total RNA from each sample, eight female wasps and eight male wasps were used to construct eac h of the 51 RNA libraries for Illumina Novaseq sequencing (Supplementary Table S4). From the 2 555 699 452 paired-end reads generated, 319 contigs aligned to RNA viruses were identified after quality control. We identif ied nine novel viruses and seven known viruses in these contigs (Fig. 1 and Supplementary Table S5). A heat map was drawn to show the abundance o f these contigs at different locations (Supplementary Fig. S1). Some viral sequences could only be classified to the order level due to fragment incompleteness. Contigs from +ssRNA, -ssRNA, and double-stranded RNA (dsRNA) viruses were identified. -ssRNA had the largest number of RNA virus contigs, with the Orthomyxoviridae family having the largest number of contig fragments. These fragments were concentrated in Zhejiang Province. For ssRNA, the Artoviridae contigs had the widest distribution at collection sites. +ssRNA was less abundant than -ssRNA. However, Narnaviridae contigs were the only ones found at all collection sites, and Potyviridae contigs were the most abundant and concentrated in Wenzhou of Zhejiang (WZZJ). In terms of collection sites, the largest number of contigs were found in WZZJ and Ningbo of Zhejiang (NBZJ). The fewest contigs were found in Guilin of Guangxi (GLGX) and Shaoyang of Hunan (SYHN).
## Genome characterization and phylogenetic analysis of nine RNA viruses
Mononegavirales: CcRLV, CcXLV, and CcALV
The complete viral genome of CcRLV is 12 322 nt in length with a Guanine-Cytosine (GC) content of 46.40%. The 3 UTR of the genome is 123 nt in length, and the 5 UTR is 218 nt in length (Fig. 2A). As is typical for rhabdoviruses, the CcRLV genome contains five ORFs arranged in the order 3 -N-P-M-G-L-5 , consistent with the characteristic 10-16 kb -ssRNA genome of this family (Supplementary Table S6). Phylogenetic analysis placed CcRLV near Hymenopteran Rhabdo-related virus 46, the species of the genus Alphahymrhavirus (Rhabdoviridae) (Fig. 2B). Furthermore, the G protein encoded by CcRLV contains six conserved cysteine r esidues, a hallmark of the Alphahymrhavirus genus (Supplementary Fig. S2) (Walker andKongsuwan 1999, Vasilakis et al. 2014). Therefore, CcRLV was identified as a novel virus belonging to the genus Alphahymrhavirus. CcRLV was detected at three sampling sites, with a concentrated distribution in Jiujiang of Jiangxi (JXJJ) (Fig. 1).
The CcXLV genome is 12 551 nt in length (excluding the PolyA tail) with a GC content of 34.17%. The 3 UTR and 5 UTRs of CcXLV are 235 and 272 nt in length (Fig. 2A). The high AU content of the UTRs (74.47% and 73.90%, respectively) may facilitate the formation of RNA secondary structures to enhance translation efficiency (Jobling and Gehrke 1987). The CcXLV genome contains four ORFs (Supplementary Table S5), with a protein order (3 -N-G1-G2-L-5 ) typical of other xinmovirids (Scarpassa et al. 2019). Encoded proteins were predicted, analysed, and compared using NCBI BLASTp (Supplementary Table S5). A phylogenetic tree constructed with representatives from 11 families, 80 official genera, and 7 unclassified genera within Mononegavirales placed CcXLV in a clade with Hymenopteran Anphe-re lated virus OKIAV71 (HARV71), the sole species of the genus Pelmivirus (Xinmoviridae), with 96% bootstrap support (Fig. 2B). Given that HARV71 was also identified in a hymenopteran wasp (Heteropelma amictum), this strengthens the association of CcXLV with Hymenoptera. Therefore, CcXLV was identified as a new virus in the genus Pelmivirus of the family Xinmoviridae. It was widely distributed and detected in 15 sampled locations, with the highest read count in Jinshan of Shanghai (JSSH) (Fig. 1).
Similarly, the complete genome of CcALV is 12 351 nt in length with 49.06% GC content. The 3 and 5 UTRs of the genome are 247 and 149 nt in length (Fig. 2A). The genome structures of CcALV are similar to that of PpNSRV-1 (an active Artoviridae virus), Table S 5). Phylogenetic analysis revealed a close relationship between CcALV and HbRLV6 of the genus Peropuvirus (Artoviridae), with >95% bootstrap support (Fig. 2B). Therefore, CcALV was identified as a novel virus in the genus Peropuvirus in the family Arthropoviridae. It was detected in 13 locations, with a distribution similar to CcXLV, and was a dominant virus in Yichang, Hubei (YCHB) (Fig. 1).
## Qinviridae: CcQ LV
The family Qinviridae contains only a viral genus Yingvirus and eight species, typically with bipartite genomes (Käfer et al. 2019). The CcQLV genome is 6549 nt long with a GC content of 47.35%. The 3 and 5 UTRs are 419 and 586 nt, respectively (Fig. 3A). Protein prediction revealed a single ORF (5544 nt) encoding a putative L protein (1847 aa) (Supplementary Table S5). CcQLV clustered with Wuhan insect virus 15 (Qinviridae) with high bootstrap support, indicating it is a novel virus belonging to Qinviridae (Fig. 3B). Multiple sequence alignment confirmed that CcQLV possesses the same conserved RdRp motif found in eight other qinviruses. While qinviruses have been reported in blattodean and dipteran insects, decapods, gastropods, and nematodes (Shi et al. 2016). The discovery of CcQLV in C. chilonis further expands the insect host range of this family to hymenopterans. CcQLV was detected in nearly half of the sampled locations but was not a dominant species in any locations (Fig. 1).
## Orthomyxoviridae: CcO LV
Five viral segments with high similarity to members of Orthomyxoviridae were identified in C. chilonis population libraries and tentatively named CcOLV. Each segment contains one ORF, putativel y encoding potential PB2, PB1, PA, GP, and a protein of unknown function, respectively (Fig. 3A). Phylogenetic analysis based on the conserved PB1 sequence shows a close r elationship with Sinu virus of the genus Thogotovirus (Fig. 3B), supporting the classification of CcOLV as a novel thogotovirus. CcOLV was detected in more than half of sampled locations, primarily concentrated near the Shanghai-Zhejiang border (Fig. 1).
## Phenuiviridae: CcPLV and CcBLV
The typical Bunyavirales genome comprises three segments (L, M, S) encoding the RdRp, envelope glycoprotein (GP), and nucleocapsid protein (NP), respectively. Here, we identified two viruses with significant similarity to Bunyavirales: CcPLV and CcBL V. All three segments (L, M, S) were identified for CcPLV, whereas only the L segment was found for CcBLV (Fig. 3A). The genome structure and coding capacity of CcPLV are consistent with typical bunyaviruses. Phylogenetic analysis based on the RdRp region placed CcPLV within the Phenuiviridae family (Fig. 3C). However, due to the incomplete M and S segments, CcPLV is provisionally classified as a novel Phenuiviridae species. The L segment of CcBLV showed high similarity in the RdRp region to Wuhan insect virus 16. Phylogenetically, CcBLV clustered near members of the genus Mobuvirus (Phenuiviridae) (Fig. 3C). As the potential existence of additional genome segments cannot be ruled out, CcBLV is also provisionally classified as a novel Phenuiviridae species. Both CcPLV and CcBLV contain eight highly conserved motifs in the core of the RdRp region, consistent with typical Bunyavirales members (Supplementary Fig. S3). Motif E, located in a conserved β-hairpin, is f lanked by poorly conserved residues (Lang et al. 2013(Lang et al. , Černý et al. 2014)). We identified a conserved proline and several downstream residues near motif E in both viruses, matching patterns observed in other phenuiviruses ( Amroun et al. 2017). CcPLV was primarily distributed in eastern Zhejiang and Shanghai, similar to CcOLV, whereas CcBL V was mainly found in Anhui and other locations (Fig. 1).
## Two novel +ssRNA viruses: CcNLV and CcVLV
Two nearly complete +ssRNA viral sequences were obtained: CcNLV, classified in the family Narnaviridae (order Wolframvirales), and CcVLV, placed in the family Virgaviridae (order Martellivirales). CcNLV was detected at all sampled sites , whereas CcVLV was abundant in Zhejiang (detected at three samples) and was the only virus found in Guilin, Guangxi (GLGX) (Fig. 1).
The complete CcNLV genome is 2971 nt in length with a GC content of 59.37% (Fig. 4A). A single ORF covers almost the entire genome, encoding a 975 aa protein. Alignment analysis shows that CcNLV is similar to Sanya narnavirus 2 in the RdRp sequence. Interestingly, an ORF of similar size (962 aa) to the forward strand protein was identified on the reverse strand (Supplementary Table S5). This reverse-strand ORF shares high similarity with a protein encoded by the complementary strand of Sanya narnavirus 2. Similar findings of additional ORFs on complementary strands with unknown function have been reported for other insect narnaviruses (Wu et al. 2020), and some narnaviruses may possess bipartite genomes (Grybchuk et al. 2018, Chiba et al. 2020, Jia et al. 2021). Notably, many insect-derived narnaviruses show evolutionary divergence from the representative narnaviruses. CcNLV clusters with Sanya narnavirus 2 (Fig. 4B), supporting its identification as a novel virus in the family Narnaviridae.
Virgaviridae is a plant viral family with baculovirus particles, of which the number of ORFs varies greatly among virus genera. The complete genome of CcVLV, 9764 nt in length and 31.93% GC content, was obtained by amplifying the viral sequence. The 3 -UTR of the genome is 204 nt in length, and the 5 -UTR is 27 nt in length (Fig. 4A). The CcVLV genome was predicted to have four ORFs , named ORF1, ORF2, ORF3, and ORF4 (Supplementary Table S5). The phylogenetic tree shows that the virus CcVLV clusters with Pimento virga-like virus ( Virgaviridae) with a high bootstrap value (Fig. 4C). It is noteworthy that Pimento virga-like virus is also insect-derived, and both viruses form a distinct clade separate from typical plant-infecting vir gaviruses. Thus, CcVLV is a new virus in the family Virgaviridae.
## Infection dynamics and viral load analysis of CcXLV and CcALV in C. chilonis
We investigated the infection status and viral load dynamics of two widely distributed and abundant -ssRNA viruses, CcXLV and CcALV. Both viruses were detected in 12 collected samples, with the highest infection rate for CcXLV observed in NBZJ and for CcALV in WHHB. Their codetection in C. chilonis populations from half of the sampling sites indicates that 'co-infection' is common (Supplementary Table S7).
The viral load of CcXLV varied with host development. The highest viral load (68 670 copies/ng) was observed on Day 1 of the larval stage, which gradually decreased during larval development but increased significantly during the cocoon and adult stages (Fig. 5A). Notably, the viral load of CcXLV did not differ significantly between male and female adults (F 6, 14 = 1 419, P < .001). CcXLV load exhibited clear tissue specificity (F 5, 12 = 1 587, P < .001), being highest in the ovary (49 121 copies/ng) and lowest in the testes (5452 copies/ng) (Fig. 5B). Viral loads in the head, thorax, and abdomen of females were 1.9, 3.3, and 3.1 times higher than in males, respectively. Among female tissues, the viral load in the h ead was approximately half that in the thorax and abdomen (F 2, 6 = 117.3, P < .001) (Fig. 5C). In adults, the highest viral load (65 196 copies/ng) occurred on Day 2 posteclosion, which decreased significantly to 18 891 copies/ng by Day 3. From Day 1 to Day 3 posteclosion, CcXLV loads w ere consistently higher in females than in males (F 5, 12 = 474.1, P < .001) (Fig. 5D).
For CcALV, the highest viral load (6014 copies/ng) was also observed on Day 1 of the larval stage, with no significant difference between the Day 7 and Day 9 larvae. The viral load decreased continuously from Day 1 larva to Day 5 cocoon, reaching a minimum of 21 copies/ng. Virus loads in adult were higher than in other stages, with females (26 406 copies/ng) showing significantly higher loads than males (18 060 copies/ng) (F 6, 14 = 478.8, P < .0001) (Fig. 5E). CcALV loads were highest in the ovary and female remnants (4104 copies/ng and 4303 copies/ng, respectively), while loads in the testis (1294 copies/ng) were significantly lo wer than in other tissues (F 5, 12 = 209.9 P < .0001) (Fig. 5F). Female adults had significantly higher CcALV loads than males. Among females' tissues, the load was highest in the thorax (20 015 copies/ng) and lowest in the head (2356 copies/ng) (F 5, 12 = 1 250, P < .001) (Fig. 5G). Viral loads in males and females were similar on Day 2 posteclosion and decreased to the lowest level by Day 3 (F 5, 12 = 198.5, P < .001) (Fig. 5H).
## Discussion and conclusion
This study identified nine novel RNA viruses in C. chilonis, characterized their phylogenetic relationships and geographic distribution, and quantified the viral loads of two widely distributed -ssRNA viruses, CcXLV and CcALV. These findings significantly expand the known div ersity of RNA viruses in hymenopteran insects.
Through high-throughput sequencing and bioinformatics, we identified 167 contigs corresponding to the nine novel RNA viruses, which were classified into eight viral families across six viral orders: Xinmoviridae, Artoviridae, Rhabdoviridae, Qinviridae, Orthomyxoviridae, Phenuiviridae, Virgaviridae, and Narnaviridae. Notably, seven of these nine viruses are -ssRNA viruses, reinforcing the close association between -ssRNA viruses and insects (Käfer et al. 2019). All RNA viruses were detected in at least two geographic locations, with sequence alignment similarities exceeding 98% between different sites. This high degree of conservation indicates minimal nucleotide variation and remarkable genome stability across distances of hundreds of kilometres, a phenomenon similarly observed in mosquito viruses from diverse regions (Pan et al. 2024).
Regarding geographic distribution and diversity, the nine RNA viruses exhibited varying patterns across China. some viruses, such as CcNLV and CcXLV, were widely distributed, while others, like CcRLV and CcVLV, had a more restricted presence, suggesting that local ecological conditions or host population traits inf luence viral distribution. Given that C. suppressalis larvae are parasitized by multiple parasitoid species, we investigated whether RNA virus diversity in C. chilonis correlates with its status as a dominant parasitoid. We observed higher viral diversity in locations like Zhejiang Province (HZZJ and NBZJ) compared to Guilin (GLGX). Intriguingly, C. chilonis was the dominant parasitoid in Hangzhou and Ningbo but not in Guilin, suggesting potential ecological linkages between RNA virus communities and the success of their parasitoid hosts. The geographical expansion of parasitoids is known to depend on factors like host availability and environmental suitability, and the presence of diverse RNA viruses may be another factor correlated with these distribution patterns (Murillo et al. 2018). Furthermore, we found that some viruses, like CcXLV, were not only widespread but also highly abundant, potentially representing dominant species within the C. chilonis RNA virome. Another virus, CcNLV (a Narnaviridae), was present in all populations, mirroring the broad distribution of narnaviruses in invertebrates from paddy ecosystems and their high prevalence in mosquitoes across different regions (Pan et al. 2024). In summary, these novel viruses enrich the known viral diversity in parasitoid wasps and provide a foundation for understanding the ecology and host associations of insect RNA viruses.
These findings have significant implications for virus-vector ecology. Sequence analysis revealed that CcXLV lacks conserved motifs found in other xinmovirids, which m ay ref lect distinct regulatory strategies in different hosts (Parry and Asgari 2018). The discovery of CcXLV confirms the close relationship between the Pelmivirus genus and parasitoid wasps, advancing our understanding of the biological characteristics of Xinmoviridae (Käfer et al. 2019). The cysteine residue count in CcRLV is consistent with other reported Alphahymrhavirus species in hymenopterans, including wasps and ants (Roche et al. 2006). Similarly, the identification of CcQLV, a new member of the poorly characterized Qinviridae family (which currently has only eight known members, all from arthropods), further expands the viral reservoir and host range known for hymenopteran insects.
We also obtained the complete genomes of two +ssRNA viruses. CcVLV was classified into Virgaviridae, a family typically associated with plants. Numerous insect-specific alpha-like viruses related to virga/nege-like viruses have been reported, and CcVLV possesses FtsJ-like domains annotated in other alphalike viral genomes (Carapeta et al. 2015, Shi et al. 2016, Nunes et al. 2017). Transcriptome Shotgun Assembly (TSA) data have previously uncovered unknown vir ga-like sequences in many hymenopterans (Kondo et al. 2019), a finding corroborated by recent studies on parasitoid RNA viruses. For instance, a ∼9000 nt virga-like virus fr om C. vestalis in China was also assigned to Virgaviridae (Caldas-Garcia et al. 2023). The other +ssRNA virus, CcNLV, has the smallest genome in our study and belongs to Narnaviridae , which are known for their simple genome structure (Hillman and Cai 2013). Narnaviruses are widespread and exhibit high infection rates in insects (Chiapello et al. 2021), a pattern confirmed in our survey of CcNLV. Although it remains unclear whether the true hosts of narnaviruses are insects or their associated fungi, some can persist at high abundance even with low fungal infection rates. This evidence suggests that CcVLV and CcNLV may have a deep, specialized association with hymenopteran hosts, with potential effects on insect biology that warrant further exploration (Wang et al. 2024).
Additionally, CcPLV and CcBLV were identified as members of Phenuiviridae (order Bunyavirales), one of the largest groups of -ssRNA viruses (Kuhn et al. 2020). Many bunyaviruses are arthropod-borne and pose significant thr eats to public health and industry (Ren et al. 2021). Our comparisons confirmed that these two viruses share characteristic features of bunyaviruses and possess unique conserved residues of Phenuiviridae; related viruses are primarily found in insects like planthoppers and mosquitoes. We also identified a multi-segmented virus, CcOLV, belonging to the Thogotovirus genus (Orthomyxoviridae). Thogotoviruses are typically tick-borne and can threaten human and animal health (Fuchs et al. 2020). These discoveries broaden our perspective on the diversity of parasitoid viruses and their potential roles in disease transmission.
Given that these viruses were found in C. chilonis, an important parasitoid wasp, their impact on wasp ecology, interactions with other parasitoids, and potential consequences for agricultural pest control could be substantial. The identification of these novel viruses adds a layer of complexity to our understanding of virushost interactions in natural ecosystems.
Finally, our results point to potential virus-virus interactions. The -ssRNA viruses, particularly CcXLV and CcALV, demonstrated the highest diversity and abundance, suggesting they are key components of the C. chilonis RNA virome. The detection of both CcXLV and CcALV in half of the surveyed regions indicates that co-infection is common in this wasp species. While the viral load of CcXLV was consistentl y higher than that of CcALV across all sampled periods, we observed a notable shift after indoor rearing: the load of CcXLV decreased while that of CcALV increased (Supplementary Fig. S4). This inverse dynamic during co-infection suggests a potential interaction between the two viruses within C. chilonis. Similar virus-virus interactions that modulate viral replication have been documented in other insects, for example, infection with Nhumirim virus (NHUV) in Culex chidesteri inhibited West Nile virus (WNV), St. Louis encephalitis virus (SLEV), and Japanese encephalitis virus (JEV) replication in mosquito cells. As a result, the content of these three f laviviruses in the host decreased significantly (Kenney et al. 2014).
Both CcXLV and CcALV were detected across all examined tissues and developmental stages of the parasitoid wasp. The viral load of CcXLV decreased continuously during the wasp's development within the host but increased significantly after adult emergence. Notabl y, the ovarian viral load was significantly higher than in other tissues, implying a potential route for vertical transmission (Wang et al. 2017). Similarly, the viral load of CcALV was significantly elevated in the ovaries and female remnants compared to other tissues, suggesting a possible female-biased tropism for this virus. The pre sence of these RNA viruses in an ecologically important parasitoid suggests they could inf luence parasitoid-host interactions (French andHolmes 2019, Zhang et al. 2019). A question of considerable interest is whether CcXLV and CcALV possess the capacity to cross-infect the host pest, C. suppressalis. Should such cross-infection occur, these viruses could play additional functional roles in mediating the ecological dynamics between the parasitoid and its host, potentially regulating parasitism efficiency, host susceptibility, or pest population dynamics. Future empirical studies are essential to validate these hypotheses and delineate the ecological significance of these viruses. Furthermore, profiling the composition and diversity of RNA viruses across different parasitoid populations will be crucial for understanding their roles in population ecology and host invasion. The coexistence of multiple viruses within individual hosts may have important implications for viral evolution and host-pathogen dynamics.
In conclusion, this study significantly expands the database of RNA viruses in parasitoid wasps and provides new insights into the biogeography and diversity of insect RNA viruses. Our findings establish a foundation for future research to explore the functional roles of these viruses in parasitoid wasp ecology and the potential tripartite interactions among viruses, parasitoids, and their host pests.
## References
1. Amroun, Priet, De Lamballerie (2017) "Bunyaviridae RdRps: Structure, motifs, and RNA synthesis mac hinery" *Crit Rev Microbiol*
2. Andrade, Arantes, Rodrigues (2018) "Ubiquitous giants: A plethora of giant viruses found in Brazil and Antarctica" *Virol J*
3. Atoni, Zhao, Hu (2020) "A dataset of distribution and diversity of mosquito-associated viruses and their mosquito vectors in China" *Sci Data*
4. Santos, Fonseca (2023) "The viromes of six ecosystem service provider par asitoid wasps" *Viruses*
5. Carapeta, Do Bem, Mcguinness (2015) "Negeviruses found in multiple species of mosquitoes from southern Portugal: Isolation, genetic diversity, and replication in insect cell culture" *Virology*
6. Černý, Bolfíková, Valdés (2014) "Evolution of tertiary structure of viral RNA dependent polymerases" *PLoS One*
7. Chen, Zhou, Chen (2018) "Fastp: An ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
8. Chiapello, Bosco, Ciuffo (2021) "Complexity and local specificity of the virome associated with tospovirus-tr ansmitting thrips species" *J Virol*
9. Chiba, Oiki, Yaguchi (2020) "Discovery of divided RdRp sequences and a hitherto unknown genomic complexity in fungal viruses" *Virus. Evolution*
10. Della, Byrne, Mullins (2020) "Characterization of potato virus Y isolates and assessment of nanopore sequencing to detect and genotype potato viruses" *Viruses*
11. Dheilly, Maure, Ravallec (2015) "Who is the puppet master? Replication of a parasitic wasp-associated virus correlates with host behaviour manipulation" *Proc R Soc B Biol Sci*
12. French, Holmes (2019) "An ecosystems perspective on virus evolution and emer gence" *Trends Microbiol*
13. Fuchs, Oschwald, Graf (2020) "Tick-transmitted thogotovirus gains high virulence by a single MxA escape mutation in the viral nucleoprotein" *PLoS Pathog*
14. Grybchuk, Akopyants, Kostygov (2018) "Viral discovery and diversity in trypanosomatid protozoa with a focus on relatives of the human parasite leishmania" *Proc Natl Acad Sci*
15. Guinet, Leobold, Herniou (2024) "A novel and diverse family of filamentous DNA viruses associated with par asitic wasps" *Virus Evol*
16. Haas, Papanicolaou, Yassour (2013) "De novo transcript sequence reconstruction from RNA-seq using the trinity platform for reference generation and analysis" *Nat Protoc*
17. Hillman, Cai (2013) "Chapter six -the family Narnaviridae: Simplest of RNA viruses" *Adv Virus Res*
18. Hogenhout, Ammar, Whitfield (2008) "Insect vector interactions with persistently transmitted viruses" *Annu Rev Phytopathol*
19. Jia, Fu, Jiang (2021) "Interannual dynamics, diversity and evolution of the virome in sclerotinia sclerotiorum from a single cr op field" *Virus. Evolution*
20. Jobling, Gehrke (1987) "Enhanced translation of chimaeric messenger RNAs containing a plant viral untranslated leader sequence" *Nature*
21. Käfer, Paraskevopoulou, Zirkel (2019) "Re-assessing the diversity of negative strand RNA viruses in insects" *PLoS Pathog*
22. Kenney, Solberg, Langevin (2014) "Characterization of a novel insect-specific flavivirus from Brazil: Potential for inhibition of infection of arthropod cells with medically important flaviviruses" *J Gen Virol*
23. Kondo, Chiba, Maruyama (2019) "A novel insect-infecting virga/nege-like virus group and its pervasive endogenization into insect genomes" *Virus Res*
24. Kuhn, Adkins, Alioto (2020) "2020 taxonomic update for phylum Negarnaviricota (Riboviria: Orthornavirae), including the large orders Bunyavirales and Mononegavirales" *Arch Virol*
25. Lacey, Grzywacz, Shapiro-Ilan (2015) "Insect pathogens as biological control agents: Back to the future" *J Invertebr Pathol*
26. Lang, Zemla, Zhou (2013) "Highly similar structural frames link the template tunnel and NTP entry tunnel to the exterior surface i n RNA-dependent RNA polymerases" *Nucleic Acids Res*
27. Langmead, Salzberg (1923) "Fast gapped-read alignment with bowtie 2" *Nat Methods*
28. Li, Shi, Tian (2015) "Unprecedented genomic diversity of RNA viruses in arthropods reveals the ancestry of negative-sense RNA viruses" *elife*
29. Liao, Xie, Shi (2022) "Over two decades of research on the marine RN A virosphere" *iMeta*
30. Mamoon-Ur-Rashid, Riaz-Ud-Din, Baloch (2025) "Supplementary bioactivities of Curcuma longa extracts and parasitoid Anisopteromalus calandrae as a postharvest eco-friendly management tactic against maize weevil, Sitophilus zeamais (Motschulsky) (coleoptera: Curculionidae)" *Entomol Res*
31. Murillo, Vanlaerhoven, García (2018) "Food web associations and effect of trophic resources and environmental factors on parasitoids expanding their host range into non-native hosts" *Entomol Exp Appl*
32. Nunes Mr, Contreras-Gutierrez, Guzman (2017) "Genetic characterization, molecular epidemiology, and phylogenetic relationships of insect-specific viruses in the taxon Negevirus" *Virology*
33. Paixao, Teixeira, Rodrigues (2018) "Zika, chikungunya and dengue: The causes and threats of new and re-emerging arboviral diseases" *BMJ Glob Health*
34. Pan, Zhao, Gou (2024) "Metagenomic analysis of individual mosquito viromes reveals the geographical patterns and drivers of viral diversity" *Nat Ecol Evol*
35. Parry, Asgari (2018) "Aedes Anphevirus: An insect-specific virus distributed worldwide in Aedes aegypti mosquitoes that has complex interplays with W olbachia and dengue virus infection in cells" *J Virol*
36. Polaszek, Vilhemsen (2023) "Biodiversity of hymenopteran parasitoids" *Curr Opin Insect Sci*
37. Qi, Ye, Zhang (2023) "Diversity of RNA viruses in agricultural insects" *Comput Struct Biotechnol J*
38. Ren, Shen, Wang (2021) "Recent advances in bunyavirus reverse genetics research: Systems development, applications, and future perspectives" *Front Microbiol*
39. Roche, Bressanelli, Rey (2006) "Crystal structure of the low-pH form of the vesicular stomatitis virus glycoprotein G" *Science*
40. Scarpassa, Debat, Alencar (2019) "An insight into the sialotranscriptome and virome of Amazonian anophelines" *BMC Genomics*
41. Shi, Lin, Tian (2016) "Redefining the invertebrate RNA virosphere" *Nature*
42. Stork (2018) "How many species of insects and other terrestrial arthropods are there on earth?" *Annu Rev Entomol*
43. Vasilakis, Castro-Llanos, Widen (2014) "Arboretum and puerto almendras viruses: Two novel rhabdoviruses isolated from mosquitoes in Peru" *J Gen Virol*
44. Walker, Kongsuwan (1999) "Deduced structural model for animal rhabdovirus glycoproteins" *J Gen Virol*
45. Wang, Fang, Wang (2017) "A novel negative-stranded RNA virus mediates sex ratio in its parasitoid host" *PLoS Pathog*
46. Wang, Yuan (2021) "Diverse RNA viruses discovered in three parasitoid wasps of the rice weevil Sitophilus oryzae" *mSphere*
47. Wang, Chao, Yan (2024) "Genetic diversity of RNA viruses infecting invertebrate pests of rice" *Sci China Life Sci*
48. Wen, Chen, Chen (2021) "The overexpression of insect endogenous microRNA in transgenic rice inhibits the pupation of Chilo suppressalis and Cnaphalocrocis medinalis" *Pest Manag Sci*
49. Wu, Pang, Cheng (2020) "Abundant and diverse RNA viruses in insects revealed by RNA-seq analysis: Ecolog ical and evolutionary implications" *mSystems*
50. Xue, Wang, Fang (2024) "Transcriptomic alterations in host parasitoid wasps resulting in extended lifespan due to PpNSRV-1 infection" *Crop Health*
51. Zhang, Chen (2019) "Expanding the RNA virosphere by unbiased metagenomics" *Annu Rev Virol*
52. Zhang, Wang, Yuan (2021) "A novel cripavirus of an ectoparasitoid wasp increases pupal duration and fecundity of the wasp's Drosophila melanogaster host" *ISME J*
53. Zhang, Li, Li (2023) "RNA virus diversity in three parasitoid wasps of tephritid flies: Insights from novel and known species" *Microbiol Spectrum* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12431657&blobtype=pdf | # Discovery and mechanistic insights of dibenzoylmethane as a broad spectrum inhibitor of coronavirus
Yuan Sun, Shu Xiaoyang, Lu 1☯, Chen, Weijuan Shang, Yumin Zhang, Gengfu Xiao, Leike Zhang
## Abstract
Coronavirus, a large family of positive-sense RNA viruses, are responsible for both mild and severe respiratory illnesses, ranging from the common cold to life-threatening conditions. Despite significant advances in vaccine and antiviral development, the high mutability of human coronaviruses (HCoVs), such as SARS-CoV-2, presents a major challenge in treating these infections. Effective, broad-spectrum antiviral drugs are urgently needed to address both current and future HCoV outbreaks. Here, we conducted high-throughput screening of a natural product library containing 3407 compounds to identify potential antiviral agents against HCoV-OC43 and HCoV-229E. We identified several natural products with inhibitory effects on HCoV-229E, HCoV-OC43, and the SARS-CoV-2 variants Delta (B.1.617.2) and Omicron (BA.5) in vitro without evident cytotoxicity. Among these, dibenzoylmethane (DBM) not only demonstrated broad-spectrum anticoronavirus activity in vitro but also effectively inhibited HCoV-OC43 replication in a BALB/c mouse model. Pharmacokinetic analysis revealed that DBM, when administered orally, maintained effective concentrations in the blood over an extended period, suggesting its suitability for oral administration.Mechanistically, DBM was found to regulate caspase-6, a host factor that suppresses interferon signalling and promotes HCoV replication. These findings highlight DBM as a promising candidate for the development of therapeutics targeting HCoVs, offering potential for treating infections by both established and emerging HCoVs.
## Introduction
Coronaviruses (CoVs) constitute a large family of single-strand and positive-sense RNA viruses. Historically, several coronaviruses, such as SARS-CoV, MERS-CoV, and SARS-CoV-2, have caused significant public health crises. In addition to these more notorious viruses, other human coronaviruses, including HCoV-OC43 and HCoV-229E, are common pathogens responsible for mild symptoms such as fever, rhinitis, pharyngitis, and abdominal discomfort [1,2]. However, these viruses can also lead to severe acute respiratory infections, such as pneumonia and bronchitis, and may be life-threatening in immunocompromised individuals, including children, adults over 65 years of age, and patients with underlying conditions such as cardiopathy, chronic obstructive pulmonary disease, diabetes, and immunosuppression [3]. Notably, SARS-and MERS-like CoVs found in bat populations have demonstrated the ability to efficiently infect primary human airway cells [4,5], underscoring the potential for future cross-species transmission of CoVs.
Currently, effective treatments for coronavirus infections are limited, particularly given the high variability of these viruses. The efficacy of existing antiviral drugs and vaccines may be diminished or even rendered ineffective as the virus undergoes mutation [6]. For example, during the COVID-19 pandemic, SARS-CoV-2 generated multiple variants that exhibited significant differences in transmissibility, pathogenicity, vaccine escape, and drug resistance [7,8]. In this context, the development of broad-spectrum antiviral drugs is highly important.
One common approach for discovering antiviral compounds is high-throughput screening (HTS) of compound libraries [9]. In contrast to the process of de novo drug development, drug repurposing presents numerous advantages, including a reduced timeline for development, sustained safety profiles, a lower incidence of adverse effects, and increased cost-effectiveness. These attributes render drug repurposing a promising strategy for discovering therapeutic options for both emerging and re-emerging infectious diseases. Natural products (NPs) encompass a vast array of biologically active compounds. Some NPs have been proven to be ideal reservoirs for repurposing against viral infection [10][11][12][13]. However, most of these NPs have shown mild to modest effects, and very few of them have been investigated or found to be effective in vivo with favourable pharmacokinetics and treatment routes.
In this study, we carried out high-throughput repurposing screening of a natural product library containing 3407 compounds to identify active antivirals against HCoV-OC43 and HCoV-229E. In total, 5 NPs inhibited both OC43 and 229E in vitro without obvious cytotoxicity. Among these compounds, DBM showed broad-spectrum anti-HCoV activity in vitro and potently suppressed HCoV-OC43 replication in BALB/c mice. Moreover, the pharmacokinetic and metabolic stability studies suggested that DBM is a safe and well-absorbed compound for oral treatment. Mechanistically, we found that DBM exerts anticoronavirus activity by regulating caspase-6, a host factor that suppresses interferon (IFN) signalling and facilitates coronavirus replication. Taken together, our results suggest that DBM is a potential candidate for combating HCoV infection.
## Results
## Identification of hits through HTS that prevent cell death caused by coronavirus infections
HTS has become an indispensable tool in modern drug discovery, enabling the rapid evaluation of large compound libraries for potential therapeutic activity. By systematically testing thousands of compounds, HTS provides a powerful means to identify candidates with specific biological effects, including antiviral properties. In this study, we developed an HTS platform to evaluate the effectiveness of compounds that protect cells from cell death caused by two common human coronaviruses, HCoV-229E and HCoV-OC43. We screened a library of 3407 compounds from the Natural Product Library Plus (MedChemExpress, HY-L021P) with the CellCounting-Lite 2.0 Reagent (Vazyme, DD1101) to assess cell viability after inoculation with the viruses in the presence or absence of natural compounds from the library (Fig 1A). Through this screening, 20 natural compounds were identified to be able to protect cells during infection with HCoV-229E and/or HCoV-OC43 (Fig 1B). We subsequently determined the 50% effective concentration (EC 50 ) values of the 20 valid compounds for inhibiting HCoV-229E in Huh-7 cells and HCoV-OC43 in RD cells by analysing the amount of viral RNA released by the infected cells. Furthermore, the 50% cytotoxic concentration (CC 50 ) was assessed with the Cell Counting Kit-8 (CCK-8, GlpBio), and compounds with high selectivity indices (SIs), including columbianadin, DBM, erythromycin estolate, ingenol 3,20-dibenzoate, and veratramine, were selected for further evaluation (Fig 1C). These compounds exhibited antiviral effects on both HCoV-229E and HCoV-OC43, suggesting their potential broad-spectrum anticoronavirus activity. Compounds such as (2S)-2'-methoxykurarinone, (R)-sulforaphane, amphotericin B, baohuoside I, cyclovirobuxine D, deserpidine, euphorbia factor L7a, levistolide A, neoglycyrol, ponicidin, sophoraflavanone G, and tetrandrine exhibited high cytotoxicity in this assay, resulting in poor SIs below 25 (S1A Fig) . Consequently, these compounds were not further assessed, as it is challenging to develop compounds with low SIs into viable drugs.
## In vitro identification of the anticoronavirus activity and cytotoxicity of the five selected compounds
We subsequently validated the inhibitory effects of columbianidin, DBM, erythromycin estolate, ingenol 3,20-dibenzoate, and veratramine against HCoV-229E, HCoV-OC43, and the SARS-CoV-2 variants Delta (B.1.617.
## Screening of natural compounds in a suckling mouse model of HCoV-OC43 infection
After confirming the antiviral activity of the five selected natural compounds in vitro, we further assessed their therapeutic efficacy in a suckling mouse model of HCoV-OC43 infection. In this experiment, suckling mice were randomly divided into compounds in Huh-7 and RD cells. To calculate the EC 50 , cells were co-incubated with a gradient dilution of the compounds 1 h prior to inoculation with the corresponding viral strains. After 24 h of infection, viral RNA copies in the supernatants were analysed using qRT-PCR. CC 50 values were tested as described in materials and methods. The experiments were repeated three times independently with similar results. https://doi.org/10.1371/journal.ppat.1013492.g001 seven groups and treated with different agents. Columbianadin has previously been reported to suppress airway inflammation in mice at a dose of 50 mpk [14]. Feeding 1% DBM in the diet was found to be safe for mice in antitumour studies [15]. Doses of erythromycin estolate exceeding 300 mpk and of veratramine exceeding 10 mpk have also been safely used in antitumour mouse models [16,17]. VV116, which has been shown to exhibit good anticoronavirus activity at a dose of 25 mpk [18], was used as the positive control compound in this experiment. A group of mice receiving menstruum orally quaque die (QD) was used as the control group. The other groups were treated with columbianadin at 50 mg/kg (mpk) QD, DBM at 400 mpk QD, erythromycin estolate at 300 mpk QD, veratramine at 10 mpk QD, ingenol 3,20-dibenzoate at 10 mpk QD, or VV116 at 25 mpk QD. Mice were inoculated with 1 × 10 4 TCID 50 of HCoV-OC43 on the first day of the experiment and then orally treated with either a placebo or the test compounds on the first and following days. The groups of veratramine-and ingenol 3,20-dibenzoate-treated mice were euthanized on Day 1 because veratramine and ingenol 3,20-dibenzoate exhibited toxicity in the suckling mice. All the mice were sacrificed on Day 5, and the brains, spinal cords, lungs, and kidneys were collected from the mice. HCoV-OC43 infection resulted in significant body weight loss on Day 5, whereas mice from the DBM-and VV116-treated groups did not show any body weight loss during the experiment (Fig 3A). The number of viral copies from collected mouse organs and tissues was analysed by qRT-PCR to detect the HCoV-OC43 nucleocapsid gene sequence. Compared with the vehicle group, the DBM group presented significantly fewer viral copy numbers in the brain, lungs, and kidneys at 5 days post infection (Fig 3B). The viral load in the brains of the DBM-treated group was approximately 6 logs lower than that in the vehicle group. Moreover, treatment with columbianidin and erythromycin estolate did not significantly reduce the number of viral RNA copies in HCoV-OC43-infected mice.
Among the five natural compounds, only DBM had a protective effect on HCoV-OC43-infected mice. Thus, a pharmacokinetic study was performed for DBM in Sprague-Dawley (SD) rats. Following oral administration of menstruum or DBM at 200 mpk, blood samples from the rats were collected at 2 min, 5 min, 15 min, 30 min, 1 h, 2 h, 4 h, 8 h, and 24 h posttreatment. The concentration of DBM in the plasma was determined by liquid chromatography with tandem mass spectrometry (LC-MS/ MS). The minimum effective concentration (MEC) line in the graph represents the EC 95 value of DBM (Fig 3C). During the test, DBM reached its Cmax of 4,760 ± 1,920 ng/mL at 2.33 ± 1.15 h, with an AUClast of 34,300 ± 14,800 ng h/mL and a T 1/2 of 2.91 h (S2A Fig) . To analyse the metabolic stability of DBM, we incubated human and mouse liver microsomes with DBM. The T 1/2 of metabolism was an average of 13.92 min in human liver microsomes and 4.84 min in mouse liver microsomes. The results revealed that DBM was more stable when in human liver microsomes than in mouse liver microsomes (S2B Fig).
## In vivo characterization of the anti-HCoV-OC43 activity of DBM
Next, we characterized the therapeutic efficacy of the DBM in vivo. In this experiment, randomized 5-6-day-old suckling BALB/c mice were orally treated with DBM at concentrations of 500, 200, and 100 mpk 2 hours after being intranasally challenged with HCoV-OC43. Menstruum and VV116 were used as negative and positive controls, respectively. All the mice were sacrificed on Day 5 for organ and sample collection (Fig 4A). The body weight changes of the mice were continuously monitored during the trial. Infection with HCoV-OC43 led to body weight loss in the vehicle-treated mice on Day 5 (Fig 4B). Moreover, the mice treated with DBM or VV116 did not experience weight loss during the experiment, indicating that DBM protected the mice from disease-related symptoms. The number of viral RNA copies from the brain, spinal cord, lungs, and kidneys were analysed via qRT-PCR, which targeted the HCoV-OC43 nucleocapsid gene sequence. On Day 5, the number of viral RNA copies in the vehicle-treated group reached approximately 10 12 copies in the brain, 10 9 copies in the spinal cord, 10 8 copies in the lungs, and 10 6 copies in the kidney (Fig 4C). In the mice treated with DBM at 500 mpk, the number of viral RNA copies within these organs and tissues was reduced by 1-3 logs compared with the vehicle-treated group. DBM treatment at different concentrations from 100 to 500 mpk resulted in a significant reduction in viral RNA copies in certain organs or tissues. The viral titres of the organs and tissues characterized by the immunoplaque assay also indicated that the administration of DBM at 500 mpk led to a reduction in viral loads (Fig 4D). Sections of mouse brains, spinal cords, lungs, and kidneys were subjected to immunofluorescence staining with antibodies targeting HCoV-OC43 nucleocapsid proteins. The results revealed that DBM treatment reduced the viral loads in certain organs and tissues of HCoV-OC43-infected mice (Figs 4E andS3A).
The cytokine storm caused by severe HCoV-OC43 infection can lead to death in mice. By analysing the gene expression of cytokines (CCL2, CXCL10, IL-1β, IL-6, and TNF-α), we found that the inflammation accompanied by HCoV-OC43 infection was severe in the central nervous system, including the brain and spinal cord. DBM administration dramatically reduced inflammation in the brain and spinal cord of infected mice (S4A Fig) . We subsequently histologically examined the pathological changes in brain and lung sections from infected mice. The results revealed that the oral administration of DBM significantly reduced the damage caused by viral infection in these organs. Compared with vehicle treatment, DBM treatment reduced the accumulation of immune effector cells (black arrows) and severe lesions (red arrows) in the brain and pulmonary fibrosis (black arrows), pulmonary oedema (red arrows) and the formation of sputum (green arrows) in the lungs (S4B Fig).
## DBM inhibits HCoV replication by downregulating caspase-6
To obtain further insight into the stage at which DBM inhibits the replication of HCoV-OC43 and other HCoVs, RD cells were incubated with compounds, including DBM, remdesivir (RDV), and chloroquine (CQ), at different stages of viral infection. Viral RNA from the supernatant was extracted and quantitatively analysed via qRT-PCR. The results revealed that the addition of the DBM at 2 hours after virus inoculation potently inhibited the replication of HCoV-OC43, whereas the incubation of DBM with cells before and during virus infection did not significantly inhibit HCoV-OC43 infection (S5A Fig) . For the viral binding assay, cells were cultured in the presence or absence of DBM at different concentrations for 1 hour before being infected with HCoV-OC43 at an MOI of 50, and then were placed on ice for 2 hours for the viral particles to bind to the membrane of the host cells, followed by washing with PBS. The detection of viral RNA from viral particles bound to cells confirmed that the administration of DBM did not interfere with the process of viral binding (S5B Fig) . In addition to binding to receptors from host cells, the spike proteins of HCoV-OC43 could also trigger membrane fusion when HCoV-OC43 spike plasmid-transfected 293T cells were seeded together with RD cells as target cells, and the addition of DBM had no effect on this process (S5C Fig) To identify the intrinsic mechanism of the distinct antiviral phenotypes of DBM, quantitative proteomics studies were conducted via LC-MS/MS to identify the proteins whose expression was upregulated or downregulated by DBM treatment (Fig 5A). Serine/arginine-rich splicing factor protein kinase-1 (SRPK1), syntaxin 12 (STX12) and ubiquinol-cytochrome c reductase hinge protein (UQCRH) from the upregulated group have been reported to be related to the infection and replication of certain kinds of viruses [19][20][21]. Nevertheless, the overexpression of these three proteins did not significantly inhibit HCoV-OC43 infection (Fig 5B and5C). Proteins downregulated by DBM treatment, including isocitrate dehydrogenase (NAD (+)) 3 noncatalytic subunit gamma (IDH3G), ALG11 alpha-1,2-mannosyltransferase (ALG11), caspase-6 (CASP6), cytochrome P450 family 1 subfamily A member 1 (CYP1A1), ElaC ribonuclease Z 2 (ELAC2), and upstream binding protein 1 (UBP1), were knocked down in host cells by siRNA transfection, resulting in a significant reduction in the relative mRNA levels of each gene compared to the control group 24 hours post-transfection (Fig 5D). The replication efficiency of HCoV-OC43 in siRNA-transfected cells was evaluated via qRT-PCR (Fig 5E). Three different siRNAs for CASP6 were used for knockdown validation to eliminate the interference of siRNA knockdown efficiency, and the results revealed that successful knockdown of CASP6 significantly inhibited HCoV-OC43 replication (Fig 5F). siRNA-mediated transfection of CASP6-1, CASP6-2 and CASP6-3 before HCoV-OC43 infection reduced the quantity of CASP6 and the HCoV-OC43 nucleocapsid protein in HCoV-OC43-infected cells, indicating that a decrease in caspase-6 in host cells can inhibit virus replication.
To explore how caspase-6 was downregulated by DBM, we blocked the main protein degradation pathway in cells, including the ubiquitin-proteasome system, the lysosomal proteolysis pathway, and the apoptotic pathway, which leads to the cleavage of caspase, with the inhibitors CQ, MG-132, and Z-VAD-FMK, respectively. The Western blot results indicated that DBM-induced caspase-6 downregulation was not related to any of the aforementioned protein degradation pathways (Fig 5G). Moreover, we observed that in DBM-treated cells, the relative level of CASP6 mRNA was decreased starting at 12 hours after compound inoculation, which explained the terminal reduction in the level of the caspase-6 protein (Fig 5H). Interestingly, for cells transfected with the caspase-6 expression plasmid with alternative promoter elements, the relative level of caspase-6 mRNA was also decreased by DBM treatment, which suggested that DBM may affect the mRNA stability of caspase-6 (Fig 5I). Eventually, the decrease in the amount of caspase-6 mRNA and protein led to the decrease in the activity of caspase-6 in DBM-treated cells (Fig 5J). To further validate the important role of caspase-6 in DBM-mediated antiviral mechanisms, we constructed a CASP6 knockout RD cell line to evaluate the antiviral activity of DBM and Z-VEID-FMK, a peptidomimetic inhibitor that covalently binds to caspase-6. The antiviral activity of both DBM A key advantage of host-targeting antivirals (HTAs) over direct-acting agents (DAAs) is their lower susceptibility to drug resistance. In this study, we serially passaged HCoV-OC43 for 17 generations in the presence of DBM, gradually increasing the DBM concentration from 0 to 4.5 times its EC 50 value. Next, we evaluated the EC 50 values of virus strains passaged with or without DBM. The results revealed no apparent differences in the EC 50 values between the DBM-treated group (0.85 μM) and the DMSO control group (0.57 μM) (S6A and S6B Fig) . In summary, caspase-6 was demonstrated to be an important intracellular host factor that could be modulated by DBM to exert its antiviral effects.
## DBM alleviates the cleavage of nucleocapsid proteins to inhibit coronavirus replication
To further explore the correlation between the DBM-induced downregulation of caspase-6 and the antiviral effect of DBM, we pretreated cells with serially diluted DBM or Z-VEID-FMK, an effective irreversible caspase-6 inhibitor that causes no reduction in caspase-6 quantity, before HCoV-OC43 inoculation. After infection, the cells were cultured overnight, and the quantities of viral nucleocapsid protein and caspase-6 were analysed by Western blot. The results indicated that both DBM and Z-VEID-FMK inhibited virus replication and that both activity inhibition and downregulation of caspase-6 inhibited coronavirus replication (Fig 6A). Although DBM demonstrated broad-spectrum antiviral effects against a wide range of coronaviruses, namely, HCoV-229E, HCoV-OC43, HCoV-NL63, and the SARS-CoV-2 variants Delta (B.1.617.2) and Omicron (BA.5), DBM exhibited no antiviral activity against other viruses, such as enterovirus EV71 (Fig 6B ), which indicated that the antiviral mechanism of DBM may be related to certain unique mechanisms of coronavirus infection. When evaluating the inhibitory effect of DBM against infection with the SARS-CoV-2 BA.5 variant, the antiviral activity was found to be cell line-dependent. DBM demonstrated efficacy in HEK293T-AT cells, but not in Vero E6 cells (Fig 6C). Vero E6 cells, which are commonly used in virological studies, are known to be deficient in IFN production. By assessing the expression of IFNB1 in SARS-CoV-2 BA.5-infected HEK293T-AT and Vero E6 cells, we observed a marked upregulation of IFNB1 expression in HEK293T-AT cells, whereas no significant change was detected in Vero E6 cells (S8A Fig) . These findings suggest that the antiviral activity of DBM is closely associated with the presence of functional innate immune signalling pathways.
was screened as significantly different, with red representing up-regulation and green representing down-regulation. B) Protein expression of SPRK1, STX12, and UQCRH in HEK-293T cells transfected with corresponding plasmids. C) Antiviral efficacy of the transient expression of SPRK1, STX12, and UQCRH. The corresponding protein expression plasmids were transfected into RD cells by transient transfection 24 h before HCoV-OC43 inoculation. D) Relative mRNA content following siRNA knockdown. The corresponding siRNAs were transfected into RD cells separately by Lipofectamine RNAiMAX, and the total RNA in the cells was extracted 24 h after transfection. E) Antiviral efficacy of knockdown of DBM downregulated proteins. HCoV-OC43 was inoculated to the mRNA knockdown cells, and the viral RNA in the supernatant was analysed after 24 h of inoculation. F) Antiviral efficacy of caspase-6 (CASP6) knockdown, as characterised by Western blotting. Three different siRNAs were used to knockdown the expression of CASP6, respectively, and HCoV-OC43 was inoculated into the cells 24 h after siRNA transfection, and samples were collected 24 h after viral inoculation. G) Quantification of caspase-6 protein levels in DBM-treated cells, in the presence or absence of protein degradation inhibitors. Cells were pretreated with 50 μM DBM for 4 h, followed by the addition of CQ, MG-132 or Z-VAD-FMK to the cell culture medium, respectively. DMSO was used as a control. H) Quantification of CASP6 mRNA in DBM-treated cells. Cells was harvested at 12h, 24h, 36h and 48h after DBM or DMSO being added to the culture medium. I) Quantification of CASP6 mRNA in DBM-treated cells following CASP6 overexpression. Cells were pretreated with DBM or DMSO for 1 h, followed by transfection of Caspase-6 protein expression plasmid into cells by transient transfection, and cell samples were harvested at 24 h after transfection. J) Measurement of caspase-6 catalytic activity in cells treated with DMSO, DBM or Z-VEID-FMK, respectively. One unit of enzyme activity defined as the amount of enzyme that can shear 1 nmol of specific polypeptide-pNA to produce 1 nmol of pNA of Caspase in one hour at 37°C when the substrate is Previous studies have demonstrated that caspase-6 cleaves coronavirus nucleocapsid proteins, generating fragments that act as IFN antagonists and facilitate viral replication [22]. On the basis of this finding, we hypothesized that DBM downregulates caspase-6, reduces the cleavage of nucleocapsid proteins, and thereby enhances IFN signalling to inhibit viral replication. To test this hypothesis, we incubated cells transfected with plasmids expressing coronavirus nucleocapsid proteins in the presence or absence of DBM or Z-VEID-FMK. To simulate the apoptotic environment typically found in coronavirus-infected cells [23], staurosporine (STS), an apoptosis inducer known to activate caspase activity, was added to the culture medium. In the STS-treated groups, apoptosis was successfully induced, resulting in partial cleavage of caspase-6 (zymogen form, P33) to its active form (P18). In the DBM-treated groups, the levels of caspase-6 (P18) were significantly reduced, leading to a decrease in the cleavage of coronavirus nucleocapsid proteins. A similar reduction in cleaved nucleocapsid proteins was observed in the Z-VEID-FMK-treated groups, attributable to the inhibition of caspase-6 activity by Z-VEID-FMK. These results demonstrate that, compared to the control, DBM treatment downregulates caspase-6 and significantly reduces the levels of cleaved HCoV-OC43 and SARS-CoV-2 nucleocapsid proteins (Fig 6D). As the reduction in cleaved nucleocapsid protein fragments could theoretically suppress IFN signalling, we sought to investigate the relationship between DBM treatment and IFN signalling. To this end, we further transfected cells with plasmids expressing both caspase-6 and coronavirus nucleocapsid proteins, and subsequently incubated the cells with or without DBM. Analysis of IFNB1 and representative IFN-stimulated genes (ISGs), such as IFIT1, IFITM3, OAS1, and TRIM22, revealed that DBM reversed the suppression of IFN signalling by caspase-6 and nucleocapsid proteins (Figs 6E andS8B). Whereas the other two apoptosis-executing caspases, caspase-3 and caspase-7, could not inhibit IFN signalling in a similar system (S8C Fig) . Furthermore, we examined the expression of IFNB1 and its downstream IFN-stimulated genes following DBM treatment in the context of HCoV-OC43 infection (S8D and S8E Fig) . The results demonstrated that DBM treatment enhanced IFN signalling in HCoV-OC43-infected cells, leading to a significant upregulation of mRNA expression of IFNB1 and associated ISGs. Additionally, the antiviral activity of DBM was markedly reduced when the cells were pretreated with ruxolitinib (a JAK1/2 inhibitor) or IFN alpha-IFNAR-IN-1 hydrochloride (an IFN-I and IFNAR inhibitor) (Fig 6F) [24,25]. These findings suggested that DBM inhibited coronavirus infection by downregulating caspase-6, thereby reducing the generation of cleaved nucleocapsid protein fragments, which are known to suppress IFN signalling (S9A Fig).
## Discussion
Coronavirus infections have had profound impacts on public health worldwide. However, there are very few broadspectrum anti-HCoV drugs and vaccines available for clinical use. The COVID-19 pandemic led to an unprecedented international research effort on essentially all aspects of SARS-CoV-2 biology, with high-efficacy vaccines and antivirals produced in record time. The worldwide efforts to combat SARS-CoV-2 have also accelerated studies of pancoronavirus inhibitors due to concerns about known HCoV infections and unknown zoonotic coronaviruses that may be transmitted to humans in the future. While most anti-SARS-CoV-2 agents, such as remdesivir, molnupiravir, and nirmatrelvir, can also inhibit other HCoVs, including OC43 and 229E, all of these approved inhibitors are DAAs. DAAs, while highly specific and potent, face significant limitations. They are prone to resistance due to viral mutations, have a narrow spectrum of activity, DBM in IFN signalling-deficient cell lines. HEK293T-AT and Vero E6 cells were pre-treated with gradient-diluted DBM for 1 hour before infection. D) DBM downregulated caspase-6 and inhibited caspase-6-mediated cleavage of HCoV-OC43 and SARS-CoV-2 nucleocapsid proteins. Cells were pretreated with gradient-diluted DBM or Z-VEID-FMK for 24 h, and then transfected with coronaviruses nucleocapsid proteins expression plasmid, and after 16 h of transfection, cells were treated with DMSO or STS. E) DBM alleviated the suppression of IFNB1 expression induced by coronaviruses nucleocapsid proteins and caspase-6. Cells were pretreated with DMSO or DBM for 24 h, subsequently, transfected with Poly(I:C) and plasmids. F) Impact of IFN signalling suppressants on the antiviral activity of DBM. Cells were pretreated with DMSO, Ruxolitinib or IFN alpha-IFNAR-IN-1 hydrochloride, respectively, for 1 hour. Subsequently, a gradient dilution of DBM was added to the medium, and then HCoV-OC43 was inoculated into the cells. Data were statistically analysed with Student's T test. *: p ≤ 0.05, **: p ≤ 0.01, ***: p ≤ 0.001, ****: p ≤ 0.0001; ns, not significant. The experiments were repeated three times independently with similar results. https://doi.org/10.1371/journal.ppat.1013492.g006 and may cause side effects in certain populations. In contrast, by targeting host cellular mechanisms, HTAs reduce the risk of resistance and can provide broad-spectrum antiviral activity. Thus, there is an urgent need for the development of new, broad-spectrum HTAs that target various sites in currently circulating and future emerging HCoVs to prepare for future outbreaks of yet unknown HCoVs.
Natural products are an enormous family of bioactive compounds derived from plants, animals, and microbes, and the most well-known pharmaceutical development of NPs is for anticancer or antiparasitic purposes, such as paclitaxel, ivermectin, and artemisinin [26]. In comparison, the contribution of NPs to the development of antivirals is less prominent. In this study, we conducted an HTS assay to simultaneously identify inhibitors of both HCoV-OC43 and HCoV-229E from a natural product library. Through two rounds of cell-based screening, columbianidin, DBM, erythromycin estolate, veratramine, and ingenol-3,20-dibenzoate were highlighted due to their low cytotoxicity and high inhibitory activity in vitro. In suckling BALB/c mice, veratramine and ingenol 3,20-dibenzoate were highly toxic at doses as low as 10 mpk, possibly because of the neurotoxic effects of veratramine and the proinflammatory effects of ingenol-3,20-dibenzoate [27,28]. Our preliminary in vivo study indicated that, among the natural products tested, DBM was the only compound that effectively inhibited HCoV-OC43 replication in mice. Further animal studies revealed that oral treatment with 100-500 mg/kg DBM significantly suppressed OC43 viral loads in a dose-dependent manner. Additionally, DBM attenuated inflammation in the central nervous system caused by viral infection. Furthermore, DBM treatment protected brain cells and reduced lung damage in HCoV-OC43-infected mice. Notably, a pharmacokinetic study in SD rats revealed high values for Cmax (4760 ng/ml) and T 1/2 (2.91 h) when DBM was administered orally, suggesting that DBM could represent a promising candidate for further exploration in antiviral drug development. This study is the first to demonstrate that DBM has anti-HCoV activity, in particular, a potent inhibitory effect against OC43 in vivo with favourable pharmacokinetics.
DBM and its derivatives exhibit remarkable potential as multifunctional therapeutic agents, with preclinical evidence highlighting their efficacy in cancer prevention and treatment across diverse models. Their biological activities are underscored by potent cytotoxic effects against cancer cells, achieved through induction of both intrinsic and extrinsic apoptotic pathways. For instance, the derivative DPBP demonstrates selective killing of melanoma cells (IC₅₀ = 6.25 μg/mL) with a high selectivity index, while intercalating into DNA to disrupt its structure [29]. In vivo, dietary DBM profoundly inhibits mammary tumorigenesis in mice, reducing tumor incidence by 97% via suppression of cell proliferation and DNA adduct formation, and similarly arrests the cell cycle in prostate cancer cells at the G₁/S phase [30]. Beyond oncology, DBM acts as a phase 2 enzyme inducer to promote carcinogen detoxification and exhibits antimutagenic properties by inhibiting DNA adduct formation. Notably, preclinical data support its safety, with no signs of hepatotoxicity or nephrotoxicity observed in treated animals [29]. These findings collectively position DBM derivatives as promising candidates for chemoprevention and therapy, particularly for skin, breast, and prostate cancers, though clinical translation requires further investigation into optimal formulations and long-term safety profiles.
The results of a series of cell-based assays proved that DBM inhibits HCoV-OC43 at the postentry and prebudding stages. LC-MS revealed that the expression levels of several host proteins were affected by DBM treatment. Some of the upregulated proteins, including SPRK1, STX12, and UQCRH, have been reported to be relevant to viral infection and replication [19][20][21]. The downregulated proteins were validated by siRNA knockdown. The reduction in caspase-6 expression inhibited viral replication in HEK293T cells, suggesting that caspase-6 is one of the host factors for OC43 propagation. The downregulation of caspase-6 was proven to be achieved at the mRNA level, as DBM treatment may decrease the quantity of CASP6 mRNA by impairing its stability. On the basis of this evidence, we believe that caspase-6 is the host factor downregulated by DBM during the antiviral process. The virus did not develop significant resistance after 17 consecutive passages in the presence of DBM. This finding suggested that DBM, as an HTA, is not susceptible to drug resistance.
Caspase-6 has been identified as an important host factor for influenza A virus (IAV), as caspase-6 deficiency increases susceptibility to IAV infection both in vitro and in vivo, as reported previously [31]. However, in the case of coronaviruses, caspase-6 facilitates virus replication by inhibiting the host IFN signalling pathway [22]. The inhibition of caspase-6, whether by decreasing the quantity or by inhibiting enzyme activity, can restrict coronavirus replication. In this study, we found that the downregulation of caspase-6 by DBM further decreased the amount of cleaved coronavirus nucleocapsid proteins, which were reported to be associated with IFN signalling suppression. After DBM treatment, the levels of intracellular IFNB1 and downstream IFN-stimulated genes were elevated, which could explain why DBM does not function well in IFN signalling-deficient cells or in the presence of IFN signalling suppressants. Given that coronaviruses possess a distinct mechanism for inhibiting IFN signalling, caspase-6 inhibitors have been shown to exhibit antiviral activity against a range of coronaviruses, including SARS-CoV, MERS-CoV, SARS-CoV-2, HCoV-229E, and HCoV-OC43 [22]. Therefore, in addition to the antiviral activity against HCoV-229E, HCoV-OC43, HCoV-NL63, and the SARS-CoV-2 variants Delta (B.1.617.2) and Omicron (BA.5) demonstrated in this study, we hypothesise that DBM may also possess broad-spectrum antiviral effects against other coronaviruses. Moreover, caspase-6 has been implicated in a range of non-viral diseases, with dysregulated caspase-6 activity linked to conditions such as non-alcoholic steatohepatitis, Alzheimer's disease, Huntington's disease, and preeclampsia [32][33][34][35]. In this study, DBM was identified as a potent inhibitor of caspase-6, suggesting its potential therapeutic application could be extended to the treatment of these diseases.
In summary, we identified a series of active compounds from a natural product library through HTS that effectively inhibited HCoV-OC43 or HCoV-229E. Among them, DBM was demonstrated to broadly inhibit HCoVs, including OC43, 229E, and SARS-CoV-2 variants, in vitro. Furthermore, we found that oral treatment with DBM potently inhibited HCoV-OC43 in suckling mice, with the desired pharmacokinetics confirmed in SD rats. Mechanistic studies revealed that the DBM inhibits HCoV-OC43 infection by downregulating the expression of caspase-6, a host factor that facilitates coronavirus replication through the inhibition of the IFN signalling pathway, and that DBM also prevents the cleavage of viral nucleocapsid proteins to suppress the expression of IFNB1 and IFN-stimulated genes. Our findings demonstrate that DBM is a promising HTA candidate for combating coronavirus.
## Materials and methods
## Ethics statement
All animal experiments were conducted in strict accordance with the guidelines for the use and care of laboratory animals. The experimental protocol was approved by the Ethics Committee of the Wuhan Institute of Virology, Chinese Academy of Sciences (approval no. WIVA25202203).
## Cell lines, viruses and compounds
Huh-7, RD, HEK293T, HEK293T-AT, and Vero E6 cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco) supplemented with 10% foetal bovine serum (FBS). MRC-5 cells were cultured in minimum essential medium (MEM) supplemented with 10% FBS. The cells were cultured at 37 °C in a 5% CO₂ incubator and routinely checked for mycoplasma contamination. The viral strains HCoV-229E (ATCC VR-740) and HCoV-OC43 (ATCC VR-1588) were provided by Wuhan University. The SARS-CoV-2 Delta variant (B.1.617.2, IVCAS6.7585) and Omicron variant (BA. 5, IVCAS6.8981) were sourced from the National Virus Resource Center. HCoV-229E, HCoV-OC43, and SARS-CoV-2 variants were propagated in MRC-5, RD, and Vero E6 cells, respectively. All experiments involving authentic SARS-CoV-2 viruses were conducted in the Biosafety Level 3 facility (BSL-3) at the Wuhan Institute of Virology, Chinese Academy of Sciences (CAS).
## HTS of the NP library
High-throughput screening of natural antiviral compounds was conducted with a luminescent cell viability assay. Host cells were plated in 96-well black/clear bottom plates 24 hours prior to the addition of the test compounds at a final concentration of 10 μM. After a 1-hour incubation, the viruses were added to the cell culture medium. At 24 hours postinfection, 100 μL of Cell Counting-Lite 2.0 Reagent (Vazyme, DD1101) was added to each well of the 96-well plates at room temperature. Following a 15-minute incubation, the luciferase activities of each sample were measured with a microplate reader. The inhibition rate was calculated on the basis of the luciferase activity.
## qRT-PCR
RNA extraction was conducted with an automatic nucleic acid extraction instrument (Vazyme, VNP-32P) with a virus DNA/RNA extraction kit 3.0 (Vazyme, RM501-01). The purified RNA samples were subsequently reverse transcribed to cDNA with the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, RR047B), after which quantitative real-time polymerase chain reactions (qRT-PCRs) were performed with TB Green Premix Ex Taq II (Takara, RR820A). The primer pairs used in the qRT-PCR tests were as follows: HCoV-229E NP, F: 5'-CAGTCAAATGGGCTGATGCA -3' and R: 5'-AAAGGGCTATAAAGAGAATAAGGTATTCT -3'; HCoV-OC43 NP, F: 5'-CGATGAGGCTATTCCGACTAGGT -3' and R: 5'-CCTTCCTGAGCCTTCAATATAGTAACC -3'; Delta variant RBD, F: 5'-CAATGGTTTAACAGGCACAGG -3' and R: 5'-CTCAAGTGTCTGTGGATCACG -3'; Omicron variant RBD. F: 5'-CAATGGTTTAAAAGGCACAGG -3' and R: 5'-CTCAAGTGTCTGTGGATCACG -3'; mouse GAPDH, F: 5'-TGGTGAAGGTCGGTGTGAAC -3' and R: 5'-GAAGGGGTC-GTTGATGGCAA -3'; mouse CCL2, F: 5'-GTGGGGCGTTAACTGCATCT -3' and R: 5'-GGTCTGAGTGGGACTCAAGG -3'; mouse CXCL10, F: 5'-GGTCTGAGTGGGACTCAAGG -3' and R: 5'-GTGGCAATGATCTCAACACG -3'; mouse IL1B, F: 5'-TTGACGGACCCCAAAAGATG -3' and R: 5'-AGAAGGTGCTCATGTCCTCA -3'; mouse IL6, F: 5'-GTTCTCTGG-GAAATCGTGGA -3' and R: 5'-TGTACTCCAGGTAGCTATGG -3'; mouse TNF-α, F: 5'-ATCGGTCCCCAAAGGGATGA -3' and R: 5'-GCTCCTCCACTTGGTGGTTT -3'; human GAPDH, F: 5'-GAAGATGGTGATGGGATTTC -3' and R: 5'-GAAGGTGAAGGTCGGAGTC -3'; human ALG11, F: 5'-CATCCATACTGCAATGCTGGTGG -3' and R: 5'-GACCGTT GACATTAACATCGCCG -3'; human CASP6, F: 5'-AGGTGGATGCAGCCTCCGTTTA -3' and R: 5'-ATGAGCCGTTCA CAGTTTCCCG -3'; human ALG11, F: 5'-CATCCATACTGCAATGCTGGTGG -3' and R: 5'-GACCGTTGACATTAACATC GCCG -3'; human CYP1A1, F: 5'-GATTGAGCACTGTCAGGAGAAGC -3' and R: 5'-ATGAGGCTCCAGGAGATAGCAG -3'; human ELAC2, F: 5'-CCAGCATCTGTGCTTGTGGACA -3' and R: 5'-CTGCGAAGGTTGTGAACTGAGG -3'; human IDH3G, F: 5'-CCAGTGGACTTTGAAGAGGTGC -3' and R: 5'-CCAGTGGACTTTGAAGAGGTGC -3'; and human UBP1, F: 5'-TCGCTTTGCCAGAGAATCACCG -3' and R: 5'-GCCGTCCTATGATACCACAATCC -3'.
## Determination of antiviral activity in vitro
Huh-7, RD, MRC-5, HEK293T-AT, and Vero E6 cells were initially seeded onto 48-well plates at a concentration of 60000 cells/well and allowed to grow overnight. One hour prior to virus inoculation, the culture medium was replaced with fresh culture medium containing compounds that had been serially diluted. The viruses were then introduced into the culture medium at the optimal multiplicity of infection (MOI). At 24 h after infection, the inoculum was collected to determine the viral RNA copy number via qRT-PCR. The efficacy of the compounds in inhibiting viral replication was assessed on the basis of the viral copy number, and the 50% effective concentration (EC50) was calculated with GraphPad Prism software 8.0. Experiments involving authentic SARS-CoV-2 viruses were conducted at the Biosafety Level 3 facility (BSL-3) at the Wuhan Institute of Virology, Chinese Academy of Sciences (CAS), whereas experiments involving HCoV-229E and HCoV-OC43 were carried out at the Biosafety Level 2 facility (BSL-2) at the same institute.
## Cytotoxicity assay
The cells were seeded into 96-well plates at a density of 40000 cells/well and subsequently cultured overnight at 37 °C in a 5% CO₂ incubator. The compounds were serially diluted before their introduction into the cell culture medium and were incubated with the cells for 24 hours. After the addition of 10 μL of Cell Counting Kit-8 (CCK-8, GlpBio), the supernatant from each well was replaced with 100 μL of fresh culture medium. After the plate was incubated for 2 hours, the absorbance at 450 nm was measured with a microplate reader. The 50% cytotoxic concentration (CC50) was calculated with GraphPad Prism software 8.0 on the basis of the absorbance measurements.
In vivo efficacy of DBM against HCoV-OC43 in suckling mice BALB/c mice were maintained and bred in a specific-pathogen-free (SPF) environment at the Laboratory Animal Center of Wuhan Institute of Virology, CAS. Pregnant mice with the same expected delivery date were acclimated in individually ventilated cages under standard conditions within an SPF environment. Food and water were available ad libitum. The mice were anaesthetized by isoflurane inhalation before any operation that may have caused potential discomfort. Before the experiment, suckling mice were randomly allocated into 6 distinct groups, namely, the healthy group, the vehicle group, the groups receiving DBM at 500 mg/kg (mpk) orally quaque die (QD), 200 mpk QD, and 100 mpk QD, and the group receiving VV116 at 25 mpk orally QD. Each group was comprised of 5-6 suckling mice. On the initiation day (Day 0) of the experiment, 5-6-day-old BALB/c suckling mice were intranasally infected with 1 × 10 4 TCID50 of HCoV-OC43. Two hours postinoculation, the mice were orally treated with either a placebo or varying doses of the drugs on the basis of their respective groupings as outlined earlier. This oral treatment regimen was continued for several days. The body weight changes of the mice were monitored daily. On Day 5, all the mice in each group were sacrificed for virological and histopathological analyses. Various organs and tissues, such as the brain, spinal cord, lungs, and kidneys, were collected from the mice on ice. Portions of the organs and tissues were homogenized with DMEM and subsequently centrifuged at 3000 rpm for 10 min at 4 °C. Viral and host RNA from the organs and tissues were extracted and analysed as described above. For histologic examination, mouse organs and tissues were promptly collected after euthanasia and preserved in 4% paraformaldehyde for more than 5 days. The tissues were subsequently embedded in 3.5-mm paraffin. Fixed tissue samples were subjected to haematoxylin-eosin (H&E) and immunofluorescence staining to detect the HCoV-OC43 antigen with anti-HCoV-OC43 nucleocapsid protein rabbit serum (ABclonal). The images were captured with a Pannoramic MIDI system (3DHISTECH, Budapest) and FV1200 confocal microscope (Olympus). For the survival experiment, the mice were treated for 7 consecutive days after infection and euthanized on Day 14. Body weight changes were recorded daily. Owing to animal discomfort, some of the mice were euthanized before Day 14. Experiments involving authentic viruses were performed in an animal biosafety level 2 facility (ABSL-2) at the Wuhan Institute of Virology, CAS.
## Immunoplaque assay
RD cells were preseeded onto 24-well plates and incubated at 37 °C for 24 hours until they reached 90% confluence. The supernatants from the homogenized organs and tissues were serially diluted 10-fold in DMEM. Subsequently, 50 μL of each dilution was added to the wells. Following a 1-hour incubation at 37 °C, the inoculum was removed, and 200 μL of overlay medium (DMEM containing 2% FBS and 1.5% carboxymethyl cellulose) was added to the wells. Four days postinfection, the cells were fixed with 4% paraformaldehyde for 2 hours, followed by three washes with phosphate-buffered saline (PBS). The fixed cells were then permeabilized with 0.5% Triton X-100 for 30 minutes and blocked with 5% bovine serum albumin (BSA) for 30 minutes at room temperature. The cells were subsequently incubated with anti-HCoV-OC43 nucleocapsid protein rabbit serum (ABclonal, SA00001-2) as the primary antibody and HRP-conjugated anti-rabbit IgG (Proteintech) as the secondary antibody, followed by staining with a DAB staining kit (TIANGEN, PA110).
## Pharmacokinetic studies of DBM
The PK studies were conducted at Wuhan Hongren Biopharmaceutical, Inc. Eight SD rats were randomly divided into three groups and fasted for 12 h prior to dosing. The first group, consisting of three rats, received 10 mg/kg DBM via intravenous (IV) injection, whereas the second group, also consisting of three rats, received 200 mg/kg DBM orally (PO). Two rats served as the blank control. Blood samples were collected into EDTA-K2 tubes and centrifuged at 11,000 rpm for 10 min. The plasma was separated and stored at -70 °C for subsequent analysis. All procedures were conducted in an ice water bath. The concentration of analyte in the plasma was determined using liquid chromatography with tandem mass spectrometry (LC-MS/MS).
## Quantitative proteomics
The quantitative proteomics studies were conducted at Wuhan SpecAlly Life Technology Co. Ltd. Huh-7 cells were seeded into 6-well plates and incubated overnight at 37 °C. The culture medium was replaced with fresh culture medium containing either 20 μM DBM or DMSO as a control. After incubation for 48 hours, the culture medium was discarded, and the cells were washed three times with PBS. The cells were gently harvested into 1.5 ml microcentrifuge tubes and incubated at 60 °C for 1 hour with lysis buffer (1% SDC/100 mM Tris-HCl, pH = 8.5/10 mM TCEP/40 mM CAA) for protein reduction and alkylation. An equal volume of ddH 2 O was added to dilute the lysate, followed by overnight protein digestion with trypsin. The next day, TFA was added to terminate the digestion. After centrifugation, the supernatant was subjected to peptide purification with a custom-made SDB-RPS desalting column. The peptide eluate was vacuum-dried and stored at -20 °C for later use. All the samples were analysed via LC-MS/MS with an UltiMate 3000 RSLCnano (Thermo) coupled with a Q Exactive HF (Thermo).
## Western blot analysis
The cells were washed three times with PBS before being lysed on ice in RIPA lysis buffer (Beyotime, P0013C) containing PMSF (Acmec, AP0100). The lysates were mixed with SDS-PAGE sample loading buffer (Beyotime, P0015) and boiled for 10 minutes. The denatured proteins were separated by SDS-PAGE and then transferred onto a polyvinylidene difluoride (PVDF) membrane (Bio-Rad, 1620177). Next, the membrane was incubated with primary antibodies, followed by horseradish peroxidase (HRP)-conjugated secondary antibodies. The protein bands were visualized with Immobilon ECL UltraPlus Western HRP Substrate (Millipore, WBULP) and scanned with a ChemiDoc Imaging System (Bio-Rad).
## Time-of-addition assay
To evaluate the effects of DBM on the life cycle of HCoV-229E and HCoV-OC43, fresh culture medium containing 0.6 μM DBM, 0.1 μM remdesivir (RDV), or 10 μM chloroquine (CQ) was supplemented with preseeded cells at different time points relative to infection (entry group: 0 to -2 hours; postinfection group: 2-12 hours; and full-time group: 0 to -12 hours). DMSO served as the control. At 24 hours postinfection, the number of intracellular virus RNA copies was quantified by qRT-PCR as previously described.
## Viral binding assay
The cells were seeded onto 24-well plates and incubated overnight at 37 °C. Serially diluted DBM was added to the wells and incubated with the cells for 1 hour before infection at an MOI of 50. Following inoculation, the plates were placed on ice for two hours. The cells were washed three times with PBS before total RNA was extracted and analysed by qRT-PCR as previously described.
## Viral entry and internalization assays
The cells were seeded at a low density onto 35 mm glass bottom dishes with 10 mm microwells (Cellvis) and incubated overnight at 37 °C. One hour before infection, the culture medium was supplemented with DBM, and chloroquine (CQ) was added to prevent fusion. The cells were then inoculated with viruses at an MOI of 50, placed on ice for 1 hour, and subsequently incubated at 37 °C. The cells were washed three times with PBS buffer and fixed with 4% paraformaldehyde for 2 hours, followed by another three washes. The fixed cells were then permeabilized with 0.5% Triton X-100 for 30 minutes and blocked with 5% bovine serum albumin (BSA) for 30 minutes at room temperature. The cells were then incubated with an anti-HCoV-OC43 nucleocapsid antibody (Abcam, ab309964) as the primary antibody and 488-conjugated anti-human IgG (Proteintech, CL488-10284) as the secondary antibody. Images were obtained with the Dragonfly High-Speed Confocal Microscope Systems (Andor). The number of intracellular virus particles was quantified with ImageJ software.
## Cell-cell fusion assays
HEK293T cells were preseeded and cotransfected with plasmids encoding the HCoV-OC43 spike protein (including the ectodomain and transmembrane domain, ECD & TMD) and eGFP. At 24 hours posttransfection, the HEK293T cells were suspended, mixed with RD cells, and seeded onto 24-well plates. The mixed cells were cultured at 37 °C for 12 hours and washed three times with PBS before being treated with FBS-free DMEM containing TPCK-treated trypsin (Sigma-Aldrich, 100 ng/mL) for 1 h. Subsequently, the cells were incubated with fresh culture medium for 16 hours before imaging. v2 plasmids were generated using the same approach. For lentiviral packaging, HEK293T cells were seeded in six-well plates and transfected with either CASP6-targeting LentiCRISPR v2 plasmids or scrambled control LentiCRISPR v2 plasmids, along with two other packaging plasmids pMD2.G and pCMV-dR8.91. Six hours post-transfection, the culture medium was replaced with maintenance medium. Following a further 42 hours of incubation, the supernatant containing the recombinant lentivirus was harvested and filtered through a sterile 0.45 μm hydrophilic PVDF membrane syringe filter (Millipore, SLHV033R). RD cells were infected with lentivirus-containing supernatant. 96 hours post-infection, cells were subjected to selection with 2 μg/ ml puromycin for 7 days. Successful knockout of caspase-6 was confirmed by Western blot analysis.
## References
1. "Transient transfections and RNA interference Plasmid transfection was performed with Lipofectamine 3000 (Invitrogen) according to the manufacturer's instructions"
2. "For RNA interference, cells were transfected with siRNAs with Lipofectamine RNAiMAX following the manufacturer's guidelines. Scramble siRNA was used as the control. The siRNAs used in this research were as follows: IDH3G-1, F: 5'-GCAAGAGUAUCGCCAAUAATT -3' and R: 5'-UUAUUGGCGAUACUCUUGCTT -3'; IDH3G-2, F: 5'-GCACGUGAGUUC-CAAUGCUTT -3' and R: 5'-AGCAUUGGAACUCACGUGCTT -3'; CYP1A1-1, F: 5'-GAUCCCAUCUGCCCUAUAUTT -3' and R: 5'-AUAUAGGGCAGAUGGGAUCTT -3'; CYP1A1-2, F: 5'-GGGUUUGACACAGUCACAATT -3' and R: 5'-UUGUGACUGUGUCAAACCCTT -3'; ELAC2-1, F: 5'-GCUUCCAAAGUGUGUACUUTT -3' and R: 5'-AAGUACACAC-UUUGGAAGCTT -3'; ELAC2-2, F: 5'-GCAGAGGGAUGCCAUUAUUTT -3' and R: 5'-AAUAAUGGCAUCCCUCUGCTT -3'; UBP1-1, F: 5'-GCACCCACCCUUUCAGUAUTT -3' and R: 5'-AUACUGAAAGGGUGGGUGCTT -3'; UBP1-2, F: 5'-GCAUCAUAAGGGUUGUAUUTT -3' and R: 5'-AAUACAACCCUUAUGAUGCTT -3'; ALG11-1, F: 5'-GCUCUCU-GUAGUGAAGAAUTT -3' and R: 5'-AUUCUUCACUACAGAGAGCTT -3'; ALG11-2, F: 5'-GGUGCCUGUGCAAGUU-GUUTT -3' and R: 5'-AACAACUUGCACAGGCACCTT -3'; CASP3-1, F: 5'-CCGACAAGCUUGAAUUUAUTT -3' and R: 5'-AUAAAUUCAAGCUUGUCGGTT -3'" *CASP*
3. "F: 5'-CUCUGUUGCAGAAGGAUAUTT -3' and R: 5'-AUAUCCUUCUGCAACAGAGTT -3'; CASP6-2, F: 5'-CUGCGCAGAUAGAGACAAUTT -3' and R: 5'-AUUGUCUCUAUCUGCGCAGTT -3'; and CASP6-3, F: 5'-GGCAAUCA-CAUUUAUGCAUTT -3' and R: 5" *CASP*
4. "Caspase-6 activity assay Caspase-6 enzyme activity was measured with the Caspase-6 Activity Assay Kit (Beyotime, C1135) following the manufacturer's protocol. The cells pretreated with DMSO, DBM, or Z-VEID-FMK were digested with trypsin and washed three times with PBS before being collected by centrifugation. The cell samples were lysed on ice for 15 minutes using the lysis solution provided in the kit"
5. Valian, Pourakbari, Ashari et al. (2022) "CASP6-targeting oligonucleotide pairs (forward sequence: 5' -CACCGAAAAGTACAAAATGGACCAC -3' and reverse sequence: 5' -AAACGTG GTCCATTTTGTACTTTTC -3') were synthesised and cloned into LentiCRISPR v2 plasmids. Scrambled control LentiCRISPR References 1"
6. Niu, Shen, Huang et al. (2020) "Non-invasive bioluminescence imaging of HCoV-OC43 infection and therapy in the central nervous system of live mice" *Antiviral Res*
7. Da Veiga, Martins, Riediger et al. (2021) "More than just a common cold: Endemic coronaviruses OC43, HKU1, NL63, and 229E associated with severe acute respiratory infection and fatality cases among healthy adults" *J Med Virol*
8. Menachery, Yount, Jr et al. (2015) "A SARS-like cluster of circulating bat coronaviruses shows potential for human emergence" *Nat Med*
9. Chen, Yang, Si et al. (2023) "A bat MERS-like coronavirus circulates in pangolins and utilizes human DPP4 and host proteases for cell entry" *Cell*
10. Kumari, Lu, Li et al. (2022) "A critical overview of current progress for COVID-19: development of vaccines, antiviral drugs, and therapeutic antibodies" *Journal of Biomedical Science*
11. Qu, Evans, Faraone et al. (2023) "Enhanced neutralization resistance of SARS-CoV-2 Omicron subvariants BQ.1, BQ.1.1, BA.4.6, BF.7, and BA.2.75.2" *Cell Host Microbe*
12. Iketani, Mohri, Culbertson et al. (2023) "Multiple pathways for SARS-CoV-2 resistance to nirmatrelvir" *Nature*
13. Von Delft, Hall, Kwong et al. (2023) "Accelerating antiviral drug discovery: lessons from COVID-19" *Nat Rev Drug Discov*
14. Lung, Lin, Yang et al. (2020) "The potential chemical structure of anti-SARS-CoV-2 RNA-dependent RNA polymerase" *J Med Virol*
15. Su, Yao, Zhao et al. (2020) "Anti-SARS-CoV-2 activities in vitro of Shuanghuanglian preparations and bioactive ingredients" *Acta Pharmacol Sin*
16. Kim, Kim, Park et al. (2018) "high-throughput screening of natural compounds of MERS-CoV entry inhibitors using a pseudovirus expressing MERS-CoV spike protein" *Int J Antimicrob Agents*
17. Leng, Xu, Hong et al. (2024) "Cepharanthine analogs mining and genomes of Stephania accelerate anti-coronavirus drug discovery" *Nat Commun*
18. Lim, Lee, Choi et al. (2014) "Inhibition of airway inflammation by the roots of Angelica decursiva and its constituent, columbianadin" *J Ethnopharmacol*
19. Shen, Khor, Hu et al. (2007) "Chemoprevention of familial adenomatous polyposis by natural dietary compounds sulforaphane and dibenzoylmethane alone and in combination in ApcMin/+ mouse" *Cancer Res*
20. Gambacorti-Passerini, Barni, Le Coutre et al. (2000) "Role of alpha1 acid glycoprotein in the in vivo resistance of human BCR-ABL(+) leukemic cells to the abl inhibitor STI571" *J Natl Cancer Inst*
21. Yang, Cheng, Zhuang et al. (2024) "Pharmacogenomic profiling of intra-tumor heterogeneity using a large organoid biobank of liver cancer" *Cancer Cell*
22. Zhang, Sun, Xie et al. (2023) "A viral RNA-dependent RNA polymerase inhibitor VV116 broadly inhibits human coronaviruses and has synergistic potency with 3CLpro inhibitor nirmatrelvir" *Signal Transduct Target Ther*
23. Du, Duan, Wang et al. (2019) "Attenuation of Equine Lentivirus Alters Mitochondrial Protein Expression Profile from Inflammation to Apoptosis" *J Virol*
24. Mosher, Fulkerson, Boyle et al. (2022) "Human Cytomegalovirus Manipulates Syntaxin 6 for Successful Trafficking and Subsequent Infection of Monocytes" *J Virol*
25. Zheng, Ren, Wang (2023) "Serine-arginine protein kinases and their targets in viral infection and their inhibition" *Cell Mol Life Sci*
26. Chu, Hou, Yang et al. (2022) "Coronaviruses exploit a host cysteine-aspartic protease for replication" *Nature*
27. Yuan, Ma, Xie et al. (2023) "The role of cell death in SARS-CoV-2 infection" *Signal Transduct Target Ther*
28. Yang, Xiang, Sun et al. (2022) "Host MOV10 is induced to restrict herpes simplex virus 1 lytic infection by promoting type I interferon response" *PLoS Pathog*
29. Oh, Santiago, Manjunath et al. (2024) "A CRISPR-Cas9 knockout screening identifies IRF2 as a key driver of OAS3/RNase L-mediated RNA decay during viral infection" *Proc Natl Acad Sci U S A*
30. Szabó, Crowe, Mamotte et al. (2024) "Natural products as a source of Coronavirus entry inhibitors" *Front Cell Infect Microbiol*
31. Larragoite, Nell, Martins et al. (2022) "Histone deacetylase inhibition reduces deleterious cytokine release induced by ingenol stimulation" *Biochem Pharmacol*
32. Cong, Guo, Tang et al. (2015) "Metabolism Study of Veratramine Associated with Neurotoxicity by Using HPLC-MSn" *J Chromatogr Sci*
33. Nascimento, De, Baeta et al. (2022) "Dibenzoylmethane derivative inhibits melanoma cancer in vitro and in vivo through induction of intrinsic and extrinsic apoptotic pathways" *Chem Biol Interact*
34. Lin, Lu, Lou et al. (2001) "Inhibition by dietary dibenzoylmethane of mammary gland proliferation, formation of DMBA-DNA adducts in mammary glands, and mammary tumorigenesis in Sencar mice" *Cancer Lett*
35. Zheng, Karki, Vogel et al. (2020) "Caspase-6 Is a Key Regulator of Innate Immunity, Inflammasome Activation, and Host Defense" *Cell*
36. Zhao, Sun, Chaggan et al. (2020) "An AMPK-caspase-6 axis controls liver damage in nonalcoholic steatohepatitis" *Science*
37. Flores, Noël, Fillion et al. (2022) "Therapeutic potential of Nlrp1 inflammasome, Caspase-1, or Caspase-6 against Alzheimer disease cognitive impairment" *Cell Death Differ*
38. Ehrnhoefer, Skotte, Reinshagen et al. (2019) "Activation of Caspase-6 Is Promoted by a Mutant Huntingtin Fragment and Blocked by an Allosteric Inhibitor Compound" *Cell Chem Biol*
39. Deng, Yu, Xiao et al. (2024) "Knocking down macrophages Caspase-6 through HMGB1 coordinates macrophage trophoblast crosstalk to suppress ferroptosis and alleviate preeclampsia" *Int Immunopharmacol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12476816&blobtype=pdf | # Real-life practice of Kelleni's protocol in treatment and post exposure prophylaxis of SARS-CoV-2 HV.1 and JN.1 subvariants
Mina Thabet Kelleni, Mina Kelleni
## Abstract
This article discusses the evolving real-world practice using nitazoxanide, nonsteroidal anti-inflammatory drugs (NSAIDs) and/or azithromycin (Kelleni's protocol) to manage the evolving manifestations of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron EG.5.1, its descendant HV.1 as well as BA.2.86 and its descendant JN.1 subvariants in Egypt in 2024. These subvariants are well-known for their highly evolved immune-evasive properties and the manifestations include some peculiar manifestations as persistent cough besides high fever in young children as well as high fever, persistent severe cough, change of voice, loss of taste and smell, epigastric pain, nausea, vomiting, diarrhea, generalized malaise and marked bone aches in adults including the high-risk groups. It's suggested that the ongoing SARS-CoV-2 evolution is continuing to mostly affect the high-risk groups of patients, to some of whom we've also successfully prescribed nitazoxanide and/or NSAIDs for post-exposure prophylaxis of all household contacts. We also continue to recommend starting the immune-modulatory antiviral Kelleni's protocol as soon as possible in the course of infection and adjusting it in a personalized manner to be more aggressive from the beginning for the high risk patients, at least until the currently encountered surge of infections subsides.
## INTRODUCTION
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) evolution is ongoing while threatening to return the world to square one [1,2]. However, Africa has adopted early treatment approach and managed to avoid most of the burden of coronavirus disease 2019 (COVID-19) [3,4]. We would like to report the real-life practice of the Egyptian Kelleni's protocol in both treatment and post-exposure prophylaxis of SARS-CoV-2 infection in 2024.
For almost four years we've been safely and effectively practicing the immune-modulatory antiviral nitazxoanide, nonsteroidal anti-inflammatory drugs (NSAIDs) and/or azithromycin (the standard Kelleni's protocol) to early manage several respiratory viral infections including all SARS-CoV-2 variants that have evolved and affected patients of all ages including pregnant ones and those suffering from various medical co-morbidities [5]. Notably, Egypt has officially acknowledged EG.5 subvariant only in August 2023 and in December 2023, Egypt has similarly acknowledged BA.2.86 and its descendant JN.1, which are currently the dominating global ones. Importantly, SARS-CoV-2 variants of interests and concern are being closely monitored, due to their highly evolved immune-evasive properties which can lead to a potential reclassification of their global COVID-19 morbidity and mortality impact.
## NEW VARIANTS OF CONCERN ARE DETECTED CLINICALLY BEFORE BEING NAMED
We'd like to report that since October 2023 until April 2024, we've been experiencing a surge of unusually encountered more virulent respiratory infections which significantly affect the high-risk groups of patients; especially young children and immune-compromised adults. Interestingly, the abrupt onset of persistent high fever reaching occasionally exceeding 40 °C was a hallmark of those high risk patients, during the first two to three days of manifestations. Notably, for several times during COVID-19 pandemic, new SARS-CoV-2 variants of concern were highly suspected clinically before being officially named. The most recently encountered manifestations in otherwise healthy adults, since March 2024, are high fever persistent marked epigastric pain, nausea, vomiting and/or diarrhea which were preceded or followed by the respiratory manifestations of rhinorrhea, persistent cough, loss of taste and smell, generalized malaise and severe bone ache. Notably, these predominant gastrointestinal manifestations were previously suggested to imply an evolutionary change of SARS-CoV-2 tropism [5].
## ADDING AMOXICILLIN/CLAVULANIC TO THE STANDARD KELLENI'S PROTOCOL IN SELECTED CASES
The high fever was efficiently relieved by regular use of NSAIDs two to three times a day, along with cold packs and it's noteworthy that we've not encountered such intensity of fever in that frequency before and we've added amoxicillin/ clavulanic acid to the standard Kelleni's protocol if the expected clinical improvement was not found at the end of the third day. Notably, this modification has adequately managed the debilitated high risk patients who might have experienced resistance to macrolides or a more severe secondary bacterial infection which is a known serious complication of COVID-19 [6].
Importantly, though Mycoplasma pneumoniae infections, characterized by persistent dry cough and prolonged manifestations, were reported to surge especially in children in different countries after the lifting of COVID-19 restrictions, yet Egypt has had almost zero COVID-19 restrictions for over two years and these symptoms were reported in both high and low risk groups of patients of all ages. Moreover, though amoxicillin/calvulanic acid efficacy is lacking as regards to Mycoplasma pneumoniae which has no cell wall, yet enhanced antimicrobial activity of azithromycin against mycoplasma pneumoniae in the presence of amoxicillin/clavulanic acid could be suggested [7] as observed clinically in some treated pediatric patients during this surge.
As per our clinical experience, it's become increasingly not uncommon for those high-risk patients to require an extended course of Kelleni's antiviral protocol including a 5-day-course of double antibiotics; azithromycin and amoxicillin/clavulanic acid in addition to the standard nitazoxanide and NSAIDs [5].
## MANAGING PERSISTENT COUGH AND SOME INTERESTING CASES
Moreover, persistent troublesome cough was quite evident and prevalent in this current surge of SARS-CoV-2 infections. It was managed by both natural antitussives and loratadine [3], but a personalized administration of cloperastine suspension or pholcodine containing syrup was required in selected severe cases. Furthermore, change of voice was occasionally the main presentation in young healthy adults who have been re-infected during this current surge, along with mild to moderate diarrhea, headache, fatigue, malaise and some patients experienced marked retro-orbital pain or persistent severe cough. These symptoms, other than cough, were effectively managed by NSAIDs with or without nitazoxanide.
Importantly, persistent cough and high fever were less frequently encountered when we included azithromycin, together with nitazoxanide and ibuprofen, in the prompt management of early SARS-CoV-2 infection in young children.
Interestingly, a young female child presented with several bouts of hematochezia without respiratory symptoms or abdominal colic. She was effectively managed using nitazoxanide and azithromycin, without NSAIDs to avoid exacerbating bleeding attacks. We advised the parents to use acetaminophen if the child developed subsequent fever, although it was not ultimately required. Noteworthy, hematochezia associated with COVID-19 has been previously described as more common in geriatric COVID-19 male patients suffering from other co-morbidities [8], yet pediatric and female patients have also been reported [9,10].
## EARLY TREATMENT USING KELLENI'S PROTOCOL REMAINS OUR BEST RECOMMENDATION
Notably, immune-compromised patients, and an adult receiving immunotherapy with nivolumab to manage his slowly regressive hepatocellular carcinoma, experienced more frequent persistent troublesome cough, marked bone aches and fatigue. However, when these patients were early managed, the standard Kelleni's protocol was sufficient. Moreover, from our clinical practice, when immune-competent young adults were prescribed nitazoxanide and NSAIDs at the first symptoms of sore throat, sneezing and early rhinorrhea the infection was effectively managed in three days in the vast majority of cases (Figure 1).
Notably, nitazoxanide is currently heavily and effectively prescribed in Egypt especially to manage the prevalent gastrointestinal manifestations of SARS-CoV-2 and other viruses [5]. Notably, in a multicentre, randomized, double-blind, placebo-controlled clinical trial, nitazoxanide was shown to improve the clinical outcome, time to hospital discharge, and to reduce the oxygen requirements and the C-reactive protein, D-dimer, and ferritin levels compared to the placebo group and no serious adverse events were observed [11]. Furthermore, in a Mexican study that included 312 patients suffering from severe COVID-19 and were treated using nitazoxanide as part of the administered regimen, only 6 patients were died [12].
## PRACTICING KELLENI'S PROTOCOL FOR SARS-COV-2 POST-EXPOSURE PROPHYLAXIS
Furthermore, we've repeatedly used NSAIDs and/or nitazoxanide for post-exposure prophylaxis of high-risk house-hold contacts and this approach has clinically proven to be safe and effective. It has sometimes prevented but mostly significantly alleviated the inevitable infection. Unfortunately, a pioneering clinical trial (NCT04435314) to assess the efficacy of nitazoxanide in SARS-CoV-2 post exposure prophylaxis was aborted due to "sponsor's strategic decision". Moreover, we suggest that nitazoxanie is not pharmacologically fit to be tested as a drug used chronically for "prevention" as some authors did[13] and in any performed clinical trial that assess post exposure prophylaxis the intensity of symptoms as well as the duration of illness should be carefully assessed.
Interestingly, during the described surge of infections, we've also prescribed nitazoxanide, without NSAIDs, to a pregnant patient in her late second trimester and most of the symptoms were alleviated within two days except for prolonged severe cough that was only relieved after adding loratadine to the herbal antitussives. Moreover, nitazoxanide was used as an immediate post-exposure prophylaxis to this pregnant patient when re-infected in her third trimester, the clinical outcome was better, and the severe cough was not even encountered. However, the recently acquired previous immunity though couldn't prevent the re-infection; it could probably have played at least a partial role in this favorable response.
## EARLY IMMUNE-MODULATORY TREATMENT SAFEGUARDS AFRICA AGAINST PANIC
Notably, in addition to recommending starting the personalized Kelleni's protocol as early as possible for the management of all current respiratory tract viral infections, especially in high-risk groups of patients, we recommend adding amoxicillin/clavulanic acid for children younger than 5 years from the onset of manifestations, at least until this surge is over.
Fortunately, all cases encountered thus far have fully recovered within one week without any post coronavirus disease complaints. However, it's quite evident from a clinical standpoint that SARS-CoV-2 evolution is ongoing and this could reignite panic particularly in countries that have not yet adopted early immune-modulation in the pharmacological management of COVID-19 [4].
## CONCLUSION
Given the ongoing circulation and evolution of SARS-CoV-2, as well as the emergence of new variants of potential clinical concern due to their expedited transmission and/or greater immunological evasion properties, especially among vulnerable populations, it's crucial to prioritize standardized post-exposure prophylaxis protocols as part of global preparedness. Kelleni's post-exposure prophylaxis protocol presents a cost-effective, readily implementable, and potentially impactful approach to minimize severe COVID-19 and hospitalization, particularly in resource-limited countries. Kelleni's protocol could significantly enhance equitable and preventive public health strategies in both current and future pandemics.
## FOOTNOTES
Author contributions: Kelleni MT is responsible for all aspects of the writing and publication process of the manuscripts.
## Conflict-of-interest statement:
## References
1. Hattab, Amer, Al-Alami et al. (2024) "SARS-CoV-2 journey: from alpha variant to omicron and its sub-variants" *Infection*
2. Hu, Guo, Si et al. (2024) "Emergence of SARS and COVID-19 and preparedness for the next emerging disease X" *Front Med*
3. Kelleni (2023) "The African Kelleni's roadmap using nitazoxanide and broad-spectrum antimicrobials to abort returning to COVID-19 square one" *Inflammopharmacology*
4. Kelleni (2024) "COVID-19 mortality paradox (United States vs Africa): Mass vaccination vs early treatment" *World J Exp Med*
5. Kelleni (2023) "Real-world practice of the Egyptian Kelleni's protocol amid changing tropism of SARS-CoV-2 omicron BA.5.2.1.7, XBB 1.5 and CH.1.1 subvariants: a multi-purpose protocol" *Inflammopharmacology*
6. Shafran, Shafran, Ben-Zvi et al. (2021) "Secondary bacterial infection in COVID-19 patients is a stronger predictor for death compared to influenza patients" *Sci Rep*
7. Onodera, Kiyota, Endo et al. (2006) "Enhancement of antimicrobial activities of cefteram or clavulanic acid/amoxicillin against cefixime-resistant Neisseria gonorrhoeae in the presence of clarithromycin or azithromycin" *J Infect Chemother*
8. Ion, Gherghinescu, Andronic et al. "Prognosis Evaluation for Patients with Abdominal Trauma Using Usual Biological Parameters" *Chirurgia (Bucur)*
9. Frič, Borzan, Borzan et al.
10. Hasbani, Taher, Jawad et al. (2021) "Henoch-Schönlein purpura: Another COVID-19 complication" *Pediatr Dermatol*
11. Rocco, Silva, Cruz et al. (2022) "Nitazoxanide in Patients Hospitalized With COVID-19 Pneumonia: A Multicentre, Randomized, Double-Blind"
12. Romero-Cabello, Romero-Feregrino, Romero-Feregrino et al. (2023) "Outpatient treatment of COVID-19: an experience with 552 cases in Mexico" *J Infect Dev Ctries*
13. Sokhela, Bosch, Hill et al. (2022) "Randomized clinical trial of nitazoxanide or sofosbuvir/daclatasvir for the prevention of SARS-CoV-2 infection" *J Antimicrob Chemother* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12569540&blobtype=pdf | # Genomic Characterization of Human Adenovirus Type 21 Strains -7 PLADs, China, 2023-2024
Yali Jin, Naiying Mao, Xueping Ma, Linqing Zhao, Liwei Sun, Jikui Deng, Shu Liang, Hongmei Xu, Xin Li, Chunyu Zhu, ; Xia, Aili Cui, Yan Zhang, Zhen Zhu
## Abstract
Introduction: Recent sentinel surveillance has revealed a rising prevalence of human adenovirus type 21 (HAdV-21) among HAdV infections in China. This study aimed to elucidate the molecular features of currently circulating HAdV-21 strains in China.Methods: Whole-genome sequencing (WGS) was performed on 23 HAdV-21 strains isolated from acute respiratory infection cases, 56.5% involving lower respiratory tract infections, across 7 Chinese sentinel surveillance provincial-level administrative divisions (PLADs) (2023)(2024). These sequences, along with 50 previously reported HAdV-21 genomes from 6 countries , were integrated into a WGS dataset for comprehensive phylogenetic, genetic variation, and recombination analyses.Results: WGS categorized the HAdV-21 strains into 3 subtypes: HAdV-21a, HAdV-21b, and historical HAdV-21p (isolated in the 1950s). HAdV-21a (1956HAdV-21a ( -2024, involving 5 of the 6 countries) and HAdV-21b (2005HAdV-21b ( -2024, involving 3 of the 6 countries) exhibited extensive spatiotemporal distributions. Recent Chinese strains (2023-2024) belonged to HAdV-21a and HAdV-21b (HAdV-21a/b), showing extremely high genetic homology with Chinese 2019 strains (genetic distance: 0.00007) and global strains (distance: <0.00040). Phylogenetic analysis confirmed that HAdV-21a/b shared a common ancestor and maintained a highly conserved genome despite decades of circulation. Sequence variation analysis identified shared and subtype-specific mutations in these two subtypes. Recombination pattern analysis further revealed that HAdV-21a/b acquired an HAdV-3-derived fragment in the E4 region (breakpoint: nt32,843).Conclusions: Recombinant HAdV-21a/b subtypes have co-circulated in China in recent years with remarkable genetic conservation. Enhanced surveillance is essential to quantify associated disease burden and guide targeted prevention and control strategies.Human adenoviruses (HAdVs) are non-enveloped, double-stranded DNA viruses (genome size 34-36 kb) in the genus Mastadenovirus (family Adenoviridae), classified into 7 species (A-G) with 116 recognized types (http://hadvwg.gmu.edu/) (1). Among these, HAdV-3, HAdV-7, HAdV-55, HAdV-4, and species C (HAdV-C) are the major epidemiologically significant causes of respiratory infections (2-6).HAdV-21 (species B), first identified in Saudi Arabia in 1956, is increasingly recognized as a significant pathogen in military and civilian populations across multiple countries (7). Although a meta-analysis of Chinese surveillance data (2009-2021) revealed a low prevalence of HAdV-21 (0.87%), recent data from 14 sentinel surveillance sites across 11 provincial-level administrative divisions (PLADs) (2023-2024) indicated a marked increase: the proportion of HAdV-21 infections rose from 1.44% (2/139) in 2023 to 8.97% (21/234) in 2024 (Fisher's exact test: P=0.003) (unpublished data) (6). However, molecular epidemiological data on HAdV-21 in China remain limited, with only 3 isolates reported before 2023 (8-9).To investigate the potential genetic correlations among the increasingly detected HAdV-21, this study performed whole-genome sequencing (WGS) of 23 HAdV-21 strains isolated from acute respiratory infections across 7 surveillance PLADs (2023-2024) and conducted phylogenetic, genetic variation, and recombination analyses by integrating these data with previously reported HAdV-21 genomes obtained from GenBank.
isolated from 7 Chinese sentinel surveillance PLADs (Gansu, Jilin, Ningxia, Liaoning, Guangdong, Chongqing, and Beijing) and confirmed by real-time polymerase chain reaction and amplification/analysis of 3 genes (penton base, hexon, and fiber). The corresponding 23 clinical cases ranged from 5 months to 68 years (median: 3 years). Among these cases, 14 (60.87%) required hospitalization. Ten cases presented with upper respiratory tract infections, while the remaining 13 had lower respiratory tract infections, including 10 with pneumonia. Co-infections occurred in 13 cases, most commonly with Haemophilus influenzae (5/13). One fatal case (Ningxia2024-669) of acute respiratory distress syndrome exhibited Mycoplasma pneumoniae co-infection (Table 1).
Viral DNA from the 23 strains was extracted using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). WGS was performed by iGeneTech Biotechnology Co., Ltd. using probe-based hybrid capture combined with next-generation sequencing. All strains achieved 100% genome coverage, each yielding >1 Gb of data at >8,000× depth. Suboptimal regions were validated by Sanger sequencing, and two strains were randomly resequenced for quality assurance. Genome annotation was performed in Geneious Prime (version 2023.2.2, Biomatters Ltd., Auckland, New database of 73 WGSs for phylogenetic and genetic variation analyses.
Sequences were aligned using ClustalW, and sequence similarity was assessed using BioEdit. A maximum likelihood phylogenetic tree was constructed using MEGA Version 7.0, with bootstrap support values (>80%) indicated at the tree nodes. Genetic mutations were identified using Snipit (https://github. com/aineniamh/snipit), focusing on variations with frequencies >80%. Recombination was analyzed using SimPlot (window size: 1,000 bp; step size: 100 bp) and Recombination Detection Program v4 (RDP4).
## RESULTS
## Genomic Characterization
All 23 HAdV-21 strains were fully assembled (genome sizes: 35,364-35,393 bp; GC: approximately 51.2%), consistent with previous reports (8)(9). Annotation using the HAdV-21 prototype strain as a reference identified 48 conserved protein-coding regions (Supplementary Table S2, available at https:// weekly.chinacdc.cn/). Sequence identity analysis among the 23 HAdV-21 strains was ≥99.7%. Strains with identical sequences were detected across different PLADs and years, e.g., the 2023 Jilin strains (Jilin2023-296 and Jilin2023-595) matched 2024 strains from 5 PLADs (Beijing2024-364, Gansu2024-215, Ningxia2024-907, Shenyang2024-163, and Shenzhen2024-077). Identical sequences were also detected within the same PLAD, such as Ningxia2024-595 and Ningxia2024-725.
The strains shared 98.8% nucleotide identity with the HAdV-21 prototype strain. Analysis of 12 functional domains, including major capsid proteins (penton base, hexon, fiber), core proteins (pV, pVII, pTP, pIVa2), minor proteins (pIX, pIIIa, pVI, pVIII), and the non-structural protein pX, showed relatively high similarity (nt: 98.4%-99.8%, aa: 97.9%-100.0%) across 11 regions, with the penton base showing lower conservation (nt: 96.3%, aa: 95.7%-95.9%).
## Phylogenetic Analysis
To determine the genetic relationships between the 23 HAdV-21 strains (2023-2024) and previously circulating Chinese and globally prevalent strains, a phylogenetic analysis was performed using the constructed WGS dataset (Figure 1). The results showed that the 73 HAdV-21 strains were categorized into 3 distinct evolutionary lineages: HAdV-21a, HAdV-21b, and HAdV-21p (bootstrap support >80%). HAdV-21a and HAdV-21b (HAdV-21a/b) exhibited closer phylogenetic relationships to each other (genetic distance=0.001) than to HAdV-21p (genetic distance=0.007 for both subtypes), indicating a shared ancestry for HAdV-21a/b.
Further analyses revealed distinct temporal and geographical distribution patterns among the 3 HAdV-21 subtypes. HAdV-21p primarily comprised a prototype strain from Saudi Arabia and a historical strain from Germany (1950s). Conversely, HAdV-21a comprised 21 strains from 7 Chinese PLADs (2023-2024), 3 Chinese strains from 2 PLADs (2019), and 24 strains from the United States, Malaysia, Germany, andSwitzerland (1956-2016). HAdV-21b included 2 strains from 2 Chinese PLADs (2024) and 21 strains from the United States andGermany (2005-2018). All 3 subtypes demonstrated minimal intra-subtype genetic variation (distance: 0.00014-0.00031). HAdV-21a/b demonstrated high genetic consistency between 2023-2024 Chinese strains and both the 2019 Chinese strains (distance: 0.00007) and global strains (distance <0.00040), strongly suggesting close genetic relationships. No significant correlation was observed between clinical severity and 2 subtypes (HAdV-21a and HAdV-21b).
## Genetic Variation
Genetic variation analysis of HAdV-21a/b was performed using WGS with the HAdV-21p prototype strain as a reference. The results showed that the 2 subtypes shared 352 specific nucleotide variants, including 64 insertions and 49 deletions, with HAdV-21a exhibiting 27 unique variants (10 deletions) and HAdV-21b displaying 36 unique variants (3 insertions and 18 deletions). The variants were distributed across the genome, with insertions and deletions concentrated in the L2 (46.90%) and E3 (23.01%) regions, respectively. Non-coding region variants were also detected in the Chinese HAdV-21a (4 sites, 1 insertion) and HAdV-21b (7 sites).
Amino acid variation analysis across the 12 functional domains identified 70 shared HAdV-21a/b substitutions in 11 proteins, excluding pX (Figure 2). 9 of 10 shared hexon gene variants were localized to hypervariable regions (HVRs), with HVR7 exhibiting the highest mutation frequency (4 sites). In the fiber gene, 2 of 3 variants occurred in the shaft region and one in the knob region. The penton base gene exhibited 23 variations, primarily clustered within the 23456
HAdV-21a
HAdV-21b
HAdV-21p RGD loop, including 15 amino acid deletions (313TEAAKAAAIAKANIV327) and 2 insertions (362EE363). Subtype-specific mutations were also detected, including pIVa2 (HAdV-21b: L196F), pTP (HAdV-21a: D45N), and penton base (HAdV-21b: E213D and S349F).
$$A dV -2 1/ K J3 64 59 2 H A d V -2 1 /K J3 6 4 5 7 8 H A d V -2 1 /K J 3 6 4 5 8 0 H A d V -2 1 /K J 3 6 4 5 7 4 H A d V -2 1 /K J 3 6 4 5 8 3 H A d V -2 1 / K J 3 6 4 5 8 5 H A d V -2 1 / K F 8 0 2 4 2 6 H A d V -2 1 /K F 5 7 7 5 9 3 H A d V -2 1 /K F 9 3 8 5 7 6 H A d V -2 1 /O R 7 5 3 0 9 7 H A dV -2 1/ O R$$
$$A dV -2 1/ M N 68 62 06 H A d V -2 1 /M W 1 5 1 2 4 3 H A d V -2 1 /O Q 5 1 8 3 1 4 N in g x ia$$
$$T A A D E T E A A K A A A I A K A N I V V A S A K S I I G G F E E E E T T R R G G Position$$
## Genetic Recombination
To assess recombination, phylogenetic analysis was performed with the HAdV-21 strains and 18 species B prototype strains (Supplementary Table S1) using 9 consecutive genomic fragments (nt1-7,000; nt7,001 -13,877; penton base; nt15,564-18,453; hexon; nt21,304-26,000; nt26,001-31,405; fiber; nt32,378-end) and WGS (Supplementary Figure S1, available at https://weekly.chinacdc.cn/). The results demonstrated that subtype classification using the nt32,378-end fragment matched the WGS-based classification, whereas the other 8 fragments failed to distinguish HAdV-21a/b. In the nt32,378-end region, HAdV-21a/b exhibited closer phylogenetic relationships with the HAdV-3 prototype strain, suggesting potential recombination in the E4 region. SimPlot and RDP4 analyses (supported by four algorithms) further confirmed identical recombination patterns in both subtypes, with a breakpoint at nt32,843 in the E4 gene (Figure 3).
## DISCUSSION
This study analyzed 23 HAdV-21 strains (2023-2024) isolated from patients with acute respiratory infections. Among these, 56.5% (13/23) had lower respiratory tract infections, including 9 pneumonia cases and 1 fatal infant case. To elucidate the genetic basis of these emerging strains, this study performed a comprehensive genomic characterization.
Phylogenetic analyses identified viral genetic subtypes and spatiotemporal transmission patterns. Globally, HAdV-21 strains are classified into 3 subtypes: HAdV-21a, HAdV-21b, and HAdV-21p. HAdV-21p, comprising historical strains from the 1950s, is no longer detected. Conversely, HAdV-21a (1956HAdV-21a ( -2024, 5 , 5 countries) and HAdV-21b (2005-2024, 3 countries) demonstrated extensive spatiotemporal distribution, reflecting stable epidemic trends and highlighting these strains as the dominant circulating subtypes in multiple countries, including HAdV-21
Best P value on recombination event FIGURE 3. Recombinant analysis of (A) HAdV-21a and (B) HAdV-21b strains. Note: Jilin2023-296 and Shenzhen2024-201 were used as representative HAdV-21a and HAdV-21b strains, respectively. Abbreviation: HAdV-21=human adenovirus type 21.
China. Furthermore, all subtypes exhibited minimal intra-subtype genetic variation (distance <0.00031), suggesting genomic stability during prolonged circulation, consistent with observations in other HAdV types, such as HAdV-4 and HAdV-55 (10)(11). This stability may result from the conserved replication mechanism of HAdV or selective equilibrium under host immune pressure. Notably, Chinese HAdV-21a/b strains (2023-2024) exhibited extremely high genetic homology and consistency with 2019 Chinese strains (distance: 0.00007) and historical global strains (distance: <0.00040), indicating that the currently circulating HAdV-21 is not a novel variant but a continuous transmission of genetically related strains. Enhanced respiratory infection surveillance following coronavirus disease 2019 (COVID-19) likely contributed to increased detection rates in recent years. Although HAdV-21 demonstrates relatively lower virulence and infectivity than other common types (e.g., HAdV-3, -7, -4, -55), it is associated with diverse clinical manifestations, including severe pneumonia, acute respiratory distress syndrome, acute flaccid paralysis, myocarditis, and fatal outcomes across age groups, necessitating urgent research on its pathogenic characteristics (8)(9)(12)(13)(14).
Phylogenetic and recombination analyses confirmed a common ancestor for HAdV-21a/b. Despite high overall genome conservation, sequence variation analysis identified both shared and subtype-specific mutations in HAdV-21a/b, representing key molecular markers for subtype differentiation and indicating continuous adaptive evolutionary pressure. Mutations in the hexon gene are predominantly localized in HVRs containing critical neutralizing antibody epitopes, suggesting that HVRs are primary adaptive targets for immune evasion. Consistent with previous studies, HAdV-21a/b exhibited 15 amino acid deletions and 2 insertions in the RGD loop of the penton base gene compared to the prototype strain. This shortened RGD loop closely resembles those of HAdV-3 and HAdV-7, which are associated with severe diseases (15). Because the RGD loop is a core functional domain mediating viral cell entry, its alteration may directly affect viral pathogenicity and infection efficiency; therefore, the functional consequences of the penton base warrant further investigation.
Recombination analysis revealed that HAdV-21a/b acquired an HAdV-3-derived fragment in the E4 region (breakpoint: nt32,843). As E4 regulates viral DNA replication, transcript splicing, and late gene expression, this recombination likely enhances viral stability and replication efficiency through functional complementation, thereby conferring transmission advantages. Genetic recombination drives HAdV evolution, where adaptive recombination patterns establish stable transmission, as exemplified by HAdV-21a and HAdV-21b persisting in circulation for approximately 70 and 20 years, respectively.
The findings in this report are subject to at least three limitations. First, HAdV-21 strains were collected from only 7 Chinese PLADs, limiting the generalizability of the findings; broader sampling in future work would improve representativeness. Second, the impact of HAdV-21 genomic differences on viral infection phenotypes remains unclear, warranting further investigation. Finally, although this study identified a significant HAdV-3 fragment-containing recombination event, functional validation was lacking, necessitating additional experiments to clarify its implications for viral pathogenicity and transmission.
In conclusion, this study identified the cocirculation of HAdV-21a and HAdV-21b in China and characterized their genomic features. Given HAdV-21's increasing prevalence, persistence of similar viruses across multiple PLADs, and potential association with severe clinical outcomes, enhanced surveillance is required to quantify disease burden in China and provide targeted prevention and control strategies.
Conflicts
## HAdV-21b
HAdV-21a
HAdV-21b
HAdV-21a&21b HAdV-21a&21b HAdV-21a&21b
SUPPLEMENTARY FIGURE S1. Maximum-likelihood phylogenetic tree constructed with 23 HAdV-21 strains from this study and 50 HAdV-21 strains from the GenBank database based on (A-I) nine fragments and (J) whole-genome sequence (WGS). The HAdV-21 prototype strain is indicated in red and HAdV-3 prototype strain is indicated in dark blue.
## References
1. Davison, Benkő, Harrach (2003) "Genetic content and evolution of adenoviruses" *J Gen Virol*
2. Dou, Chen, Song et al. (2025) "Epidemiological characteristics and genomic analysis of respiratory adenovirus in Jining City from February 2023 to July 2024" *BMC Genomics*
3. Wu, Zhang, Liang et al. (2022) "Comparative analysis between genotypes of adenovirus isolates from hospitalized children with acute respiratory tract infections and clinical manifestations in Wuhan" *Virol Sin*
4. Mahmood, Ahmed, Farooq et al. (2024) "Molecular characterization of human adenoviruses associated with pediatric respiratory infections in Karachi, Pakistan" *BMC Infect Dis*
5. Abbasi, Shafiei-Jandaghi, Shadab et al. (2023) "Phylogenetic characterization of rhinovirus and adenovirus in hospitalized children aged ≤ 18 years with severe acute respiratory infection in Iran" *Iran J Microbiol*
6. Liu, Xu, Li et al. (2023) "Prevalence of human infection with respiratory adenovirus in China: a systematic review and meta-analysis" *PLoS Negl Trop Dis*
7. Sdjr, Mccomb, Murray et al. (1959) "Adenoviruses isolated from Saudi Arabia. I. Epidemiologic features" *Am J Trop Med Hyg*
8. Ye, Han, Zhu et al. (2020) "First identification of human adenovirus subtype 21a in China with MinION and illumina sequencers" *Front Genet*
9. Liu, Zhang, Cai et al. (2022) "Human adenovirus subtype 21a isolates from children with severe lower respiratory illness in China" *Front Microbiol*
10. Wang, Feng, Duan et al. (2024) "Human adenovirus type 4 (HAdV-4) associated acute respiratory tract infection in children & genetic characteristics of HAdV-4 in China: a prospective multicenter study" *BMC Infect Dis*
11. Hang, Kajon, Graf et al. (2020) "Human adenovirus type 55 distribution, regional persistence, and genetic variability" *Emerg Infect Dis*
12. Jia, Wan, Xiao et al. (2025) "Rapid and sensitive detection of human adenovirus types 3 and 7 using CRISPR-Cas12b coupled with multiple cross displacement amplification" *Infect Med*
13. Abd-Jamil, Teoh, Hassan et al. (2010) "Molecular identification of adenovirus causing respiratory tract infection in pediatric patients at the University of Malaya Medical Center" *BMC Pediatr*
14. Pfortmueller, Barbani, Schefold et al. (2019) "Severe acute respiratory distress syndrome (ARDS) induced by human adenovirus B21: report on 2 cases and literature review" *J Crit Care*
15. Hage, Huzly, Ganzenmueller et al. (2014) "A human adenovirus species B subtype 21a associated with severe pneumonia" *J Infect*
16. (2009) "HAdV-21b 35 NHRC 91447 USA 2007 KJ364590 HAdV-21b 36 NHRC 64589 USA 2007 KJ364582 HAdV-21b"
17. Gomen "HAdV-B106 67 Human mastadenovirus B114 DEU 2023 OR853835 HAdV-B114 SUPPLEMENTARY TABLE S2-1. Genomic annotations of strains Jilin2023-296, Jilin2023-595, Ningxia2024-533, and Ningxia2024-595"
18. Large T Antigen (3761)
20. (8495) *5 kDa protein*
21. (0380) *DNA-binding protein*
22. *L2 penton protein*
23. Dna "579 L4 100 kDa hexon-assembly associated protein"
24. "kDa protein 27"
25. (0194) "3 kDa protein 30"
26. (1250) "Genomic annotations of strains Ningxia2024-608, Ningxia2024-618, Ningxia2024-669, and Ningxia2024-671. Gene Encoded product Ningxia2024-608 Ningxia2024-618 Ningxia2024-669 Ningxia2024-671 E1A 28.4 kDa protein 575-11"
27. Large T Antigen (3841)
29. (8345) *5 kDa protein*
30. (0388) *DNA-binding protein*
31. "Continued Gene Encoded product Ningxia2024-608 Ningxia2024-618 Ningxia2024-669 Ningxia2024-671 L2 penton protein 13"
32. Dna (0194) "582 L4 100 kDa hexon-assembly associated protein"
33. (0197) "3 kDa protein 30"
34. (1963) "Genomic annotations of the strains Ningxia2024-674, Ningxia2024-725, Ningxia2024-747, and Ningxia2024-754. Gene Encoded product Ningxia2024-674 Ningxia2024-725 Ningxia2024-747 Ningxia2024-754 E1A 28.4 kDa protein 574-1,155"
35. Large T Antigen (3841)
37. "Continued Gene Encoded product Ningxia"
38. E2b Dna Polymerase (8475) *5 kDa protein*
39. (0388) *DNA-binding protein*
40. *L2 penton protein*
41. Dna "579 L4 100 kDa hexon-assembly associated protein"
42. (0194) "3 kDa protein 30"
43. (1963) "Gene Encoded product Ningxia2024-868 Ningxia2024-907 Ningxia2024-944 Beijing2024-330 E1A 28.4 kDa protein 575-1,156"
44. Large T Antigen
46. (8485) *5 kDa protein*
47. (0380) *DNA-binding protein*
48. *L2 penton protein*
49. Dna (0199) "571 L4 100 kDa hexon-assembly associated protein"
50. (0195) *L5 agnoprotein*
51. (1963) "Gene Encoded product Beijing2024-364 Beijing2024-378 Shenzhen2024-077 Shenzhen2024-201 E1A 28.4 kDa protein 575-1,156"
52. Large T Antigen
54. E2b Dna Polymerase (8495) *5 kDa protein*
55. (0379) *DNA-binding protein*
56. *L2 penton protein*
57. Dna
58. (0195) "3 kDa protein 30"
59. (1156) "Shenyang2024-163, and Gansu2024-215. Gene Encoded product Chongqing2024-069 Shenyang2024-163 Gansu2024-215 E1A 28.4 kDa protein 575"
60. Large T Antigen
62. (3886) *5 kDa protein*
63. (0388) *DNA-binding protein*
64. "Continued Gene Encoded product Chongqing2024-069 Shenyang2024-163 Gansu2024-215 L1 52 kDa protein 10"
65. *L2 penton protein*
66. Dna "580 L4 100 kDa hexon-assembly associated protein"
67. (0195) "3 kDa protein 30" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12604196&blobtype=pdf | # Unsupervised classification of influenza virus surface spikes from cryo-EM reconstructions using head-to-stem width ratio
Younes Benkarroum
## Abstract
these spikes is therefore essential for understanding viral morphology and guiding the development of vaccines and antiviral drugs [5][6][7].High-resolution three-dimensional (3D) reconstructions of influenza virions, obtained through electron microscopy (EM) and particularly cryo-electron tomography (cryo-ET), have enabled direct visualization of spike distributions and structural variation [8,9]. These reconstructions provide the spatial resolution necessary to distinguish HA and NA morphologies, but manual annotation of spikes is time-consuming, subjective, and difficult to reproduce. Automated, reliable methods are Virology Journal
## Introduction
Influenza viruses pose a major global health challenge due to their high mutation rates and frequent antigenic shifts. The two major surface glycoproteins-hemagglutinin (HA) and neuraminidase (NA)-play central roles in viral infectivity, host specificity, and immune system evasion [1][2][3][4]. Accurate identification and classification of thus required to accelerate large-scale structural virology studies.
Recent progress in automated spike classification has been propelled by machine learning (ML) and deep learning (DL) techniques applied to both sequence and structural data. These strategies fall broadly into two categories:
• Sequence-Based Approaches: Convolutional neural networks (CNNs) trained on hydrophobicityencoded sequences have achieved strong host classification performance [10]. Other studies have employed CNNs [11] and XGBoost regression models [12] to estimate antigenic distances and predict evolutionary behavior using sequence data and assay results. These approaches, however, rely heavily on fully annotated genomes-often unavailable in early outbreak stages.
## •
Imaging-Based and Structural Approaches: Structural techniques, such as cryo-ET have enabled spatial mapping of surface spikes. For example, Huang et al. [13] embedded CNN modules into the EMAN2 framework to annotate glycoproteins in 3D EM reconstructions, while McMahon et al. [14] used super-resolution microscopy and Fourierbased analysis to study spike distribution and pleomorphism. Despite their utility, these ML-based approaches depend on labeled datasets and are often computationally complex or less biologically interpretable.
In this work, we propose an unsupervised and interpretable approach to spike classification based on a simple but robust geometric feature: the ratio of head width to stem width. This biologically motivated metric captures the club-like morphology of NA [15] compared with the more slender HA [15,16] and can be extracted automatically from reconstructed virions following segmentation. Figure 1 illustrates the structural differences between HA and NA spikes, showing their relative positions on the viral lipid bilayer envelope. In these gray-value images, darker regions correspond to higher Coulomb potentials, triangular HA spikes (e.g., white arrowhead) are distinguishable from square NA spikes (e.g., white arrowhead with black border). Images reproduced from [16] enabling visualization of the spikes. Proteins have higher Coulomb potential than surrounding ice in cryogenic EM preparations, which creates this contrast.
Most importantly, our technique not only reduces subjectivity and improves reproducibility, but also serves as a labeling engine to support the training and validation of supervised ML models. Rather than competing with modern deep learning pipelines, this work enables them by supplying high-confidence annotations from raw EM data. By bridging traditional image analysis with emerging ML frameworks, our contribution supports scalable, reliable classification essential to virological research and vaccine development.
## Materials and methods
This section describes the experimental dataset, the reconstruction of influenza B virus (B/Lee/40) particles from cryo-EM tilt series, and the computational pipeline for spike segmentation, feature extraction, and classification. These methods form the basis for the subsequent evaluation on both real reconstructions and simulated virus phantoms presented in the Results section.
## Data collection and virus reconstruction
The influenza B virus (B/Lee/40) used in this study was grown and amplified in embryonated chicken eggs. Each egg was inoculated with 0.1 mL of diluted allantoic fluid and incubated at 33 °C for 75 h.
For data acquisition, images were recorded at 50,000× magnification with an underfocus of 8 ± 0.5 μm via a JEOL 3200FSC electron microscope (JEOL, West Chester, PA) operating at 300 kV. To enhance contrast, an energy filter (20 eV slit width) was used to eliminate non elastically scattered electrons. Serial EM software [17] was used for tilt series acquisition on a 4096 × 4096 pixel CCD camera (Gatan Inc., Pleasanton, CA), with a pixel edge length of 0.88 nm. The specimen was imaged at angles ranging from -60° to + 60° in 2° increments, yielding 61 projection images. The low-dose imaging mode limited total specimen dosage to 60e/Å 2 over the entire tilt series. Figure 2 shows representative projection images from a cryo-electron tomography tilt series of influenza B virus (B/Lee/40). Each panel shows multiple virions imaged at tilt angles of -60°, 0°, and + 60°. These raw projections are used as input for reconstruction algorithms, which generate 3D density maps of individual particles for subsequent spike segmentation and classification. Scale bar: 250 nm.
To reconstruct the influenza virus, we applied three different techniques to compare their outputs and determine the most effective method for generating highquality 3D reconstructions, which is crucial for spike classification. The techniques used were the algebraic reconstruction technique (ART), the simultaneous iterative reconstruction technique (SIRT), and weighted backprojection (WBP). ART and SIRT were implemented using the optimal parameters previously identified and published in [18], which have been shown to offer the best balance between preserving the edges of the reconstructed objects and minimizing local oscillations within the structures. The detailed steps for these reconstructions are outlined in [19]. Figure 3 presents the central slice from each reconstruction of an isolated virus particle, all of which are derived from the same projection data. The main acquisition and reconstruction parameters are summarized in Table 1.
Consistent with findings in previous studies [20,21], our observations indicate that WBP reconstruction is inferior. In the reconstructed slices shown in Fig. 3, ART outperforms SIRT. However, as reported in the literature [21] and [22] (Chap.12), increasing the number Fig. 2 Projection images of influenza B virus (B/Lee/40) particles acquired at tilt angles of -60°, 0°, and + 60°. Scale bar: 250 nm of iterations in SIRT could improve its reconstruction quality to approach that of ART. Despite this potential improvement, the computational expense of SIRT is already an order of magnitude greater than that of ART. Given these considerations, we selected ART as the preferred reconstruction technique, as it produces more effective 3D reconstructions from cryo-electron microscopy images of virus particles than alternative methods do.
Examining the central slice of the reconstructed 3D scene obtained via ART, as shown in Fig. 3(a), reveals critical structural details. The slice clearly displays 14-nm-long surface spikes, an 8-nm-thick envelope matrix, and discrete RNPs inside the virion. The penetration of surface proteins into the matrix is well resolved, and close visual inspection reveals two distinct surface protein morphologies: one with a near-uniform density and thickness and another with a club-like structure featuring a denser top.
While this classification provides valuable insights, it remains inherently subjective, relying on human judgment shaped by knowledge, experience, and perception. To eliminate these inconsistencies and enhance the reliability of our analysis, we opted to automate the classification of surface proteins. Automated classification ensures reproducibility by consistently applying the same criteria under identical conditions. Compared with manual classification, this process enables a computer to categorize protein spikes based on their shape, making classification more reliable and efficient while reducing time and costs. The following sections outline the steps taken to automate the classification procedure.
## Segmentation of surface spikes
Segmentation is a crucial step in image processing that involves partitioning an image into meaningful regions for further analysis; see [23]. In this study, segmentation is used to isolate virus structures from reconstructed 3D volumes, enabling the precise identification of surface proteins and other structural components. The goal of segmentation is to increase the accuracy and efficiency of feature extraction by distinguishing relevant biological structures from background noise.
This section outlines the key steps involved in the segmentation process. First, we define the region of interest (ROI) to focus on the most informative parts of the reconstruction. Next, we present the theoretical framework guiding our segmentation approach, followed by a detailed explanation of the segmentation algorithm. Finally, we describe experiments and results that validate the effectiveness of our segmentation method.
## Identification of the region of interest
The first step in the segmentation process involves isolating the relevant region from the 3D reconstruction. Since the virus has a spherical shape [16], we assume that the The outer sphere encapsulates the entire virus, while the inner sphere is large enough to include the bilayer envelope surrounding the matrix but small enough to avoid cutting off the stems of the HA and NA spikes.
To determine the radii of these spheres, we used an interactive software tool developed in [19] to assist in isolating virus particles during data processing. This software enables precise manual adjustments of circles around virus particles in 2D micrograph images, allowing fine-tuning of their radius and positioning. In this study, we applied the same software to reconstruct slices to determine the center and radii of the outer and inner spheres. In the reconstruction shown in Fig. 3(a), these radii were measured at 83.60 nm and 59.84 nm, respectively. The output of this method is illustrated in Fig. 4. On the left, a gray-value image of the slice from Fig. 3(a) displays two superimposed concentric circles, representing the intersections of the spheres with the plane of the slice. On the right, the processed gray value image shows the effect of setting pixel values outside the annular region to zero, effectively isolating the region of interest. We note that Fig. 4 depicts a reconstructed slice from a real influenza B particle. The concentric circles and annular mask were applied solely as computational tools to isolate spike regions and do not represent an assumption of spherical virus morphology.
Since the reconstructed values tend to be less reliable in slices further from the central section, we limited our region of interest to the central slices. A total of 21 slices were included to ensure that full spikes were captured at various orientations, given that a single spike extends across approximately 10 slices. The outcome of this process serves as the input for further processing, as described in the subsequent subsections.
## Segmentation theory
In this study, we initially applied the simplest segmentation method: thresholding. This technique generates binary 3D scenes by assigning a value of zero to voxels below a chosen threshold and a value of one to all others. However, due to noise in the input data, the results were not satisfactory. To address this limitation, we adopted a more advanced segmentation approach-the fuzzy connectedness technique [24][25][26], which is detailed in the following subsections.
The fuzzy connectedness (FC) technique is an image analysis method used for segmentation, where objects or regions are defined based on their degree of connectivity rather than strict boundaries. It leverages fuzzy set theory to assign a connectivity strength between pixels, reflecting their likelihood of belonging to the same object. The key idea is that each pixel is connected to others with a certain affinity, which is determined by properties such as intensity similarity, spatial proximity, and Fig. 4 Gray-value slice from a reconstructed influenza B virus (B/Lee/40) particle: (a) inner and outer concentric circles superimposed on the slice, and (b) processed image in which pixels outside the annulus were set to zero to isolate the region of interest. The circles are analysis aids and do not imply a spherical virus model texture [25]. Instead of a binary classification, where a pixel either belongs or does not belong to a region, FC assigns a continuous membership value, allowing for more flexible and robust segmentation.
FC typically begins with seed points selected manually or automatically, which serve as references for computing the strongest fuzzy connectivity paths between pixels [24,26]. The algorithm then expands regions based on the highest connectedness values, ensuring that segmentation is smooth and resistant to noise. Unlike hard-threshold methods, FC considers global connectivity, meaning that segmentation decisions are influenced by the entire image rather than just local properties. This technique is particularly useful in medical imaging, where accurate segmentation of structures such as organs, tumors, and tissues is required despite variations in intensity, contrast, or partial volume effects.
## Segmentation algorithm
In this general approach, we consider an arbitrary set V, whose elements, referred to as spels (short for spatial elements), can represent various entities, such as pixels in an image or voxels in a 3D volume. In our specific application, V consists of all voxels within the region of interest, as defined in the previous subsection. The goal is to partition V into multiple objects in a fuzzy manner-that is, rather than assigning a spel to a single object definitively, we assign it a membership grade between 0 and 1. A value of 0 indicates that the spel certainly does not belong to the object, whereas a value of 1 indicates definite membership.
A sequence of spels is called a chain, with links representing ordered pairs of consecutive spels within the sequence. The strength of a chain is determined by the weakest link within it. The fuzzy connectedness of a spel c to another spel d within a set A is then defined as the strength of the strongest chain in A connecting c and d.
Each of the M objects has its own definition of link strength, denoted by ψ m , which serves as the affinity function for the m th object. Additionally, each object has its own set of seed spels, denoted V m . An object is defined as the collection of spels that are connected entirely within it to one of its own seed spels with greater strength than to any seed spels of other objects. The detailed specifications of the multiobjective fuzzy segmentation (MOFS) algorithm [24] are provided below. In this approach, a fuzzy segmentation of V is represented as a function σ, which maps each spel c ∈ V to an (M + 1)-dimensional vector:
Here, σ c m represents the grade of membership of spel c in the m th object, whereas σ c 0 is always defined as:
The key feature of the MOFS algorithm is that it computes the grade of membership for each spel across all objects and assigns the spel to the object for which its membership grade is highest.
$$σ c = (σ c 0 , σ c 1 , . . . , σ c M )$$
$$σ c 0 = max 1≤ m≤ M σ c m Algorithm 1: MOFS algorithm$$
## Segmentation experiments
In the following experiments, we apply the MOFS algorithm to the 3D region of interest identified at the beginning of this subsection to segment the virus spikes. The segmentation targets M = 3 objects: the spikes (foreground), the background, and the remaining envelope matrix (interior). The fuzzy spel affinity function is defined based on statistical properties of the links within regions designated by the user as belonging to these three objects.
The affinity functions ψ m and seed sets V m (for 1 ≤ m ≤ 3) are specified as follows. Using an interactive seed-selection application (Fig. 5), we manually select spels in each slice, identifying them as belonging to the m th object. The set V m is then formed by these selected spels along with their twenty-six neighbors (eight from the same slice and eighteen from adjacent slices) that fall within the region of interest.
We define g_m as the mean and h_m as the standard deviation of the average brightness for all face-adjacent pairs of spels in V m . Similarly, e_m and f_m represent the means and standard deviations of the absolute brightness differences for these pairs. The affinity function ψ m (c,d) is defined as 0 if c and d are not face-adjacent. Otherwise, it is given by [ρ g_m,h_m (g) + ρ e_m,f_m (e)]/2, where g is the mean brightness and e is the absolute brightness difference between c and d. The function ρ r, s (x) represents the Gaussian probability density function with mean r and standard deviation s, scaled so that its peak value is 1.
The interactive seed-selection application allows users to select a slice and activate an object m by pressing the corresponding button labeled with the object's name. Once activated, the shape of the mouse cursor changes to indicate that seed selection is in progress. Three distinct colors are used to mark the selected spels: green for the foreground, red for the background, and yellow for the interior objects. Figure 5 illustrates this process for the 11th slice of the region of interest.
As seeds are selected, the brightness statistics for object m-specifically g_m, h_m, e_m, and f_m, which are used in the affinity functions-are computed in real time and displayed in the right panel of the application.
Manual selection of seed spels is a tedious and timeconsuming task that also raises concerns about reproducibility. In our illustrative example, more than 3,000 seeds must be assigned for the following reasons: the virus's region of interest contains 44 spikes, and approximately four seeds per spike (two for the foreground and two for the background) must be selected across 21 slices.
To address this challenge, we enhanced the interactive application by incorporating a feature for the automated generation of seed spels. This automation significantly reduces the time and effort required for seed selection while improving reproducibility. The benefits of this feature will become even more evident in the evaluation methodology (Data collection and virus reconstruction), where seed selection must be performed for 30 different simulated viruses.
We now provide a detailed explanation of how seed spels were selected for our illustrative example in the MOFS algorithm. Using the interactive seed-selection application, we processed each slice of the region of interest, as illustrated for Slice 11 in Fig. 5. The majority of foreground and background seeds were automatically generated by pressing the Generate Seeds button, which assigns a foreground seed at each 2D local brightness minimum and a background seed at each 2D local brightness maximum. In some cases, additional background seeds were manually added. However, all interior object seeds were selected manually, amounting to approximately 500 seed spels for the entire 3D region of interest.
During automatic seed generation, some interior object spels were initially misclassified as foreground seeds because their brightness values were local minima within the interior object. To correct this, the interactive application provides an option to reassign misclassified spels by clicking on them with the Interior Select Seeds button activated. Once reassigned, their color changes from green, indicating the foreground, to yellow, indicating the interior.
In Fig. 5, the green and most of the red marks represent automatically generated seeds, corresponding to local minima and local maxima of brightness values, respectively. A small number of red marks were manually added to the background. The yellow marks indicate manual reassignments from foreground to interior within the envelope matrix.
Another important feature of our interactive application is its ability to save the seed locations into a text file and reload them later for further modifications. This functionality ensures that previously selected seeds can be preserved and revisited if additional seeds need to be added. It enhances flexibility in the segmentation process by allowing adjustments without requiring the user to start the seed selection from scratch.
The MOFS algorithm takes as input the set V of spels in the 3D region of interest, the sets of seed spels V m , and the affinity functions ψ m (1 ≤ m ≤ 3). The output is an array that associates each spel c ∈ V with a four-dimensional vector
, where σ c m (1 ≤ m ≤ 3) represents the grade of membership of the spel c in the m th object. The value of σ c 0 is determined as max( σ c 1 , σ c 2 , σ c 3 ). The foreground object, which corresponds to the structural system of virus spikes, is defined as the finite collection of spels c whose vectors satisfy the condition σ c 0 = σ c 1 . This foreground segmentation result is visualized in Fig. 6 using the molecular visualization software UCSF Chimera, which applies a surface smoothing feature to enhance the clarity of the displayed spikes.
$$σ c = ( σ c 0 , σ c 1 , σ c 2 , σ c 3 )$$
## Feature extraction
To achieve our primary goal of classifying surface spikes, we first need to extract the individual spike structures from the foreground object. By examining Fig. 6, we hypothesize that the individual spike structures can be isolated by partitioning the foreground object into distinct components based on face-adjacency (as described in [27]). In cases where the separation between spikes is insufficient to extract them individually, we can address this issue by revisiting the interactive seed selection application and adding additional background seeds to ensure proper separation.
Before proceeding with the classification of individual surface spikes, we perform a preliminary step in which each spike structure is rotated to an approximately upright position. Specifically, we align the direction from the stem to the head of the spike with the positive y-axis of the coordinate system. The coordinate system used here is defined such that the x-axis points to the right within each slice, the y-axis points upward, and the z-axis is perpendicular to the slice plane, pointing toward the observer. The detailed procedure for this rotation, particularly for our illustrative example, is carried out using a custom-designed program. We refer to this procedure as the reorientation process, and its steps are outlined below.
1. The user selects a slice that is perpendicular to the z-axis in which the spike to be extracted is clearly visible. Let s (where 1 ≤ s ≤ 44) denote the index of the spike to be extracted, with 44 being the total number of spike structures present in the foreground (refer to Fig. 6). Let I k represent the chosen slice and c k its z-coordinate. I k is a 2D binary image of size 200 × 200 pixels, where each pixel has a value of 1 if it corresponds to a location in the foreground (i.e., part of one of the spike structures) and 0 otherwise. This is illustrated in Fig. 7, where the blue cross marks the slice center at the point (0,0,c k ), which is referred to as the slice center in the following steps. 2. The user selects and clicks on a pixel inside the spike s within the 2D image I k . This pixel serves as the reference point for the subsequent rotation. 3. The application calculates the angle α s required for counterclockwise rotation around the slice center. This angle is needed to move the selected pixel to the positive side of the vertical line passing through the slice center. Specifically, the goal is to rotate the pixel to the position (0, p, c k ), where p is the distance Fig. 6 Surface-smoothed display of the viral spike structure system segmented within the region of interest by the MOFS algorithm, using the seeds selected as illustrated in Fig. 5 between the rotation center (the slice center) and the center of the selected pixel. (where i ≠ k), of the foreground object, applying the same rotation and cropping process. 7. The resulting slices, J i , are stacked together in their original order to form the rotated volume of the foreground object. This process effectively simulates a counterclockwise rotation of the entire 3D foreground object by the angle α s around the z-axis. 8. The rotated spike s is defined as the component of the foreground object that, when partitioned using face-adjacency (see [27]), contains the rotated pixel at (0, p, c k ). Since both HA and NA spikes extend radially from the membrane (see [16]), the resulting spike structure will be approximately upright, with the positive y-axis direction aligning with the radial extension from the stem to the head of the spike. 9. The entire procedure (steps 1 to 8) is repeated for each individual spike structure in the foreground object.
It is important to note that the majority of the computations described above are carried out by the application. The user's main task is to select a slice and click on a spike to extract its structure. A user may opt to use a single slice to extract all spike structures, which would involve making 44 clicks in the chosen slice (one click for each spike). Figure 7 demonstrates this concept. In this figure, the 23rd spike is clearly visible. Panel (a) shows the original slice, whereas Panel (b) presents the rotated version. This process is applied individually to each spike to obtain its rotated structure, and the results of this procedure serve as inputs for the next processing task.
To automate the classification of the segmented structures corresponding to individual spikes, we need to identify distinguishing features between the two types of spikes. In our search for appropriate features, we referred to the work of Harris et al. [16], who reported the following: both HA and NA spikes extend radially from the membrane, ending in bulbous heads (see Fig. 1). NA spikes are slightly longer than HA spikes and can be distinguished in longitudinal sections by their shorter head (the head of HA has a distinctive bi-lobed "peanut" In transverse sections, NA spikes have a square profile, while HA spikes have a triangular profile (see Fig. 1(d)). These observations suggest that it is possible to define a feature vector for each spike structure, which can then be used to classify the spikes as either HA or NA.
However, discussions with biologists led us to consider the possibility of using a single feature for reliable classification. This feature is the ratio of the width of the spike's head to the width of its stem. It appears that this ratio is greater for NA spikes than for HA spikes, as evident from a comparison of Figs. 1(b) and 1(c). The methodology for extracting this feature from each segmented spike structure closely follows the approach outlined in Subsection 9.7.1 of [23], which distinguishes various types of bone structures in the human foot and ankle. The procedure involves first identifying the centroid and the first principal axis for each spike structure, as obtained through the reorientation process described earlier in this subsection. The first principal axis for each spike structure was determined using principal component analysis (PCA) as described in [28,29], and centroids were calculated using the MATLAB function regionprops. This function measures several properties, including area, centroid, and bounding box, for each connected component in an image. We now describe the technical steps involved in calculating the "ratio of the width of the spike's head to the width of its stem" for any given spike.
As mentioned earlier, we assume that the spike structure has already been reoriented, with the positive y direction running approximately from the stem to the head of the spike. Additionally, we assume that the first principal axis and centroid for the spike structure have been identified. The first principal axis is shown in Fig. 8(a), and the centroid is marked by a blue cross in Figs. 8(b), 8(c), and8(d). We now describe how the spike structure is rotated so that the centroid remains fixed while the first principal axis is aligned with the y-axis, with the positive y direction extending from the stem to the head of the spike. This rotation is performed in two stages, as illustrated in Figs. 8(b) and 8(c), with further details provided below.
For each point along the first principal axis, the crosssectional area of the spike is defined as the area of intersection between the spike and a plane perpendicular to the first principal axis that contains that point. The maximal value of these cross-sectional areas is referred to as the "width of the spike's head. " The "width of the stem" is defined as the median value of all cross-sectional areas between the centroid and the point at the extreme bottom (furthest from the head) of the first principal axis. The locations where these widths are measured are marked in Fig. 8(d) in green for the head and yellow for the stem.
We now describe the procedure for rotating a reoriented spike structure to align its first principal axis with the y-axis while keeping the centroid fixed. This procedure consists of two actions, as illustrated in Fig. 8. The steps are performed in a local coordinate system with its origin at the centroid, and its axes x l , y l , and z l are parallel to the global coordinate axes x, y, and z, respectively.
The first action involves rotating the spike structure by less than π/2 around the x l axis to bring the first principal axis into the plane defined by (x l , y l ). This rotation, which involves transforming a set of voxels from the original cubic grid into a new set of voxels, is performed in a manner similar to the reorientation process described earlier in this subsection. Figure 8(b) shows the (x l , y l )-slice of the 23rd spike, with the first principal axis and centroid (marked in blue) after the first rotation. Note that, owing to the prior reorientation of the spike structure and the rotation being less than π/2, the positive y-direction still extends approximately from the stem to the head of the spike.
The second action is a rotation of the resulting spike structure by less than π/2 around the z l -axis to align the first principal axis with the y l -axis. Figure 8(c) shows the (x l , y l )-slice of the 23rd spike, displaying the first principal axis after this second rotation.
Let R head/stem (s) denote the feature we are seeking for the spike s. This feature is the ratio of the spike's head cross-sectional area to its stem cross-sectional area. In Fig. 8(d), the green and yellow lines show the locations of the cross-sectional areas for the spike's head and stem in the 23rd spike structure. The results for other spike structures (specifically, the 11th, 12th, and 16th spikes) are presented in Fig. 9. The feature values for the four spike structures illustrated in Figs. 8 and9 are R head/stem (23) = 4.42, R head/stem (11) = 18.95, R head/stem (12) = 19.66, and R head/stem (16) = 3.43.
## Classification strategy
The feature presented in the last subsection appears to be sufficient for the classification task. This feature assigns a single numerical value to each spike structure, and we believe that the values associated with NA spikes are consistently greater than those associated with HA spikes. If this assumption holds, classification reduces to finding an optimal threshold that effectively separates the two sets of numbers. The methodology for determining this threshold is outlined below.
To establish the optimal threshold, we employ Fisher's linear discriminant [30]. First, we arrange all the distinct feature values obtained in the previous subsection in ascending order. We then consider threshold candidates positioned halfway between each pair of consecutive values in this sequence. Each candidate threshold partitions the dataset into two classes: class c 1 for values below the threshold and class c 2 for values above it. The optimal Fig. 9 Feature extraction for (a) the 11th, (b) the 12th, and (c) the 16th spike structures threshold, denoted as τ, is the candidate that maximizes Fisher's linear discriminant, which is defined as:
where m i and s 2 i represent the mean and the variance of class c i , respectively.
Figure 10(a) plots Fisher's linear discriminant values for each threshold candidate, whereas Fig. 10(b) displays the R head/stem values in ascending order. The optimal threshold, τ = 8.29, is marked by the red line in Fig. 10(b).
Figure 11 illustrates a reoriented vertical planar section for each spike structure s, along with its computed feature value R head/stem (s). Using this feature and the threshold τ, we classify spike structures into two groups: Figure 12 visually illustrates the classification of all spike structures based on the threshold τ = 8.29: spikes in class c 1 are displayed in yellow, whereas those in class c 2 are shown in red. Based on discussions with biologists and an analysis of both the morphological characteristics and distribution of spike structures in each class, we strongly believe that class c 1 corresponds to HA spikes, while class c 2 corresponds to NA spikes. This assertion will be further substantiated in the next section.
$$J = |m 1 -m 2 | 2 s 2 1 + s 2 2$$
$$c 1 = {s | R head/stem (s) ≤ τ } c 2 = {s | R head/stem (s) ≤ τ }$$
## Results
We first applied the pipeline to experimental influenza B virus reconstructions to demonstrate feasibility on real data. To provide quantitative validation under controlled conditions with known ground truth, we then generated simulated virus phantoms and applied the same classification pipeline. This dual evaluation demonstrates both the biological relevance of the method and its quantitative accuracy under controlled conditions with ground truth.
## Application to real EM data
Using ART technique, we reconstructed influenza B virus (B/Lee/40) particles from tilt-series data and segmented individual spikes with the fuzzy connectedness algorithm. As shown in Fig. 3(a), the reconstructions reveal the bilayer envelope, ribonucleoprotein complexes (RNPs), and surface spikes with clear morphological variation. Application of the head-to-stem width ratio separated the spikes into two classes, c 1 and c 2 , consistent with the expected presence of both HA and NA on the viral surface. However, because the true identities of individual spikes cannot be directly established in experimental reconstructions, we could not calculate classification purity or definitively label c 1 as HA and c 2 as NA. Nevertheless, the emergence of two distinct morphological groups provides strong evidence that the approach captures biologically meaningful variation. To confirm the correspondence between classes c 1 and c 2 and the true HA and NA spike types, we next generated simulated virus phantoms with known spike labels and
## Evaluation on simulated virus phantoms
To provide quantitative validation, we generated 3D test phantoms of viruses with randomly arranged HA and NA conformations, maintaining an approximate ratio of 75% HA to 25% NA spikes. These phantoms were created via the jSNARK software package (jsnark.sourceforge.net/). The reconstruction region consists of a 200 × 200 × 200 voxel grid.
Since this section on computer simulations, we do not specify the exact physical size of a voxel; instead, we denote the edge length of a voxel as σ. The viral envelope was simulated as a sphere with a radius of R = 70σ, while the spikes were modeled using cylindrical and ellipsoidal shapes:
• HA Spikes: Heads were represented as ellipsoids with semi-axes 3.5σ, 3.5σ, and 6σ (see Fig. 13(a)), and stems were modeled as cylinders with a radius of 2σ and height of 10σ. • NA Spikes: Heads were modeled as ellipsoids with semi-axes 6σ, 6σ, and 3σ (see Fig. 13(b)), while stems were represented as cylinders with a radius of 1.5σ and height of 16σ. Harris et al. [16] visualized the 3D structure of a A H3N2 strain X-31 virus via cryo-EM tomography and determined that a typical 120 nm diameter influenza virion can contain up to 375 surface spikes, although the actual count may be lower because of the presence of bare spots. On this basis, we placed 337 randomly distributed spikes in each phantom, all extending radially from the virus surface.
The spike distribution was handled as follows: 45 spikes were placed equidistantly along the perimeter (2πR) of the central slice, and spikes in other slices were positioned proportionally, maintaining a consistent ratio of spike count to perimeter length. This concept is illustrated in Fig. 14(a), where the red line represents a lateral view of the central slice, and the blue lines mark the positions of the slices at polar angles ± θ.
The spikes were distributed across 11 slices, specifically at angles of θ = 0°, ± 15°, ± 30°, ± 45°, ± 60°, and ± 75°. Next, we used jSNARK to generate projection images from the phantoms, mimicking the real-world projection images obtained during data collection and processing prior to 3D reconstruction, as described in [19]. Specifically, the specimen angles ranged from -60° to + 60° in 2° increments, resulting in 61 projection images, each measuring 200 × 200 pixels.
To better simulate real experimental conditions, Gaussian noise with a mean of 0 and variance of 1.1 was added to the mathematical projections. This adjustment ensures that the quality of spike reconstructions from the simulated noisy data (Fig. 15(b)) closely resembles that of real data (Fig. 3(a)).
## Reconstruction
In this subsection, we focus on the process of reconstructing virus phantoms from simulated projection data using the algebraic reconstruction technique (ART). As in Sect. 2, we employed ART with the optimal parameters identified in [18]. Using ART ensures accurate reconstruction of the spike structures within the virus phantoms, even in the presence of noise introduced during data collection, which simulates real-world conditions. For clarity, the central slices of both the phantom and its 3D reconstruction are shown in Fig. 15.
## Performance evaluation
It is possible to apply the virus spike classification procedure outlined in Sect. 2 to the output of the reconstruction algorithm. This will classify the spike structures in the reconstruction as belonging to either class c 1 or c 2 . Intuitively, each spike structure in the reconstruction "originates" from a spike in the test phantom, which we created. This allows us to label each spike either an HA spike or an NA spike, providing the ground truth for our evaluation methodology. Technically, the labeling process works as follows: the spike structure is part of the foreground object (identified by face-adjacency) obtained through the segmentation method described in Sect. 2. For each spike structure, we examine the corresponding voxels in the 3D scene and compare them with the phantom definition. If more of those voxels belong to an HA spike than to an NA spike, the spike structure in the reconstruction is labeled HA; otherwise, it is labeled NA.
To evaluate the classifier's performance, we use a metric called classification purity (CP) [31,32]. As described in [32,33], this evaluation measure is computed as follows: for each reconstruction, a 2 × 2 contingency table is created (see Table 2). The rows correspond to the two types of spikes (HA and NA), while the columns correspond to the classifier's predicted classes (c 1 and c 2 ), as defined in Sect. 2. The numbers in the table represent the number of spikes that simultaneously belong to the row's spike type and the column's predicted class. Table 2 shows the classification purity for the reconstruction displayed in Fig. 15(b).
Ideally, each class should consist solely of spikes of one type. Thus, we define classification purity (CP) in % as 100 times the sum of the maximum values in each column divided by the total sum of all entries in the array. A more effective classification procedure results in a higher classification purity value.
For the array in Table 2, the classification purity is CP = 100× (33+11) 33+1+0+11 %, that is 97.78%. To obtain statistically significant results, we repeated the procedure with 30 randomly generated phantoms. By "random, " we mean both the distributions of HA and NA spikes and the noise in the simulated projections. The ART reconstruction algorithm was applied to each of the 30 datasets, followed by the application of the virus spike classification procedure to all the reconstructions. The classification purity was calculated for each reconstruction. The average classification purity across all 30 reconstructions was 97.48%, which is considered a high figure of merit (FOM).
An important observation is that the high classification purity value (such as 97.48%) reinforces the conclusion drawn at the end of Sect. 2, namely, that classes c 1 and c 2 correspond to HA and NA spikes, respectively.
## Discussion
This study demonstrates that an unsupervised, geometry-based classifier can distinguish influenza hemagglutinin (HA) and neuraminidase (NA) spikes in cryo-EM reconstructions with high accuracy. Application to experimental influenza B virus (B/Lee/40) reconstructions confirmed the feasibility of the approach on real data, while benchmarking on simulated phantoms with known ground truth yielded an average classification purity of 97.5%. Together, these results show that the head-to-stem width ratio provides a robust and interpretable discriminator of spike morphology, enabling automated classification without prior training data.
Our work also complements recent machine learning (ML) and deep learning (DL) approaches for spike analysis. Sequence-based ML models, including convolutional neural networks and regression methods, have been used to predict viral evolution and antigenic properties [10][11][12]. Imaging-based ML efforts, such as CNN-enhanced cryo-EM pipelines [13] and super-resolution microscopy combined with ML analysis [14], highlight the potential for automated spike annotation. However, these methods require large annotated datasets and often act as black boxes with limited biological interpretability. By contrast, our unsupervised, morphology-driven method provides interpretable classifications without prior training data. Most importantly, it generates high-confidence labels that can be used to train supervised ML systems, thereby bridging traditional image analysis with data-driven frameworks.
While the method was developed and tested specifically for cryo-electron tomography (cryo-ET) reconstructions, its underlying principle is not inherently restricted to this modality. Any imaging technique that provides 3D density maps with sufficient resolution to capture spike morphology could, in principle, support application of the head-to-stem ratio descriptor. Examples include subtomogram averaging pipelines or advanced fluorescence-based super-resolution techniques. However, these applications remain to be validated, and the current study should be considered as primarily optimized for cryo-ET data.
This study also has limitations. Performance was evaluated primarily on synthetic phantoms, and future work should extend to more diverse virion morphologies and additional experimental datasets. Expanding beyond central slices, incorporating heterogeneous particle populations, and testing robustness under varying imaging conditions will provide a fuller validation. Furthermore, while Fig. 4 illustrates the use of concentric circles to define an annular mask, these were applied solely as computational tools for isolating spike regions and do not imply that influenza virions are spherical in morphology. Nonetheless, the combination of segmentation, geometric analysis, and synthetic benchmarking provides a reproducible framework that can support both immediate spike classification tasks and the longer-term development of ML-based virology pipelines.
## Conclusion
We developed an unsupervised and interpretable method for classifying influenza surface glycoproteins directly from cryo-EM reconstructions. Application to experimental influenza B virus (B/Lee/40) demonstrated feasibility on real data, while evaluation on simulated phantoms with known ground truth confirmed high classification accuracy. By combining fuzzy connectedness segmentation with a biologically motivated geometric descriptor-the head-to-stem width ratio-the approach provides robust spike discrimination without the need for annotated training data. Beyond immediate application to influenza, the method also generates high-confidence labels that can support supervised ML development, thereby bridging traditional image analysis with emerging data-driven frameworks. Together, these features make it a scalable tool for structural virology and influenza surveillance.
## References
1. Chen, Lee, Steinhauer et al. (1998) "Structure of the hemagglutinin precursor cleavage site, a determinant of influenza pathogenicity and the origin of the labile conformation" *Cell*
2. Wang, Cheng, Lu et al. (2008) "Crystal structure of unliganded influenza B virus hemagglutinin" *J Virol*
3. Wang, Tao (2010) "Influenza: Molecular Virology"
4. Burmeister, Ruigrok, Cusack (1992) "The 2.2 A° resolution crystal structure of influenza B neuraminidase and its complex with sialic acid" *EMBO*
5. Ruigrok, Krijgsman, Ronde-Verloop et al. (1985) "Natural heterogeneity of shape, infectivity and protein composition in an influenza A (H3N2) virus preparation" *Virus Res*
6. Bucher, Kharitonenkov, Zakomirdin et al. (1980) "Incorporation of influenza virus M-protein into liposomes" *J Virol*
7. Roberts, Lamb, Compans (1998) "The M1 and M2 proteins of influenza A virus are important determinants in filamentous particle formation" *J Virol*
8. Frank (2006) "Introduction: principles of electron Tomography, in electron tomography: methods for three-dimensional visualization of structures in the cell. 2nd ed"
9. Frank (2006) "Three-dimensional electron microscopy of macromolecular assemblies: visualization of biological molecules in their native state"
10. Chrysostomou, Alexandrou, Nicolaou et al. (2021) "Classification of influenza hemagglutinin protein sequences using convolutional neural networks" *Annu Int Conf IEEE Eng Med Biol Soc*
11. Forghani, Khachay (2020) "Convolutional neural network based approach to in silico non-anticipating prediction of antigenic distance for influenza virus" *Viruses*
12. Li, Li, Shang et al. (2024) "A sequence-based machine learning model for predicting antigenic distance for H3N2 influenza virus" *Front Microbiol*
13. Huang, Song, Xu et al. (2022) "Quantitative structural analysis of influenza virus by cryo-electron tomography and convolutional neural networks" *Structure*
14. Mcmahon, Andrews, Groves et al. (2023) "High-throughput super-resolution analysis of influenza virus pleomorphism reveals insights into viral spatial organization" *PLoS Pathog*
15. Wilson, Skehel, Wiley (1981) "Structure of the hemagglutinin membrane glycoprotein of influenza virus at 3 A° resolution" *Nature*
16. Harris, Cardone, Winkler et al. (2006) "Influenza virus pleiomorphy characterized by cryo-electron tomography" *Proc Natl Acad Sci*
17. Mastronarde (2005) "Automated electron microscope tomography using robust prediction of specimen movements" *J Struct Biol*
18. Benkarroum, Herman, Rowland (2015) "Blob parameter selection for image representation" *J Opt Soc Am A Opt Image Sci Vis*
19. Benkarroum, Gottlieb, Katz et al. (2014) "Computational methods for electron tomography of influenza virus. in Computational methods for Three-Dimensional microscopy reconstruction"
20. Carazo, Herman, Sorzano et al. (2006) "Algorithms for three-dimensional reconstruction from the imperfect projection data provided by electron microscopy"
21. Sorzano, Marabini, Boisset et al. (2001) "The effect of overabundant projection directions on 3d reconstruction algorithms" *J Struct Biol*
22. Herman (2009) "Quadratic optimization methods"
23. Fishman, Kuszyk (2000) "3D imaging: musculoskeletal applications"
24. Carvalho, Herman, Kong (2005) "Simultaneous fuzzy segmentation of multiple objects" *Discrete Appl Math*
25. Ciesielski, Herman, Kong (2016) "General theory of fuzzy connectedness segmentations" *J Math Imaging Vis*
26. Herman, Carvalho (2001) "Multiseeded segmentation using connectedness" *IEEE Trans Pattern Anal Mach Intell*
27. Herman (1998) "Boundary tracking"
28. Jackson (1991) "A user's guide to principal components"
29. Jolliffe (2002) "Principal component analysis and factor analysis"
30. Fukunaga (1990) "Parametric classifiers. in Introduction to statistical pattern recognition. 2nd ed"
31. Tan, Steinbach, Kumar (2006) "Cluster analysis: basic concepts and algorithms. in Introduction to data mining"
32. Sikhakolli, Sikhakolli (2023) "Effective purity method for measuring the clustering accuracy and its illustration" *Int J Comput Appl*
33. Herman, Kalinowski (2008) "Classification of heterogeneous electron microscopic projections into homogeneous subsets" *Ultramicroscopy* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282453&blobtype=pdf | # Mutational landscapes of NITD008-resistant EV71 variants revealed through population sequencing
Fang Yu, Qiu-Yan Zhang, Zhe-Rui Zhang, Cheng-Lin Deng, Bo Zhang, Virologica Sinica
## Dear Editor,
Enterovirus 71 (EV71) is the main pathogen of hand, foot, and mouth disease (HFMD), which is a serious public health threat, especially in the Asia-Pacific region (Wu et al., 2013). Due to the lack of effective antivirals for treatment, supportive therapy remains to be the primary measure for severe infections of EV71. EV71 belongs to the genus Enterovirus in the family of Picornavirridae. EV71 encodes a polyprotein that is proteolytically cleaved into four structural proteins and seven nonstructural proteins, i.e., VP1 to VP4, 2A to 2C, and 3A to 3D. Moreover, an alternative encoding strategy of harboring a novel open reading frame encoding a short peptide has been recently reported in gut epithelial cells infected with some enteroviruses (Lulla et al., 2019). Structural proteins play a key role in the packaging and maturation of virus particles, while non-structural proteins are mainly involved in the replication process of virus. Among them, the 3D polymerase (3D pol ) protein functions as an RNA-dependent RNA polymerase (RdRP) that is essential for viral RNA synthesis (Wu et al., 2010).
Nucleoside or nucleotide analogues have been well accepted as therapeutic agents and play a dominant role in the treatment of a number of viral infections (Lou et al., 2014). NITD008 is an adenosine nucleoside analog that has been reported to selectively inhibit viruses in the family Flaviviridae and EV71 (Deng et al., 2014;Yin et al., 2009). It is well known that the triphosphate form of NITD008 competes with natural adenosine triphosphate substrates to incorporate into the growing RNA chain and terminate RNA elongation (Yin et al., 2009). However, the role of NITD008 as a mutagen in EV71 infection remains unclear. To clarify this, we evaluated the effect of NITD008 on virus complexity by analyzing the mutational landscape of wild type (WT) and NITD008-resistant EV71 populations in treated or untreated NITD008.
In our previous study, two NITD008-resistant isolates were selected by passaging WT EV71 with increasing concentrations of NITD008, and resistance mutations of 3A and/or 3D were identified (Deng et al., 2014). To test the mutational landscape of NITD008-resistant EV71 viruses induced by NITD008, we selected several mutants (3A V75A , 3D V63A þ M393L , 3A V75A -3D V63A þ M393L ) for deep sequencing. First of all, referring to our previous methods (Deng et al., 2014), the drug concentration of NITD008 was selected as 1 μM, which showed significant inhibition effect for WT virus and no cytotoxicity to Vero cells. Using the same concentration of DMSO as control, Vero cells were infected with WT, 3A V75A , 3D V63A þ M393L , and 3A V75A -3D V63A þ M393L at a multiplicity of infection (MOI) of 0.1. At 72 hours post-infection (hpi), the three NITD008-resistant viruses (3A V75A , 3D V63A þ M393L , 3A V75A -3D V63A þ M393L ) showed significant resistance relative to WT virus in cytopathic effect (CPE) (Fig. 1A and Supplementary Fig. S1), viral titer and RNA copy number (Fig. 1B), as previously reported (Deng et al., 2014). Subsequently, the RNA samples were used for DNBSEQ™ T7 sequencing platform across the full genome and analyzed by the ViVan pipeline (Isakov et al., 2015). We mined the deep sequence data for minority variants above 1% frequency to calculate the number of minority variants, which can be used as metrics of viral population diversity (Stapleford et al., 2015). Each sample achieved mean depth of more than 100,000 and 100% genome sequence coverage (Supplementary Table S1).
Firstly, the mutation spectra of proteins in WT population and NITD008-resistant populations were compared. Across the whole genome, WT maintained the highest number of mutations regardless of NITD008 induction (73 or 159 mutations), while NITD008-resistant populations had lower levels of mutations (51/49/45 or 123/94/59 mutations) (Fig. 1C andD). Notably, the mutation number of 3D gene in WT population increased remarkably from 1 to 51 (51-fold) upon NITD008 treatment, providing direct evidence that NITD008 targets viral polymerase during EV71 replication, which is consistent with in vitro experiments (Yin et al., 2009). Additionally, in WT population following NITD008 treatment, there was a 17-fold increase in mutation number of 2C gene. 2C is a highly conserved and multifunctional membrane protein of picornaviruses, which exhibits ATPase and helicase activity in vitro (Supplementary Fig. S2, boxes A, B andC) (Wang et al., 2020). Besides, EV71 2C has also been found to contain ATP-independent RNA chaperoning activity that can destabilize RNA duplexes (Xia et al., 2015). Further analysis showed that these mutations were located in the vicinity of motifs A, B, and C (Supplementary Fig. S2), thus we speculated that
## References
1. Deng, Yeo, Ye et al. (2014) "Inhibition of enterovirus 71 by adenosine analog NITD008" *J. Virol*
2. Fang, Wang, Wang et al. (2021) "Antiviral peptides targeting the helicase activity of enterovirus nonstructural protein 2C" *J. Virol*
3. Graci, Gn€ Adig, Galarraga et al. (2012) "Mutational robustness of an RNA virus influences sensitivity to lethal mutagenesis" *J. Virol*
4. Isakov, Borderia, Golan et al. (2015) "Deep sequencing analysis of viral infection and evolution allows rapid and detailed characterization of viral mutant spectrum" *Bioinformatics*
5. Lulla, Dinan, Hosmillo et al. (2019) "An upstream proteincoding region in enteroviruses modulates virus infection in gut epithelial cells" *Nat. Microbiol*
6. Rozen-Gagnon, Stapleford, Mongelli et al. (2014) "Alphavirus mutator variants present host-specific defects and attenuation in mammalian and insect models" *PLoS Pathog*
7. Smith, Blanc, Surdel et al. (2013) "Coronaviruses lacking exoribonuclease activity are susceptible to lethal mutagenesis: evidence for proofreading and potential therapeutics" *PLoS Pathog*
8. Stapleford, Rozen-Gagnon, Das et al. (2015) "Viral polymerase-helicase complexes regulate replication fidelity to overcome intracellular nucleotide depletion" *J. Virol*
9. Wang, Wang, Zhao et al. (2020) "The structure, function, and mechanisms of action of enterovirus non-structural protein 2" *C. Front. Microbiol*
10. Wu, Zhao, Pan et al. (2013) "Patterns of polymorphism and divergence in the VP1 gene of enterovirus 71 circulating in the Asia-Pacific region between 1994 and 2013" *J. Virol. Methods*
11. Wu, Lou, Miao et al. (2010) "Structures of EV71 RNA-dependent RNA polymerase in complex with substrate and analogue provide a drug target against the hand-foot-and-mouth disease pandemic in China" *Protein Cell*
12. Xia, Wang, Wang et al. (2015) "Human enterovirus nonstructural protein 2C ATPase functions as both an RNA helicase and ATP-independent RNA chaperone" *PLoS Pathog*
13. Yin, Chen, Schul et al. (2009) "An adenosine nucleoside inhibitor of dengue virus" *Proc. Natl. Acad. Sci. U S A*
14. Yu (2025) *Virologica Sinica* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12568186&blobtype=pdf | # First Serological Evidence of Crimean-Congo Hemorrhagic Fever Virus Infections in Croatia: A Multispecies Surveillance Approach Emphasising the Role of Sentinel Hosts
Crimean-Congo Hemorrhagic Fever, Gorana Miletic, Ivona Coric, Snjezana Kovac, Alenka Skrinjaric, Magda Kamber Taslaman, Margarita Bozikovic, Ljubo Barbic, Viktor Masovic, Jelena Prpic, Lorena Jemersic, Vladimir Stevanovic
## Abstract
Crimean-Congo hemorrhagic fever virus (CCHFV) is a tick-borne zoonotic pathogen of growing public health concern in southeastern Europe. This study provides the first serological evidence of CCHFV circulation in Croatia, based on testing 1473 serum samples from farm and companion animals, including sheep, horses, cattle, goats, dogs, and cats. A total of 109 samples (7.4%) tested positive for CCHFV antibodies using a commercially available enzyme-linked immunosorbent assay (ELISA) kit. The highest seroprevalence was recorded in sheep (28.3%), followed by horses (4.3%) and a single cat (0.5%), with no antibodies detected in cattle, goats, or dogs. Almost all seropositive animals originated from coastal and subcoastal Croatia, where Hyalomma ticks are present. Only two seropositive cases were detected in continental areas. Sheep samples from several farms in Zadar County showed intra-farm seropositivity rates of up to 85.7%, suggesting localised virus circulation likely influenced by vector distribution and farm-level practices. No viral ribonucleic acid (RNA) was detected by quantitative reverse transcription polymerase chain reaction (qRT-PCR), consistent with the transient nature of viremia in most animal hosts. These findings confirm the silent circulation of CCHFV in Croatia and reinforce the need for targeted, regionally adapted surveillance strategies that integrate multiple hosts and support early warning systems aligned with the One Health concept.
## 1. Introduction
Crimean-Congo hemorrhagic fever virus (CCHFV) is an RNA virus classified in the Orthonairovirus haemorrhagiae species, genus Orthonairovirus, a member of the family Nairoviridae [1]. It is a pathogen that poses a significant threat to public health, with a human case fatality rate of up to 30% [2]. The virus was first identified during World War II in the Crimean region. It was later found to be antigenically similar to a virus isolated in 1956 in what is now the Democratic Republic of Congo. This led to the naming of the virus as Crimean-Congo hemorrhagic fever virus [2,3]. Since its discovery, CCHFV has been reported in countries across Africa, Asia, Eastern Europe, and the Mediterranean basin [4][5][6].
The distribution of the virus closely correlates with the presence of ticks from the genus Hyalomma, which are considered the primary vector and potential reservoir of CCHFV [7]. The virus replicates in a variety of mammalian and avian hosts in endemic regions. Domestic animal hosts typically include small ruminants; however, serological evidence of infection has also been detected in equids, cattle, and wild animals [8][9][10].
Humans and laboratory mice are the only species that develop clinical disease following infection [11]. The initial symptoms in humans are nonspecific, resembling a common cold. This early phase is particularly critical for human-to-human transmission, making Crimean-Congo hemorrhagic fever (CCHF) a significant nosocomial threat in endemic areas [12]. A considerable proportion of confirmed human cases is associated with occupational exposure, especially among individuals working on farms or in slaughterhouses, with cattle frequently implicated [13,14]. Infected animals, although asymptomatic, undergo a brief viremic phase during which their body fluids pose a substantial risk for zoonotic transmission [11]. The risk is heightened by the fact that infection in animals often goes unnoticed. Consequently, detection of infection in animals usually occurs during outbreak investigations in humans or through targeted research, primarily via serological testing for virus-specific antibodies. The subclinical nature of the infection and the short duration of the viremic period, combined with close human-to-animal contact in high-risk settings, make early detection challenging and limit the effectiveness of outbreak prevention [8]. Considering these risks, there is a clear need for improved surveillance systems in risk areas, particularly those that incorporate more animal species that could serve as sentinels during the early stages of virus circulation.
Certain regions of Croatia, due to the abundance of vectors and potential reservoir hosts, are at considerable risk for disease emergence. In response to the slow yet steady spread of the virus across the Balkan Peninsula, such as the appearance in bordering Bosnia and Herzegovina in 2022 [15], the Croatian public health community has remained aware of the potential introduction of CCHFV within national borders [4,7,16,17]. As a proactive step, a surveillance study was launched to assess viral activity in multiple domestic and companion animal species.
## 2. Materials and Methods
For this study, Croatia was divided into four broader regions based on geographical position and the documented presence of the Hyalomma marginatum tick species [18]: Littoral Croatia and Dalmatia, where Hyalomma ticks are present, and Eastern and Central Croatia, where their presence has not been documented to date. Initially, we tested 304 horse, 304 dog, and 196 cat serum samples, along with 276 samples from cattle and 57 from goats. For most samples, data on species, sex, age, and geographical location were available (Table 1). These samples were collected during 2024 as part of the CROOH project ("Establishment of a coordinated disease surveillance system in Croatia in accordance with the One Health approach", Project No. 101132755).
In the second phase of the study, 184 archived sheep serum samples collected in one county in Dalmatia (Zadar County) in 2024 and 152 collected in Dubrovnik-Neretva County in Dalmatia in 2023 were tested. Sample size calculations for the first round of testing were performed using the RiBESS software version 0.1.2 [19], based on an expected seroprevalence of 30% in domestic ruminants and 10% in horses and pet animals, as sentinel species, based on previously reported seroprevalences in neighbouring countries and studies in respective sentinel animals [8,20,21]. Archived sheep samples originated from previous diagnostic and surveillance activities. Their numbers were determined by availability rather than sample size calculation. Archived sheep samples were limited to the central and south coastal counties of Zadar and Dubrovnik in the Dalmatia region (Figure 1). Since all samples used in this research were remnant serum samples collected by veterinarians during routine diagnostics and national surveillance for other diseases, not specifically for this study, no animal study permit was required.
All samples were stored at -80 • C until testing.
Serum samples were tested for anti-CCHFV nucleoprotein antibodies using the ID Screen ® CCHF Double Antigen Multi-species ELISA kit (IDvet, Grabels, France), according to the manufacturer's instructions [22,23]. This ELISA employs a double-antigen sandwich (DAS) format, allowing for the detection of total antibodies (IgG, IgM, and IgA). It does not rely on species-specific secondary antibodies, which makes it convenient for multispecies testing. According to the manufacturer, the assay has a specificity of 100% and a sensitivity of 98.9% across multiple animal species, with no cross-reactivity with related nairoviruses, including Hazara virus, Dugbe virus, and Nairobi Sheep Disease Virus. However, due to antigenic similarity, cross-reactivity with Aigai virus cannot be ruled out. Optical density (OD) values were read at 450 nm using a TECAN Sunrise ELISA microplate reader (Tecan Trading AG, Männedorf, Switzerland). The test was valid if the mean OD value of the positive control was greater than 0.350 and the OD ratio of the positive and negative controls was greater than three. The S/P% ratio was calculated for each sample by dividing the OD of the sample by the OD of the positive control. Samples that had S/P% greater than 30% were considered positive. All samples were tested in duplicate.
Serum samples that tested positive were subsequently tested by quantitative (q) reverse-transcription (RT) PCR. RNA extraction was performed using the IndiMag Pathogen Kit (Indical, Leipzig, Germany), with automated processing carried out on the KingFisher Flex nucleic acid extraction system (Thermo Fisher Scientific, Waltham, MA, USA). Isolates that were not immediately subjected to qPCR were stored at -20 • C until use. Molecular analysis was conducted using a commercial one-step RT-PCR kit, CCHF Virus One-Step RT-qPCR Kit (NZYtech, Lisboa, Portugal) [24]. This assay enables simultaneous reverse transcription and amplification in a single reaction tube, designed to detect a wide range of CCHFV genetic variants with high specificity. According to the manufacturer's documentation, the test targets conserved regions of the viral genome and has a detection limit of approximately 10 copies per reaction. Real-time PCR reactions were performed using the CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) according to the manufacturer's protocol. Samples with a Ct (cycle threshold) value below 40 in the FAM channel were considered positive.
Statistical analyses were performed using Statistica v.14 (TIBCO Software Inc., Palo Alto, CA, USA, 2020) and MedCalc Odds Ratio Calculator v.23 (MedCalc Software Ltd., Ostend, Belgium, 2025). Exact binomial 95% confidence intervals (CIs) were calculated for species-specific prevalence estimates. Descriptive statistics were used to summarise seropositivity by species, age group, sex, and geographical region. Categorical variables were compared using Pearson's chi-square test or Fisher's exact test when expected frequencies were low. Odds ratios (ORs) with 95% CIs were calculated to describe associations between potential risk factors and seropositivity. For sheep, logistic regression with age as a continuous variable was applied to assess the effect of age on seropositivity. A significance level of p < 0.05 was used.
## 3. Results
## 3.1. Serological Results in Different Animal Species
Out of 1473 serum samples tested, 109 were positive for CCHFV nucleocapsid antibodies. These results represent the first serological evidence of CCHFV infection in Croatia. Among the species tested, the highest seroprevalence was observed in sheep at 28.3% (95%CI 23.5-33.4), followed by 4.3% in horses (95%CI 2.3-7.2), and 0.5% in domestic cats (95%CI 0.01-2.8). All samples collected from cattle (n = 276), goats (n = 57), and dogs (n = 304) tested negative (Table 2).
## 3.2. Geographic Distribution of Seropositive Animals
A total of 788 animals were tested from inland Croatia, and 685 from coastal and subcoastal regions. Only two samples, one feline and one equine, tested positive in the inland region (Eastern Croatia and Central Croatia). On the other hand, 107 samples comprising 95 sheep and 12 horses, tested positive in the coastal/subcoastal region (Littoral Croatia and Dalmatia) (Figure 2). Notably, all seropositive sheep were sampled in 2024 and originated from one county in Dalmatia (Zadar County). In this county, 184 animals were tested across 14 farms. Eight farms were negative, while six had multiple seropositive cases. Within the positive farms, the intra-farm positivity rate ranged from 63.6% to 85.7% (Figure 3). In contrast, all sheep tested from the Dubrovnik-Neretva County (n = 152) in Dalmatia were seronegative.
## 3.3. Association of Age and Sex with Seroprevalence in Sheep and Horses
Due to a small number of positive cats, age-related analysis was performed only for sheep from Zadar County and horses.
Among the 184 sheep samples, 16 were lambs (<1 year), 17 were juveniles (1-2 years), and 151 were adults (>2 years). Seroprevalence rates were 6.3% (1/16; 95% CI: 0.2-30.2%) in lambs, 11.8% (2/17; 95% CI: 1.5-36.4%) in juveniles, and 60.9% (92/151; 95% CI: 52.5-68.9%) in adults. A Chi-square test revealed significantly higher seroprevalence in adult sheep compared to younger animals (χ 2 = 29.24; p < 0.001). When age was considered as a continuous variable, the likelihood of seropositivity increased by approximately 1.35 times for every year of age (OR = 1.35; 95%CI: 1.21-1.53). Information on sex was available for 183 sheep, with six males and 177 females. Among these, one male and 94 females tested positive. Although more females were seropositive, the difference was not significant (p = 0.65, Fisher's exact test).
In horses, age was available for 196 animals, including eight foals (<1 year), 46 juveniles (1-3 years), 104 adults (4-14 years), and 38 seniors (>14 years. Seropositivity was highest among foals (2/8; 25.0%), followed by juveniles (5/46; 10.9%), and lower in adults (4/104; 3.8%) and seniors (1/38; 2.6%). Pairwise comparison using Fisher's exact test did not reveal statistically significant differences. Information on sex was available for 226 horses: 152 females and 74 males. Among these, seven females and six males tested positive. No significant influence of sex on test outcome was found (OR = 0.55; 95% CI: 0.18-1.69; p = 0.362).
## 3.4. Molecular Testing
All seropositive samples were subsequently tested by RT-PCR to determine the presence of CCHFV RNA. All samples tested negative.
## 4. Discussion
In the past decade, CCHFV has continued to spread across the Balkan countries [7]. Although human cases were not confirmed in all countries or areas, CCHFV circulation was reported based on serological screening of animal sera. Romania and Bulgaria regularly report high seroprevalence rates in small ruminants and cattle, even in areas without documented human cases, indicating the silent circulation of the virus [25,26]. Of particular concern for Croatia is Bosnia and Herzegovina, which shares a long natural border. This country reported its first seropositive animals in 2022 through retrospective testing of sheep samples collected in 2018 [15]. Subsequent testing of other domestic species, including cattle, also yielded seropositive results [27], further supporting evidence of undetected viral circulation. In addition, a recent study from Hungary found the highest livestock seroprevalence (1.8%) in the south-central region, close to the Croatian border [28].
Being surrounded by countries with confirmed endemicity and the presence of competent vectors, Croatia was considered a high-risk area for disease emergence. Considering the Hungarian border to the northeast, a targeted serosurvey of sheep in continental Croatia was previously conducted [29]. All tested animals were seronegative; however, they originated from areas where Hyalomma ticks have not yet been reported.
This study expanded surveillance to coastal and subcoastal regions of the country, where the presence of Hyalomma ticks had already been documented, resulting in the first serological evidence of CCHFV infections in Croatia. The highest seroprevalence was observed in sheep (28.3%, 95% CI 23.5-33.4), followed by horses (4.3%, 95% CI 2.3-7.2), and a single domestic cat (0.5%, 95% CI 0.01-2.8). These results align with trends reported in other endemic countries such as Bulgaria and Turkey, where sheep often show higher seroprevalence rates compared to other livestock species [30,31]. However, unlike those studies, which also documented positive goats and high seroprevalence in cattle, no seropositive cattle or goats were detected here, likely due to the sampling of these species solely from inland regions.
While seroprevalence in horses is less frequently studied, varying rates have been reported from endemic areas, including 16.7% in Benin and as high as 70.3% in Senegal [7,20]. Horses have proven to be highly effective sentinel animals for monitoring CCHFV circulation in regions where the virus is emerging [32], as was the case in our study. Wide distribution, outdoor housing, and documented strong humoral immune responses make them valuable sentinels for early detection of pathogens. Recent immunogenetic research has shown that horses possess unusually high antibody diversity due to their unique immune gene structure, which enables an immune response to low-level and sporadic exposure [33]. This may be important for identifying virus circulation in areas where vector abundance is low, thus providing an early warning system in regions considered at risk but not yet endemic. Unlike livestock species that may serve as amplifying hosts, horses typically develop detectable, long-lasting antibodies without substantial viremia, making them excellent indicators of environmental viral exposure without contributing significantly to onward transmission [34].
Interestingly, the aforementioned Senegalese study also reported a 6.9% seroprevalence in domestic dogs, a species that rarely tests positive for CCHFV and was negative in our survey. Evidence of infection in domestic cats is even scarcer. A recent study from Namibia reported low seroprevalence in cats (1.7%) but a higher rate in dogs (11.5%), attributing this to the free-roaming behaviour of dogs in that setting [21]. In contrast, dog ownership in Croatia is tightly regulated, with few free-roaming animals, which likely limits their exposure to infected ticks. As frequently described in the literature, differences in seroprevalence across species may reflect not only variations in host susceptibility and immune response but also differences in animal management practices, vector exposure, and the regional distribution of competent tick species [35].
The geographic distribution of seropositive cases in this study was markedly uneven. Cattle, sheep, and goat samples were available from limited areas, but surveillance in horses and pet animals involved the whole country. Only two seropositive animals, one horse and one cat, were detected among 788 animals tested in continental Croatia. The strong spatial clustering of cases suggests that the absence of seropositivity in certain areas can be more closely linked to the limited presence of competent tick vectors rather than host species distribution. These isolated findings may represent sporadic transmission events, potentially facilitated by less commonly implicated tick species, such as Rhipicephalus or Dermacentor, which are present in inland regions [18,36]. Alternatively, the undocumented movement of animals from endemic areas could explain these detections. For non-endemic but high-risk countries, surveillance strategies focused on areas with confirmed or ecologically plausible vector presence, rather than merely host density, may provide a more effective framework for early detection and prevention of CCHFV.
Interestingly, significant clustering was observed on a farm level. In Dalmatian Zadar County, six out of fourteen tested sheep farms had high and relatively uniform seropositivity rates, ranging from 63.6% to 85.7%. This finding suggests that, in addition to distribution and clustering of infected vectors, factors such as livestock management practices, pasture maintenance, and overall housing could play a role in facilitating CCHFV transmission. Similar patterns have been reported in other endemic regions. In Afghanistan, studies have highlighted the strong clustering of seropositive animals on specific farms, attributed to the communal use of pastureland, which led to increased tick exposure and virus transmission [37]. These observations underscore the importance of considering not only ecological but also anthropogenic factors when designing surveillance and control strategies for CCHFV. Interestingly, none of the 152 sheep tested from Dubrovnik-Neretva County were seropositive. As the southernmost Dalmatian county, it is likely to have suitable conditions for CCHFV circulation, including the presence of competent tick vectors. Differences in animal husbandry may explain the absence of seropositive results compared to Zadar County, which could include more frequent housing, less use of shared pastures, or generally lower exposure to tick habitats.
Studies in endemic areas consistently found that older animals had significantly higher odds of CCHFV seropositivity, likely due to cumulative exposure to tick-infested pastures [37][38][39]. One study noted that infection rates markedly increased after two years of age, coinciding with when young animals began full-time grazing [40]. Our results in sheep align with these observations, where adults demonstrated markedly higher antibody prevalence compared to younger age groups. Sheep over two years of age were 23.4 times more positive than lambs and 11.7 times more positive than juveniles up to the age of two. This supports the hypothesis that prolonged environmental exposure increases the likelihood of infection. Although seroprevalence in tested horses appeared to be highest in younger age groups, particularly foals and juveniles, statistical analysis did not show significance. This is likely influenced by the small number of seropositive animals (n = 13), which limited the power to detect differences between groups.
Although RT-PCR detected no viral RNA in any of the seropositive animals, the serological results provide strong evidence for prior exposure and silent viral circulation in Croatia. Antibody detection, particularly in asymptomatic animals, is a well-established early indicator of virus presence and remains a crucial surveillance tool for pathogens like CCHFV, which cause transient viremia and subclinical infection in most animal hosts. However, in regions where antigenically related nairoviruses (e.g., Aigai virus) co-circulate, like the southern Balkans, some degree of serological cross-reactivity cannot be excluded [41,42]. Although the Aigai virus has not been reported in Croatia or neighbouring countries to date, its circulation in more distant parts of the Balkans suggests that such cross-reactivity remains possible. Additional serological testing, as well as molecular confirmation with follow-up sequencing, would be valuable for differentiating closely related viruses. These findings provide a basis for future targeted vector studies and emphasise the importance of maintaining an integrated, species-diverse surveillance system aligned with the One Health approach.
## 5. Conclusions
This study provides the first serological evidence of CCHFV infection in Croatia. Antibodies were detected in sheep, horses, and a cat, while all tested cattle, goats, and dogs were seronegative. The seroprevalence was highest in sheep, particularly among older individuals. Most seropositive animals were located in coastal and subcoastal regions where the principal vector, Hyalomma marginatum, is established. These findings demonstrate the silent circulation of CCHFV among multiple animal species in Croatia, confirming the value of a multispecies surveillance approach. However, molecular confirmation of the virus in animals or vectors is still lacking.
## 6. Research Limitations
This study has several limitations. The distribution of sampled animals varied by species, with cattle and goats originating exclusively from continental regions, and sheep restricted to two coastal counties. This limited our ability to compare seroprevalence across several species and areas. The small number of seropositive animals in some species, particularly cats and horses, restricted the statistical power of analyses between groups, especially when determining the influence of age and sex. The potential co-circulation of some antigenically related nairoviruses, specifically Aigai virus, raises the possibility that detected antibodies may, in part, reflect cross-reactive responses rather than actual CCHFV exposure, notably since molecular confirmation is lacking.
## References
1. Walker, Siddell, Lefkowitz et al. (2020) "Changes to virus taxonomy and the International Code of Virus Classification and Nomenclature ratified by the International Committee on Taxonomy of Viruses"
2. Whitehouse (2004) "Crimean-Congo hemorrhagic fever" *Antivir. Res*
3. Ergönül (2006) "Crimean-Congo haemorrhagic fever" *Lancet Infect. Dis*
4. Spengler, Bergeron, Spiropoulou (2019) "Crimean-Congo hemorrhagic fever and expansion from endemic regions" *Curr. Opin. Virol*
5. Messina, Pigott, Golding et al. (2015) "The global distribution of Crimean-Congo hemorrhagic fever" *Trans. R. Soc. Trop. Med. Hyg*
6. Belobo, Kenmoe, Kengne-Nde et al. (2021) "Worldwide epidemiology of Crimean-Congo Hemorrhagic Fever Virus in humans, ticks and other animal species, a systematic review and meta-analysis" *PLoS Negl. Trop. Dis*
7. Dreshaj, Ahmeti, Ramadani et al. (2016) "Current situation of Crimean-Congo hemorrhagic fever in Southeastern Europe and neighboring countries: A public health risk for the European Union" *Travel Med. Infect. Dis*
8. Yessinou, Farougou, Olopade et al. (2025) "Seroprevalence and risk factors of Crimean-Congo hemorrhagic fever exposure in wild and domestic animals in Benin" *Viruses*
9. Spengler, Bergeron, Rollin (2016) "Seroepidemiological studies of Crimean-Congo hemorrhagic fever virus in domestic and wild animals" *PLoS Negl. Trop. Dis*
10. Mustafa, Ayazi, Mohareb et al. (1940) "Crimean-Congo hemorrhagic fever" *Emerg. Infect. Dis*
11. Mertens, Schmidt, Ozkul et al. (2013) "The impact of Crimean-Congo hemorrhagic fever virus on public health" *Antivir. Res*
12. Akuffo, Brandful, Zayed et al. (2016) "Crimean-Congo hemorrhagic fever virus in livestock ticks and animal handler seroprevalence at an abattoir in Ghana" *BMC Infect. Dis*
13. Monsalve-Arteaga, Alonso-Sardón, Bellido et al. "Seroprevalence of Crimean Congo hemorrhagic fever in humans in the World Health Organization European region: A systematic review" *PLoS Negl*
14. Šatrović, Softić, Žuko et al. (2022) "First evidence of Crimean Congo haemorrhagic fever virus circulation in Bosnia and Herzegovina" *Vet. Med. Sci*
15. Portillo, Palomar, Santibáñez et al. "Epidemiological aspects of Crimean Congo hemorrhagic fever in Western Europe: What about the future? Microorganisms 2021"
16. Shahhosseini, Wong, Babuadze et al. (1907) "Crimean-Congo hemorrhagic fever virus in Asia" *Africa and Europe. Microorganisms*
17. Krčmar (2012) "Hard ticks (Acari, Ixodidae) of Croatia" *Zookeys*
18. "A framework to substantiate absence of disease: The risk based estimate of system sensitivity tool (RiBESS)"
19. Mangombi, Roqueplo, Sambou et al. (2020) "Seroprevalence of Crimean-Congo hemorrhagic fever in domesticated animals in northwestern Senegal. Vector Borne Zoonotic Dis"
20. De Villiers, Molini, Van Zyl et al. (2025) "Serological evidence of Crimean-Congo haemorrhagic fever in domestic animals from eight regions of Namibia" *Acta Trop*
21. Id, Id Screen, Cchf Double Antigen et al. (2017) "Instructions for Use; Version CCHFDA ver 0917 EN; ID.vet: Grabels"
22. Sas, Comtet, Donnet et al. (2018) "A novel double-antigen sandwich ELISA for the species-independent detection of Crimean-Congo hemorrhagic fever virus-specific antibodies" *Antivir. Res*
23. Nzytech (2025) "Instructions for Use (IFU) MD03481_IFU_en_V2501: Crimean-Congo Haemorrhagic Fever Virus One Step RT qPCR Kit"
24. Ceianu, Panculescu-Gatej, Coudrier et al. (2012) "First serologic evidence for the circulation of Crimean-Congo hemorrhagic fever virus in Romania. Vector Borne Zoonotic Dis"
25. Christova, Panayotova, Groschup et al. (2018) "High seroprevalence for Crimean-Congo haemorrhagic fever virus in ruminants in the absence of reported human cases in many regions of Bulgaria" *Exp. Appl. Acarol*
26. Goletić, Šatrović, Softić et al. (2022) "Serologic and molecular evidence for circulation of Crimean-Congo hemorrhagic fever virus in ticks and cattle in Bosnia and Herzegovina" *Ticks Tick Borne Dis*
27. Deézsi-Magyar, Dénes, Novák et al. "First broad-range serological survey of Crimean-Congo hemorrhagic fever among Hungarian livestock" *Viruses*
28. Barbić, Stevanović, Maurić Maljković et al. (2025) "Vilibić Cavlek, T. Seroprevalence study for selected zoonotic vector borne pathogens in sheep from endemic areas of Croatia" *Front. Vet. Sci*
29. Barthel, Mohareb, Younan et al. (2014) "Seroprevalance of Crimean-Congo haemorrhagic fever in Bulgarian livestock" *Biotechnol. Biotechnol. Equip*
30. Nurettin, Engin, Sukru et al. "The seroprevalence of Crimean Congo hemorrhagic fever in wild and domestic animals: An epidemiological update for domestic animals and first seroevidence in wild animals from Türkiye" *Vet. Sci*
31. Celina, Italiya, Tekkara et al. "Crimean-Congo haemorrhagic fever virus in ticks, domestic, and wild animals" *Front. Vet. Sci*
32. Navas, Manso, Martins et al. (2022) "The major role of junctional diversity in the horse antibody repertoire" *Mol. Immunol*
33. Spengler, Estrada Peña, Garrison et al. (2016) "A chronological review of experimental infection studies of the role of wild animals and livestock in the maintenance and transmission of Crimean-Congo hemorrhagic fever virus" *Antivir. Res*
34. Nasirian, Pouriayevali (2019) "Crimean-Congo hemorrhagic fever (CCHF) seroprevalence: A systematic review and meta analysis" *Antivir. Res*
35. Tsapko, Volynkina, Evchenko et al. (2022) "Detection of Crimean-Congo hemorrhagic fever virus in ticks collected from South Russia" *Ticks Tick Borne Dis*
36. Raheemi, Afsheen, Abbas et al. "Serosurveillance of Crimean-Congo hemorrhagic fever virus antibodies in livestock as a reservoir for human infection in Afghanistan"
37. Munyua, Odoyo, Mulwa et al. "Seroepidemiology of Crimean-Congo hemorrhagic fever virus (CCHFV) in cattle across three livestock pastoral regions in Kenya" *Dairy*
38. Schulz, Barry, Stoek et al. "Crimean-Congo hemorrhagic fever virus antibody prevalence in Mauritanian livestock (cattle, goats, sheep and camels) is stratified by the animal's age" *PLoS Negl*
39. Sidiq, Abdelrahman, Salih et al. (2021) "A seroepidemiological survey of Crimean-Congo hemorrhagic fever among cattle in North Kordufan State" *Sudan. Trop. Anim. Health Prod*
40. Papa, Marklewitz, Paraskevopoulou et al. "History and classification of Aigai virus (formerly Crimean-Congo haemorrhagic fever virus genotype VI)" *J. Gen*
41. Maze, Booth, Limon et al. (2025) "Serological cross-reactivity between Crimean-Congo haemorrhagic fever virus and Nairobi sheep disease virus glycoprotein C" *Front. Immunol*
42. "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=PMC12548464&blobtype=pdf | # Hepatitis C virus NS3/4A protease cleaves SPG20, a key regulator of lipid droplet turnover, to promote lipid droplet formation
Chieko Matsui, Putu Yuliandari, Lin Deng, Takayuki Abe, Ikuo Shoji
## Abstract
Hepatitis C virus (HCV) assembles in close proximity to lipid droplets (LDs), which play important roles in HCV RNA replication. HCV infection often causes the accumulation of large LDs in hepatocytes. However, the molecular mechanism under lying HCV-induced large LD formation is poorly understood. It has been reported that the SPG20/Spartin protein associates with the LD surface and plays a crucial role in LD turnover by recruiting the ubiquitin ligase Itch to promote the ubiquitindependent degradation of adipophilin (ADRP), which protects LDs from lipase-mediated degradation. To elucidate the mechanism underlying HCV-induced large LD formation, we investigated the SPG20 protein's role in LD formation in HCV J6/JFH1-infected Huh-7.5 cells. Immunoblot analysis revealed that HCV infection promoted SPG20 protein cleavage. Transfection of increasing amounts of NS3/4A, but not the inactive NS3/4A mutant, resulted in SPG20 cleavage, implicating the NS3/4A protease in this cleavage. Site-directed mutagenesis suggested that the NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 . The SPG20 protein was co-immunoprecipitated with the LD-attached protein TIP47. Increasing amounts of NS3/4A protease, but not inactive NS3/4A, decreased the co-precipitation of SPG20 with TIP47. The siRNA-mediated knockdown of Itch in Huh-7.5 cells restored ADRP levels, suggesting that Itch mediates ubiquitylation-depend ent ADRP degradation. Immunofluorescence staining of HCV-infected cells revealed that ADRP was localized mainly around LDs in HCV-infected cells, whereas cytosolic ADRP was decreased. We propose that the HCV NS3/4A protease specifically cleaves SPG20 and inhibits Itch-mediated ubiquitin-dependent degradation of LD-associated ADRP, thereby promoting the formation of large LDs. IMPORTANCE HCV infection often promotes the formation of large LDs in HCV-infec ted cells. However, the molecular mechanism underlying large LD formation is poorly understood. LD turnover is regulated by SPG20, Itch E3 ligase, and ADRP. To elucidate the mechanism underlying the formation of large LDs induced by HCV infection, we investigated the roles of SPG20, Itch, and ADRP in large LD formation. The HCV NS3/4A protease specifically cleaves SPG20 and disrupts Itch recruitment to LD-associated ADRP. Therefore, LD-associated ADRP can escape and protects LDs from lipase-mediated degradation, thereby promoting LD growth. We propose that HCV NS3/4A proteasemediated cleavage of SPG20 contributes to a previously uncharacterized mechanism underlying the formation of large LDs in HCV-infected cells. These findings may lead to a better understanding of how the virus forms large LDs in infected cells. KEYWORDS hepatitis C virus, NS3/4A protease, SPG20, lipid droplet, Itch, adipophilin H epatitis C virus (HCV) is a positive-sense, single-stranded RNA virus belonging to the Hepacivirus genus of the Flaviviridae family (1). HCV infection often causes chronic hepatitis, liver cirrhosis, and hepatocellular carcinoma (HCC). The HCV genome consists of a 9.6 kb RNA encoding a polyprotein of 3,010 amino acids (aa). The polyprotein
is cleaved co-translationally and post-translationally into at least 10 proteins by viral proteases and cellular signalases. The viral proteins required for RNA replication include NS3, NS4A, NS4B, NS5A, and NS5B. The NS3/4A serine protease is responsible for the processing of HCV proteins (1). Highly potent direct-acting antivirals (DAAs) against the NS3/4A protease, NS5A protein, and NS5B polymerase have been developed, representing a breakthrough in HCV therapy (2). DAA therapy results in a >90% sustained virological response (SVR) rate. However, the occurrence of HCC following SVR in patients has been reported, and the molecular mechanism underlying post-SVR HCC is poorly understood (3).
Fatty liver is observed more often in hepatitis C patients than in the general population or in hepatitis B patients (4). These clinical findings suggest that HCV infection may directly cause fatty changes in patients' hepatocytes. However, the molecular mechanisms underlying HCV-induced fatty changes are poorly understood (5,6). Lipid droplets (LDs) are intracellular organelles that consist of cholesterol esters and triacylglycerol. LDs in hepatocytes play important roles in HCV RNA replication and the production of infectious HCV (7)(8)(9). LDs are associated with phospholipid membranes and the PAT family proteins. The PAT family of protein consists of perilipin (PLIN1), adipophilin (ADRP/PLIN2), tail-interacting protein of 47 kDa (TIP47/PLIN3), S3-12 (PLIN4), and MLDP (PLIN5/OXPAT) (10,11). These proteins are important for the biogenesis, stabilization, and degradation of LDs (12). Among them, ADRP is the most abundant LD-associated protein in hepatocytes (13). SPG20, which is also known as Spartin, plays an important role in LD turnover (14)(15)(16). SPG20 was initially identified as the target gene of autosomal recessive hereditary spastic paraplegia in Troyer syndrome (17). Frameshift mutations in the SPG20 gene cause Troyer syndrome. SPG20 harbors three domains: a microtubule-interacting and trafficking domain (MIT), a plant-senescence domain (PSD), and a ubiquitin-binding region (UBR) (18). The SPG20 PSD domain is responsible for its association with LDs (14). SPG20 interacts with several Nedd4 family ubiquitin ligases, such as atrophin-1-inter acting protein 4 (AIP4/Itch) (15,19), atrophin-1-interacting protein 5 (AIP5/WWP1) (14,15), and atrophin-1-interacting protein 2 (AIP2/WWP2). Nedd4 family ubiquitin ligases possess an N-terminal C2 domain responsible for subcellular localization and two to four WW domains that are important for binding to the substrates (20). The association of SPG20 with an adaptor protein promotes ubiquitin ligase activity (21). SPG20 interacts with Itch via protein-protein interactions between the WW domain of Itch and the PPxY motif of SPG20. SPG20 recruits Itch to ADRP, resulting in the ubiquitylation and degradation of ADRP. ADRP protects LDs from lipase-mediated degradation, whereas degradation of ADRP results in disruption of growth of LDs (22). We previously reported that HCV infection induces JNK activation (23,24). We also reported that HCV-induced JNK activation results in Itch phosphorylation and activation (25).
In this study, we sought to identify the previously uncharacterized mechanism underlying the formation of large LDs in HCV-infected cells. We provided evidence that the HCV NS3/4A protease specifically cleaves SPG20 and inhibits the Itch-medi ated ubiquitin-dependent degradation of LD-associated ADRP, thereby promoting the formation of large LDs.
## RESULTS
## HCV infection increases the size and diameter but not the number of LDs
To examine the size and number of LDs in HCV-infected cells, we performed immuno fluorescence staining using a BODIPY lipid probe. Immunofluorescence staining revealed that the LDs in HCV J6/JFH1-infected Huh-7.5 cells (Fig. 1A, 2nd panel, left) were larger than those in mock-infected cells (Fig. 1A, first panel, left). To determine the role of HCV infection in LD formation, we treated the cells with the NS3/4A protease inhibitor VX950. Treatment of the cells with VX950 resulted in a reduction in LD surface area (Fig. 1A, third panel, left). These results suggest that HCV infection increases LD surface area. We used ImageJ software to measure LD surface area and diameter. We observed that the LDs in HCV-infected cells were larger and had greater diameters than those in mockinfected cells (Fig. 1B through D). On the other hand, treatment of the cells with the NS3/4A inhibitor VX950 reduced LD surface area and diameter. We counted the LDs in HCV-infected cells and in mock-infected cells via a BODIPY lipid probe under a micro scope. The number of LDs per cells was not increased in HCV-infected cells (Fig. 1E). These results suggest that HCV infection increases the size and diameter of LDs, but not their number.
## HCV NS3/4A protease activity is responsible for the cleavage of SPG20
To determine the effects of HCV-infection on the SPG20 protein, which is known as a key regulator of LD turnover, we performed immunoblot analysis of the SPG20 protein in HCV-infected Huh-7.5 cells. The SPG20 band decreased, and the short form of SPG20 was detected in HCV-infected cells at 3-, 5-, and 7-days post-infection (Fig. 2A, upper panel, lanes 6, 8, and 10). This result suggested that SPG20 was cleaved in HCV-infected cells.
To examine whether HCV proteases are involved in the cleavage of SPG20, we examined the effects of either the HCV NS2 protease or the NS3/4A protease on the cleavage of SPG20. We transfected cells with pCAG-HA-SPG20 together with either pCAG-FLAG-NS2 or pCAG-FLAG-NS2-3ΔC. Immunoblot analysis revealed that FLAG-NS2 was processed in the cells transfected with pCAG-FLAG-NS2-3ΔC, suggesting that the NS2 protease was active and cleaved NS2-3ΔC at the NS2/NS3 cleavage site (Fig. 2B, lower panel, lane 3). However, the short form of SPG20 was not detected in the presence of the HCV NS2 protease (Fig. 2B, upper panel, lane 3), suggesting that the HCV NS2 protease is not responsible for the cleavage of SPG20. To determine whether the NS3/4A protease is involved in the cleavage of SPG20, Huh-7.5 cells were transfected with FLAG-SPG20 and increasing amounts of HA-NS3/4A. The overexpression of HA-NS3/4A resulted in the appearance of short forms of SPG20, suggesting that the HCV NS3/4A protease plays a role in the cleavage of SPG20 (Fig. 2C, upper panel, lane 3). To determine whether SPG20 is cleaved by the NS3/4A protease, FLAG-SPG20 was co-expressed with either HA-NS3/4A or the inactive mutant HA-NS3(S139A)/4A in Huh-7.5 cells. Immunoblot analysis demonstrated that cleavage of SPG20 occurred only in the presence of wild-type NS3/4A and not in the presence of HA-NS3(S139A)/4A (Fig. 2D, upper panel, lanes 3 and 4). These results suggest that HCV NS3/4A protease activity is responsible for the cleavage of SPG20.
## HCV-induced cleavage of SPG20 is inhibited by the NS3/4A protease inhibitor
To determine whether HCV infection results in the cleavage of SPG20, we examined the effects of either the specific NS3/4A protease inhibitor VX950 or the NS5A inhibitor daclatasvir (DCV) on the cleavage of SPG20. Huh-7.5 cells were infected with HCV J6/ JFH1 with or without either VX950 treatment or DCV treatment. When the cells were treated with VX950, the cleaved form of SPG20 disappeared (Fig. 2E, upper panel, lane 6). We also treated the cells with the NS5A inhibitor DCV. Treatment of HCV-infected cells with DCV reduced the level of cleaved SPG20 (Fig. 2E, upper panel, lane 8). These results suggest that HCV infection is, indeed, responsible for the cleavage of SPG20. We then examined the cleavage of exogenous SPG20 by NS3/4A in a transient expression system. Huh-7.5 cells were co-transfected with the HA-NS3/4A plasmid and the FLAG-SPG20 plasmid, and the cells were treated with the NS3/4A protease inhibitor VX950 or TMC435. The cleavage of exogenous SPG20 was inhibited by either VX950 or TMC435 (Fig. 2F, upper panel, lanes 7, 8, and 9), whereas DCV had no inhibitory effect on the cleavage of SPG20 (Fig. 2F, upper panel, lane 10). These results indicate that HCV-induced cleavage of SPG20 is dependent on NS3/4A protease activity.
## NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562
To determine the cleavage sites on SPG20 by the NS3/4A protease, we introduced substitution mutations at possible cleavage sites on SPG20. There were at least four possible cleavage sites on SPG20 by HCV NS3/4A, considering the consensus sequence of the NS3/4A protease cleavage site (Fig. 3B, bold, underlined). To detect cleaved SPG20, we constructed a plasmid expressing SPG20 as a C-terminally FLAG-tagged protein. The cells expressed either SPG20-FLAG or HA-NS3/4A, and the cell lysates were immunopreci pitated with FLAG M2 beads. Immunoprecipitation analysis revealed two cleaved SPG20 bands with molecular masses of approximately 22 kDa and 19 kDa (Fig. 3C, upper panel, lane 4, arrows) in the presence of HCV NS3/4A but not in the presence of HCV NS3(S139A)/4A (Fig. 3C, upper panel, lane 5).
We generated a series of SPG20-FLAG plasmids carrying a substitution mutation at a possible cleavage site. Immunoprecipitation analysis revealed that the 22 kDa fragment of SPG20-FLAG was decreased in the cells expressing SPG20 C504A-FLAG (Fig. 3D, upper panel, lane 4). In addition, the 19 kDa fragment of SPG20-FLAG was markedly decreased in the cells expressing SPG20 C562A-FLAG (Fig. 3D, upper panel, lane 5). On the other hand, two cleaved bands of SPG20 were detected in other mutants, namely, SPG20 C499A-FLAG and SPG20 T587A-FLAG (Fig. 3D, upper panel, lanes 3 and 6). These results suggest that the NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 . To further examine whether the NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 , we generated a plasmid expressing the double mutants SPG20 C504A/C562A-FLAG. Immunoprecipitation analysis revealed that cleaved SPG20-FLAG bands were not detected in the cells expressing C504A/C562A-FLAG (Fig. 3E, first panel, lane 4). These results indicate that the NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 .
## Ubiquitin ligase Itch mediates the polyubiquitylation of ADRP
We previously reported that HCV-induced JNK activation results in the phosphorylation of Itch (25). We examined the phosphorylation of Itch in HCV J6/JFH1-infected cells. The phosphorylation of Itch at Thr222 was increased in HCV-infected cells, indicating that HCV infection induces JNK activation and Itch activation (Fig. 4A, first panel, lanes 2 and 4).
To identify an E3 ligase for the ubiquitylation of ADRP in Huh-7.5 cells, Huh-7.5 cells were co-expressed with Myc-ADRP and a plasmid encoding HA-tagged ubiquitin together with either FLAG-Itch, FLAG-WWP1, or FLAG-WWP2. A cell-based ubiquitylation assay revealed that overexpression of FLAG-Itch promoted polyubiquitylation of ADRP protein (Fig. 4B, right, upper panel, lane 2). On the other hand, neither WWP1 nor WWP2 promoted polyubiquitylation of the ADRP protein (Fig. 4B, right, upper panel, lanes 3 and 4). In addition, siRNA-mediated knockdown of Itch inhibited the polyubiquitination of ADRP (Fig. 4C, right, upper panel, lane 4). Furthermore, overexpression of FLAG-Itch increased polyubiquitylation of ADRP, whereas the overexpression of the inactive mutant FLAG-Itch C868A failed to promote the polyubiquitylation of ADRP (Fig. 4D, right, upper panel, lanes 2 and 3). These results suggest that Itch specifically mediates the polyubiqui tuylation of ADRP in Huh-7.5 cells.
To determine whether Itch interacts with ADRP, Huh-7.5 cells were co-expressed with FLAG-Itch together with HA-ADRP. Immunoprecipitation analysis revealed that HA-ADRP was coimmunoprecipitated with FLAG-Itch using anti-HA PAb (Fig. 4E, fourth panel, lane 4). Conversely, immunoprecipitation analysis revealed that HA-ADRP was coimmunopre cipitated with FLAG-Itch using anti-FLAG MAb (Fig. 4F, fourth panel, lane 4). These results suggest that Itch interacts with ADRP in Huh-7.5 cells. We concluded that the ubiquitin ligase Itch mediates the polyubiquitylation of the ADRP protein in Huh-7.5 cells.
## HCV infection induces the ubiquitin-dependent proteasomal degradation of ADRP
To determine whether HCV infection affects the expression levels of the ADRP protein, we performed immunoblot analysis of the endogenous ADRP protein in HCV J6/JFH1infected Huh-7.5 cells. Immunoblot analysis revealed that the protein level of ADRP was markedly lower in HCV-infected cells than in mock-infected control cells (Fig. 5A, first panel, lanes 2, 4, and 6). On the other hand, TIP47 protein levels did not change (Fig. 5A, second panel, lanes 2, 4, and 6). The proteasome inhibitor MG132 restored the protein level of ADRP (Fig. 5B, first panel, lanes 6 and 8). These results suggest that HCV-infection promotes proteasomal degradation of the ADRP protein.
To determine whether Itch plays a role in HCV-induced proteasomal degradation of the ADRP protein, Itch was knocked down by siRNA (Fig. 5C). siRNA-mediated knock down of Itch restored the level of ADRP in HCV-infected cells (Fig. 5C, first panel, lane 4). These results suggest that HCV infection promotes Itch-mediated ubiquitin-dependent degradation of ADRP.
To determine whether HCV is involved in the degradation of ADRP, we assessed the effects of the NS3/4A protease inhibitor VX950 or the NS5A inhibitor DCV on the levels of ADRP in HCV-infected cells. Treatment of HCV-infected cells with either VX950 or daclatasvir restored the protein levels of ADRP (Fig. 5D, first panel, lanes 6 and 8). These results suggest that HCV infection induces the degradation of ADRP.
## ADRP accumulates around LDs in HCV-infected cells
To examine the subcellular localization of ADRP in HCV-infected Huh-7.5 cells, we performed immunofluorescence staining with a BODIPY lipid probe. In mock-infected cells, ADRP was localized in the cytoplasm (Fig. 6, upper panel, ADRP). However, ADRP was localized mainly on the surface of LDs in HCV-infected cells (Fig. 6, lower panel, Merge). These results suggest that HCV infection promotes the accumulation of ADRP on the surface of LDs.
## NS3/4A protease cleaves SPG20 and inhibits the SPG20-TIP47 interaction
We investigated why ADRP proteins on the surface of LDs are not degraded by the ubiquitin-proteasome pathway in HCV-infected cells even though the Itch-mediated ubiquitin-dependent degradation of ADRP is promoted. SPG20 is known to interact with TIP47 at the surface of LDs. The region ranging from aa 433 to aa 584 on SPG20 (Fig. 3A) was found to be important for the interaction with TIP47 (26). Therefore, we hypothesize that the NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 , resulting in the inhibition of the SPG20-TIP47 interaction, thereby disrupting the recruitment of Itch to LD-attached ADRP (Fig. 7A).
To determine whether SPG20 interacts with TIP47 in Huh-7.5 cells, we transfected cells with pCAG-SPG20-FLAG together with pCAG-HA-TIP47. Immunoprecipitation analysis revealed that HA-TIP47 was coimmunoprecipitated with SPG20-FLAG using anti-FLAG MAb (Fig. 7B, right panel, fourth panel, lane 4). Conversely, immunoprecipitation analysis revealed that SPG20-FLAG was coimmunoprecipitated with HA-TIP47 using anti-TIP47 PAb (Fig. 7C, fourth panel, lane 4). These results suggest that SPG20 interacts with TIP47 in Huh-7.5 cells.
To further confirm that the NS3/4A protease cleaves SPG20 and inhibits SPG20-TIP47 interaction, we performed immunoprecipitation analysis. Immunoprecipitation analysis revealed that the SPG20-FLAG was not co-immunoprecipitated with HA-TIP47 in the presence of wild-type NS3/4A (Fig. 7D, second panel, lane 3). However, SPG20-FLAG was co-immunoprecipitated with HA-TIP47 in the presence of inactive NS3/4A mutant (Fig. 7D, third panel, lane 4). We further investigated the interaction between endogenous TIP47 and transfected SPG20-FLAG. Immunoprecipitation using anti-TIP47 rabbit PAb revealed that SPG20-FLAG was co-precipitated with endogenous TIP47 (Fig. 7E, third panel, lane 2). On the other hand, in the presence of HA-HCV NS3/4A, SPG20-FLAG was not co-precipitated with endogenous TIP47 (Fig. 7E, third panel, lane 4). We further examined whether the TIP47-SPG20 interaction was restored when SPG20 C504A, which is not cleaved by NS3/4A, was used in HCV J6/JFH1-infected Huh-7.5 cells. The SPG20-TIP47 interaction was inhibited by HCV infection when SPG20-FLAG was used (Fig. 7F, third panel, lane 2), whereas the SPG20-TIP47 interaction was restored when SPG20 C504A-FLAG was used (Fig. 7F, third panel, lane 3). These results suggest that SPG20 cleavage by HCV NS3/4A inhibits the SPG20-TIP47 interaction.
These results suggest that the NS3/4A protease cleaves SPG20, thereby inhibiting the SPG20-TIP47 interaction.
Taken these results together, we propose a model in which the HCV NS3/4A protease cleaves SPG20 at Cys 504 and Cys 562 and inhibits the SPG20-TIP47 interaction, thereby inhibiting the recruitment of Itch to LD-attached ADRP, preventing the ubiquitindependent degradation of LD-attached ADRP and promoting the formation of large LDs (Fig. 8). and anti-GAPDH mouse MAb. The level of GAPDH served as a loading control. (B) Huh-7.5 cells were plated at 1.5 × 10 6 cells per 10 cm dish and cultured for 12 h. The cells were transfected with pCAG-myc-ADRP and either pCAG-FLAG-Itch, pCAG-FLAG-WWP1, or pCAG-FLAG-WWP2, together with a plasmid encoding HA-tagged ubiquitin. At 48 h after transfection, the cells were harvested. Cell lysates were immunoprecipitated with anti-c-Myc mouse MAb, and bound proteins were immunoblotted with anti-HA rabbit PAb, anti-c-Myc MAb, and anti-FLAG M2 mouse MAb. (C) Huh-7.5 cells were plated at a density of 1.5 × 10 6 cells per 10 cm dish and cultured for 12 h. The cells were transfected with either Itch-specific siRNA or universal negative control siRNA. At 24 h after siRNA transfection, cells were transfected with pCAG-myc-ADRP and pCAG-FLAG-Itch, together with a plasmid encoding HA-tagged ubiquitin. At 48 h after transfection, the cells were harvested. Cell lysates were immunoprecipitated with anti-c-Myc MAb, and bound proteins were immunoblotted with anti-HA PAb, anti-c-Myc MAb, and anti-FLAG MAb. (D) Huh-7.5 cells were plated at a density of 1.5 × 10 6 cells per 10 cm dish and cultured for 12 h. The cells were transfected with pCAG-Myc-ADRP and either pCAG-FLAG-Itch, or pCAG-FLAG-Itch C868A, together with a plasmid encoding HA-tagged ubiquitin. At 48 h after transfection, the cells were harvested. The cell lysates were immunoprecipitated with anti-c-Myc MAb, and the bound proteins were immunoblotted with anti-HA PAb, anti-c-Myc Mab and anti-FLAG MAb. (E) Huh-7.5 cells were plated at 1.5 × 10 6 cells per 10 cm dish and cultured for 12 h. The cells were transfected with pCAG-FLAG-Itch together with pCAG-HA-ADRP. At 48 h after transfection, the cells were harvested. Cell lysates were immunoprecipitated with anti-HA PAb or normal rabbit IgG, and bound proteins were immunoblotted with the indicated antibodies. (F) Cell lysates were immunoprecipitated with anti-FLAG MAb or normal mouse IgG, and bound proteins were immunoblotted with the indicated antibodies. The immunoblots are representative of three independent experiments that yielded similar results.
## The asterisk indicates nonspecific bands (B-D).
## DISCUSSION
LDs play important roles in HCV RNA replication and the production of HCV particles (6,27). HCV infection often causes the accumulation of large LDs in hepatocytes. In this study, we aimed to elucidate the molecular mechanism underlying large LD formation induced by HCV infection. Here, we demonstrated that HCV infection promoted the formation of large LDs (Fig. 1A through D). However, the number of LDs per cell was not increased in HCV-infected cells (Fig. 1E). We also demonstrated that SPG20 was cleaved in HCV-infected cells (Fig. 2A). SPG20 was cleaved in the presence of the HCV NS3/4A protease (Fig. 2C) but not the HCV NS2 protease (Fig. 2B). An inactive mutant of the HCV NS3/4A protease abolished the cleavage of SPG20 (Fig. 2D). These findings suggest that The immunoblots are representative of three independent experiments that yielded similar results.
the HCV NS3/4A protease is responsible for the cleavage of SPG20. We detected only one specific cleaved form of SPG20 with anti-SPG20 goat PAb presumably due to the limitation of the antibody epitopes (Fig. 2A). We could detect several forms of SPG20 using either FLAG-SPG20 or SPG20-FLAG following the detection with anti-FLAG MAb (Fig. 2C through F).
On the basis of the consensus sequence for the HCV NS3/4A protease cleavage site, there were at least four possible cleavage sites on SPG20 by the HCV NS3/4A protease (Fig. 3B). Site-directed mutagenesis revealed that Cys 504 and Cys 562 of SPG20 are impor tant for cleavage by the HCV NS3/4 protease (Fig. 3D andE). Cys 504 and Cys 562 of SPG20 reside within the SPG20-TIP47 interaction region. Therefore, we hypothesized that HCV NS3/4A cleaves SPG20 at Cys 504 and Cys 562 , thereby inhibiting the interaction between SPG20 and TIP47. The NS3 protein has been detected at the LD surface (8,28), consistent with the notion that the HCV NS3/4A protease cleaves the LD-associated protein SPG20.
ADRP plays an important role in LD biogenesis. SPG20 is known to play a key role in the turnover of lipid droplets (14,16). A recent study (29) demonstrated that SPG20 functions as a lipophagy receptor. SPG20 inhibition in cultured human neurons or murine brain neurons enhanced LD formation, suggesting that impaired LD metabolism is involved in the development of Troyer syndrome.
Nedd4 family E3 ligases contain WW domain (30). This domain is responsible for substrate recognition via the PPxY motif. Hopper et al. (16) reported that SPG20 interacts with the WW domains of Itch via the PPxY motif of SPG20. The interaction between SPG20 and Itch increases the enzymatic activity of the E3 ligase of Itch. SPG20 is an adapter of Itch for the polyubiquitylation of the LD-attached ADRP. We demonstrated that Itch promoted polyubiquitylation of ADRP in Huh-7.5 cells. However, neither WWP1 nor WWP2 promoted the polyubiquitylation of ADRP (Fig. 4B). ADRP was co-immunopre cipitated with Itch (Fig. 4E andF). These data suggest that Itch is responsible for the ubiquitin-dependent degradation of ADRP in Huh-7.5 cells.
Itch is directly activated by phosphorylation of the proline-rich region at S199, S232, and T222 via JNK protein kinase and conformational changes (31). Itch phosphorylation is important for the interaction between the WW domain and HECT domain of Itch, which leads to a conformational change from the closed inactive form to the open active form. This conformational change in Itch is important for the polyubiquitylation of its substrates. We previously reported that Itch promotes the release of HCV particles via polyubiquitylation of VPS4A, leading to the activation of the ROS/JNK/Itch/VPS4A signaling pathway for the release of infectious HCV particles (25). We confirmed that Itch activation with its phosphorylation at Thr222 was induced via JNK activation in HCVinfected cells (Fig. 4A).
The protein levels of ADRP were decreased in HCV J6/JFH1-infected cells (Fig. 5A). Treatment of the cells with MG132, a proteasome inhibitor, restored the protein levels of ADRP (Fig. 5B), suggesting that the protein level of ADRP is decreased in HCV-infected cells via the ubiquitin-proteasome pathway. In mock-infected cells, ADRP was localized pCAG-HA-TIP47. The cell lysates were immunoprecipitated with anti-TIP47 MAb, and the bound proteins were immunoblotted with the indicated antibodies.
(E) Huh-7.5 cells were transfected with pCAG-SPG20-FLAG together with or without pCAG-HA-NS3/4A as indicated. At 48 h after transfection, the cells were harvested. The cell lysates were immunoprecipitated with anti-TIP47 rabbit PAb (lanes 2 and 4) or isotype control rabbit IgG (lanes 1 and 3) as negative controls.
The bound proteins were immunoblotted with the indicated antibodies. (F) Huh-7.5 cells were transfected with the plasmids as indicated and then infected with HCV J6/JFH1 at an m.o.i. of 2. At 48 h after infection, the cells were harvested. The cell lysates were immunoprecipitated with anti-c-Myc MAb, and the bound proteins were immunoblotted with the indicated antibodies. The immunoblots are representative of three independent experiments that yielded similar results. in both the cytoplasm and the nucleus (Fig. 6). On the other hand, ADRP was co-local ized with the surface of LDs in HCV-infected cells. We provided evidence that NS3/4Amediated cleavage of SPG20 inhibited SPG20-TIP47 interaction, thereby disrupting the recruitment of Itch to LDs (Fig. 7). LDs are highly dynamic organelles with rapid turnover. ADRP protects LDs from lipase-mediated breakdown. Under normal condition, ADRP is ubiquitylated by Itch and gets degraded via the proteasome pathway, resulting in a decrease in the amount of LD-associated ADRP. ADRP-lacking LDs are sensitive to lipase-mediated breakdown and get smaller. However, under HCV-infection, HCV NS3/4A protease specifically cleaves SPG20 and inhibits the Itch-mediated ubiquitin-dependent degradation of LD-associated ADRP. Therefore, LD-associated ADRP protects LDs from lipase-mediated breakdown, resulting in the formation of large LDs (Fig. 8). LD-binding proteins are classified into two categories, as summarized by Kory et al. (32). Class I proteins, such as oleosin and glycerol-3-phosphate acyltransferase 4 (GPAT4), are translated at the ER and localize to both the ER and the LD surface. Class II proteins, such as PAT (perilipin/ADRP/TIP47)-domain proteins, harbor a central 11-mer repeat-con taining domain that is predicted to fold into amphipathic and hydrophobic helices upon membrane binding. The C-terminal four-helix bundle of ADRP and TIP47 participates in LD binding. In this study, we clarified the role of ADRP in the formation of large LDs.
Among the various HCV genotypes, genotype 3 is most strongly associated with hepatic steatosis and exerts most pronounced effects on LDs in hepatocytes. This has been consistently supported by both clinical observations and experimental evidence -particularly studies showing that core proteins from genotype 3 have a greater capacity to induce LD accumulation compared to those from other genotypes (33,34). Future studies are required to elucidate how the interplay between NS3/4A-mediated mechanism and the HCV core protein-mediated mechanism contributes to steatogenic changes in HCV-infected hepatocytes.
Other flaviviruses, such as dengue virus (DENV) and Zika virus (ZIKV), are also known to utilize LDs for replication (35). DENV NS2B3 interacts with Rab18, which plays a role in regulating ADRP function. The interaction between DENV NS2B3 and Rab18 is needed for transporting NS3 to the viral replication sites (36). It will be intriguing to investigate a role of SPG20, ADRP, Ich, and JNK in cells infected with other flaviviruses.
In conclusion, we propose a novel mechanism in which the HCV NS3/4A pro tease cleaves SPG20 and inhibits the SPG20-TIP47 interaction, thereby preventing the ubiquitin-dependent degradation of ARDP on the surface of LDs and promoting the formation of large LDs in HCV-infected cells. Understanding the molecular mechanism underlying the HCV-induced formation of large LDs may shed new light on the treatment of steatosis and chronic liver diseases caused by HCV infection. Validating key findings in more physiologically relevant systems, such as primary human hepatocytes, liver organoids, or in vivo models, would strengthen the conclusions and help assess the broader relevance of the proposed pathway in the context of human liver disease.
## MATERIALS AND METHODS
## Cell culture
The human hepatoma cell line Huh-7.5 (37) was kindly provided by Dr. Charles M. Rice (The Rockefeller University, NY). The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (High Glucose) with L-glutamine (Wako, Osaka, Japan) supplemented with 50 IU/mL penicillin, 50 µg/mL streptomycin (Gibco, Grand Island, NY), 10% heatinactivated fetal bovine serum (Biowest, Nuaillé, France), and 0.1 mM nonessential amino acids (Invitrogen, New York, NY) at 37°C in a 5% CO 2 incubator. The cells were transfected with plasmid DNA via FuGene6 transfection reagents (Promega, Madison, WI).
The pFL-J6/JFH1 plasmid, encoding the entire viral genome of a chimeric strain of HCV-2a, JFH1 (38), was kindly provided by Charles M. Rice. The HCV genomic RNA was synthesized in vitro via the use of pFL-J6/JFH1 as a template and was transfected into Huh-7.5 cells by electroporation (39)(40)(41). The virus produced in the culture supernatant was used for infection experiments (39).
## Expression plasmids
To express NS3/4A, the cDNA fragment of nt 3,420 to 5,474 derived from the HCV Con1 strain was amplified by polymerase chain reaction (PCR). The specific primers used for PCR were as follows: sense primer, 5′-TAACTCGAGCGCGCCTATTACGGCCTACTC -3′; antisense primer, 5′-AAAGCGGCCGCTCAGCACTCTTCCATCTCATCGA-3′. The amplified PCR product was purified, cloned, and inserted into the XbaI-NotI site of pCAG-HA. The plasmids pCAG-FLAG-NS2-3ΔC and pCAG-HA-NS3(S139A)/4A were kindly provi ded by Dr. Suzuki (Hamamatsu University School of Medicine, Shizuoka, Japan). The plasmids pCAG-FLAG-SPG20, pCAG-FLAG-Itch, pCAG-FLAG-WWP2, pCAG-HA-ADRP, and pCAG-HA-TIP47 were constructed by amplification of cDNA fragments by PCR. The cDNA fragments were amplified via the use of pBluescriptR-SPG20, pCMV-SPORT6-WWP2, pOTB7-ADRP, and pDNR-Dual-TIP47 as templates. These plasmids were purchased from the PlasmID database (Harvard Medical School, MA). The expression plasmid pCAG-FLAG-WWP1 was previously described (42). The specific primers used for PCR were as follows: SPG20 sense primer, 5′-TCGAGCTCAGCGGCCGCCATGGAGCAAGAGCCACAA -3′; SPG20 antisense primer, 5′-AGTGAATTCGCGGCCGCTCATTTATCTTTCTTCTT-3′, WWP1 sense primer, 5′-TCGAGCTCAGCGGCCGCCATGGCCACTGCTTCACCA-3′; WWP1 antisense primer, 5′-AGTGAATTCGCGGCCGCTCATTCTTGTCCAAATCC-3′, WWP2 sense primer, 5′-TC GAGCTCAGCGGCCATGGCATCTGCCAGCTCT-3′; WWP2 antisense primer, 5′-AGTGAATTCG CGGCCTTACTCCTGTCCAAAGCCC-3′, ADRP sense primer, 5′-TCGAGCTCAGCGGCCGCCAT GGCATCCGTTGCAGTT-3′; ADRP antisense primer, 5′-AGTGAATTCGCGGCCGCTTAATGAGT TTTATGCTC-3′, TIP47 sense primer, 5′-TCGAGCTCAGCGGCCGCCATGTCTGCCGACGGGGC -3′; TIP47 antisense primer, 5′-AGTGAATTCGCGGCCGCCTACTTCTTCTCCTCCGG-3′. These amplified PCR products were subsequently purified. Each of them was inserted into the Not I site of pCAG-FLAG or pCAG-HA via an In-Fusion HD-Cloning Kit (Clontech, Mountain View, CA). To express ADRP as a myc-tagged protein, the cDNA fragment was amplified by PCR, cloned, and inserted into the Kpn I/Bgl II site of pCAG-MCS2. The specific primers used for PCR were as follows: ADRP sense primer, 5′-AAGGTACCATGGAGCAAAAGCTCATT TCTGAAGAGGACTTGATGGCATCCGTTGCAGTTGAT-3′;
ADRP antisense primer, 5′-TTAGATCTTTAATGAGTTTTATGCTCAGATCG-3′. The cDNA fragment for the pCAG-SPG20-FLAG plasmid was amplified by PCR, cloned, and inserted into the Not I site of pCAG-MCS2. The specific primers used for PCR were as follows: SPG20-FLAG sense primer, 5′-AAGCGGCCGCACCATGGAGCAAGAGCCACAAAAT-3′; SPG20-FLAG antisense primer, 5′-TTGCGGCCGCTTACTTATCGTCGTCATCCTTGTAATCTT TATCTTTCTTCTTTGCCTC-3′.
The point mutations of SPG20 were introduced into pCAG-SPG20-FLAG via over lap extension using PCR. The specific primers used for PCR were as follows: SPG20 C499A sense primer, 5′-CTGGTTGATGGAGTTGCCACTGTAGCAAATTGC-3′ ; SPG20 C499A antisense primer, 5′-GCAATTTGCTACAGTGGCAACTCCATCAACCAG-3′ (C499A). SPG20 C504A sense primer, 5′-TGCACTGTAGCAAATGCCGTTGGAAAAGAA-3′ ; SPG20 C504A antisense primer, 5′-TTCTTTTCCAACGGCATTTGCTACAGTACG-3′. SPG20 C562A sense primer, 5′-GAATGTGCAGCTAAAGCCATCGTTAACAAT-3′ ; SPG20 C562A antisense primer, 5′-ATTGTTAACGATGGCTTTAGCTGCACATTC-3′, SPG20 T587A sense primer, 5′-AATGCAG GAGAAGCTGCCCACCATGCGGTG-3′ ; SPG20 T587A antisense primer, 5′-CACCGCATGGTG GGCAGCTTCTCCTGCATT-3′. The sequences of the inserts were extensively confirmed by sequencing (Eurofins Genomics, Tokyo, Japan).
## Antibodies
The mouse monoclonal antibodies (MAbs) used in this study were anti-FLAG (M2) MAb (F3165, Sigma, St. Louis, MO), anti-HA MAb (H-3663, Sigma), anti-NS3 MAb (MAB8691, Millipore, Billerica, MA), anti-c-Myc MAb (9E10, Santa Cruz Biotechnology, Santa Cruz, CA), anti-TIP47 MAb (B-3, Santa Cruz Biotechnology), anti-ADRP MAb (03-610102, ARP American Research Products, Waltham, MA), anti-Itch mouse MAb (611198, BD Bioscien ces, Franklin Lakes, NJ), and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) MAb (MAB374, Millipore). The rabbit MAbs used in this study were anti-SAPK/JNK (56G8) rabbit MAb (9258, Cell Signaling Technology, Beverly, MA). The rabbit polyclo nal antibody (PAbs) used in this study was anti-HA PAb (H-6908, Sigma), anti-DDDDK tag (FLAG) PAb, anti-ADFP (ADRP) PAb (ab108323, Abcam, Cambridge, MA), anti-Perili pin3 (TIP47) PAb (ab47639, Abcam), anti-Itch rabbit PAb (SAB4200036, Sigma), anti-phos pho-Itch (Thr222) rabbit PAb (AB10050, Millipore), and anti-phospho-SAPK/JNK (Thr183/ Tyr185) rabbit PAb (9251, Cell Signaling Technology). The goat PAb used in this study was anti-SPG20 goat PAb (sc-49521, Santa Cruz Biotechnology). Horseradish peroxidase (HRP)-conjugated antibodies, including anti-mouse IgG and anti-rabbit IgG (both from Cell Signaling Technology) as well as donkey anti-goat IgG (Santa Cruz Biotechnology), were used.
## Immunoblot analysis
Immunoblot analysis was performed essentially as described previously (40,42,43). The cell lysates were separated by 10% or 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride mem brane (Millipore). The membranes were incubated with primary antibodies, followed by incubation with peroxidase-conjugated secondary antibody. The positive bands were visualized using enhanced chemiluminescence (ECL) Western blotting detection reagents (GE Healthcare, Buckinghamshire, UK).
## Immunoprecipitation
Cultured cells were lysed with a buffer containing 150 mM NaCl, 10 mM Tris-HCl (pH 7.4), 0.1% SDS, 1% sodium deoxycholic acid, 1% Triton X-100, and cOmplete Protease Inhibitor Cocktail Tablets (Roche Diagnostics, Indianapolis, IN). The lysate was centri fuged at 12,500 × g for 15 min at 4°C, and the supernatant was immunoprecipitated with appropriate antibodies. Immunoprecipitation was performed as described previously (44). Briefly, the cell lysates were immunoprecipitated with anti-FLAG M2 affinity gel (Sigma) or protein A-Sepharose 4 fast flow (GE Healthcare) and incubated with appropri ate antibodies at 4°C overnight. After being washed with lysis buffer five times, the immunoprecipitates were analyzed by immunoblotting.
## siRNA transfection
HCV-infected Huh-7.5 cells or uninfected control cells (3.0 × 10 5 cells per12-well plate) were transfected with 20 pmol of either Itch-specific siRNA duplexes (Qiagen, Valen cia, CA) or MISSION siRNA Universal Negative Control (SIC-001, Sigma Genosys) using Lipofectamine RNAiMAX transfection reagent (Life Technologies, Carlsbad, CA) according to the manufacturer's instructions and cultured for 48 h. The Itch-siRNA target sequences were as follows: 5′-ATGGGTAGCCTCACCATGAUU-3′.
## Immunofluorescence staining
Huh-7.5 cells cultured on glass cover slips were fixed with 4% paraformaldehyde at room temperature for 15 min. After being washed with PBS, the cells were permeabilized with PBS containing 0.1% Triton X-100 for 15 min at room temperature and incubated in PBS containing 1% bovine serum albumin for 60 min to block nonspecific reactions. The cells were incubated with Can Get Signal Immunostain Solution A (TOYOBO, Osaka, Japan) containing mouse anti-NS3 antibody (Millipore) at room temperature for 60 min. The cells were washed four times with PBS and incubated with Can Get Signal Immunos tain Solution A containing Alexa Fluor 405-conjugated anti-mouse immunoglobulin G (IgG) (Life Technologies) with BODIPY 493/503 (Molecular Probes, Eugene, OR) at room temperature for 60 min. The cells were washed four times with PBS and then observed under an LSM700 confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany).
## Quantification and measurement of LD
To analyze the size and the number of LDs, the cells were analyzed with an LSM700 confocal microscope. The cells selected as the region of interest (ROIs) were processed by thresholding. ROIs were generated via a free-hand selection tool. The images were used to quantify the LD surface area per cell. The LDs per cell were counted from the same images. The size and the number of LDs were analyzed via ImageJ (Ver. 1.48).
## Statistics
The results are expressed as the mean ± standard error. Statistical significance was determined by Student's t-test. P-values < 0.01 (**) were considered significant.
## References
1. Ray, Bailey, Thomas et al. (2013) "Hepatitis C virus"
2. Li, Clercq (2017) "Current therapy for chronic hepatitis C: the role of direct-acting antivirals" *Antiviral Res*
3. Hayes, Zhang, Zhang et al. (2018) "Molecular mechanisms of hepatocarcinogenesis following sustained virological response in patients with chronic hepatitis C virus infection" *Viruses*
4. Machado, Oliveira, Pinto (2011) "Hepatic steatosis in hepatitis B virus infected patients: meta-analysis of risk factors and comparison with hepatitis C infected patients" *J Gastroenterol Hepatol*
5. Loizides-Mangold, Clément, Alfonso-Garcia et al. (2014) "HCV 3a core protein increases lipid droplet cholesteryl ester content via a mechanism dependent on sphingolipid biosynthesis" *PLoS One*
6. Vieyres, Pietschmann (2019) "HCV pit stop at the lipid droplet: refuel lipids and put on a lipoprotein coat before exit" *Cells*
7. Akil, Peng, Omrane et al. (2016) "Septin 9 induces lipid droplets growth by a phosphatidylinositol-5-phosphate and microtubule-dependent mechanism hijacked by HCV" *Nat Commun*
8. Miyanari, Atsuzawa, Usuda et al. (2007) "The lipid droplet is an important organelle for hepatitis C virus production" *Nat Cell Biol*
9. Vogt, Camus, Herker et al. (2013) "Lipid droplet-binding protein TIP47 regulates hepatitis C Virus RNA replication through interaction with the viral NS5A protein" *PLoS Pathog*
10. Wolins, Quaynor, Skinner et al. (2005) "S3-12, adipophilin, and TIP47 package lipid in adipocytes" *Journal of Biological Chemistry*
11. Yamaguchi, Matsushita, Motojima et al. (2006) "MLDP, a novel PAT family protein localized to lipid droplets and enriched in the heart, is regulated by peroxisome proliferator-activated receptor alpha" *J Biol Chem*
12. Pawella, Hashani, Eiteneuer et al. (2014) "Perilipin discerns chronic from acute hepatocellular steatosis" *J Hepatol*
13. Masuda, Itabe, Odaki et al. (2006) "ADRP/adipophilin is degraded through the proteasome-dependent pathway during regression of lipid-storing cells" *J Lipid Res*
14. Eastman, Yassaee, Bieniasz (2009) "A role for ubiquitin ligases and Spartin/SPG20 in lipid droplet turnover" *J Cell Biol*
15. Edwards, Clowes, Tsang et al. (2009) "Endogenous spartin (SPG20) is recruited to endosomes and lipid droplets and interacts with the ubiquitin E3 ligases AIP4 and AIP5" *Biochem J*
16. Hooper, Puttamadappa, Loring et al. (2010) "Spartin activates atrophin-1-interacting protein 4 (AIP4) E3 ubiquitin ligase and promotes ubiquitination of adipophilin on lipid droplets" *BMC Biol*
17. Patel, Cross, Proukakis et al. (2002) "SPG20 is mutated in Troyer syndrome, an hereditary spastic paraplegia" *Nat Genet*
18. Karlsson, Washington, Dimitrova et al. (2014) "The role of spartin and its novel ubiquitin binding region in DALIS occurrence" *Mol Biol Cell*
19. Ingham, Colwill, Howard et al. (2005) "WW domains provide a platform for the assembly of multiprotein networks" *Mol Cell Biol*
20. Wang, Liu, Chen et al. (2019) "A multi-lock inhibitory mechanism for fine-tuning enzyme activities of the HECT family E3 ligases" *Nat Commun*
21. Riling, Kamadurai, Kumar et al. (2015) *J Biol Chem*
22. Nguyen, Lee, Mun et al. (2019) "Nterminal acetylation and the N-end rule pathway control degradation of the lipid droplet protein PLIN2" *J Biol Chem*
23. Lin, Shoji, Ogawa et al. (2011) "Hepatitis C virus infection promotes hepatic gluconeogenesis through an NS5A-mediated, FoxO1-dependent pathway" *J Virol*
24. Deng, Chen, Tanaka et al. (2015) "HCV upregulates bim through the ROS/JNK signalling pathway, leading to bax-mediated apoptosis" *J Gen Virol*
25. Deng, Liang, Ariffianto et al. (2022) "Hepatitis C virus-induced ROS/JNK signaling pathway activates the E3 ubiquitin ligase itch to promote the release of HCv particles via polyubiquitylation of VPS4A" *J Virol*
26. Tokunaga, Shiheido, Hayakawa et al. (2013) "Hereditary spastic paraplegia protein spartin is an FK506-binding protein identified by mRNA display" *Chem Biol*
27. Bley, Schöbel, Herker (2020) "Whole lotta lipids-from HCV RNA replication to the mature viral particle" *Int J Mol Sci*
28. Rouillé, Helle, Delgrange et al. (2006) "Subcellular localization of hepatitis C virus structural proteins in a cell culture system that efficiently replicates the virus" *J Virol*
29. Chung, Park, Lai et al. (2023) "The troyer syndrome protein spartin mediates selective autophagy of lipid droplets" *Nat Cell Biol*
30. Pirozzi, Mcconnell, Uveges et al. (1997) "Identification of novel human WW domain-containing proteins by cloning of ligand targets" *J Biol Chem*
31. Gallagher, Gao, Liu (2006) "Activation of the E3 ubiquitin ligase Itch through a phosphorylation-induced conformational change" *Proc Natl Acad Sci*
32. Kory, Farese, Jr et al. (2016) "Targeting fat: mechanisms of protein localization to lipid droplets" *Trends Cell Biol*
33. Rubbia-Brandt, Quadri, Abid et al. (2000) "Hepatocyte steatosis is a cytopathic effect of hepatitis C virus genotype 3" *J Hepatol*
34. Abid, Pazienza, De Gottardi et al. (2005) "An in vitro model of hepatitis C virus genotype 3a-associated triglycerides accumulation" *J Hepatol*
35. Hsia, Liu, Haynes et al. (2024) "Lipid droplets: formation, degradation, and their role in cellular responses to flavivirus infections" *Microorganisms*
36. Tang, Lin, Liao et al. (2014) "Rab18 facilitates dengue virus infection by targeting fatty acid synthase to sites of viral replication" *J Virol*
37. Blight, Mckeating, Rice (2002) "Highly permissive cell lines for subgenomic and genomic hepatitis C virus RNA replication" *J Virol*
38. Lindenbach, Evans, Syder et al. (2005) "Complete replication of hepatitis C virus in cell culture" *Science*
39. Bungyoku, Shoji, Makine et al. (2009) "Efficient production of infectious hepatitis C virus with adaptive mutations in cultured hepatoma cells" *J Gen Virol*
40. Deng, Adachi, Kitayama et al. (2008) "Hepatitis C virus infection induces apoptosis through a bax-triggered, mitochondrion-mediated, caspase 3-dependent pathway" *J Virol*
41. Wakita, Pietschmann, Kato et al. (2005) "Production of infectious hepatitis C virus in tissue culture from a cloned viral genome" *Nat Med*
42. Shirakura, Murakami, Ichimura et al. (2007) "E6AP ubiquitin ligase mediates ubiquitylation and degradation of hepatitis C virus core protein" *J Virol*
43. Matsui, Shoji, Kaneda et al. (2012) "Hepatitis C virus infection suppresses GLUT2 gene expression via downregulation of hepatocyte nuclear factor 1α" *J Virol*
44. Matsui, Sianipar, Minami et al. (2015) "A single-amino-acid mutation in hepatitis C virus NS5A disrupts physical and functional interaction with the transcription factor HNF-1α" *J Gen Virol*
45. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12665423&blobtype=pdf | # Ubiquitously expressed transcript isoform 2 (UXT-V2) restricts HSV-2 replication by targeting glycoprotein B for degradation through ubiquitin-proteasome pathway
Chuntian Li, Yuncheng Li, Ranqing Cheng, Miaomiao Li, Mudan Zhang, Zhiyuan Zhu, Ping Yang, Qinxue Hu, Yalan Liu, Virologica Sinica
## Abstract
Herpes simplex virus 2 (HSV-2) is a major pathogen causing neonatal herpes and increasing the risk of human immunodeficiency virus 1 (HIV-1) infection. However, the mechanisms underlying host restriction of HSV-2 infection are still not fully understood. The ubiquitously expressed transcript isoform 2 (UXT-V2), an α-type prefoldin protein, functions as a versatile transcription factor associated with numerous human tumors, but its role in viral infection remains unclear. In this study, we found that ectopic expression of UXT-V2 significantly inhibited HSV-2 replication, while knockout of endogenously expressed UXT-V2 promoted HSV-2 proliferation. Further analysis revealed that UXT-V2 restricts HSV-2 replication independent of its role in regulating NF-κB. In the context of HSV-2 infection or in viral glycoprotein B (gB)-transfected cells, UXT-V2 facilitates K48-linked ubiquitination of gB, leading to its degradation via the proteasome pathway, thereby inhibiting viral replication. Furthermore, we identified that UXT-V2 interacts with gB, recruiting the E3 ligase TRIM21 to facilitate K48-linked ubiquitination of gB. HSV-2, in turn, reduces the abundance of UXT-V2 proteins both in vitro and in mice, highlighting the complexity of HSV-2-host interactions. Collectively, our findings, for the first time, demonstrate an anti-HSV-2 role of UXT-V2, unveiling a novel host immune defense mechanism involved in regulating glycoprotein homeostasis.
## INTRODUCTION
Herpes simplex virus (HSV-2), a double-stranded DNA virus belonging to the subfamily of Alphaherpesvirinae (Byrne et al., 2018), is the major cause of genital herpes, and its infection can increase the risk of HIV-1 infection (Suazo et al., 2015). HSV-2 primarily infects the genital epithelium and can be transmitted to the peripheral nervous system, where it establishes life-long latent infection (Groves, 2016;Li et al., 2018). An estimated 491.5 million people were living with HSV-2 infection in 2016, equivalent to 13.2% of the world's population aged 15-49 years (James et al., 2020). Currently, there is no available HSV-2 vaccine, and the treatment options for HSV-2 infection remain limited, largely due to an insufficient understanding of the interaction between HSV-2 and the host. The envelope glycoproteins of HSV-2 play important roles during the process of viral entry/egress and cell-to-cell spread (Liu et al., 2023). Among these envelope glycoproteins, HSV glycoprotein B (gB) plays key roles in viral entry, cell fusion, and cell-to-cell spread (Cheshenko and Herold, 2002). It directly mediates the fusion between perinuclear virus particles and the outer nuclear membrane (Wright et al., 2009), mobilizes lipid rafts (Bender et al., 2003), initiates interaction with CD1d in the endoplasmic reticulum (ER), and stably associates with it throughout CD1d trafficking (Rao et al., 2011). Additionally, gB helps maintain ER homeostasis through its interaction with PERK (Mulvey et al., 2007) and diverts HLA-DR into the exosome pathway (Temme et al., 2010). To date, the regulation of gB within host cells remains largely unknown.
Several host factors have been identified that restrict HSV replication, particularly HSV-1, through various mechanisms. For example, human myxovirus resistance protein B (MxB) inhibits the delivery of incoming HSV-1/2 DNA to the nucleus (Crameri et al., 2018), TRIM23 restricts HSV-1 by mediating autophagy (Sparrer et al., 2017), and USP18 limits HSV-1 infection by deubiquitinating STING (Zhang et al., 2016). Additionally, IFI16 restricts HSV-1 replication by accumulating on the HSV-1 genome, repressing HSV-1 gene expression, and modulating histone modifications either directly or indirectly (Everett et al., 2014). Furthermore, the cellular protein E-Syt1 negatively modulates HSV-1 release into the extracellular milieu, cell-to-cell viral spread, and viral entry (El Kasmi et al., 2018). However, our understanding of host factors that restrict HSV-2 replication remains limited.
UXT gene is located in Xp11, a densely populated gene region harboring a variety of disease loci (Schroer et al., 1999). The UXT gene consists of seven exons and gives rise to at least two naturally occurring alternatively spliced transcripts, namely UXT-V1 and UXT-V2 (Huang et al., 2012). UXT-V1 has 169 amino acids and is predominantly found in the cytoplasm, while the short form UXT-V2, also known as UXT, ART-27, or STAP1, has 157 amino acids expressed in both the cytoplasm and nucleus (Yang et al., 2009;Chen et al., 2013). UXT-V2 is widely expressed in human tissues (Schroer et al., 1999). Most of the studies to date have been focused on its biological functions and physiological significance. For example, UXT-V2 expression has been found to be elevated in numerous human tumor tissues, with increasing evidence associating it with carcinogenesis and tumor progression (Sanchez-Morgan et al., 2017). UXT-V2 can directly bind to the N-terminus of the androgen receptor (AR), facilitating the regulation of androgen-responsive genes (Markus et al., 2002;Taneja et al., 2004). UXT-V2 significantly increases caspase 8 activity and enhances SARM-induced apoptosis by activating the extrinsic pathway and inducing mitochondrial depolarization (Sethurathinam et al., 2013). UXT-V2 has the capability to interact with estrogen receptor, leading to the inhibition of transcriptional activity associated with estrogen-responsive genes (Sanchez-Morgan et al., 2017). In addition, UXT-V2 has also been reported to positively regulate the Treg-cell function, cell transformation, and cell viability (Zhao et al., 2005;Sethurathinam et al., 2013;Li et al., 2014;Huehn and Beyer, 2015;Su et al., 2016). It is known that UXT-V2 interacts with NF-κB subunit p65 to regulate the transcription of NF-κB-dependent genes (Sun et al., 2007). UXT-V2 can sustain NF-κB activity, thereby inhibiting spontaneous activation of Epstein-Barr virus (EBV), which belongs the Gammaherpesvirinae subfamily. Although UXT-V2 is known to be necessary for maintaining EBV latency (Chang et al., 2012), its antiviral activity against HSV-2 remains unknown. Nevertheless, the function of UXT-V2 in viral infection has yet to be elucidated.
In this study, we have demonstrated that UXT-V2 promotes K48linked ubiquitination of gB, leading to proteasome-dependent degradation, thereby inhibiting HSV-2 replication. Conversely, HSV-2 infection reduces the abundance of UXT-V2 protein. This study informs the complexity of HSV-2-host interactions, providing insight into the role of UXT-V2 as a viral restriction factor and the mechanisms underlying viral countermeasures.
## RESULTS
## UXT-V2 inhibits HSV-2 replication during viral infection
Although it is known that UXT-V2 is associated with many human tumors, the function of UXT-V2 in viral infection remains unclear. To explore whether UXT-V2 affects HSV-2 replication, HeLa cells were transfected with different doses of empty vector or plasmid expressing His-UXT-V2, followed by infection with HSV-2 at an MOI of 0.2 PFU/ cell. At 12, 18, and 24 h post-infection (hpi), the cells were collected for qRT-PCR, Western blotting, and plaque assays. The results showed that overexpression of UXT-V2 significantly downregulated the relative abundance of ICP0 mRNAs (Fig. 1A), the protein expression levels of HSV-2 envelope glycoprotein B and D (gB and gD) (Fig. 1B), and viral yields (Fig. 1C) at all tested time points.
Furthermore, the inhibitory effect of UXT-V2 on HSV-2 proliferation was also detected in ARPE-19 cells, another target cells of HSV-2 (Supplementary Fig. S1A, S1B and S1C). The anti-HSV-2 effect of UXT-V2 was further confirmed by UXT-V2 knockdown using UXT-V2 siRNA. We confirmed the effectiveness of the siRNA against UXT-V2 by monitoring the levels of its mRNA and protein (Supplementary Fig. S2A), showing that the yield of HSV-2 increased after the expression of endogenous UXT-V2 was knocked down in ARPE-19 cells (Supplementary Fig. S2B).
To further confirm the significance of UXT-V2 in virus restriction in human cells, we generated UXT-V2 knockout (KO) HeLa cell line (Fig. 1D). The CCK-8 cell counting assay indicated that UXT-V2 deficiency did not limit cell proliferation (Fig. 1E). UXT-V2 KO cells were infected with HSV-2 at an MOI of 0.2 PFU/cell. At 12, 18, and 24 hpi, the cells were collected for qRT-PCR, Western blotting and plaque assays. As shown in Fig. 1, knockout of UXT-V2 significantly upregulated the relative abundance of ICP0 mRNAs (Fig. 1F), the protein expression level of HSV-2 ICP5 and gB (Fig. 1G) and viral yields (Fig. 1H), further confirming that endogenously expressed UXT-V2 inhibits HSV-2 replication. These results indicate that the presence of UXT-V2 in HSV-2infected cells is essential for suppressing virus replication.
## UXT-V2 inhibits HSV-2 replication via NF-κB independent pathway
UXT has two alternative splicing isoforms. Additionally, we uncovered that UXT-V1 has the capability to suppress HSV-2 replication (Fig. 2A). UXT-V1 has been reported to facilitate the virus-induced activation of NF-κB and IRF3 (Huang et al., 2012), while UXT-V2 interacts with NF-κB subunit p65 to regulate the transcription of NF-κB-dependent genes (Sun et al., 2007). To explore the potential mechanism underlying the suppression of HSV-2 replication by UXT-V1 and UXT-V2, we examined whether either isoform is required for the activation of IRF3 or NF-κB during viral infection. Reporter gene assays were performed in HeLa cells using NF-κB-Luc and PRD(III-I) 4 -Luc reporter systems, respectively. As shown in Fig. 2B, in the context of HSV-2 infection, UXT-V1 but not UXT-V2 increased the activity of NF-κB-Luc or PRD (III-I) 4 -Luc. Furthermore, overexpression of UXT-V2 did not affect NF-κB activation in HEK293T cells (Fig. 2C).
We also found that UXT-V1 inhibited Japanese encephalitis virus (JEV) and Zika virus (ZIKV) replication (Fig. 2D andF), whereas UXT-V2 had no such inhibition (Fig. 2D-G). Moreover, an inhibition of HSV-2 replication induced by UXT-V2 was evident in Vero E6 cells, a cell line incapable of producing type I IFNs. It is noteworthy that the replication of HSV-2 was enhanced in UXT-V2 knockdown cells (Fig. 2H). These collective results suggest that UXT-V1 inhibits various viruses, likely by facilitating the activation of NF-κB and IRF3 signaling pathways, while UXT-V2 inhibits HSV-2 replication, likely via a NF-κB independent pathway.
## UXT-V2 inhibits HSV-2 replication by downregulating HSV-2 gB expression
To investigate the molecular mechanisms by which UXT-V2 suppresses HSV-2 replication, we hypothesized that UXT-V2 might interact with HSV-2 viral proteins to inhibit the replication of HSV-2. Coimmunoprecipitation (Co-IP) analysis was performed by adding UXT-V2 antibody to HeLa cells infected with HSV-2 for 24 hours, followed by sample collection and liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) analysis. We found that a number of HSV-2 viral proteins, including gB, UL37, and UL47, could be fished out by the UXT-V2 antibody (Supplementary Table S2). Proteins with higher scores are listed in Fig. 3A. Subsequently, Co-IP was performed to validate the interaction between UXT-V2 and gB, demonstrating that both exogenously and endogenously expressed UXT-V2 can interact with gB (Fig. 3B andC). To localize the region of UXT-V2 that interacts with gB, we utilized Alphafold 3 to simulate the interactions between gB and various forms of UXT-V2, including its wildtype, several point mutation variants (L32P/L50P/L59P, L50A/L59A, L50P/L59P, T3E, T3V) and deletion mutation variants (Δ21-59 and 20-146) (Sun et al., 2007;Chang et al., 2012;Yoon et al., 2021). The wild type UXT-V2 exhibited an interaction with gB, which is consistent with the co-immunoprecipitation results. In contrast, none of the UXT-V2 mutations demonstrated any substantial interaction with gB (Supplementary Fig. S3). We speculate that a three-dimensional motif may be involved in this interaction, or alternatively, the entire protein might contribute to forming a functional surface for gB interaction. In addition, Western blotting showed that UXT-V2 could reduce the expression of gB (Fig. 3B).
To further investigate whether UXT-V2 reduced the expression of gB, plasmids expressing various Flag-tagged HSV-2 proteins and plasmid expressing His-UXT-V2 were co-transfected into HEK293T cells. At 24 h post-transfection (hpt), the expression levels of Flag-tagged HSV-2 proteins were measured by Western blotting. It was observed that UXT-V2 significantly reduced the abundance of gB (Fig. 3D). The relative expression level of gB was also examined at different time points after co-transfection with plasmid expressing His-UXT-V2 and plasmid expressing Flag-gB. UXT-V2 expression resulted in a significant reduction in gB levels. As shown in Fig. 3E, the levels of gB started to decline upon detection of gB via Western blotting. Furthermore, although UXT-V2 inhibited HSV-2 gB expression, UXT-V1 had no effect on the expression level of HSV-2 gB (Fig. 3F). This observation was further confirmed by UXT-V2 knockdown. The expression level of gB increased after endogenous UXT-V2 expression was knocked down in ARPE-19 cells (Fig. 3G).
We subsequently examined whether UXT-V2 reduces gB expression during HSV-2 infection. HeLa cells were transfected with plasmid expressing His-UXT-V2 prior to infection with HSV-2 at an MOI of 5. UXT-V2 significantly reduced gB expression without altering the expression level of another HSV-2 envelop glycoprotein gD (Fig. 3H). A similar inhibitory effect on the expression of gB and gD was observed in Fig. 1. UXT-V2 inhibits HSV-2 replication. A-C HeLa cells were transfected with different doses of empty vector or plasmids expressing His-UXT-V2 (0.2, 0.5 or 1 μg), followed by infection with HSV-2 at an MOI of 0.2 PFU/cell. At 12, 18, and 24 hpi, the cells were collected for qRT-PCR (A), Western blotting (B), and plaque assay (C). D UXT-V2 in sgControl and sgUXT-V2 HeLa cells was assessed by immunoblot analysis. E sgControl and sgUXT-V2 HeLa cells were cultured for 24 h. Cell proliferation was assessed using the CCK-8 cell counting assay. F-H WT and UXT-V2-KO HeLa cells were infected with HSV-2 at an MOI of 0.2 PFU/cell. At 12, 18, and 24 hpi, the cells were collected for qRT-PCR (F), Western blotting (G), and plaque assay (H). For graphs, data shown are mean ± SD of three independent experiments with each condition performed in triplicate (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001). For images, one representative experiment out of three is shown. Fig. 2. UXT-V2 inhibits HSV-2 replication via NF-κB independent pathway. A ARPE-19 cells were transfected with an empty vector or different doses of plasmids expressing His-UXT-V1 (0.5 or 1 μg) or His-UXT-V2 (0.5 or 1 μg), followed by infection with HSV-2 at an MOI of 0.2 PFU/cell at 6 hpt for 18 h. The virus yields were measured by plaque assay. B HeLa cells were transfected with 0.3 μg PRD(III-I) 4 -Luc or NF-κB-Luc reporter plasmid along with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V1 or His-UXT-V2, followed by infection with HSV-2 at an MOI of 0.2 PFU/cell at 6 hpt for 18 h. The luciferase activity was measured to determine the activation fold. HEK293T cells were transfected with 0.3 μg NF-κB-Luc reporter plasmid along with an empty vector or 1 μg plasmid expressing His-UXT-V2 for 24 h, followed by treatment with or without 10 ng/mL TNF-α for 6 h. Luciferase activity was measured to determine the activation fold. C HEK293T cells were transfected with 0.3 μg NF-κB-Luc reporter plasmid along with an empty vector or 1 μg plasmid expressing His-UXT-V2 for 24 h, followed by treatment with or without 10 ng/mL TNF-α for 6 h. Luciferase activity was measured to determine the activation fold. D BHK-21 cells were transfected with an empty vector or different doses of plasmids expressing His-UXT-V1 (0.5 or 2 μg) or His-UXT-V2 (0.5 or 2 μg), followed by infection with JEV at an MOI of 0.5 PFU/cell at 6 hpt. The supernatants and cells were harvested, and the virus yields were measured by plaque assay at 24 hpi. E BHK-21 cells were transfected with an empty vector or 1.5 μg plasmid expressing His-UXT-V2, followed by infection with JEV at an MOI of 0.5 PFU/cell at 6 hpt. The supernatants and cells were harvested, and the virus yields were measured by plaque assay at 8, 16, 24 and 36 hpi. F Vero E6 cells were transfected with an empty vector or different doses of plasmids expressing His-UXT-V1 (0.5 or 2 μg) or His-UXT-V2 (0.5 or 2 μg), followed by infection with ZIKV at an MOI of 0.5 PFU/cell at 6 hpt. The supernatants and cells were harvested and the virus yields were measured by plaque assay at 24 hpi. G Vero E6 cells were transfected with an empty vector or 1.5 μg plasmid expressing His-UXT-V2, followed by infection with ZIKV at an MOI of 0.5 PFU/cell at 6 hpt. The supernatants and cells were harvested, and the virus yields were measured by plaque assay at 8, 16, 24 and 36 hpi. H Vero E6 cells transfected with the control or the two UXT shRNAs, respectively, were infected with HSV-2 at an MOI of 0.2 PFU/cell for 24 h. The virus yields were measured by plaque assay. For graphs, data shown are mean ± SD of three independent experiments with each condition performed in triplicate (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant). For images, one representative experiment out of three is shown. were co-transfected with 2 μg empty vector or 2 μg His-UXT-V2 together with 2 μg plasmid expressing Flag-gB. At 24 hpi, cells were harvested and lysed, and the extracts were subjected to IP using the anti-Flag MAb or control IgG. C HeLa cells were infected with or without HSV-2 at an MOI of 0.2 PFU/cell. At 24 hpi, cells were harvested and lysed, and the extracts were subjected to IP using the anti-gB MAb. D HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V2 together with 0.5 μg plasmid expressing various Flag-tagged HSV-2 proteins for 24 h. The expression of Flag-tagged HSV-2 proteins were measured by Western blotting. E HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V2 together with 0.5 μg plasmid expressing Flag-gB for 12 or 18 h. Cells were harvested and the expression of Flag-gB was detected by Western blotting. F HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V1 or His-UXT-V2 together with 0.5 μg plasmid expressing Flag-gB for 12 or 18 h. Cells were harvested and the expression of Flag-gB was measured by Western blotting. G ARPE-19 cells were transfected with 60 nM NC or #4 (UXT-V2 siRNA), followed by transfection with an empty vector or different doses of plasmids expressing Flag-gB (0.5 or 2 μg) for 24 h. Cells were harvested and the expression of Flag-gB was detected by Western blotting. H HeLa cells were transfected with an empty vector or 2 μg plasmid expressing His-UXT-V2, followed by infection with HSV-2 at an MOI of 5 PFU/cell. At 6 or 8 hpi, cells were harvested, and the expression level of HSV-2 gB or gD was analyzed by Western blotting. I WT and UXT-V2-KO HeLa cells were infected with HSV-2 at an MOI of 5 PFU/cell. At 6 or 8 hpi, cells were harvested and the expression level of HSV-2 gB or gD was measured by Western blotting. J HeLa cells were cotransfected with an empty vector or different doses of plasmids expressing His-UXT-V2 (0.5 or 2 μg) together with 0.5 μg plasmid expressing Flag-gB, followed by infection with HSV-2 at an MOI of 0.2 PFU/cell at 6 hpt for 18 h. Virus yields were measured by plaque assay. The graph shows compensating efficiency of gB on HSV-2 yields. For graphs, data shown are mean ± SD of three independent experiments with each condition performed in triplicate (***P < 0.001; ****P < 0.0001; ns, not significant). For images, one representative experiment out of three is shown.
UXT-V2-KO HeLa cells (Fig. 3I). It is known that HSV-2 gB plays a crucial role as an envelope protein in HSV-2 absorption, internalization, release, and spread. We reintroduced Flag-gB-expressing plasmids into HeLa cells, showing that gB could indeed rescue HSV-2 replication (Fig. 3J). These findings together indicate that UXT-V2 restricts HSV-2 replication by downregulating gB expression.
## UXT-V2 downregulates the expression of HSV-2 gB through proteasomal pathway
UXT-V2 is recognized as a transcriptional regulation cofactor, modulating the transcriptional activity of various genes (Thomas et al., 2018). Subsequntly, the abundance of gB mRNA in the His-UXT-V2-transfected cells was evaluated to determine whether the reduction of gB protein resulted from a specific decrease in mRNA expression. The overexpression of UXT-V2 had no effect on HSV-2 gB mRNA expression (Fig. 4A), suggesting that UXT-V2-induced gB reduction was not due to decreased synthesis of gB mRNA. To investigate whether proteasomal or/and lysosomal pathways are involved in UXT-V2-induced decrease of gB, we utilized the proteasome inhibitor MG132 and the lysosome inhibitor chloroquine diphosphate (CQ) to assess their inhibitory effects. As shown in Fig. 4B, treatment with CQ significantly restored gB expression levels in HEK293T cells transiently expressing UXT-V2. Similarly, MG132 also had an effect on restoring gB. It has been reported that the conformation of HSV-1 gB changes at low pH. As CQ can alter the pH of cells (Stampfer et al., 2010;Weed et al., 2018), we speculate that CQ may impact the abundance of gB. We therefore conducted experiments by treating Flag-gB-transfected cells with CQ and MG-132, respectively. As shown in Fig. 4C, in the absence of UXT-V2 co-transfection, CQ treatment significantly increased the protein level of gB, whereas MG132 had no effect on the expression of gB. To eliminate the possibility of CQ affecting pH changes, another lysosome inhibitor, 3-Methyladenine (3-MA), which has almost no effect on pH, was further utilized to assess its inhibitory effects. As shown in Fig. 4D, treatment with CQ or MG132 significantly restored gB expression levels in HEK293T cells transiently expressing UXT-V2, whereas 3-MA had no effect on restoring gB. These results suggest that the abundance of HSV-2 gB is influenced by low pH, wihle UXT-V2 induces gB degradation likely through the proteasomal pathway rather than the lysosomal pathway.
## UXT-V2 promotes K48-linked polyubiquitination of HSV-2 gB
UXT-V2 belongs to the family of prefoldin proteins, which can facilitate the ubiquitination of their substrates (Abe et al., 2013). To determine whether UXT-V2 facilitates the ubiquitination of cellular proteins, HEK293T cells were co-transfected with empty vector or different doses of plasmids expressing His-UXT-V2 and plasmid expressing HA-Ub. Ubiquitinated cellular proteins were assessed by Western blotting. As shown in Fig. 5A, expression of UXT-V2 resulted in a dose-dependent increase of the level of ubiquitinated cellular proteins. We also confirmed that UXT-V2 facilitated the endogenous ubiquitination of cellular proteins (Fig. 5B). To further determine whether UXT-V2 induces the ubiquitination of HSV-2 gB, HEK293T cells were co-transfected with empty vector or plasmid expressing His-UXT-V2 together with plasmid expressing Flag-gB. Cell lysates were immunoprecipitated with an anti-Flag MAb, followed by immunoblot for gB using the anti-Flag MAb, or for ubiquitinated gB using an anti-Ub Mab. As shown in Fig. 5C, UXT-V2 enhanced the ubiquitination of HSV-2 gB.
It is well established that diverse classes of ubiquitin chains are implicated in driving distinct biological outcomes. HSV-1 gB ubiquitination has been reported to involve in part, K63, a residue implicated in endocytosis rather than proteasome-dependent degradation (Calistri et al., 2007). To further elucidate the type of ubiquitination facilitated by UXT-V2 on HSV-2 gB, HEK293T cells were transfected with plasmids expressing His-tagged UXT-V2, Flag-tagged gB and HA-tagged Ub-WT, HA-tagged Ub-K48 or HA-tagged Ub-K63. The results revealed that UXT-V2 promotes K48-linked rather than K63-linked ubiquitination of the HSV-2 gB (Fig. 5D). To clarify whether gB was ubiquitinated in infected cells, HeLa cells were transfected with plasmids expressing Fig. 4. UXT-V2 induces gB degradation through proteasomal pathways. A HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V2 together with 0.5 μg plasmid expressing Flag-gB for 12 h or 18 h. Cells were harvested and the transcripts of gB were detected by relative quantitative PCR. B HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V2 together with 0.5 μg plasmid expressing Flag-gB in the absence or presence of CQ (20 or 100 μM) or MG132 (1 or 2 μM). At 24 hpt, the expression of Flag-gB was detected by Western blotting. C HEK293T cells were transfected with 1 μg plasmid expressing Flag-gB in the absence or presence of CQ (20 or 100 μM), 3-MA (5 or 20 μM) or MG132 (1 or 2 μM). At 24 hpt, the expression of Flag-gB was detected by Western blotting. D HEK293T cells were co-transfected with 1.5 μg empty vector or 1.5 μg plasmid expressing His-UXT-V2 together with 0.5 μg plasmid expressing Flag-gB in the absence or presence of CQ (20 or 100 μM), 3-MA (5 or 20 μM) or MG132 (1 or 2 μM). At 24 hpt, the expression of Flag-gB was detected by Western blotting. For graphs, data shown are mean ± SD of three independent experiments with each condition performed in triplicate (ns, not significant). For images, one representative experiment out of three is shown.
His-tagged UXT-V2 and HA-tagged Ub-WT, HA-tagged Ub-K48 or HA-tagged Ub-K63, followed by infection with HSV-2 at an MOI of 2 PFU/cell at 6 hpt for 18 h, and immunoprecipitation with the MAb to gB. As shown in Fig. 5E, UXT-V2 prompted K48-linked ubiquitination of HSV-2 gB in infected cells. Taken together, these results suggest that UXT-V2 induces the K48-linked ubiquitination of HSV-2 gB, leading to gB degradation through proteasomal pathway.
Fig. 5. UXT-V2 promotes K48-linked polyubiquitination of HSV-2 gB. A HEK293T cells were co-transfected with 0.5 μg empty vector or different doses of plasmids expressing His-UXT-V2 (0.2, 0.5 or 1.5 μg) together with 0.5 μg plasmid expressing HA-Ub. At 24 hpt, the expression of His-UXT-V2 and ubiquitinated cellular proteins was detected by Western blotting. B HEK293T cells were transfected with 0.5 μg empty vector or different doses of plasmids expressing His-UXT-V2 (0.2, 0.5 or 1.5 μg). At 24 hpt, the expression of His-UXT-V2 and ubiquitinated cellular proteins was detected by Western blotting. C HEK293T cells were co-transfected with 5 μg empty vector or 5 μg plasmid expressing His-UXT-V2 together with 4 μg plasmid expressing Flag-gB for 24 h. Cells were immunoprecipitated with the anti-Flag MAb. D HEK293T cells were co-transfected with 2 μg empty vector or 2 μg plasmid expressing His-UXT-V2 together with 2 μg plasmid expressing Flag-gB and 2 μg plasmid expressing HA-Ub, HA-K48 or HA-K63 for 24 h. Cells were immunoprecipitated with the indicated antibodies. E HeLa cells were co-transfected with 2 μg empty vector or 2 μg plasmid expressing His-UXT-V2 together with 2 μg plasmid expressing HA-Ub, HA-K48 or HA-K63, followed by infection with HSV-2 at an MOI of 2 PFU/cell at 6 hpt for 18 h. Cells were immunoprecipitated with the indicated antibodies. For images, one representative experiment out of three is shown. C. Li et al. Virologica Sinica 40 (2025) 778-792
## UXT-V2 promotes K48-linked polyubiquitination of HSV-2 gB by recruiting the E3 ubiquitin ligase TRIM21
Having demonstrate that UXT-V2 primarily mediates the degradation of gB through the proteasomal pathway, we next examined E3 ubiquitin ligases likely involved in ubiquitination. After analyzing the Co-IP/LC-MS/MS data (Supplementary Table S3), six E3 ubiquitin ligases were fished out. Among them, tripartite motif-containing protein 21 (TRIM21) exhibits the highest score and the greatest number of identified peptides (Fig. 6A). Given that TRIM21 plays a multifaceted role in limiting viral replication during various viral infections, we sought to determine whether TRIM21 is potentially involved in UXT-V2induced gB degradation. Accordingly, HEK293T cells were transfected with plasmids expressing His-tagged UXT-V2, Flag-tagged gB, and HA-Fig. 7. HSV-2 infection downregulates the expression of UXT-V2. A ARPE-19 cells were infected with or without HSV-2 at an MOI of 0.2 PFU/cell at 0, 6, 12, 24 and 30 hpi. The transcripts of gB and UXT-V2 were detected by qPCR. B ARPE-19 cells were infected with or without HSV-2 at an MOI of 0.2 PFU/cell. The cells were collected at 0, 3, 6, 12, 18, and 24 hpi, and the expression of endogenous UXT-V2 was detected by Western blotting. C ARPE-19 cells were transfected with 1 μg plasmid expressing His-UXT-V2, followed by infection with or without HSV-2 at an MOI of 0.2 PFU/cell at 6 hpt. The cells were collected at 0, 12, 18, 24, or 36 hpi, and the expression of His-UXT-V2 was detected by Western blotting. D ARPE-19 cells were infected with or without HSV-2 at an MOI of 0, 0.02 or 1 PFU/cell. At 24 hpi, the expression of endogenous UXT-V2 was detected by Western blotting. E Vero cells were infected with or without ZIKV at an MOI of 0.5 PFU/cell. The cells were collected at 8, 12 or 24 hpi, and the expression of endogenous UXT-V2 was detected by Western blotting. F TZM-bl cells were infected with or without HIV-1 at an MOI of 1 PFU/cell. The cells were collected at 0, 3, 6, 12, 18, and 24 hpi, and the expression of endogenous UXT-V2 was detected by Western blotting. G Six-to eight-week-old female BALB/c mice were challenged intravaginally (i.vag.) with HSV-2 (G strain). The mice were sacrificed when genital ulceration and severe inflammation were observed at 7 dpi. Tissue sections with a thickness of 4 μm were employed for the immunofluorescence staining of UXT-V2 and HSV-2 ICP5. H The Mean Fluorescence Intensity (MFI) values of UXT-V2 were obtained from individual GFP + or GFP-cells. For this quantification, individual cells were selected using the freehand selection tool in ImageJ software following the borders of each individual cell. Over 10 cells were selected for quantification in ImageJ software. For graphs, data shown are mean ± SD of three independent experiments with each condition performed in triplicate. For images, one representative experiment out of three is shown. tagged TRIM21. Co-IP and Western blotting analyses demonstrated that HSV-2 gB physically interacts with both UXT-V2 and TRIM21 (Fig. 6B). To further validate the TRIM21-mediated polyubiquitination of HSV-2 gB, we established TRIM21 KO HeLa cell lines (Fig. 6C). Notably, the TRIM21-mediated K48-linked polyubiquitination of HSV-2 gB was abolished in UXT-V2 knockdown HeLa cells, as evidenced by the loss of this modification in the TRIM21 KO HeLa cell lines (Fig. 6D). These findings collectively indicate that UXT-V2 promotes K48-linked polyubiquitination of HSV-2 gB by recruiting the E3 ubiquitin ligase TRIM21.
## Dynamics of UXT-V2 expression in response to HSV-2 infection
Having identified UXT-V2 as an antiviral protein capable of inhibiting HSV-2 proliferation, we hypothesized that HSV-2 may have evolved specific countermeasures to overcome this restriction and facilitate its replication. We next sought to determine whether HSV-2 infection affects UXT-V2 expression. ARPE-19 cells were infected with or without HSV-2 at an MOI of 0.2 PFU/cell, and the dynamics of UXT-V2 were examined. The results demonstrated that the mRNA levels of gB and UXT-V2 increased as the infection progressed (Fig. 7A). The protein level of UXT-V2 was also examined at different time points following HSV-2 infection, showing that HSV-2 infection resulted in a significant decrease of protein level of UXT-V2 (Fig. 7B). The amount of UXT-V2 began to decrease at 12 hpi and continued to significantly decrease at 24 hpi. We subsequently examined whether HSV-2 infection inhibits the expression of exogenous UXT-V2. As shown in Fig. 7C, similar to endogenous UXT-V2, the protein level of exogenous UXT-V2 decreased in the context of HSV-2 infection.
ARPE-19 cells were infected with different titers of HSV-2, and the cells were collected at 24 hpi for subsequent analyses. Western blotting results informed a significant decrease of endogenous UXT-V2 with the increase of virus titers (Fig. 7D). Likewise, HSV-1 infection also inhibited UXT-V2 expression (Supplementary Fig. S4), whereas infection with two unrelated viruses HIV-1 or ZIKV had no such effect (Fig. 7E andF). The observed increase in mRNA expression level and decrease in protein level of UXT-V2 along with HSV-2 infection suggest that HSV-2 likely promotes the degradation of UXT-V2 protein.
Furthermore, six-to eight-week-old female BALB/c mice were challenged intravaginally (i.vag.) with HSV-2 to investigate the potential interaction between the virus and UXT-V2 in vivo. The vaginal tissues were collected for immunofluorescence staining, revealing a colocalization of UXT-V2 and HSV-2 ICP5 in the mouse vaginal tissues (indicated by white arrows in Fig. 7G). Moreover, the expression level of HSV-2 at an MOI of 0.5 PFU/cell in the absence or presence of 100 μg/mL phosphonoacetic acid (PAA). At 8, 12 and 24 hpi, the expression of HSV-2 gD and endogenous UXT-V2 was detected by Western blotting. For images, one representative experiment out of three is shown.
UXT-V2 in HSV-2 infected cells was lower than that in uninfected cells (Fig. 7H). These findings are in agreement with the results from cell lines, indicating that HSV-2 infection leads to the reduction of UXT-V2 expression.
## HSV-2 inhibits UXT-V2 expression depending on virus replication
Having demonstrated that HSV-2 infection reduces UXT-V2 expression, we next examined the potential underlying mechanism. As shown in Fig. 7A, HSV-2 infection reduced the protein level of UXT-V2 but did not affect its mRNA level. There was no detectable effect on UXT-V2 expression when cells were treated with a translational inhibitor cycloheximide (CHX) (Fig. 8A), while the expression level of UXT-V2 in HSV-2-infected cells was significantly decreased at 12 hpi, indicating that HSV-2 inhibits UXT-V2 expression depending on viral replication. To further validate the findings, ARPE-19 cells were infected with UV-HSV-2 or live HSV-2 at an MOI of 0.5 PFU/cell, and the cells were collected at 8, 12 and 24 h after treatment. Western blotting results showed that the expression level of UXT-V2 significantly decreased at 24 hpi in HSV-2-infected cells while UXT-V2 protein level had no change following an incubation with UV-HSV-2 (Fig. 8B). A comparable experiment was performed in the presence of phosphonoacetic acid (PAA), a well-documented inhibitor of viral DNA polymerase. As shown in Fig. 8C, the reduction of UXT-V2 expression was abolished and the viral protein gD was decreased in the presence of PAA. Taken together, these results indicate that HSV-2 inhibits UXT-V2 expression in a replication-dependent manner.
## DISCUSSION
The UXT gene encodes two splicing variants, namely UXT-V1 and UXT-V2 (Huang et al., 2012). UXT-V1 is implicated in mediating the formation of antiviral signalosomes (Huang et al., 2012), while UXT-V2 is involved in enhancing NF-κB signaling (Sun et al., 2007). UXT-V2 is ubiquitously expressed in human and mouse tissues. To date, most of the studies on UXT-V2 have been focusing on its roles in tumor immune responses (Markus et al., 2002;Taneja et al., 2004;Zhao et al., 2005;Chen et al., 2013;Sanchez-Morgan et al., 2017;Zeng et al., 2019). The function of UXT-V2 in viral infection remains unclear. In the current study, we revealed the critical role of UXT-V2 in inhibiting HSV-2 replication by targeting gB for degradation. In response, HSV-2 suppresses UXT-V2 expression to counteract its antiviral effects, further confirming the negative correlation between UXT-V2 expression and HSV-2 replication (Schematic diagram in Fig. 9). For the first time, our findings document a considerable association between HSV-2 infection and UXT-V2.
UXT-V2 belongs to α-type prefoldin protein family (Enunlu et al., 2011), which plays crucial roles in assembling cytoskeleton and other cytoplasmic complexes, as well as preventing functional proteins by inhibiting protein aggregation and promoting proteolytic degradation (Abe et al., 2013). α-type prefoldin proteins can also facilitate the ubiquitination of their substrates. For instance, prefoldin PFDN5 facilitates c-Myc degradation by recruiting the ubiquitin E3 ligase MM-1 (Kimura et al., 2007), while prefoldin UXT-V2 promotes the ubiquitination of AR via VHL (von Hippel-Lindau) to regulate AR-dependent transactivation (Chen et al., 2013). Meanwhile, when analyzing the signaling pathway following UXT knockdown, the differentially expressed genes were found to be specifically enriched in the gene ontology related to ubiquitin-mediated proteolysis, cell cycle, and p53 signaling (Su et al., 2016). Consistent with these results, we observed that treatment with the proteasome inhibitor MG132 significantly restored gB expression levels. Furthermore, we found that UXT-V2 promotes the ubiquitination of HSV-2 gB. Therefore, we conclude that UXT-V2 induce gB degradation through the ubiquitin-proteasome pathway. In general, UXT-V2 facilitates the ubiquitination of its substrates to exert its transcriptional activation capacity (Chen et al., 2013), and the induction of protein degradation represents a novel property of UXT-V2. HSV (HSV-1/2) envelope glycoprotein B is essential for HSV replication by forming a bridge between viral and cellular membranes to promote fusion (Cooper and Heldwein, 2015;Luo et al., 2015). HSV-2 gB can also mediate the fusion of the virion envelope with the outer nuclear membrane (NM) during HSV-2 egress (Wright et al., 2009). In addition, HSV-2 gB facilitates cell-to-cell spread of the virus (Fan et al., 2002;Hensler et al., 2014). Therefore, the abundance of gB could be important for HSV-2 infection. It is known that F-box only protein 2 (FBXO2) suppresses EBV replication by targeting gB for degradation (Chang et al., 2012). In the current study, UXT-V2 was identified as a restriction factor for HSV-2 replication. It facilitates the ubiquitination and subsequent degradation of gB via the ubiquitin-proteasome pathway, thereby inhibiting HSV-2 replication. It is known that K63-linked ubiquitin chains regulate the trafficking of gB in the endosomal pathway, and cellular ubiquitination of HSV-1 gB correlates with increased release of exosomal material (Calistri et al., 2007). We demonstrate that UXT-V2 facilitates K48 ubiquitination of HSV-2 gB, a residue implicated in proteasome-dependent degradation. Further investigations are warranted to explore the impact of UXT-V2 and the type ubiquitination of gB on the viral life cycle.
The impact of UXT-V2 on NF-κB signaling pathways is complex. Previous study has shown that the interaction between UXT-V2 and p65 is dependent on external stimuli, while overexpression of UXT-V2 did not markedly affect NF-κB activation nor detectable basal NF-κB binding activity (Sun et al., 2007). In agreement, we reveal that overexpression of UXT-V2 does not affect NF-κB activation (Fig. 2C), while exogenous UXT-V2 inhibits HSV-2 replication via a NF-κB independent pathway. Although UXT-V2 restricts HSV-2 replication, HSV-2 can still propagate in UXT-V2-expressing cells, indicating the presence of viral countermeasures. Others have reported that the Epstein-Barr Virus BGLF4 mediates the phosphorylation of UXT-V2 in the lytic cycle (Chang et al., 2012). It is well known that viruses have developed distinct strategies to inhibit the expression of host cell proteins in order to facilitate virus replication (Wang et al., 2012;Wang et al., 2014;Xiao et al., 2016). UXT-V2 is a long half-life protein that is stably expressed in resting cells (Huang et al., 2011). In our study, HSV-2 infection led to an increase in UXT-V2 mRNA levels but a decrease at protein level, indicating the complexity of HSV-2-UXT-V2 interactions. Given the established associations between UXT-V2 and autophagy (Pan et al., 2020), our results suggest that HSV-2 infection suppreses UXT-V2 expression. Although UXT-V2 has been demonstrated to play a pivotal role in enhancing the mTOR signaling pathway and suppressing autophagy (Pan et al., 2020), its specific impact on autophagic processes in HSV-2-infected cells remains unclear. Beyond the scope of this study, further investigation is needed to elucidate the mechanisms by which UXT-V2 promotes signaling pathways and suppresses autophagy in HSV-2-infected cells.
## CONCLUSIONS
In conclusion, we demonstrate that UXT-V2 suppresses HSV-2 replication, playing an antiviral role during HSV-2 infection. Furthermore, we show that UXT-V2 induces gB degradation through ubiquitinproteasome pathway, whereas HSV-2 in turn reduces the abundance of UXT-V2 protein. Our findings collectively shed light on the complexity of HSV-2-host immune interactions, providing insight into the role of UXT-V2 as a viral restriction factor and the mechanisms underlying viral countermeasures.
## MATERIALS AND METHODS
## Cell lines and viruses
Human embryonic kidney 293T (HEK293T) cell line, baby hamster kidney cell line BHK-21, African green monkey kidney cell line Vero and Vero E6, and human cervical epithelial cell line HeLa were purchased from the American Type Culture Collection. TZM-bl cell line was obtained from the AIDS Research and Reference Reagent Program, Division of AIDS, National Institutes of Health. HEK293T, BHK-21, Vero, Vero E6, HeLa and TZM-bl were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS), 100 Units/mL penicillin and 100 Units/mL streptomycin at 37 • C in 5% CO 2 . ARPE-19 cell line was purchased from the American Type Culture Collection and maintained in a 1:1 mixture of Dulbecco's modified Eagle's medium and F12 medium (Gibco) supplemented with 10% fetal bovine serum (FBS), 100 Units/mL penicillin and 100 Units/ mL streptomycin at 37 • C in 5% CO 2 .
HSV-2 (strain G) was obtained from LGC standards. HSV-1 (strain F) was kindly provided by Professor Chunfu Zheng (University of Calgary). Both HSV-2 and HSV-1 were propagated in African green monkey kidney cells (Vero). Virus stock was stored at -80 • C before used for infection. To obtain UV-inactivated virus (UV-HSV-2), the HSV-2 suspension was placed in 10 cm dishes and exposed to UV light for 15 min. The Zika virus (ZIKV) strain (Zika virus/SZ01/2016/China, GenBank: KU866423.2) was propagated in Vero cells culturing in DMEM medium containing 2% FBS. Virus titer was determined by a plaque-forming assay on Vero cells as previously described (Tan et al., 2018).
## Antibodies
The commercial antibodies used in this study include: anti-Flag monoclonal antibody (1:5000 for immunoblotting, Sigma, St. Louis, MO; F3165), anti-His monoclonal antibody (1:2500 for immunoblotting, Sigma A7058), anti-HA polyclonal antibody (1:5000 for immunoblotting, Proteintech Group, Wuhan, China; S1064), anti-STING monoclonal antibody (1:2000 for immunoblotting, Cell Signaling Technology, Massachusetts, USA; #13647), anti-Ub polyclonal antibody (1:1000 for immunoblotting, Cell Signaling Technology 3936S), anti-UXT-V2 monoclonal antibody (1:1000 for immunoblotting, Sigma QC16645 and 1:1000 for immunoblotting, Mreda M198774), anti-TRIM21 antibody (1:2000 for immunoblotting, Proteintech Group, Wuhan, China; 83530-3-RR), anti-HSV1 + HSV2 gB antibody (1:2000 for immunoblotting, Abcam ab6506), and anti-HSV ICP5 Major Capsid Protein antibody (1:3000 for immunoblotting, Abcam Ab6508).
## Plasmid constructions and transfection
The ORFs of various HSV-2 (strain G) viral genes [UL2, UL3, UL4, UL5, UL6, UL7, UL8, UL9, UL11, UL12, UL13, UL14, UL15, UL17, UL18, UL19, UL20, UL24, UL31, UL41, UL45, UL46, UL47, UL48, UL49, UL54 (ICP27), UL55, RL2 (ICP0), UL27 (gB), UL44 (gC), US6 (gD), US8 (gE), US4 (gG), UL22 (gH), UL10 (gL), US5 (gJ), UL10 (gM), and UL53 (gK) ] were cloned into pcDNA3.1(+) as previously described (Zhang et al., 2015;Guan et al., 2019). Human UXT-V1 and UXT-V2 genes with His tag were cloned into pcDNA3.1(+). Human TRIM21 gene with HA tag was cloned into pcDNA3.1(+). All the plasmids were verified by DNA sequencing analysis (Sunny Biotechnology, Shanghai, China). pLenti-CRISPR.v2 was from Addgene. Transfection of plasmids into cells was carried out using Lipofectamine 2000 (Life Technology, Selangor, Malaysia. 11668019) according to the manufacture's protocol.
## Luciferase reporter assay
HEK293T cells pre-seeded on 24-well plates (2 × 10 5 cells/well) were co-transfected with 1 μg empty vector or 1 μg plasmid expressing His-UXT-V2 and reporter plasmid NF-κB-Luc. At 24 hpt, cells were treated with 10 ng/mL TNF-α for 6 h. Cell lysates were used for measuring Firefly luciferase activities using the Luciferase Reporter Assay System (E1501; Promega) according to the manufacturer's instructions. For some experiments, HeLa cells were transfected with reporter plasmid PRD(III-I) 4 -Luc (The PRD III-I region of IFN-β promoter is specifically recognized and bound by activated members of the IRF family; Thanos and Maniatis, 1995) or reporter plasmid NF-κB-Luc, followed by infection with HSV-2 at an MOI of 0.2 PFU/cell. At 18 hpi, cells were treated with 10 ng/mL TNF-α for 6 h. Cell lysates were used for measuring Firefly luciferase activities using the Luciferase Reporter Assay System (E1501; Promega) according to the manufacturer's instructions.
## Viral infection
The viral infection assay was carried out as previously described (Jenssen et al., 2008;Wang et al., 2018). Briefly, the cell cultures were washed with PBS for three times, followed by the addition of HSV-2 (MOI = 0.2 PFU/cell), JEV (MOI = 0.5 PFU/cell) or ZIKV (MOI = 0.5 PFU/cell) for an incubation at 37 • C for 1 h. After the removal of viruses, cells were washed once with PBS and maintained in fresh medium supplemented with 5% FBS. The samples were collected at different hours post-infection (hpi).
## RNA extraction, reverse transcription, and real-time PCR
Total RNA was isolated from the cells using the TRIzol® Reagent (Invitrogen) following the manufacturer's protocol. cDNA was then synthesized with M-MLV reverse transcriptase (Promega) and random hexamer primers (TaKaRa). The newly synthesized cDNA was used as the template for the amplification of a highly specific nucleotide region of target genes in quantitative PCR (qPCR) assay. Relative quantitative PCR was performed using a SYBR Green Real-Time PCR Master Mix (Toyobo) Dye and an ABI StepOne real-time PCR system (Applied Biosystems). The final reaction conditions were as follow: 95 • C for 1 min, followed by 40 cycles of 95 • C for 15 s, 60 • C for 15 s, and 72 • C for 45 s. The difference in gene expression was calculated on the basis of 2 -ΔΔCT values. The primers used in this study were shown in Supplementary Table S1.
## Western blotting
The collected cells were lysed in the lysis buffer (1 M Tris-HCl, 150 mM NaCl, 200 mM EDTA.2Na, 1% Triton X-100) supplemented with protease inhibitor cocktail (1:100, MedChemExpress, HY-K0010). The proteins in supernatants were separated by 12% sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto 0.45 μm polyvinylidene difluoride membranes (Millipore). Nonspecific binding was blocked using 5% non-fat milk in TBST for 1 h at room temperature. The membranes were incubated with appropriate antibodies overnight at 4 • C and then washed three times with TBST, followed by incubation for 1 h with HRP conjugated secondary antibody (1:10,000, BA1050, Boster). After three washes with TBST, the bands were visualized by exposure to FluorChem HD2 Imaging System (Alpha Innotech) following the addition of chemiluminescent substrate (SuperSignal® West Dura Extended Duration Substrate, 34075, Thermo Scientific Pierce).
## Co-immunoprecipitation
HEK293T cells were seeded on 10 cm bottom culture dishes and transfected with 5 μg plasmid expressing Flag-gB followed by infection with or without HSV-2 strain G at an MOI of 0.2 PFU/cell. At 24 h posttransfection (hpt), cells were harvested and lysed on ice with 1 mL lysis buffer. For each IP, 500 μL lysates were incubated with anti-Flag MAbs or control mouse IgG antibody, and 30 μL fresh Dynabeads protein G (Invitrogen, 10009D) for at least 4 h at 4 • C. The complexes were washed three times with 1 mL of lysis buffer and then subjected to Western blotting analysis.
HeLa cells were seeded on 10 cm culture dishes overnight and transfected with an empty vector or 5 μg plasmid expressing HA-TRIM21, followed by infection with HSV-2 strain G at an MOI of 0.2 PFU/cell. At 24 hpt, cells were harvested and lysed on ice with 1 mL lysis buffer. For each IP, 500 μL lysates were incubated with anti-HSV-2 gB MAbs or control mouse IgG antibody, and 30 μL fresh Dynabeads protein G (Invitrogen, 10009D) for at least 4 h at 4 • C. The complexes were washed three times with 1 mL of lysis buffer and then subjected to Western blotting analysis.
## Liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS)
HeLa cells were infected with or without HSV-2 at an MOI of 0.2 PFU/cell. The cells were collected at 24 hpt and lysed on ice with 1 mL lysis buffer. Lysis buffer lysates were then incubated with anti-UXT-V2 MAbs, and 30 μL fresh Dynabeads protein G (Invitrogen, 10009D) for at least 4 h at 4 • C. The complexes were washed three times with 1 mL of lysis buffer before used for MS/MS analysis (Shanghai Bioprofile Technology, Shanghai, China).
## Mice challenging and tissue harvesting
Six-to eight-week-old female BALB/c mice were housed in specific pathogen-free conditions. Five to seven days prior to challenge, mice were subcutaneously injected with Depo-Provera in the neck ruff (medroxyprogesterone acetate, 2 mg/mouse). Mice (n = 3/group) were anesthetized with pentobarbital sodium and challenged intravaginally (i.vag.) with 10 μL/mouse HSV-2 (G strain) at a concentration of 2 × 10 7 PFU/mL. Following the challenge, the weight and clinical symptoms of all mice were monitored every day. Mice were sacrificed when genital ulceration and severe inflammation were observed. The vaginal tissues of each mouse were harvested and immersed in 10% neutral buffered formalin for 24 h to ensure optimal fixation.
## Immunofluorescence staining of cells in tissues
Following fixation, the tissues underwent a transition into 70% ethanol and were subsequently embedded in paraffin. Tissue sections with a thickness of 4 μm were employed for the immunofluorescence staining of UXT-V2 and HSV-2 ICP5. Images were collected under a Pannoramic MIDI system (3DHISTECH, Thermo) using Pannoramic scanner software and analyzed by ImageJ (NIH).
## RNA interference
UXT-V2 siRNAs were purchased from QIAGEN (1027416). ARPE-19 cells were seeded on 6-well plates. Negative control siRNA (NC) or UXT-V2 siRNAs were transfected into ARPE-19 cells using HiPerFect Transfection Reagent (QIAGEN, 301705) according to the manufacturer's instruction. At 24 hpt, cells were harvested, and then subjected to qPCR and Western blotting analysis.
The target sequences of UXT shRNA were: 5 ′ -CCT TCA ACT GAG AAA TGT CAT-3 ′ and 5 ′ -GCT GTA ACT TCT TCG TTG ACA-3 ′ . The shRNA constructs were designed by using the pLKO.1 vector according to the manufacturer's instructions. HEK293T cells were co-transfected with 1 μg pCMV-VSVg, 3 μg pCMV-Gag/pol and 4 μg pLKO.1 for 72 h to produce lentiviruses. For shRNA knockdown of UXT, lentiviruses and 7 μg/mL puromycin were added to Vero E6 or HeLa cells pre-seeded in culture dishes. After 12 h incubation, the medium was replaced with fresh complete growth media for another 36 h. Cells were harvested, and then subjected to Western blotting assay to measure the expression level of UXT-V2.
## Generation of UXT-V2 and TRIM21 knockout (KO) cell lines using CRISPR-Cas9
To generate the UXT-V2-KO or TRIM21-KO cell line, the designed UXT sgRNAs or TRIM21 sgRNAs were constructed into pLentiCRISPR.v2 as described previously (Li et al., 2015). On day 0, HEK293T cells were seeded in 10-cm dishes at a density of 4 × 10 6 cells per dish. The cells were then co-infected with 3 μg sgRNA-inserted pLentiCRISPR.v2, 1.5 μg/dish psPAX2, and 0.5 μg/dish pMD2.G. On day 3, the culture media containing the viruses were harvested to infect HeLa cells. The infected cells were then subjected to selection in culture medium supplemented with puromycin (4 μg/mL) for an additional seven days. Single clonal knockout cells were generated by serial dilution and conformed through Sanger sequencing and Western blotting analysis. The sgRNA sequence used to confirm targeting specificity is as follows: sgUXT: 5 ′ -GCGAGACTTGCAAAAGGTGC-3 ′ , sgTRIM21: 5 ′ -AGGTGTTTCGTCCTTTCCA CAAGAT-3 ′ .
## Statistical analysis
All experiments were repeated at least three times. Data are presented as mean values ± SD of three independent experiments. Data analyses were performed with GraphPad Prism 7 software (GraphPad). Comparison between two groups was analyzed by a two-tailed unpaired Student's t-test, whereas comparisons among more than two groups were analyzed by one-way ANOVA with the Tukey test. *P < 0.05 was considered statistically significant.
## References
1. Abe, Takahashi-Niki, Takekoshi et al. (2013) "Prefoldin plays a role as a clearance factor in preventing proteasome inhibitor-induced protein aggregation" *J. Biol. Chem*
2. Bender, Whitbeck, Ponce De Leon et al. (2003) "Specific association of glycoprotein B with lipid rafts during Herpes simplex virus entry" *J. Virol*
3. Byrne, Gantt, Coombs (2018) "Effects of spatiotemporal HSV-2 lesion dynamics and antiviral treatment on the risk of HIV-1 acquisition" *PLoS Comput. Biol*
4. Calistri, Sette, Salata et al. (2007) "Intracellular trafficking and maturation of Herpes simplex virus type 1 gB and virus egress require functional biogenesis of multivesicular bodies" *J. Virol*
5. Chang, Wang, Doong et al. (2012) "Epstein-Barr virus BGLF4 kinase downregulates NF-kappaB transactivation through phosphorylation of coactivator UXT" *J. Virol*
6. Chen, Chen, Zhang et al. (2013) "Regulation of the transcriptional activation of the androgen receptor by the UXT-binding protein VHL" *Biochem. J*
7. Cheshenko, Herold (2002) "Glycoprotein B plays a predominant role in mediating herpes simplex virus type 2 attachment and is required for entry and cell-to-cell spread" *J. Gen. Virol*
8. Cooper, Heldwein (2015) "Herpesvirus gB: a finely tuned fusion machine" *Viruses*
9. Crameri, Bauer, Caduff et al. (1980) "MxB is an interferon-induced restriction factor of human herpesviruses" *Nat. Commun*
10. El Kasmi, Khadivjam, Lackman et al. (2018) "Extended synaptotagmin 1 interacts with Herpes simplex virus 1 glycoprotein M and negatively modulates virus-induced membrane fusion" *J. Virol*
11. Enunlu, Ozansoy, Basak (2011) "Alfa-class prefoldin protein UXT is a novel interacting partner of amyotrophic Lateral Sclerosis 2 (Als2) protein" *Biochem. Biophys. Res. Commun*
12. Everett, Johnson, Bottero et al. (2014) "IFI16 restricts HSV-1 replication by accumulating on the HSV-1 genome, repressing HSV-1 gene expression, and directly or indirectly modulating histone modifications" *PLoS Pathog*
13. Fan, Grantham, Smith et al. (2002) "Truncation of Herpes simplex virus type 2 glycoprotein B increases its cell surface expression and activity in cell-cell fusion, but these properties are unrelated" *J. Virol*
14. Groves (2016) "Genital herpes: a review" *Am. Fam. Physician*
15. Guan, Zhang, Fu et al. (2019) "Herpes simplex virus type 2 Immediate Early Protein ICP27 inhibits IFN-beta production in mucosal epithelial cells by antagonizing IRF3 activation" *Front. Immunol*
16. Hensler, Tomaszewski, Rappocciolo et al. (2014) "Human herpesvirus 8 glycoprotein B binds the entry receptor DC-SIGN" *Virus Res*
17. Huang, Chen, Zhou et al. (2011) "UXT-V1 protects cells against TNF-induced apoptosis through modulating complex II formation" *Mol. Biol. Cell*
18. Huang, Liu, Ge et al. (2012) "UXT-V1 facilitates the formation of MAVS antiviral signalosome on mitochondria" *J. Immunol*
19. Huehn, Beyer (2015) "Epigenetic and transcriptional control of Foxp3+ regulatory T cells" *Semin. Immunol*
20. James, Harfouche, Welton et al. (2016) "Herpes simplex virus: global infection prevalence and incidence estimates" *Bull. World Health Organ*
21. Jenssen, Sandvik, Andersen et al. (2008) "Inhibition of HSV cell-to-cell spread by lactoferrin and lactoferricin" *Antivir. Res*
22. Kimura, Nagao, Fujioka et al. (2007) "MM-1 facilitates degradation of c-Myc by recruiting proteasome and a novel ubiquitin E3 ligase" *Int. J. Oncol*
23. Li, Guan, Du et al. (2015) "Inhibition of HIV-1 infection of primary CD4+ T-cells by gene editing of CCR5 using adenovirus-delivered CRISPR/Cas9" *J. Gen. Virol*
24. Li, Gao, Shao et al. (2018) "Gallid herpesvirus 1 initiates apoptosis in uninfected cells through paracrine repression of p53" *J. Virol*
25. Li, Wang, Jiang et al. (2014) "UXT is a novel regulatory factor of regulatory T cells associated with Foxp3" *Eur. J. Immunol*
26. Liu, Zhu, Li et al. (2023) "Migrasomes released by HSV-2-infected cells serve as a conveyance for virus spread" *Virol. Sin*
27. Luo, Hu, He et al. (2015) "Contribution of N-linked glycans on HSV-2 gB to cell-cell fusion and viral entry" *Virology*
28. Markus, Taneja, Logan et al. (2002) "Identification and characterization of ART-27, a novel coactivator for the androgen receptor N terminus" *Mol. Biol. Cell*
29. Mulvey, Arias, Mohr (2007) "Maintenance of endoplasmic reticulum (ER) homeostasis in herpes simplex virus type 1-infected cells through the association of a viral glycoprotein with PERK, a cellular ER stress sensor" *J. Virol*
30. Pan, Yin, Wang et al. (2020) "Mice deficient in UXT exhibit retinitis pigmentosa-like features via aberrant autophagy activation" *Autophagy*
31. Rao, Pham, Kulkarni et al. (2011) "Herpes simplex virus 1 glycoprotein B and US3 collaborate to inhibit CD1d antigen presentation and NKT cell function" *J. Virol*
32. Sanchez-Morgan, Kirsch, Trackman et al. (2017) "UXT is a LOX-PP interacting protein that modulates Estrogen receptor alpha activity in breast cancer cells" *J. Cell. Biochem*
33. Schroer, Schneider, Ropers et al. (1999) "Cloning and characterization of UXT, a novel gene in human Xp11, which is widely and abundantly expressed in tumor tissue" *Genomics*
34. Sethurathinam, Singh, Panneerselvam et al. (2013) "UXT plays dual opposing roles on SARM-induced apoptosis" *FEBS Lett*
35. Sparrer, Gableske, Zurenski et al. (2017) "TRIM23 mediates virus-induced autophagy via activation of TBK1" *Nat. Microbiol*
36. Stampfer, Lou, Cohen et al. (2010) "Structural basis of local, pH-dependent conformational changes in glycoprotein B from herpes simplex virus type 1" *J. Virol*
37. Su, Li, Lei et al. (2016) "The EZH1-SUZ12 complex positively regulates the transcription of NF-kappaB target genes through interaction with UXT" *J. Cell Sci*
38. Suazo, Tognarelli, Kalergis et al. (2015) "Herpes simplex virus 2 infection: molecular association with HIV and novel microbicides to prevent disease" *Med. Microbiol. Immunol*
39. Sun, Tang, Lou et al. (2007) "UXT is a novel and essential cofactor in the NF-kappaB transcriptional enhanceosome" *J. Cell Biol*
40. Tan, Zhang, Sun et al. (2018) "ZIKV infection activates the IRE1-XBP1 and ATF6 pathways of unfolded protein response in neural cells" *J. Neuroinflammation*
41. Taneja, Ha, Swenson et al. (2004) "ART-27, an androgen receptor coactivator regulated in prostate development and cancer" *J. Biol. Chem*
42. Temme, Eis-Hubinger, Mclellan et al. (2010) "The herpes simplex virus-1 encoded glycoprotein B diverts HLA-DR into the exosome pathway" *J. Immunol*
43. Thomas, Mita, Ha et al. (2018) "Role of the unconventional prefoldin proteins URI and UXT in transcription regulation" *Adv. Exp. Med. Biol*
44. Thanos, Maniatis (1995) "Virus induction of human ifnβ gene expression requires the assembly of an enhanceosome" *Cell*
45. Wang, Hour, Lai et al. (2018) "Epigallocatechin-3-gallate inhibits the early stages of Japanese encephalitis virus infection" *Virus Res*
46. Wang, Fang, Li et al. (2012) "Foot-and-mouth disease virus 3C protease cleaves NEMO to impair innate immune signaling" *J. Virol*
47. Wang, Chi, Wei et al. (2014) "Influenza A virus-induced degradation of eukaryotic translation initiation factor 4B contributes to viral replication by suppressing IFITM3 protein expression" *J. Virol*
48. Weed, Dollery, Komala Sari et al. (2018) "Acidic pH mediates changes in antigenic and oligomeric conformation of Herpes simplex virus gB and is a determinant of cell-specific entry" *J. Virol*
49. Wright, Wisner, Hannah et al. (2009) "Fusion between perinuclear virions and the outer nuclear membrane requires the fusogenic activity of herpes simplex virus gB" *J. Virol*
50. Xiao, Ma, Wang et al. (2016) "Downregulation of protein kinase PKR activation by porcine reproductive and respiratory syndrome virus at its early stage infection" *Vet. Microbiol*
51. Yang, Zhao, Wu et al. (2009) "The single-macro domain protein LRP16 is an essential cofactor of androgen receptor" *Endocr. Relat. Cancer*
52. Yoon, Choi, Kim et al. (1955) "UXT chaperone prevents proteotoxicity by acting as an autophagy adaptor for p62-dependent aggrephagy" *Nat. Commun*
53. Zeng, Xiang, Zhang et al. (2019) "Ubiquitous expressed transcript promotes tumorigenesis by acting as a positive modulator of the polycomb repressive complex 2 in clear cell renal cell carcinoma" *BMC Cancer*
54. Zhang, Liu, Wang et al. (2015) "HSV-2 immediate-early protein US1 inhibits IFN-beta production by suppressing association of IRF-3 with IFN-beta promoter" *J. Immunol*
55. Zhang, Zhang, Zhang et al. (2016) "USP18 recruits USP20 to promote innate antiviral response through deubiquitinating STING/MITA" *Cell Res*
56. Zhao, Wang, Zhang et al. (2005) "UXT is a novel centrosomal protein essential for cell viability" *Mol. Biol. Cell* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12347268&blobtype=pdf | # Transcriptional Analysis of Spodoptera frugiperda Sf9 Cells Infected with Daphnis nerii Cypovirus-23
Wendong Kuang, Jian Yang, Jinchang Wang, Chenghua Yan, Junhui Chen, Xinsheng Liu, Chunhua Yang, Zhigao Zhan, Limei Guan, Jianghuai Li, Tao Deng, Feiying Yang, Guangqiang Ma, Liang Jin
## Abstract
Daphnis nerii cypovirus-23 (DnCPV-23) is a new type of cypovirus that has a lethal effect on many species of Sphingidae pests. DnCPV-23 can replicate in Spodoptera frugiperda Sf9 cells, but the replication characteristics of the virus in this cell line are still unclear. To determine the replication characteristics of DnCPV-23 in Sf9 cells, uninfected Sf9 cells and Sf9 cells at 24 and 72 h after DnCPV-23 infection were collected for transcriptome analysis. Compared to uninfected Sf9 cells, a total of 188 and 595 differentially expressed genes (DEGs) were identified in Sf9 cells collected at 24 hpi and 72 h, respectively. KEGG analyses revealed that 139 common DEGs in two treatment groups were related to nutrition and energy metabolism-related processes, cell membrane integrity and function-related pathways, detoxification-related pathways, growth and development-related pathways, and so on. We speculated that these cellular processes might be manipulated by viruses to promote replication. This study provides an important basis for further in-depth research on the mechanism of interaction between viruses and hosts. It provides additional basic information for the future exploitation of DnCPV-23 as a biological insecticide.
## 1. Introduction
Daphnis nerii cypovirus-23 (DnCPV-23) was initially isolated from naturally diseased larvae of Daphnis nerii (D. nerii), which belongs to the order Lepidoptera and family Sphingidae, and is a worldwide pest [1] that mainly damages the leaves of the ornamental plant oleander and the medicinal plants Rauvolfia vomitoria Afzel and Catharanthus [1,2]. Compared with other cypoviruses, this new type of cypovirus has different electrophoretic migration patterns and conserved terminal sequences [1,3,4]. In addition to D. nerii, DnCPV-23 can infect and induce death in various harmful insects of the Sphingidae family. However, as a potential biopesticide, the molecular mechanism of the interaction between DnCPV-23 and its hosts remains unclear. The study of the interaction mechanism between DnCPV-23 and hosts is essential for understanding virus virulence and virus genome function, screening host molecules that promote virus replication, and screening complex drugs that improve virus insecticidal efficacy.
We previously performed a transcriptome analysis of DnCPV-23-infected and uninfected D. nerii midguts [1]. DnCPV-23 can infect wild insects such as D. nerii, but the lifecycle of these insects is strongly affected by season and cannot be artificially reproduced at present. Therefore, it is essential to study the effects of DnCPV-23 on cell lines. In addition, DnCPV-23 infection of the D. nerii midgut occurs in the form of occlusion-derived viruses (ODVs) in vivo, and the midgut cell composition is complex. However, the Sf9 cell line can be infected in vitro with free virus particles of DnCPV-23. Based on different infection forms and cell types, we speculate that the transcriptome differs between cultured cell lines and D. nerii midguts. Our previous results revealed that DnCPV-23 can be effectively replicated and passaged in Sf9 cells [5]. However, the molecular mechanism underlying DnCPV-23 infection in Sf9 cells has not been fully elucidated. Moreover, there are few reports on the replication of insect RNA viruses in heterologous cell lines. Perina nuda virus (PnV) can establish persistent infection in a heterologous Lymantria xylina cell line, NTU-LY. Various methods were used to investigate virus replication in NTU-LY cells [6]. Moreover, the honeybee virus deformed wing virus (DWV) can infect the heterologous Lepidopteran haemocytic cell line P1 cells [7]. This study reports the replication of another RNA virus, DnCPV-23, in the heterologous cell line Sf9. Our work provides an important foundation for in-depth research on the replication of RNA viruses in heterologous cells and for establishing a sustained infection cell model for insect RNA viruses. To investigate how Sf9 cells react to DnCPV-23 infection, we used a high-throughput sequencing approach to assess the impact of DnCPV-23 infection on global gene expression in Sf9 cells and analysed the host factors that may affect virus replication. In this study, uninfected Sf9 cells and DnCPV-23-infected Sf9 cells were collected at 24 h and 72 h after infection for transcriptome analysis to investigate changes in the gene expression profiles of Sf9 cells after virus infection and identify the host signalling pathways that may affect virus replication.
We speculate that upregulating some processes (including nutrition and energy metabolism, cell membrane integrity and functions, detoxification, and growth and development) after viral infection is beneficial for DnCPV-23 replication. For example, juvenile hormone acid O-methyltransferase-like (JHAMT) can promote virus proliferation by prolonging insect pupation time; virus replication can be promoted by inhibiting host apoptosis through upregulating the expression of cytochrome P450 9e2 (CYP9E2); and Sequestosome 1 (p62/SQSTM1) affects the lifecycle of viruses by regulating autophagy. Through this study, we have identified some host genes that may affect the replication of DnCPV-23. In future work, we will further verify the interaction between these genes and DnCPV-23, which provides an important theoretical basis for understanding the interaction between viruses and hosts, virus replication in heterologous cell lines, and viral biological insecticide development.
## 2. Results
## 2.1. Viral Infection of Cells
The virus-to-cell ratio is 0.2 polyhedra per cell. Prior to transcriptome analysis, qRT-PCR was used to assess the mRNA levels of the DnCPV-23 S1 gene in infected cells (samples collected at 24 h and 72 h post-infection (hpi)) and in uninfected cells. The results revealed higher relative expression of viral gene mRNA in infected cells than in uninfected cells (Figure 1).
## 2.2. Overview of Transcriptome Sequencing
All of the samples were sequenced independently. Transcriptome sequencing analysis of 15 samples resulted in a total of 81.39 Gb of clean data. The clean data of each sample ranged from 6.41 to 7.04 G. The Q30 of each sample ranged from 94.12% to 94.42%, and the average GC content was 45.96%. The genome alignment of each sample was obtained by mapping reads to the reference genome, with a mapping rate of 81.42% to 82.46%. A total of 595 DEGs were identified, with 506 being upregulated and 89 being downregulated between the CPV_72h samples and the control samples.
## 2.3. Effects of Viral Infection on the Transcriptome Expression of Sf9 Cells
As shown in Figure 2, PCA1 accounted for 75.91%, and PCA2 accounted for 5.75%. Therefore, the percentage of the total of the two was 81.66%, thus accounting for a high proportion and representing the overall population to a large extent. The principal component analysis revealed a clear separation of the samples among the three groups (Figure 2A), indicating that the samples had good repeatability. Compared with uninfected cells, in infected cells, the number of upregulated genes at 24 hpi and 72 hpi was 161 and 506, respectively, and the number of downregulated genes at 24 hpi and 72 hpi was 27 and 89, respectively (Figure 2B). In addition, a heatmap of the gene expression data is presented in Figure 2C. The results suggested that these DEGs could be used to distinguish the samples. The results revealed that viral infection could influence Sf9 gene expression at different time points.
## 2.4. Analysis of DEGs
We analysed 139 DEGs shared by the two treatment groups for further analysis (Figure 3A). KEGG functional enrichment analysis was performed using the common DEGs between the two treatment groups to identify the relevant biological pathways associated with these DEGs. Based on the enrichment score obtained from the KEGG analysis, we identified the 20 most enriched pathways. These pathways play important roles in nutrition and energy metabolism-related processes (for example, "linoleic acid metabolism" and "vitamin digestion and absorption"), cell membrane integrity and function-related pathways (for example, "glycerophospholipid metabolism" and "ether lipid metabolism"), detoxification-related pathways (for example, "ABC transporters"), growth and development-related pathways (for example, "insect hormone biosynthesis"), and insect immune responses (for example, "arachidonic acid metabolism") (Figure 3B). We selected 50 DEGs with the highest expression levels from these 139 DEGs for heatmap analysis. A heatmap of the gene expression data is presented in Figure 3C. The results suggested that the identified DEGs could distinguish the samples and that DnCPV-23 infection influenced gene expression in Sf9 cells. Figure 3C shows the top 50 genes with the highest expression levels among the overlapping 139 DEGs, and the relevant information of these genes is presented in Table 1.
## 2.5. qRT-PCR Validation of DEGs
To verify the reliability of the transcriptome data and the DEG results obtained by RNA-seq, thirteen DEGs were selected for qPCR analysis. As shown in Figure 4, the foldchange values for the infected cells (CPV_72_1) vs. the uninfected cells (Con_1) obtained via the qPCR analysis were consistent with the values obtained via RNA-seq for all of the selected genes.
## 3. Discussion
To analyse the important signalling pathways and genes that may be involved in the interaction between DnCPV-23 and Sf9 cells, we selected 139 common genes between 24 h DEG and 72 h DEG for analysis. The KEGG analysis revealed that many processes changed in Sf9 cells after virus infection, including nutrition and energy metabolism-related processes (such as "linoleic acid metabolism" and "vitamin digestion and absorption"), cell membrane integrity and function-related pathways (such as "glycerophospholipid metabolism" and "ether lipid metabolism"), detoxification-related pathways (such as "ABC transporters"), and growth and development-related pathways (such as "insect hormone biosynthesis"). The KEGG analysis indicated that these processes or pathways might have important roles in the interactions between DnCPV-23 and Sf9 cells. Furthermore, these processes or pathways also changed in the midgut of Daphnis nerii after DnCPV-23 infection [1].
DnCPV-23 infection can affect the expression of critical genes involved in energy metabolism, such as glycerol kinase-like (GK; Gene ID: LOC118276660) and trehalose transporter 1 (Tret1; Gene ID: LOC118271571, LOC118279019). Glycerol kinase (GK) is an enzyme that catalyses the formation of glycerol 3-phosphate from ATP and glycerol [8]. GK2 is a glycerol biosynthesis gene that plays a crucial role in survival during the cold period [9]. HCV/HBV infection reduces the expression of miR-451a, which inhibits GK expression, and miR-451a attenuates hepatitis C virus replication by targeting glycerol kinase. Furthermore, the supplementation of miR-451a could impede lipid deposition, reduce steatohepatitis, and inhibit HCV replication in the liver [10]. After virus infection, two Tret1 transcripts were significantly upregulated. Tret1 is a specific and high-capacity facilitated transporter of trehalose, which is the major sugar found in insect haemolymph fluid, providing energy and promoting growth, metamorphosis, stress recovery, chitin synthesis, and virus replication in insects [11][12][13]. In BmN cells, BmNPV infection promotes the expression of trehalose hydrolysis and transport-related genes to facilitate BmNPV proliferation through the trehalose-PI3K-Akt pathway in the midgut [14]. Moreover, trehalose could facilitate virus replication and shedding of dengue virus (DENV) in Aedes aegypti cells. It is possible that trehalose increases DENV2 infection in Aag2 cells by promoting autophagy to prolong cell survival and enhance virus maturation, and enhancing virus cell entry through the modification of cell membranes [12].
In addition to affecting energy metabolism, DnCPV-23 infection also affected insect metamorphosis, day-night rhythm, and oocyte development-related genes. Juvenile hormone esterase-like (JHE, Gene ID: LOC118263046) was upregulated, induced by DnCPV-23 infection. JHE is the primary juvenile hormone (JH)-specific degradation enzyme that plays a crucial role in regulating JH levels. Depletion of JHE resulted in the extension of Bombyx mori larval stages [15] or larval mortality in Telchin licus (Lepidoptera) [16]. JHE activity is downregulated in larval Adoxophyes honmai following dhoNPV and AdorNPV infection, and small interfering RNAs were proposed to play a role in the downregulation of JHE gene expression in baculovirus-infected caterpillars [17]. The Mamestra brassicae Multiple Nucleopolyhedroviruses (MbMNPV) infection of H. armigera larvae inhibits the expression of JHE following the upregulation of JH titre. The upregulation of JH titre prevents the pupation of H. armigera and promotes MbMNPV replication through the JH-Met-Kr-h1 signalling pathway [18]. DnCPV-23 infection promoted the expression of juvenile hormone acid O-methyltransferase-like (JHAMT; Gene ID: LOC118265567). Juvenile hormone (JH) titres in insects are regulated by synthesis with JHAMT and catabolism with JHE [19]. JH could regulate metamorphosis, reproduction, diapause, and polyphenisms [20,21]. The knockdown of JHAMT reduced JH titre, leading to larval death in Aedes aegypti [20] and changing the phenotype of S. frugiperda [22]. During the early stages of infection, Zika virus (ZIKV) regulates the expression of some ribosomal protein genes (RpL23 and RpL27) through the JH-Met-Tai signalling pathway. Moreover, the upregulated expression of RpL23 and RpL27 could facilitate the translation of viral proteins and promotes ZIKV infection in Aedes aegypti [23]. Further research is needed to investigate the detailed effects of these host genes on DnCPV-23 infection.
After viral infection, the expressions of many detoxification-related genes were upregulated, such as cytochrome P450 9e2 (CYP9E2), cytochrome P450 6B7 (CYP6B7), cytochrome P450 6B6 (CYP6B6), glutathione S-transferase (GST), UDP-glucuronosyltransferase (UGT), and multidrug resistance-associated protein 4-like (MRP4/ABCC4). CYP9E2 (Gene ID: LOC118264059, LOC118264054, LOC118262727), CYP6B7 (Gene ID: LOC118279736, LOC118279835) and CYP6B6 (Gene ID: LOC118263048, LOC118262642) are involved in xenobiotic detoxification in insects. In Helicoverpa armigera, CYP6B6 plays crucial roles in the transformation of esfenvalerate and capsaicinoids [24,25]. In H. armigera, CYP6B7 is known as an important detoxification gene in response to fenvalerate [26,27]. In honeybees, thiacloprid can be specifically metabolised by CYP9E2 [28]. In addition, the CYP9E2 gene of Bombyx mori can be regulated by the microRNA bmo-miR-31-5p to inhibit apoptosis and promote BmNPV proliferation [29]. In the future, we will further analyse the role of these CYP450 genes in virus replication. GST (Gene ID: LOC118261931) is a group of enzymes associated with detoxification and plays an important role in detoxifying endogenous and exogenous compounds [30]. The expression of GST could be regulated by viral infection [31,32]. GST may modify lipids to inhibit viral infection, or regulate inflammatory-like mediators to exert antiviral effects [33]. However, GST could promote BmNPV proliferation through regulating glutathione (GSH) levels [34]. DnCPV-23 infection also induced the expression of another detoxification enzyme, UGT2B15-like (Gene ID: LOC118278849). UDP-glucuronosyltransferases (UGTs) could transform various exogenous and endogenous compounds, which play a critical role in detoxification and homeostasis in insects [35]. Moreover, UGTs could regulate viral infection. It is speculated that UGT33D1 reduces oxidative stress to promote BmNPV infection [36]. DnCPV-23 replication could promote the expression of MRP4/ABCC4 (Gene ID: LOC118262225). MRP4 (multidrug resistance-associated protein 4) is a member of the ATP-binding cassette (ABC) transporter C subfamily, which are transporters of various substrates including endogenous compounds, xenobiotic compounds, and nutrients [37,38]. HIV-1 infection induced a significant increase in MRP4 expression in human macrophages. The increase in MRP4 expression may favour the efflux of antiretroviral drugs in macrophages to promote virus replication [39].
DnCPV-23 could upregulate the expression of Sequestosome 1 (p62/SQSTM1, Gene ID: LOC118273096, LOC118273082). BmCPVs can induce mitophagy, which is a form of autophagy, through the interaction of VP4 with host Tom40 to promote self-replication [40]. Additionally, BmCPV replication could be attenuated by vsp21-induced autophagy [41]. Thus, the impact of autophagy on virus replication may be opposite in different situations. Many studies have shown that autophagy is a double-edged sword for virus replication [42][43][44]. We used autophagy inhibitors to analyse the effects of blocking different stages of the autophagy process on virus replication. Compared with the control (DMSO), treatment with Baf A1 significantly restricted virus replication, whereas treatment with 3-MA or CQ did not affect virus replication (Supplementary Figure S1). The results suggested that maintaining the acidic pH and protein degradation ability of lysosomes might be crucial for virus replication [45,46]. In addition, we found that DnCPV-23 infection promoted the expression of Sequestosome 1 (SQSTM1/p62), an adaptor protein of selective autophagy, to promote the degradation of aggregate-prone proteins. SQSTM1 is a wellknown autophagy cargo receptor (ACR) that plays a dual role during pathogen infection; it can promote the degradation of viral proteins, aiding in the elimination of the virus, but can also be manipulated by viruses to promote mitochondrial degradation, thereby suppressing immune responses [47]. A previous study reported that Baf A1 treatment can decrease the protein level of SQSTM1 [48]. Above all, we speculate that the DnCPV-23 infection of host cells promotes autophagy and self-replication by upregulating SQSTM1 expression, which can be blocked by Baf A1. However, in this study, we did not explore it thoroughly. To confirm this conclusion, further research is needed to analyse the impact of lysosomal pH on DnCPV-23 replication and elucidate whether Baf A1 affects the replication of DnCPV-23 by downregulating the expression of SQSTM1. In the future, we will (1) use CRISPR/Cas9 to knock out SQSTM1 in Sf9 to determine whether Baf A1 inhibits DnCPV replication through SQSTM1; (2) use immunofluorescence, electron microscopy scanning, Western Blotting, and RT qPCR to determine which stage of the virus Baf A1 specifically affects:
(3) use knockdown, overexpression, or inhibitor inhibition assays to analyse the host factors involved in Baf A1 affecting virus replication [49][50][51].
In Sf9 cells, we revealed substantial differences in the transcription of genes related to energy metabolism, growth, development, and detoxification processes induced by DnCPV-23 replication. We speculate that these cellular processes might be regulated by DnCPV-23 to promote its replication. In addition, we found that autophagy-related genes also undergo changes, and we need further research to determine the role of autophagy in the DnCPV-23 replication. In these pathways, we analysed some specific host genes that may be involved in DnCPV-23 replication. This study provides important preliminary work for future validation experiments. This study analysed the differentially expressed genes (DEGs) of Sf9 cells after DnCPV-23 infection for the first time, and identified host genes that may be involved in DnCPV-23 replication. Our work enriched the research on the interaction between cypovirus and hosts. However, there has been no definitive and credible validation of the role of relevant host genes in virus replication, making it difficult for readers to determine which host genes are truly involved in the lifecycle of DnCPV-23. In summary, the results obtained in this study provide an important foundation for further in-depth research on the mechanism of interaction between DnCPV-23 and its hosts.
## 4. Materials and Methods
## 4.1. Cell Line, Virus Stock, Inhibitors, and Antibodies
Sf9 cells, an ovary-derived cell line of Spodoptera frugiperda, were cultured at 28 • C in Grace's insect medium (Gibco, Waltham, MA, USA) supplemented with 10% foetal bovine serum (Gibco, USA). DnCPV-23 was initially isolated from the larvae of D. nerii and propagated in D. nerii larvae [1]. The suspension of DnCPV-23 polyhedra used for infecting Sf9 cells was stored at 4 • C in the dark. 3-Methyladenine (3-MA) (Selleck, Houston, TX, USA) was used at 5 mM, chloroquine (CQ) (MedChemExpress, Monmouth Junction, NJ, USA) was used at 40 µM, and bafilomycin A1 (Baf A1) (MedChemExpress) was used at 20 nM. For immunoblotting, rabbit polyclonal antibodies against the DnCPV-23 protein encoded by the viral S1 gene (ABclonal Biotechnology, Woburn, MA, USA) were used at a 1:500 dilution, and mouse monoclonal antibodies against tubulin (Abbkine, Atlanta, GA, USA) were used at a 1:10,000 dilution.
## 4.2. Virus Inoculation
To promote virus infection in cells, purified DnCPV-23 polyhedra at a concentration of approximately 2 × 10 7 /mL were treated with 0.2 M Na 2 CO 3 -NaHCO 3 (pH 10.8), after which the pH was adjusted to 7.4 with 1 M Tris-HCl (pH 6.8) buffer. The free virus particles were stored at -80 • C. Sf9 cells were seeded in 6-well plates, and when they reached a confluency of approximately 75%, they were infected with free DnCPV-23 virus particles (50 µL of the viral suspension containing free DnCPV-23 particles without polyhedral particles). After incubation for 12 h, the cell culture medium was replaced with Grace medium containing 10% (v/v) FBS, and the cells were harvested 24 h and 72 h after infection. Uninfected cells were used as controls.
## 4.3. RNA Extraction, Library Preparation, and RNA-Seq
Control Sf9 cells and Sf9 cells infected with DnCPV-23 were collected at 24 h and 72 h postinfection. All RNA sequencing (RNA-seq) procedures were conducted by the Oebiotech Company (Shanghai, China). Total RNA was extracted from Sf9 cells using TRIzol reagent (Invitrogen, Waltham, MA, USA) in accordance with the manufacturer's protocols. The RNA integrity and concentrations were assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Nine RNA samples (including three uninfected samples and nine infected samples) with confirmed RNA integrity were used to construct the libraries. The cDNA libraries were prepared using a TruSeq RNA Sample Preparation Kit (Illumina, San Diego, CA, USA) in accordance with the manufacturer's protocols. Thereafter, the obtained cDNA libraries were sequenced on the Illumina HiSeq2500 platform, which generated paired-end raw reads of 150 bp.
## 4.4. RNA-Seq Data Analysis
To obtain clean reads, raw reads in fastq format were first processed using fastp 1, and the low-quality reads were removed. The clean reads were mapped to the reference genome ZJU_Sfru_1.0 GCF_011064685.1 using HISAT2 (version 2.2.4) [52]. Gene expression levels were normalised as fragments per kilobase of transcript per million mapped reads (FPKM) values. Principal component analysis (PCA) was performed using R (v 3.2.0) to evaluate the biological duplication of samples.
Differentially expressed genes (DEGs) were analysed using DESeq2. A Q value < 0.05 and a fold change > 2 or fold change < 0.5 were set as the thresholds for DEGs. Hierarchical cluster analysis of DEGs was performed using R (v 3.2.0) to investigate the expression patterns of genes in different groups and samples. Furthermore, all DEGs were subjected to GO and KEGG analyses to screen for significantly enriched terms via R (v 3.2.0).
## 4.5. Quantitative Real-Time PCR
Quantitative real-time PCR (qRT-PCR) was used to analyse the expression level of the DnCPV-23 S1 gene in transcriptome samples and verify the DEGs identified by RNA-seq. Total RNA was isolated from samples for transcriptomic analysis using TRIzol reagent (Life Technologies, Carlsbad, CA, USA) and then treated with DNase I (Fermentas, Glen Burnie, MD, USA). Total RNA (500 ng per sample) was reverse-transcribed into complementary DNA (cDNA) using a PrimeScript RT Reagent Kit (Takara, San Jose, CA, USA). Then, qRT-PCR was performed using Talent qPCR PreMix SYBR Green (Tiangen, Beijing, China) on a QuantStudio™ 7 Flex Real-Time PCR System (Applied Biosystems™, Waltham, MA, USA). One cycle was added for melting curve analysis for all of the reactions to verify the product specificity. Relative expression levels of genes were calculated using the 2 -∆∆CT method, and GAPDH was used as a reference for normalisation [53]. All of the primers for the target genes are listed in Table 2.
## 4.6. Cell Treatments
Sf9 cells were seeded (1 × 10 5 cells per well) in a 24-well plate and allowed to attach overnight. The next day, the cells were pretreated with the autophagy inhibitors 3-MA, CQ, and Baf A1 for 2 h prior to DnCPV-23 infection. The final concentrations of 3-MA, CQ, and Baf A1 were 5 mM, 40 µM, and 20 nM, respectively. Two hours later, 10 µL of the viral suspension containing free DnCPV-23 particles without polyhedral particles was added to the cells for 6 h. After the incubation period, the medium was aspirated, and fresh cell culture media containing autophagy inhibitors were added to the cells, which were subsequently cultured for 24 h and 48 h.
## 4.7. Western Blotting
Cells were collected by centrifugation at 300× g for 5 min at 4 • C, and total protein was extracted using RIPA lysis buffer (Beyotime, Shanghai, China). The cells were placed on ice for 20 min, and the supernatant was obtained by centrifugation at 6000× g for 6 min at 4 • C. Samples were mixed with 5× SDS loading buffer and boiled at 100 • C for 7 min. For Western blot analysis, after SDS-PAGE, proteins were transferred to polyvinylidene fluoride (PVDF) membranes, which were incubated overnight at 4 • C with the appropriate primary antibodies. The membranes were subsequently rinsed three times with TBST (each for 10 min) and incubated with horseradish peroxidase-conjugated goat anti-mouse IgG or goat anti-rabbit IgG as the secondary antibody for 2 h at room temperature. Protein bands were visualised using an enhanced chemiluminescence (ECL) Western blot detection kit and analysed using the ChemiScope 3000 mini system (Clinx, Shanghai, China).
## 4.8. Statistical Analysis
Statistical analysis was performed using GraphPad Prism 6.0. The values are presented as the mean plus or minus the standard error of the mean (SEM). The unpaired two-tailed t-test was used to compare one factor among different groups. Statistically significant differences are indicated as * p < 0.05, ** p < 0.01, or *** p < 0.001. Data with p values < 0.05 were considered statistically significant.
Funding. C.Y. (Chenghua Yan): Writing-review and editing, Supervision, Investigation, Formal analysis, Conceptualisation. W.K. and L.J.: Writing-review and editing, Supervision, Resources, Funding. Others (J.Y., J.W., J.C., X.L., C.Y. (Chunhua Yang), Z.Z., L.G., J.L.; T.D., F.Y. and G.M.): Methodology and Suggestions. All authors have read and agreed to the published version of the manuscript.
## References
1. Kuang, Yan, Zhan et al. (2021) "Transcriptional responses of Daphnis nerii larval midgut to oral infection by Daphnis nerii cypovirus-23" *Virol. J*
2. Sun, Chen, Gao et al. (2017) "Comparative mitochondrial genome analysis of Daphnis nerii and other lepidopteran insects reveals conserved mitochondrial genome organization and phylogenetic relationships" *PLoS ONE*
3. Green, White, Rao et al. (2007) "Biological and molecular studies of a cypovirus from the black fly Simulium ubiquitum (Diptera: Simuliidae)" *J. Invertebr. Pathol*
4. Zhou, Qin, Xiao et al. (2014) "Genomic and biological characterization of a new cypovirus isolated from Dendrolimus punctatus" *PLoS ONE*
5. Kuang, Zhan, Guan et al. "Study on proliferation characteristics of Daphnis nerii cypovirus-23 in Sf9 Cells" *J. Agric. Biotechnol. 2021*
6. Nai, Lo (2020) "Persistent PnV (Perina nuda virus) infection in a heterologous Lymantria xylina cell line, NTU-LY" *Biocontrol Sci. Technol*
7. Erez, Chejanovsky (2020) "Infection of a lepidopteran cell line with deformed wing virus" *Viruses*
8. Agosto, Mccabe (2006) "Conserved family of glycerol kinase loci in Drosophila melanogaster" *Mol. Genet. Metab*
9. Vatanparast, Sajjadian, Park (2022) "Glycerol biosynthesis plays an essential role in mediating cold tolerance the red imported fire ant" *Solenopsis invicta. Arch. Insect Biochem. Physiol*
10. Majumdar, Chowdhury, Chakraborty et al. (2025) "MiR-451a attenuates hepatic steatosis and hepatitis C virus replication by targeting glycerol kinase" *J. Transl. Med*
11. Kikawada, Saito, Kanamori et al. (2007) "Trehalose transporter 1, a facilitated and high-capacity trehalose transporter, allows exogenous trehalose uptake into cells" *Proc. Natl. Acad. Sci*
12. Marten, Haslitt, Martin et al. (2025) "Trehalose supports the growth of Aedes aegypti cells and modifies gene expression and dengue virus type 2 replication" *PLoS Pathog*
13. Kaur, Singh, Sharma (2024) "Unlocking Trehalose's versatility: A comprehensive Journey from biosynthesis to therapeutic applications" *Exp. Cell Res*
14. Wang, Zhu, Wang et al. (2023) "Trehalose hydrolysis and transport-related genes promote Bombyx mori nucleopolyhedrovirus proliferation through the phosphoinositide 3-kinase-Akt signalling pathway in BmN cell" *Dev. Comp. Immunol*
15. Zhang, Liu, Shiotsuki et al. (2017) "Depletion of juvenile hormone esterase extends larval growth in Bombyx mori" *Insect Biochem. Mol. Biol*
16. Reis, Noriega, Alves et al. (2022) "Why is oral-induced RNAi inefficient in Diatraea saccharalis? A possible role for DsREase and other nucleases" *Pestic. Biochem. Physiol*
17. Saito, Kamita, Hammock et al. (2015) "Juvenile hormone (JH) esterase activity but not JH epoxide hydrolase activity is downregulated in larval Adoxophyes honmai following nucleopolyhedroviruses infection" *J. Insect Physiol*
18. Yang, Dai, Zhang et al. (2024) "Mamestra brassicae multiple nucleopolyhedroviruses prevents pupation of Helicoverpa armigera by regulating juvenile hormone titer"
19. Takatsuka, Nakai, Shinoda (2017) "A virus carries a gene encoding juvenile hormone acid methyltransferase, a key regulatory enzyme in insect metamorphosis" *Sci. Rep*
20. Gu, Yang, Liu et al. (2025) "RNAi-mediated knockdown of juvenile hormone acid methyltransferase depresses reproductive performance in female" *Pestic. Biochem. Physiol*
21. Zhu, Liu, Chen et al. "Inter-organelle communication dynamically orchestrates juvenile hormone biosynthesis and female reproduction" *Natl. Sci*
22. Wang, Yang, Zhang et al. (2025) "Insecticide-induced sublethal effect in the fall armyworm is mediated by miR-9993/miR-2a-3p-FPPS/JHAMT-JH molecular module" *Pestic. Biochem. Physiol*
23. Shi, Wen, Chang et al. "Juvenile hormonesensitive ribosomal activity enhances viral replication in Aedes aegypti" *Msystems 2021*
24. Tian, Liu, Yuan et al. (2017) "CYP6B6 is involved in esfenvalerate detoxification in the polyphagous lepidopteran pest, Helicoverpa armigera. Pestic" *Biochem. Physiol*
25. Xiong, Yu, Li et al. (2025) "CYP6B6 mediated adaptation to capsaicinoids in the generalist Helicoverpa armigera and specialist H. assulta: Transcriptional response and metabolic detoxification" *Int. J. Biol. Macromol*
26. Huang, Zheng, Wu et al. (2024) "Functional study on candidate regulators mediating the expression of CYP6B7 induced by fenvalerate in a susceptible strain of Helicoverpa armigera" *Pestic. Biochem. Physiol*
27. Huang, Zheng, Wu et al. (2023) "A comparative study of transcriptional regulation mechanism of Cytochrome P450 CYP6B7 between resistant and susceptible strains of Helicoverpa armigera" *J. Agric. Food Chem*
28. Manjon, Troczka, Zaworra et al. (2018) "Unravelling the molecular determinants of bee sensitivity to neonicotinoid insecticides" *Curr. Biol*
29. Yang, Liu, Zhu et al. (2024) "MicroRNA bmo-miR-31-5p inhibits apoptosis and promotes BmNPV proliferation by targeting the CYP9e2 gene of Bombyx mori" *Pest Manag. Sci*
30. Li, Liu, Lv et al. (2025) "Molecular mechanism of λ-cyhalothrin detoxification by a delta-class Glutathione S-Transferase (PxGSTD3) from" *J. Agric. Food Chem*
31. Haas, Seiler, Nguyen et al. (2025) "Regulation of detoxifying enzymes expression and restriction of picorna-like virus infection by natural polysaccharide extracts in" *Virology*
32. Gui, Hou, Liu et al. (2009) "Effects of insect viruses and pesticides on Glutathione S-Transferase activity and gene expression in Bombyx mori" *J. Econ. Entomol*
33. Nebert, Dalton (2006) "The role of cytochrome P450 enzymes in endogenous signalling pathways and environmental carcinogenesis" *Nat. Rev. Cancer*
34. Cao, Kong, Ling et al. (2024) "Bmo-miR-3351 modulates glutathione content and inhibits BmNPV proliferation by targeting BmGSTe6 in Bombyx mori" *Insect Sci*
35. Fan, Cui, Yang et al. (2023) "UDP-glucuronosyltransferase is involved in susceptibility of Chironomus kiiensis Tokunaga, 1936 (Diptera: Chironomidae) to insecticides" *Ecotoxicol. Environ. Saf*
36. Zhu, Han, Hong et al. (2024) "Characterization of the UDP-glycosyltransferase UGT33D1 in silkworm" *Insect Mol. Biol*
37. Shang, Ding, Niu et al. "microRNA maintains nutrient homeostasis in the symbiont-host interaction"
38. Huang, Lu-Bo, Haddad (2014) "A Drosophila ABC transporter regulates lifespan" *PLoS Genet*
39. Jorajuria, Dereuddre-Bosquet, Naissant-Storck et al. (2004) "Differential expression levels of MRP1, MRP4, and MRP5 in response to human immunodeficiency virus infection in human macrophages" *Antimicrob. Agents Chemother*
40. Zhu, Pan, Zhang et al. (2022) "Bombyx mori cypovirus (BmCPV) induces PINK1-Parkin mediated mitophagy via interaction of VP4 with host Tom40" *Dev. Comp. Immunol*
41. Zhang, Zhu, Pan et al. (2023) "BmCPV replication is suppressed by the activation of the NF-kappaB/autophagy pathway through the interaction of vsp21 translated by vcircRNA_000048 with ubiquitin carboxyl-terminal hydrolase" *Insect Biochem. Mol. Biol*
42. Choi, Bowman, Jung (2018) "Autophagy during viral infection-A double-edged sword" *Nat. Rev. Microbiol*
43. Chen, Tu, Ding et al. (2023) "The role of autophagy in viral infections" *J. Biomed. Sci*
44. Nardacci, Ciccosanti, Marsella et al. (2017) "Role of autophagy in HIV infection and pathogenesis" *J. Intern. Med*
45. Kreutzberger, Sanyal, Saminathan et al. (2022) "SARS-CoV-2 requires acidic pH to infect cells" *Proc. Natl. Acad. Sci*
46. Doyle, Busey, Iobst et al. (2024) "Endosomal fusion of pH-dependent enveloped viruses requires ion channel TRPM7" *Nat. Commun*
47. Li, Li, Zhao et al. (2024) "Coronavirus M protein promotes mitophagy over virophagy by recruiting PDPK1 to phosphorylate SQSTM1 at T138" *Nat. Commun*
48. Pesti, Barczánfalvi, Dulka et al. (2024) "Bafilomycin 1A affects p62/SQSTM1 autophagy marker protein level and autophagosome puncta formation oppositely under various inflammatory conditions in cultured rat microglial cells" *Int. J. Mol. Sci*
49. Tan, Wang, Liu et al. (2018) "Inhibition of silkworm vacuolar-type ATPase activity by its inhibitor Bafilomycin A1 induces caspase-dependent apoptosis in an embryonic cell line of silkworm" *Arch. Insect Biochem. Physiol*
50. Mauvezin, Neufeld (2015) "Bafilomycin A1 disrupts autophagic flux by inhibiting both V-ATPase-dependent acidification and Ca-P60A/SERCA-dependent autophagosome-lysosome fusion" *Autophagy*
51. Song, Korolkova (1374) "Bafilomycin A1 inhibits HIV-1 infection by disrupting lysosomal cholesterol transport" *Viruses*
52. Jin, Shan, Peng et al. "Downregulation of a transcription factor associated with resistance to Bt toxin Vip3Aa in the invasive fall armyworm"
53. Livak, Schmittgen (2001) "Analysis of relative gene expression data using real-time quantitative PCR and the 2(T)(-Delta Delta C) method" *Methods*
54. "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=PMC12607698&blobtype=pdf | # Validation of H5 influenza virus subtyping RT-qPCR assay and low prevalence of H5 detection in 2024-2025 influenza virus season
David Bacsik, Margaret Mills, Luke Monroe, Cassey Spring, Ailyn Perez-Osorio, Jonathan Reed, Ferric Fang, Lori Bourassa, Pavitra Roychoudhury, Katharine Crawford, Kevin Snekvik, Alexander Greninger
## Abstract
A sustained outbreak of H5N1 influenza virus among wild fowl and domestic livestock has caused more than 70 zoonotic infections in humans in North America, including two deaths. The United States Centers for Disease Control and Prevention has recommended rapid H5 subtyping for all hospitalized cases with influenza A virus infection to enable prompt initiation of antiviral treatment, as well as infection prevention and implementation of public health measures to control spread. To address these needs, we developed a qualitative multiplex RT-qPCR assay to subtype H5 influenza virus in nasal, nasopharyngeal, and conjunctival specimens with a limit of detection of 250 copies/mL. No cross-reactivity was observed with other common respiratory viruses, including seasonal H3N2 and H1N1 influenza A viruses. We retrospec tively subtyped 590 influenza A virus-positive clinical specimens with Ct values less than 31 processed by University of Washington labs between March 2024 and February 2025, including 512 specimens collected during the 2024-2025 influenza season, and detected no H5 positives. After clinical implementation, we performed 150 clinically ordered H5 subtyping tests between February and April 2025 and again detected no positives. This work enhances clinical pandemic preparedness activities and highlights the exceedingly low prevalence of H5N1 influenza virus during the 2024-2025 respiratory season.
IMPORTANCEThe spread of H5N1 influenza virus in the United States has led to the culling of almost 200 million birds, infected cow herds across 17 states, and resulted in 70 human infections as of July 2025. Rapid PCR subtyping of H5 influenza virus is critical to inform hospital infection prevention and public health to enable containment of viral transmission. Here, we report the design, validation, and clinical implementation of a qualitative multiplex H5-subtyping RT-qPCR assay for nasopharyngeal, nasal, and conjunctival swab specimens. Additionally, we offer the largest reported study of H5 subtyping of influenza A virus-positive specimens in the United States to date. No H5 infections were detected in 740 samples collected between March 2024 and April 2025 from patients with confirmed influenza A virus infection in a large academic medical system in Seattle, WA.
circulates in wild bird populations and has caused repeated introductions into livestock, including domestic poultry (4) and dairy cattle (5).
Human infections of both genotypes have been detected, with the first outbreak-rela ted case identified in April 2024 (6). To date, more than 70 cases have been reported in North America (7). Although most cases have not required hospitalization, three cases of critical illness have been reported, including deaths in Louisiana and Durango, Mexico (8)(9)(10). Conjunctivitis has been the most common symptom reported (11). Exposure to infected livestock is a significant risk factor, and farm workers have been disproportion ately impacted (11). In a serological survey of dairy workers on H5N1-infected farms, 7% had evidence of recent infection with a subtype H5 influenza virus (12), suggesting that the prevalence of infection in this population could be much higher than has been recognized.
Prompt identification of H5N1 influenza virus infection is necessary to provide appropriate medical care and to contain potential further spread of the virus. The Centers for Disease Control and Prevention (CDC) recommends that all hospitalized patients with influenza A virus infections have subtyping performed within 24 h of admissionespecially patients who are critically ill (13). Because H5N1 influenza virus can cause severe illness, and because circulating strains remain sensitive to antiviral medications, treatment with oseltamivir is indicated for symptomatic patients with suspected H5N1 influenza virus infections (14). Furthermore, airborne precautions and patient isolation are recommended for infection control (15).
In addition to informing clinical management, molecular diagnosis of H5N1 improves influenza surveillance. Though efforts have been made to increase surveillance of non-subtypeable influenza, much of the current surveillance infrastructure has focused on monitoring epidemiologically linked cases with known exposure to infected animals, and it is possible that cases without recognized risk factors have gone undetected (16). This pattern has occurred during previous outbreaks, including SARS-CoV-2 (17). To provide unbiased monitoring of H5N1 infections among all influenza cases, integration within health systems is required. To increase testing capacity, the CDC has issued a call (18) requesting the development and implementation of laboratory-developed tests to detect subtype H5 influenza virus (19)(20)(21).
Here, we report our validation of a qualitative multiplex RT-qPCR assay for H5 influenza virus subtyping in nasopharyngeal, nasal, and conjunctival swabs. We have used the assay to test 740 influenza A virus-positive specimens processed by the University of Washington Laboratories between March 2024 and April 2025, detecting no cases of H5 influenza virus to date.
## MATERIALS AND METHODS
## Clinical specimens and associated metadata
Residual deidentified nasopharyngeal swabs, nasal swabs, and conjunctival swabs from the UW Virology Laboratory were used for validation. Retrospective subtyping was performed by testing available residual influenza A virus-positive respiratory specimens from UW Medicine with collection dates of March 2024 or later. These dates were chosen based on the date of the first human H5N1 infection reported in the United States (22). Available samples had been previously collected for genomic surveillance and had Ct values less than 31. After assay validation and implementation at UW Medicine, test results and metadata were obtained from the laboratory information system. This study was approved by the UW Institutional Review Board with a consent waiver (STUDY00010205).
## H5 RNA templates
Because clinical specimens containing H5N1 influenza virus were not readily available during initial test validation, we employed three alternative sources of H5 RNA: two sets of synthetic RNA templates and a collection of H5-positive RNA samples extracted from animal specimens. The first set of synthetic RNAs was transcribed in vitro, refer red to as "IVT RNA templates. " The H5 RNA was transcribed from a gBlock containing the HA coding sequence of the clade 2.3.4.4b H5N1 virus strain A/white-tailed eagle/ Hokkaido/20220322001/2022 (File S1) using the New England Biosystems HiScribe T7 RNA Synthesis Kit (E2040). The M RNA was prepared as previously described (23). A second set of synthetic RNA was obtained from the National Institute of Standards and Technology (NIST) as H5N1 (Avian Influenza) Synthetic RNA Fragments (Research-Grade Test Material 10263). The NIST RNA set includes synthetic RNAs encoding the HA, M, and NA genes of the clade 2.3.4.4b H5N1 virus strain A/American Wigeon/South Carolina/22/2021 in a background of 5 ng/µL Jurkat cell RNA.
Authentic H5N1-positive RNA samples (n = 7) were obtained from the Washington State University Animal Disease Diagnostic Laboratory. The samples were previously extracted from a variety of animal-source specimens, including mammalian respiratory swabs, mammalian tissue, bulk milk, and avian oropharyngeal/cloacal swabs (Table S1). All were known to be influenza A virus-positive by prior RT-qPCR and known to contain H5N1 RNA by prior whole-genome sequencing.
## Inactivated H5N1 viruses
Additional validation was performed using inactivated H5N1 viruses representing the B3.13 and D1.1 genotypes. For genotype B3.13, inactivated H5N1 reference material was obtained from BEI Resources (NR-59886), containing tissue-culture adapted strain A/bovine/Ohio/B24OSU-439/2024 and cell lysate that had been inactivated by gammairradiation.
For genotype D1.1, H5N1 influenza virus isolate A/Washington/239/2024 was obtained from the CDC, passaged, and inactivated. Before passaging, 3.5 × 10 6 MDCK cells were seeded in a T75 flask in growth media consisting of MEM with Earle's salts and 2 mM glutaMAX (Thermo 42360) supplemented with 10% vol/vol heat-inactivated FBS (Thermo A3840002), 1% vol/vol Pen-Strep (Thermo 151401), and 1% vol/vol NEAA (Thermo 111400) and incubated overnight. The next day, virus inoculum was prepared by diluting the virus stock in infection media containing MEM with Earle's salts and 2 mM glutaMAX (Thermo 42360), 0.3% wt/vol BSA (Sigma-Aldrich A8412), 1% vol/vol Pen-Strep (Thermo 151401), 1% vol/vol NEAA (Thermo 111400), and 0.4 µg/mL TPCKtreated trypsin (Thermo NC9783694). The cells were washed with plain MEM to remove FBS and infected at an MOI of 0.01 based on the titer provided with the stock (2.0 × 10 8 TCID50/mL) in 2 mL of infection media. The virus was allowed to adsorb for 1 h, redistributing the inoculum every 20 min. After adsorption, 8 mL of infection media was added for a final volume of 10 mL. At 2 days post-infection, with more than 90% of the cell monolayer exhibiting cytopathic effect, the supernatant was collected, separated from cellular debris by centrifugation at 500 × g for 10 min, and aliquoted. Virus stocks were titered by TCID50 assay on MDCK cells using eight replicates per dilution, and cytopathic effect was scored at 5 days post-infection. TCID50/mL estimates were determined using the Spearman-Karber method (24).
The virus was inactivated using a validated procedure approved by the University of Washington Institutional Biosafety Committee. Briefly, viral supernatants were heat-trea ted in 400 µL aliquots at 65°C for 30 minutes. For each aliquot, inactivation was confirmed by testing 100 µL of the inactivated material by TCID50 assay. Positive controls (not heat-treated) and negative controls (no virus) were included with each assay.
## RNA extraction
RNA was extracted from clinical specimens on a Roche MagNA Pure 96 instrument using the Viral NA Small Volume Kit (06543588001). For each sample, 200 µL of specimen was used as input and RNA was eluted in 50 µL of elution buffer. An exogenous RNA template was included with the lysis buffer as an internal control (Table 1).
## Reverse transcription quantitative PCR reaction conditions
Multiplex reverse transcription quantitative PCR (RT-qPCR) was performed using the AgPath-ID One-Step RT-PCR Kit (AM1005). For each reaction, 5 µL of extracted RNA was added to a reaction mix containing 12.5 µL AgPath Master Mix, 1 µL AgPath Enzyme, and primers and probes at specified concentrations (see Table 1); the total reaction volume was 25 µL. Reactions were run on ABI 7500 Thermocyclers. Cycle parameters were 45°C for 10 min, 95°C for 10 min, and 45 cycles of 95°C for 15 s, followed by 60°C for 45 s. ROX normalization and automatic baseline were used for all fluorophores. The threshold values were 0.18 for the H5 target (FAM), 0.16 for the influenza A target (VIC), and 0.1 for the internal control target (Cy5). Ct values less than 45 were considered positive.
## Reverse transcription droplet digital PCR reaction conditions
Reverse transcription droplet digital PCR (RT-ddPCR) was performed with the BioRad One-Step RT-ddPCR Advanced Kit for Probes (1864022) using the H5-2 and M primers and probes from the RT-qPCR assay (Table 1). Each 25 µL reaction was prepared by combining 5 µL of extracted RNA with 20 µL of master mix containing 6.25 µL Super Mix, 2.5 µL RT enzyme, 1.25 µL DTT, primers at a final concentration of 900 nM for H5-2 and 450 nM each for M, and probes at a final concentration of 250 nM. 20 µL of the total reaction mix was used for droplet generation using a BioRad Automated Droplet Generator, and 40 µL of droplet suspension was transferred to a 96-well plate. Reactions were run on a BioRad C1000 Touch thermocycler, using the following conditions: 50°C for 60 min, 95°C for 10 min, 40 cycles of 95°C for 30 s, followed by 60°C for 1 min, and 98°C for 10 min. Fluorescence was measured on a BioRad QX600 Digital Droplet Reader using the absolute quantification method. Samples with more than 50% positive droplets fell outside of the quantitative range and were excluded from analysis.
The absolute concentration of the IVT RNA and NIST RNA templates was meas ured by RT-ddPCR as described above. The absolute copy number of the unitless BEI H5N1 reference material and the inactivated H5N1 virus was measured following RNA extraction as described above.
## Computational analysis of primer and probe sequences
To assess the similarity between the chosen oligonucleotides and contemporary North American viruses, computational analysis was performed. All H5 HA sequences collected in North America between 1 January 2024 and 31 January 2025 were downloaded from NCBI (25). The sequences were aligned using MAFFT (26). Binding regions were annotated, and each primer or probe was analyzed independently. Sequences that lacked coverage at one or more nucleotides in the binding site were excluded. The number of mismatches was counted; ambiguous nucleotides were conservatively treated as mismatches.
## Limit of detection
The limit of detection (LOD) of the complete assay, including extraction, was assessed using inactivated H5N1 virus (A/Washington/239/2024) spiked into pooled, influenza virus-negative nasopharyngeal swab specimens. The initial LOD was estimated using a 10-fold dilution series ranging from 1:10 2 to 1:10 8 , with RNA extracted from four replicates per concentration. Additional replicates were prepared for comparison with the complete CDC assay protocol (see "Comparison to CDC Influenza A/H5 Subtyping Kit" below). The LOD was confirmed using a 4-fold serial dilution around the initial limit, with RNA extracted from 20 replicates of each concentration. The 95% LOD was reported as the lowest concentration with at least 19 of the replicates yielding a positive result.
The absolute LOD of the multiplex RT-qPCR reaction was determined using IVT H5 (stock 1.07 × 10 6 copies/µL by RT-ddPCR using H5-2 primers and probes) and NIST M (1.47 × 10 5 copies/µL by RT-ddPCR using M primers and probes) RNA templates. Ten microliters of each RNA template was combined and diluted in 80 µL of AE buffer (10 mM Tris-HCl and 0.5 mM EDTA at pH 9.0; Qiagen #19077). This stock was diluted 10-fold in AE buffer for initial LOD determination and 2-fold around the initial LOD for confirmation. Twenty replicates of each concentration were tested, and the 95% LOD was reported as the lowest concentration with at least 19 of the replicates yielding a positive result.
## Assay sensitivity and matrix compatibility
Assay sensitivity was assessed using H5N1-positive RNA samples extracted from animal specimens (n = 7) and contrived H5N1-positive specimens spiked with inactivated B3.13 virus (n = 69) or D1.1 virus (n = 160). To prepare low-positive specimens containing genotype B3.13 virus, inactivated A/bovine/Ohio/B24OSU-439/2024 virus, which was received unitless from BEI, was added at a ratio of 1:50,000 by volume to the nasal and conjunctival specimens, and 1:10,000 to the nasopharyngeal specimens. Specimens with at least 220 µL volume were tested individually, while specimens with less volume were pooled and aliquoted. In total, 29 nasal specimens were prepared (20 individual and 9 pooled); 20 conjunctival specimens were prepared (7 individual and 13 pooled); and 20 nasopharyngeal specimens were prepared (20 individual and 0 pooled).
To prepare H5N1-positive specimens containing genotype D1.1 virus, inactivated A/ Washington/239/2024 virus was added to pooled nasopharyngeal, nasal, or conjunctival specimens or to AE buffer. Three levels of H5N1-positive specimens were prepared with each matrix, targeting low (400 TCID50/mL), medium (20,000 TCID50/mL), and high (1,000,000 TCID50/mL) titers of H5N1 virus. Twenty replicates were prepared of the low-concentration specimens, and 10 replicates were prepared of the medium-and high-concentration specimens. Confidence intervals (95% CIs) for measures of agree ment were calculated using the Clopper-Pearson exact method.
## Assay specificity
Three types of H5-negative specimens were used to assess specificity: nasopharyngeal swabs (n = 20), nasal swabs (n = 29), and conjunctival swabs (n = 18). The respira tory specimens were confirmed as negative for influenza A/B virus by Cepheid Xpert Xpress CoV-2/Flu/RSV plus or Hologic Panther Fusion Flu A/B/RSV assays. Conjunctival swabs were collected for HSV and/or VZV testing. Specificity was further evaluated using specimens known to contain common respiratory viruses by clinical testing via Cepheid or Hologic respiratory panel tests or whole-genome sequencing. These included influenza A virus subtypes H3N2 (n = 21) and H1N1 (n = 17), influenza B virus (n = 3), respiratory syncytial virus (n = 7), and SARS-CoV-2 (n = 10).
## Comparison to CDC influenza A/H5 subtyping kit
Since only one RT-qPCR assay is FDA-cleared for H5N1 testing, we compared our multiplex RT-qPCR assay with the CDC's Influenza A/H5 Subtyping Kit (Version 4, #FluIVD03-11). To assess the relative performance of the complete assays, includ ing extraction, replicate samples were prepared containing inactivated A/Washing ton/239/2024 virus spiked into pooled nasopharyngeal specimens (see "Limit of Detection" above). RNA was extracted following the CDC H5 Subtyping Kit protocol, which differs from the extraction protocol used in our multiplex assay. Using a Roche MagNA Pure 96 instrument with the Viral NA Small Volume Kit (06543588001), 100 µL of specimen was combined with 350 µL of External Lysis Buffer (06374913001) and the entire volume (450 µL) was used as input. RNA was eluted in 100 µL of elution buffer. The CDC RT-qPCR reactions were run according to the manufacturer's protocol using the Invitrogen SuperScript III Platinum One-Step Quantitative RT-PCR System on the ABI 7500 FAST thermocycler. For each singleplex reaction in the CDC kit, 5 µL of nucleic acid template was added to a 20 µL reaction mix containing 12.5 µL of 2× PCR master mix, 5.5 µL water, 0.5 µL enzyme, and 1.5 µL combined primer/probe mix prepared according to the kit protocol; total volume was 25 µL.
To compare the performance of the RT-qPCR reactions on isolated RNA, the CDC H5 Subtyping reactions and the novel multiplex reactions were run in parallel on the same set of isolated RNA templates. For the IVT RNA and NIST RNA templates, 10-fold dilution series were prepared by combining H5 and M synthetic RNA at a ratio of 1:1 by volume and serially diluting the template mix in AE buffer. Inactivated A/bovine/Ohio/ B24OSU-439/2024 virus was initially diluted 1:100 in PBS, and a 10-fold dilution series in PBS was prepared from this stock with RNA extracted as described above. The absolute copy number of each template stock was determined by RT-ddPCR after dilution (and extraction, in the case of the inactivated virus). RT-qPCR reactions were run as described for each assay.
## RESULTS
## Design and computational analysis of primer and probe sequences
A three-channel multiplex RT-qPCR reaction was designed for qualitative detection of subtype H5 influenza virus. The primary target was the H5 influenza virus HA gene. To reduce the risk of false-negative results due to viral evolution (27) or reagent failure (28), primers and probes were selected to target two non-overlapping regions of the gene. One H5 target (H5-1) generated a 149 bp product from positions 1,481 to 1,629 of the HA CDS (GenBank accession no. LC730539; Fig. S1) using oligonucleotides adapted from designs by Shu et al. (29). The other H5 target (H5-2) used oligonucleotides designed by Sahoo et al. (19) and adapted from sequences originally designed by the Hong Kong Centre for Health Protection (30). This target generated a 144 bp product from positions 1,101 to 1,244 of the HA CDS. As a pan-influenza A virus control, a highly conserved region of the influenza A M gene was targeted using primers and probes designed by the CDC National Center for Immunization and Respiratory Diseases (31). Oligonucleotides targeting an exogenous RNA template encoding the TMP1 gene from the marine species Podocornye carnea were included as an internal control (32).
Because materials for validation were highly limited when assay development began, the primers and probes were designed to bind to a reference material containing inactivated H5N1 virus from 2009 (BEI Resources NR-59421) with a divergent HA sequence. As a result, genotype B3.13 viruses carry two mismatches in H5 Probe 1 and no mismatches in H5 Probe 2; and genotype D1.1 viruses carry three mismatches in H5 Probe 1 and one mismatch in H5 Probe 2 (Fig. S1). Despite these sequence differences, the probes remain functional with contemporary templates. The same mismatches are present in the IVT RNA target, which is reliably detected by either probe alone (Table S2).
The similarity between the oligonucleotide sequences and contemporary North American H5 influenza virus sequences was assessed computationally. A total of 2,788 H5 HA sequences collected in North America between 1 January 2024, and 31 January 2025, were retrieved from NCBI GenBank (Table 2). Of the six oligonucleotides targeting H5 HA, five matched recent sequences closely, with two or fewer mismatches in 99% of sequences. One oligonucleotide (H5 Probe 1) displayed two mismatches in 92% of recent H5 sequences, and three or more mismatches in 8% of sequences-predominantly genotype D1.1 strains.
## Estimates of template copy number by Ct value
To determine the relationship between the Ct value produced by the RT-qPCR reaction and the absolute copy number of template molecules, a dilution series of RNA extrac ted from inactivated H5N1 virus A/Washington/239/2024 was tested by RT-qPCR and RT-ddPCR in parallel (Fig. S2). Dilutions between 1:10 4 and 1:10 6 yielded RT-ddPCR results in the quantitative range, providing a range of three orders of magnitude for comparison (Table S3). A strong linear relationship was observed for both the H5 target (R 2 = 0.99) and the M target (R 2 = 0.98). While the clinical assay is designed to be qualitative and not quantitative, the absolute copy number of templates in the multiplex RT-qPCR reaction can be estimated from the Ct value using a linear model fit to these data (Table S4).
## Limit of Detection
The absolute LOD of the RT-qPCR reaction was measured using a 2-fold serial dilution of H5 and M IVT RNA templates in AE buffer (Table 3). The 95% LOD (LOD95) was measured by identifying the lowest concentration with at least 19 out of 20 replicates detected (Table S5). The LOD95 for the H5 target was 4.7 copies per reaction and 36.3 copies per reaction for the M target.
The LOD of the complete assay including extraction was measured using inacti vated H5N1 virus (A/Washington/239/2024) diluted in pooled nasopharyngeal swab specimens, with independent replicates starting at extraction (Table 4). The LOD95 for the H5 target was 24 TCID50/mL and the LOD95 for the M target was 390 TCID50/mL (Tables S6 andS7). Based on the Ct values observed, we estimate an absolute LOD95 in nasopharyngeal specimens of 5.0 copies per reaction for the H5 target (250 copies/mL) and 12.3 copies per reaction for the M target (615 copies/mL), which agree closely with the absolute limits measured using synthetic RNA templates in buffer. The assay displayed linear performance across 5 logs of input with R 2 values of 0.998 for the H5 target and 0.996 for the M target (Fig. 1). Efficiency was 99.0% for the H5 target and 104.5% for the M target.
## Sensitivity and matrix compatibility
Sensitivity was assessed across a range of virus concentrations, matrices, and genotypes using clinical specimens spiked with inactivated H5N1 virus. For the B3.13 genotype, we generated low-concentration H5-positive samples by adding inactivated A/bovine/ Ohio/B24OSU-439/2024 virus to influenza virus-negative nasal and conjunctival swab specimens at a ratio of 1:50,000 by volume and to nasopharyngeal specimens at a ratio S8). Of the conjunctival swabs, 19 out of 20 samples tested positive for the H5 target, demonstrating 95% positive agreement.
For the M target, all 20 of the nasopharyngeal swab specimens tested positive with a mean Ct value of 34.6 (estimated 48.7 copies/reaction), demonstrating 100% positive agreement. Twenty five of the 29 nasal swab specimens tested amplified the M target with a mean Ct value of 37.3 (estimated 6.7 copies per reaction), demonstrating 86% positive agreement. Eighteen of the 20 conjunctival swab specimens amplified the M target with a mean Ct value of 37.2 (estimated 7.2 copies per reaction), demonstrating 90% positive agreement.
For the D1.1 genotype, all three matrices and AE buffer were spiked with inactivated H5N1 virus (A/Washington/239/2024) to create contrived positives at low, medium, and high titer (Table 6). The mean Ct values for the H5 target were 33.7 for the low-con centration specimens, 28.2 for the medium-concentration specimens, and 22.2 for the high-concentration specimens (estimated at 100, 4,500, and 284,000 copies per reaction, respectively). For the M target, the mean Ct values were 36.3, 30.7, and 24.8, respectively (estimated at 15, 850, and 64,000 copies per reaction, respectively). All specimens tested positive for the H5 target, demonstrating 100% positive agreement across all matrices and virus concentrations (Table S9). For the M target, one low-positive nasopharyng eal swab specimen falsely tested negative, demonstrating 95% positive agreement; all other matrices demonstrated 100% positive agreement. No strong matrix effects were observed. For all concentrations, Ct values were similar across specimen types, including AE buffer (Fig. 2). We obtained 7 animal RNA samples that were known to contain authentic H5N1 RNA by whole-genome sequencing and included both genotype B3.13 (n = 2) and genotype D1.1 (n = 1) genomes. All 7 samples amplified the H5 target with a mean Ct value of 24.4, demonstrating 100% positive agreement (Table S1). One B3.13 sample failed to amplify the pan-influenza A virus M target and had a late H5 Ct value of 38.2. Analysis of this sample's sequence (BioSample SAMN41892145) revealed two mismatches in H5 Probe 1, one mismatch in M Reverse Primer 1, and two mismatches in M Reverse Primer 2. Given that the same mismatch profiles were present in many samples that successfully amplified both targets (e.g., BioSample SAMN46070075), the high Ct values are likely a result of sample degradation, rather than primer mismatch.
## Specificity
Three types of clinical specimens were evaluated for compatibility with the multi plex RT-qPCR assay (Table 7). Residual nasopharyngeal, nasal, and conjunctival swab specimens known to be influenza A virus-negative were extracted and tested. All specimens tested negative for both the H5 and M targets and tested positive for the internal control (Table S10). Residual specimens known to contain other common respiratory pathogens were tested to assess specificity (Table 7). Importantly, we included non-H5 influenza A virus subtypes H3N2 (n = 21) and H1N1 (n = 17) identified by whole-genome sequencing (Table S11). We also included specimens containing influenza B virus (n = 3), respiratory syncytial virus (n = 7), and SARS-CoV-2 (n = 10). None of the specimens produced a positive result for the H5 target, demonstrating 100% negative agreement with prior clinical testing. The influenza A samples appropriately amplified the IAV M target, while samples containing other pathogens did not.
## Accuracy and precision
Combining all true positive and true negative results, the accuracy of the H5 target was 99.7% (95% CI: 98.6-100%) and the accuracy of the M target was 97.7% (95% CI: 95.6-99.0%) (Table 8).
The precision of the multiplex RT-qPCR reaction was evaluated across 6 logs of input using a pool of IVT H5 RNA and NIST M RNA (Table S12). Templates were combined and diluted in AE buffer in a 10-fold series with four replicates per concentration. For the H5 target, the maximum coefficient of variation was 3.3% at a mean Ct value of 35.4 (estimated 58.0 copies per reaction). For the M target, the maximum coefficient of variation was 3.6% at a mean Ct value of 36.4 (estimated 13.0 copies per reaction).
Day-to-day and operator-to-operator variation were assessed by including a positive control reaction on each PCR plate (Table S13). Thirteen independent PCR plates run by 3 different operators over 5 days were analyzed. Mean Ct values were 30.4 for the H5 target (estimated 984 copies per reaction) and 28.3 for the M target (estimated 4,920 copies per reaction). Across all experiments, the total coefficient of variation was 2.1% for the H5 a Sensitivity was assessed using contrived clinical specimens spiked with inactivated genotype D1.1 virus (A/Washington/239/2024) at low, medium, or high titer. Summary statistics were calculated using detected samples; non-detected samples were excluded.
target and 2.4% for the M target. The day-to-day variation was 1.8% for the H5 target and 2.3% for the M target. The operator-to-operator variation was 1.1% for the H5 target and 1.2% for the M target.
## Comparison to CDC Influenza A/H5 Subtyping Kit
The assay's performance was benchmarked against the CDC's Influenza A/H5 Subtyping Kit to determine its suitability for surveillance and clinical applications. A comparison was made using the complete protocol of each assay, including extraction. In parallel, contrived positive specimens used to measure the initial LOD were extracted following each assay's protocol. The CDC assay consistently produced higher Ct values than the multiplex assay (Fig. 1). The multiplex assay detected low-concentration samples more frequently than the singleplex CDC assay for both the H5 and pan-influenza A targets. a Specificity was assessed using influenza virus-negative specimens and residual specimens containing common respiratory pathogens. Samples without a positive result were labelled not detected ("NDET").
The initial LOD of the CDC assay was estimated between 209 and 2,090 TCID50/mL for the H5a target, between 2,090 and 20,900 TCID50/mL for the H5b target, and between 209 and 2,090 TCID50/mL for the InfA target (Table S14). On confirmatory analysis using the more precise 4-fold dilution series, the LOD95 was estimated higher, between 3,120 and 12,500 TCID50/mL for the H5a target, between 780 and 3,120 TCID50/mL for the H5b target, and between 3,120 and 12,500 TCID50/mL for the InfA target (Table S15).
A comparison was also made of the RT-qPCR reactions on identical RNA templates. Ten-fold serial dilutions of the IVT RNA, NIST RNA, and inactivated virus templates were prepared, and the same aliquots were used for testing in parallel (Fig. 3). In this case, the CDC H5 Subtyping kit produced lower Ct values than the multiplex assay when tested on the same extracted RNA templates, especially for the pan-influenza A target, where Ct values differed by approximately 5 units. At the lower extreme of template concentra tion, the CDC H5 Subtyping Kit detected several samples that were not detected by the multiplex assay, suggesting higher sensitivity of the CDC RT-qPCR reaction. Above concentrations of 10 copies per reaction, corresponding to 500 copies/mL, both assays detected all samples tested. Across this 12-week period, the median number of H5 subtyping test orders was 9 per week (IQR: 3.5-18.8) and the maximum was 32 per week. Most test orders (n = 104) came from emergency departments or were ordered for patients admitted to an inpatient service (n = 45); one order was made from an outpatient clinic. The median age of patients who received an H5 subtyping test as part of clinical care was 62 years old (IQR: 46.5-75), which is significantly older than patients in the retrospective subtyping cohort (P < 0.01 by Wilcoxon rank-sum test) (Table 10). There were also more samples collected from male patients in this cohort than in the retrospective subtyping cohort (52.0% vs 50.3%), although this difference was not statistically significant (P = 0.75, chi-squared test).
## DISCUSSION
Here, we report the design, validation, and clinical implementation of an RT-qPCR assay to subtype H5 influenza A virus in nasopharyngeal, nasal, and conjunctival swab specimens. The assay is sensitive, with a LOD of 250 copies/mL, and specific, with no cross-reactivity to common respiratory pathogens, including other influenza A virus subtypes. Recently, several other groups have validated and implemented laboratorydeveloped tests for H5 subtyping, though little published performance characteristic data are available for comparison (19)(20)(21). The sensitivity of our assay is comparable to that of the assay recently published by Sahoo et al., with both assays reporting limits of detection below 5 copies per reaction (19).
It is difficult to compare these values to the LOD reported for the CDC's H5 subtyping kit, ~250 EID50/mL, as this value is reported in units of egg infectious dose rather than absolute copy number. In both tissue culture models (31) and human infections (32), the number of template molecules detectable by RT-qPCR does not consistently correlate with viral titer because the ratio of non-infectious viral particles varies greatly between individual infections. However, when the same specimens were tested side-by-side, our multiplex assay demonstrated higher sensitivity than the CDC H5 Subtyping Kit. This difference was due to the efficiency of RNA extraction, as the CDC RT-qPCR reaction demonstrated higher sensitivity when identical isolated RNA templates were tested. In addition to high sensitivity, the multiplex format of our assay requires just 1 reaction per test, simplifying handling and reducing the cost of the test compared to the 4-well design of the CDC assay.
Given the preponderance of conjunctivitis reported in human cases related to the current outbreak, we validated the assay's compatibility with conjunctival swab specimens. Although the CDC H5 subtyping kit is authorized for use with conjunctival swab specimens (33), to the best of our knowledge, no published validation studies have included this specimen type. We also validated the assay's ability to detect both circulating genotypes that have caused human cases, B3.13 and D1.1.
Following validation, we tested 740 specimens from UW Medicine collected between March 2024 and April 2025 and identified no infections with subtype H5 virus. Impor tantly, we used the test to distinguish between H5 influenza virus and seasonal influenza viruses during the 2024-2025 influenza virus season, and influenza virus sequencing data were available for 157 of these specimens. This is the first surveillance study performed during a period of high seasonal influenza virus activity (33) since the outbreak began and complements a recent study by Adams et al. that detected no H5N1 infections during a period of low seasonal influenza virus activity (34). Our data set includes both retrospective testing of residual specimens (n = 590) and clinical testing ordered during routine medical care after the assay was implemented at UW Medicine (n = 150). This surveillance was performed in Washington State, which to date has experienced the second-highest number of human cases of H5N1 influenza infection in the United States.
Our results come with several limitations. First, the assay is designed to maximize the sensitivity of the H5 targets at the expense of the M target. Because of this, the M target is less sensitive, and we have observed false-negative results on samples that are known to be influenza A virus-positive. This limitation is ameliorated by the ability of all influenza tests to detect H5N1 as influenza virus, and the indication of the subtyping assay to be used with known influenza virus-positive specimens. Second, the probe "H5 Probe 1" carries mismatches to the majority of recent H5 sequences from North America. This occurred because initially, material for validation was extremely limited, and probes were designed to bind the only inactivated H5N1 reference material available (BEI NR-59421), which encoded an HA sequence from 2009. Since that time, updated inactivated H5N1 reference material (BEI NR-59886) has become available and was used for validation, highlighting the importance of making inactivated viral material available that can be used for BSL-2 validation activities. While our probe functions with contem porary H5N1 genotypes, future evolution might diminish probe binding and necessitate a revised probe sequence. Third, while the exogenous internal control we add during extraction controls for RT-qPCR reaction performance, it does not control for specimen adequacy (e.g., confirm that a specimen was loaded into extraction), and an endogenous control target would be useful for this purpose in the future. Finally, although our surveillance data represent the largest set of influenza A virus-positive specimens tested for H5 in the literature since the outbreak began, almost all tested specimens originated from Seattle, Washington. Sampling from more rural regions, where patients are more likely to have exposure to livestock or avian species, is critical for both rapid detection of zoonotic infections and assessing point prevalence in a potentially higher-risk region. With engagement and collaboration with clinicians and public health departments in rural areas, the clinically reportable H5 subtyping assay described here can help in these efforts.
Our data are most useful for understanding the prevalence of H5 infections during the 2024-2025 annual influenza virus season in a single academic medical system. We detected no cases of subtype H5 infection within the population of patients infected with influenza A viruses. As such, we did not detect evidence of any evolutionary events that can occur in the presence of co-circulation, such as recombination between H5 viruses and seasonal influenza A viruses (1). Our conclusions are in agreement with data from both the national influenza surveillance program, which has identified very few cases through non-specific respiratory disease surveillance (7), and the recent Adams et al. study, which found no H5 cases by screening clinical respiratory samples collec ted outside of the annual influenza season (34). Taken together, these data support current public health guidance indicating that H5 infection risk remains almost entirely associated with exposure to infected animals. Focused testing of patients with epidemio logical risk factors, such as farm workers, and prompt treatment with antiviral medica tions remain cornerstones of clinical management (35). However, as H5N1 influenza virus is expected to spread in animal species in the United States for the foreseeable future, integration of routine H5 testing into health systems remains critical to monitor for future changes to the epidemiology of this outbreak.
## References
1. Kandeil, Patton, Jones et al. (2023) "Rapid evolution of A(H5N1) influenza viruses after intercontinental spread to North America" *Nat Commun*
2. Worobey, Gangavarapu, Pekar et al. (2024) "Preliminary report on genomic epidemiology of the 2024 H5N1 influenza A virus outbreak in U.S. cattle (Part 1 of 2) -Influenza virus / H5N1-global. Virological"
3. Peacock, Moncla, Dudas et al. (2025) "The global H5N1 influenza panzootic in mammals" *Nature*
4. Ison, Marrazzo (2025) "The emerging threat of H5N1 to human health" *N Engl J Med*
5. (2025) "APHIS confirms D1.1 genotype in dairy cattle in Nevada"
6. (2024) "Highly pathogenic avian influenza A (H5N1) virus infection reported in a person in the U.S. Available from: htt ps"
7. (2025) *Centers for Disease Control. H5 Bird Flu: current situation. Avian Influenza*
8. Jassem, Roberts, Tyson et al. (2025) "Critical illness in an adolescent with influenza A(H5N1) virus infection" *N Engl J Med*
9. (2025) "Louisiana Department of Health"
10. (2025) "Avian influenza A"
11. Garg, Reinhart, Couture et al. (2025) "Highly pathogenic avian influenza A(H5N1) virus infections in humans" *N Engl J Med*
12. Mellis, Coyle, Marshall et al. (2024) "Serologic evidence of recent infection with highly pathogenic avian influenza A(H5) virus among dairy workers -Michigan and Colorado, June-August Full-Length Text Journal of Clinical Microbiology November"
13. *MMWR Morb Mortal Wkly Rep*
14. (2025) "Accelerated Subtyping of Influenza A in Hospitalized Patients" *Centers for Disease Control*
15. (2024) "Interim guidance on the use of antiviral medications for treatment of human infections with novel influenza a viruses associated with severe human disease"
16. (2024) "Interim guidance for infection control within healthcare settings when caring for confirmed cases, probable cases, and cases under investigation for infection with novel influenza a viruses associated with severe disease"
17. Pinsky, Bradley (2024) "Opportunities and challenges for the U.S. laboratory response to highly pathogenic avian influenza A(H5N1)" *J Clin Virol*
18. Bedford, Greninger, Roychoudhury et al. (2020) "Cryptic transmission of SARS-CoV-2 in Washington state" *Science*
19. (2024) "Lab Advisory: CDC open call to industry for influenza A(H5) diagnostic test development and validation"
20. Sahoo, Morante, Huang et al. (2024) "Multiplex Dual-Target Reverse Transcription PCR for Subtyping Avian Influenza A(H5) Virus" *Emerg Infect Dis*
21. (2025) "Quest diagnostics offers test to detect avian influenza virus"
22. (2025) "Labcorp launches H5 Bird Flu test in the U.S., now available for order through physicians. 2025. Labcorp"
23. (2024) "Update: human infection with highly pathogenic avian influenza A(H5N1) Virus in Texas"
24. Kuypers, Wright, Ferrenberg et al. (2006) "Comparison of real-time PCR assays with fluorescent-antibody assays for diagnosis of respiratory virus infections in children" *J Clin Microbiol*
25. Karakus, Crameri, Lanz et al. (2018) "Propagation and titration of influenza viruses" *Methods Mol Biol*
26. Sayers, Bolton, Brister et al. (2022) "Database resources of the national center for biotechnology information" *Nucleic Acids Res*
27. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
28. Metzger, Lienhard, Hmb et al. (2021) "PCR performance in the SARS-CoV-2 Omicron variant of concern" *Swiss Med Wkly*
29. (2025) "Class 2 device recall CDC, influenza A/H5 subtyping kit"
30. Shu, Lindstrom (2019) "Compositions and methods for detection and discrimination of emerging influenza virus subtypes"
31. (2021) "Information for molecular diagnosis of influenza virus, 7th revision"
32. (2023) "CDC's influenza SARS-CoV-2 multiplex assay" *Centers for Disease Control*
33. Bruce, Huang, Perchetti et al. (2020) "Direct RT-qPCR detection of SARS-CoV-2 RNA from patient nasopharyngeal swabs without an RNA extraction step" *PLoS Biol*
34. (2025) "Pacific northwest respiratory virus epidemiology data"
35. Adams, Devlin, Klontz et al. "Combing the haystacks: The search for highly pathogenic avian influenza virus using a combined clinical and research-developed testing strategy"
36. Cdc, Highly (2025) "Pathogenic Avian Influenza A(H5N1) Virus: Interim Recommen dations for Prevention, Monitoring, and Public Health Investigations" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12424472&blobtype=pdf | # Monitoring of Influenza and Influenza-Like Viruses in the 2022/2023 and 2023/2024 Epidemic Seasons Using the SENTINEL and NON-SENTINEL Surveillance Systems in Poland
Acdeg Lidia, Bernadeta Brydak, D Aleksander, Masny Cd, Anna Poznańska Bf, Karol Szymański, B Katarzyna, Kondratiuk Bf, Emilia Czajkowska, B Bartosz, Mańkowski Befg, Katarzyna Łuniewska
## Abstract
Background:The SENTINEL influenza surveillance system has been used in Poland since 2004, incorporating both epidemiological and virological monitoring of influenza viruses. SENTINEL works in cooperation with general practitioners, 16 Voivodship Sanitary Epidemiological Stations (VSES), and the National Influenza Centre (NIC). NON-SENTINEL samples are collected from places that do not participate in the SENTINEL program. This enables continuous observation of virological and epidemiological situation in the country. The aim of the study was to conduct a comparative analysis of circulating influenza and influenza-like viruses during the 2022/2023 and 2023/2024 epidemic seasons, as monitored through the SENTINEL and NON-SENTINEL surveillance systems in Poland.
Material/Methods:The study material consisted of nasal and throat swabs collected by the VSES. Molecular biology methods were employed for virus detection and identification. The analyses were based on data obtained from the SENTINEL system for the 2022/2023 and 2023/2024 epidemic seasons.
Results:During the 2022/2023 epidemic season, co-dominance of the A/H1N1/pdm09 and A/H3N2/subtypes was noted, whereas the 2023/2024 epidemic season in Europe was dominated by the A/H1N1/pdm09 subtype. Among the influenza-like viruses, SARS-CoV-2 and RSV were the most frequently detected in both epidemic seasons.
Conclusions:The influenza surveillance system in Poland enables continuous, weekly monitoring of the epidemiological situation throughout each season. This allows for a rapid response in the event of the emergence of new influenza virus variants. The results obtained as part of influenza surveillance in Poland are consistent with those observed in other European countries.
## Introduction
Influenza is a contagious respiratory infection, known since the 16 th century, that spreads quickly during outbreaks. Influenza is caused by types A and B viruses, and occasional pandemics are triggered exclusively by influenza A viruses [1]. In the 20 th century, there were 3 significant influenza pandemics in the world: in 1918-1919 Spanish flu, caused by the A/H1N1/ subtype, the 1957-1958 Asian flu pandemic, caused by the A/H2N2/subtype, and the 1968-1969 Hong Kong pandemic caused by the A/H3N2/subtype [2]. The rapid spread of the disease is attributed to both the increased mobility of people and the growing volume of international travel, as well as the presence of animal reservoirs, which contribute to the emergence of new strains and subtypes of the influenza virus [3].
The World Health Organization (WHO) recognized the farreaching consequences of the Spanish flu pandemic in the 20 th century, not only in terms of health outcomes, estimated to have caused 50-100 million deaths, but also the severe economic impact. Therefore, global virological and epidemiological research program was established in the form of international influenza surveillance [4]. The WHO Global Influenza Surveillance Network has been continuously improved and in 2011 it was renamed the Global Influenza Surveillance and Response System (GISRS) [3].
Currently, the network consists of 7 International Reference Centres worldwide, along with 152 National Influenza Centres (NICs) in 130 countries [5]. In Poland, the NIC is located at the National Institute of Public Health and the National Institute of Hygiene -National Research Institute (NIPH, NIH-NRI). In the 2004/2005 epidemic season, after meeting WHO's requirements, Poland established the SENTINEL Influenza Surveillance System. The system, which includes cooperation with 16 VSES, monitors circulating respiratory viruses [6]. The materials for the studies consist of throat and nasal swabs, as well as bronchial washings. A model for weekly epidemiological reporting was developed for collecting patient interviews, and training in molecular biology techniques of real-time RT-PCR and RT-PCR was provided. The VSES laboratories were equipped with appropriate equipment and mandated their active participation in SENTINEL supervision, as well as to support the development of this surveillance system at a provincial level, particularly by establishing cooperation with family medicine physicians, whose involvement was essential to its functioning. Clinical surveillance of influenza was conducted by the European Influenza Surveillance Network (EISN) and consisted of weekly reporting of confirmed and suspected influenza cases during the epidemic season by participating physicians [7].
Currently, due to the digitalization of data in Poland, family physicians no longer complete the MZ-55 Report on suspected cases of influenza and influenza-like viruses, which was previously submitted to the Epidemiology Department of the National Institute of Public Health, National Institute of Hygiene -National Research Institute. Instead, data are now transmitted directly to IT systems as part of medical events to the eHealth Center system. Reporting of this document has been suspended from July 1, 2023 [8]. The quality of virological and epidemiological surveillance for seasonal influenza and highly pathogenic influenza viruses depends on national government policies, economic conditions, and the public health and economic threats posed by this pathogen.
The aim of this study was to comparatively analyze circulating influenza viruses and influenza-like viruses during the 2022/2023 and 2023/2024 epidemic seasons, based on data from the SENTINEL and NON-SENTINEL surveillance systems in Poland.
## Material and Methods
The study was approved by the Local Bioethics Committee at the National Institute of Public Health (opinion 2/2024 from 10.04.2024). All procedures followed national surveillance protocols under GISRS.
## Surveillance System
Poland has been a part of the Global Influenza Surveillance and Response System (GISRS). Since the 2004/2005 influenza season, surveillance has been conducted through the SENTINEL system, which tracks both virological and epidemiological data. NIC acts as the coordinator of the system. Monitoring takes place throughout the entire influenza season, with the highest activity usually occurring from January to April, consistent with seasonal patterns in the northern hemisphere. Data are collected on a weekly basis and include: • epidemiological data: the number of patients showing symptoms that meet the criteria for influenza-like illness (ILI).
• virological data: laboratory analysis of collected samples using molecular biology techniques.
The system includes data from both SENTINEL and NON-SENTINEL sources. The group of doctors participating in SENTINEL surveillance, who perform these activities on an honorary basis, should constitute 1-5% of practicing doctors in the country. They are responsible for collecting swabs and gathering information on the patient's health status using paperbased reports. it to the system. Only the NIC, as the coordinator of the system, has access to the entered data. Currently, data are aggregated and forwarded as weekly reports to the TESSy platform. NON-SENTINEL samples are collected from hospitals, public institutions, and private healthcare facilities that do not participate in the SENTINEL program. The SENTINEL and NON-SENTINEL Surveillance Systems cover 15 respiratory viruses, namely: influenza virus type A and B, RS virus type A and B, parainfluenza type 1, 2, 3, 4, human metapneumovirus (hMPV), adenovirus, rhinovirus, coronavirus 229E/NL63 and OC43/HKU1, and enterovirus. Since the onset of the SARS-CoV-2 pandemic, the SARS-CoV-2 virus has been included in the SENTINEL and NON-SENTINEL system reporting. Within the scope of epidemiological surveillance, VSES report weekly case numbers in 7 age groups: 0-4, 5-9, 10-14, 15-25, 26-44, 45-64, and over 65 years. Since the 2013/2014 season, all reporting has been conducted via an online platform (https://sentinel.pzh.gov.pl/artyku-ly/podglad.php?id_artykulu=29), where employees of NIC are the administrators and coordinators. They have access to the epidemiological and virological data entered by the employees of VSES. This continuous reporting supports global monitoring and ensures a rapid response if a new influenza subtype with pandemic potential emerges.
## Material
The analyses concerned data submitted to the SENTINEL system during the 2022/2023 and 2023/2024 epidemic seasons. Samples were collected from patients from all over the country. Swabs with transport medium or saline were used. During the 2022/2023 epidemic season, a total of 3324 samples were tested in the SENTINEL system and 5328 in NON-SENTINEL system. During the 2023/2024 epidemic season, 2613 samples were tested in the SENTINEL system and 10362 samples were tested in the NON-SENTINEL system.
Twice during the epidemic season (in January and April), VSES are required to send positive samples to the National Influenza Center-coordinator of the SENTINEL surveillance system. Further analyses are carried, including subtyping and assessment of the resistance of viruses to antiviral drugs (Oseltamivir and Zanamivir).
## Methods
The research was conducted in VSES, hospital laboratories as well as in the laboratory of the NIC. NIC acting as a reference center, was responsible for controlling the analysis performed by VSES. The results were subsequently weekly entered into the SENTINEL system by VSES, along with annotations about the patient's age, the testing method used, and the test outcome. Data results were downloaded weekly from the system and analyzed in NIC. Only NIC has access to the results entered to the system by VSES.
## Analysis Performed in VSES
The VSES do not use standardized testing protocols. Instead, they employ various IVD tests based on molecular biology methods, in accordance with the manufacturer's recommendations (Table 1). material in NIC. According to the recommendations of the manufacturer, 200 μL of the suspension was taken. An inventory of isolation reagents used in other laboratories was not prepared.
## Viral RNA Isolation in NIC
## Typing and Subtyping Influenza Viruses in NIC
In NIC primers and probes from the International Reagent Resource were used in the analyses. The SuperScript Platinum III (Seegene) was used for the analysis. Reactions were performed in a CFX OPUS 96 Real-Time PCR system under the following reaction conditions: RNA was subjected to reverse transcription (at 50°C for 30 minutes). The obtained DNA was subjected to the initial denaturation process (1 cycle at 95°C for 2 minutes), followed by 45 cycles of amplification denaturation at 95°C for 15 seconds, annealing at 55°C for 10 seconds and elongation at 72°C for 20 seconds. Overall volume of the reaction was 20 µL (5 µL of genetic material and 15 µL of mixed primers, probes, buffer, water, and enzyme). Positive controls were viruses that were selected for the vaccines:
## Statistical Analysis
The numbers and proportions of positive samples for different viruses (influenza virus type or subtype, influenza-like virus type) according to season (2022/2023 and 2023/2024) and surveillance system (SENTINEL, NON-SENTINEL) were analyzed using descriptive statistics methods.
Comparisons between data from both analyzed surveillance systems or the epidemiological seasons regarding the frequency of confirmation of particular viruses (in the case of influenza, also their types/subtypes) were conducted using the chi-square test. A significance level of 0.05 was assumed in all analyses.
## Results
## Epidemic Season 2022/2023
During the 2022/2023 epidemic season, a total of 3324 samples were tested in the SENTINEL system, of which 1047 (31.5%) were confirmed as positive for influenza infection, as shown in Figure 1. From all confirmed cases, 505 were determined to be influenza type A and were 351 influenza type B. Within detections of influenza type A, 92 cases were determined to be A/H1N1/pdm09 subtype and 99 were A/H3N2/subtype. In the NON-SENTINEL part of the surveillance system, 5328 samples were tested among the confirmed cases, and the following were registered: 1625 cases of unsubtyped type A viruses, 61 cases of A/H1N1/pdm09 viruses, 36 A/H3N2/, and 297 type B viruses. In both SENTINEL and NON-SENTINEL, influenza-positive cases display different percentages: influenza type A accounting for 69.5% of registered cases, influenza type B for 21.1%, subtype A/H1N1/pdm09: 5.0%, and A/H3N2/: 4.4%.
The NON-SENTINEL system more frequently confirmed influenza viruses (37.9% vs 31.5%), identified type A viruses (85.3% vs 66.5%), and recorded unsubtyped type A viruses (80.5% vs 48.5%). All differences were statistically significant (chisquare test, p<0.001).
## Influenza-Like Viruses During the 2022/2023 Epidemic Season
During the 2022/2023 season, RSV was the predominant influenza-like virus, with 164 cases in the SENTINEL system and 66 in NON-SENTINEL, together accounting for 91.3% of detections (Figure 2). Other viruses, found only in SENTINEL, included adenovirus (13 cases), Mpv (4), rhinovirus (3), and PIV-4 (2). The overall detection rate was significantly higher in SENTINEL (5.6%) than in NON-SENTINEL (1.2%) (p<0.001).
## Epidemic Season 2023/2024
During the 2023/2024 epidemic season, 2613 samples were tested in the SENTINEL system, with 874 (33.4%) confirmed cases: 536 unsubtyped type A viruses, 246 A/H1N1/pdm09, 35 A/ H3N2/, and 57 type B viruses (Figure 3). In the NON-SENTINEL system, 10 362 samples were tested, with 5872 confirmed cases: 4749 unsubtyped type A viruses, 641 A/H1N1/pdm09, 18 A/H3N2/, and 464 type B viruses. The dominant virus types were unsubtyped type A (78.3%), A/H1N1/pdm09 (13.1%), type B (7.7%), and A/H3N2/ (0.8%).
The NON-SENTINEL system more frequently confirmed influenza viruses (56.7% vs 33.4%) and recorded unsubtyped type A viruses (80.9% vs 61.3%). The differences were statistically significant (chi-square test, p<0.001).
## Influenza-Like Viruses During the 2023/2024 Epidemic Season
During the 2023/2024 epidemic season, a total of 841 cases of influenza-like viruses were confirmed across the SENTINEL and NON-SENTINEL systems (Figure 4). The dominant pathogen was SARS-CoV-2, accounting for 87.6% of all detections. In the SENTINEL system, 723 cases were confirmed, with SARS-CoV-2 (86.6%) and RSV (11.5%) being the most common. Other viruses included PIV-4 (5 cases), Rh-4 and Adv (4 each), and Mpv (1 case). In the NON-SENTINEL system, 111 cases were confirmed, with SARS-CoV-2 comprising 94.1%, followed by RSV (2.5%), Rh (1.7%), and both PIV-4 and Mpv (0.8%).
A statistically significant difference was observed in detection rates between systems (p<0.001), with 27.7% positivity in SENTINEL vs 1.1% in NON-SENTINEL. Additionally, the proportion of SARS-CoV-2 among confirmed cases was significantly higher in the NON-SENTINEL system (94.1% vs 88.6%, p=0.022).
## Comparison of the 2022/2023 and 2023/2024 Seasons
In the 2023/2024 season, the number of samples tested decreased by 21.4% in the SENTINEL system, while it nearly doubled (+94.5%) in the NON-SENTINEL system. Despite this, the detection rate of influenza viruses in SENTINEL remained stable (33.4% vs 31.5% in 2022/2023), whereas in NON-SENTINEL it increased significantly (56.7% vs 37.9%; p<0.001).
For influenza-like viruses, the trend was reversed: SENTINEL showed a marked increase in detection (from 5.6% to 27.7%; p<0.001), mainly due to SARS-CoV-2, while NON-SENTINEL remained virtually unchanged (1.2% vs 1.1%).
The structure of detected influenza viruses also shifted. In both systems, the share of type A viruses increased significantly (SENTINEL: 66.5% ® 93.5%; NON-SENTINEL: 85.3% ® 92.1%; p<0.001), reducing the proportion of type B viruses. The share of unsubtyped type A viruses rose in SENTINEL (48.5% ® 61.3%; p<0.001) but remained stable in NON-SENTINEL (80.5% ® 80.9%). However, their relative share within type A viruses decreased in both systems (SENTINEL: 72.6% ® 65.6%, p=0.004; NON-SENTINEL: 94.4% ® 87.8%, p<0.001).
As in previous years, the peak of infections occurred between January and March, with the highest number of confirmed influenza cases recorded in week 2 of 2022/2023 and week 8 of 2023/2024 (Figure 5).
## Discussion
According to WHO data, approximately one billion cases of seasonal influenza are recorded worldwide each year, of which 3-5 million are severe [9].
The results obtained in the study correspond to the results observed across Europe. During the 2022/2023 and 2023/2024 epidemic seasons, the dominance of the influenza virus type A was also recorded in other European countries. Notably, in the 2022/2023 epidemic season, Europe experienced a new pattern of seasonal dynamics, characterized by 2 peaks during the epidemic season. The first peak concerned the dominance of influenza type A, while the second peak saw influenza type B becoming dominant [10,11]. This pattern mirrors the results obtained in Poland, with the first peak occurring in week 2 and the second in week 10. In the case of influenza type A, co-dominance of the A/H1N1/pdm09 and A/H3N2/subtypes was noted, which is confirmed by the results obtained in Poland. In the 2023/2024 epidemic season in Europe, a clear dominance of the A/H1N1/pdm09 subtype was recorded [12].
During the analyzed epidemic seasons, influenza-like viruses were also recorded, with the most common viruses being SARS-CoV-2 and RSV. This is likely due to the high popularity of rapid antigen tests, which can detect influenza A and B, RSV, and SARS-CoV-2 from a single sample. This type of test appeared on the market after the outbreak of the COVID-19 pandemic and was a supplement to molecular diagnostics. At first, they were dedicated to medical personnel and used in emergency departments; they gradually became widely available in pharmacies and grocery stores and were intended for home use. Furthermore, in accordance with the Minister of Health's regulation of January 5, 2023, primary healthcare providers were authorized to perform rapid tests free of charge [13]. Thanks to this, primary care physicians began to receive confirmation of the occurrence of influenza-like viruses, which enriched the data reported by them in epidemiological reports.
The most cost-effective and efficient method of preventing influenza is vaccination. The vaccination rate remains disappointingly low, despite the fact that the vaccine is free of charge for many patient groups. Information on the refund percentage for a given age group is provided at the beginning of the epidemic season. In the 2022/2023 epidemic season, the vaccination rate in Poland was 5.1%, and in the 2023/2024 season, it increased slightly to 5.5% [14]. In recent years, inactivated quadrivalent split or subunit vaccines have been available in the country and were administered intramuscularly.
The NON-SENTINEL system consistently reported higher detection rates of influenza viruses, likely due to broader testing coverage and inclusion of hospitalized patients. In contrast, the SENTINEL system showed a more balanced detection of both influenza and influenza-like viruses, particularly SARS-CoV-2 in 2023/2024. This comparative analysis highlights the complementary roles of the SENTINEL and NON-SENTINEL systems in capturing the dynamics of influenza and influenza-like viruses. The observed differences underscore the importance of maintaining both systems for comprehensive surveillance and timely public health response.
## Limitation of the Study
The work is based solely on the analyses of samples that were reported to the sentinel influenza surveillance system. Not all tested samples are reported to the system; therefore, the number of patients studied in Poland in given seasons could be much higher.
## Conclusions
The influenza surveillance system in Poland enables continuous, weekly monitoring of the epidemiological situation throughout each season. The SENTINEL and NON-SENTINEL systems, despite methodological differences, provided complementary insights into virus circulation patterns. This allows for a rapid response in the event of the emergence of new influenza virus variants or other respiratory viruses. The results obtained as part of influenza surveillance in Poland are consistent with those observed in other European countries. Influenza is a huge threat, especially for patients from high-risk groups, and causes many deaths every season. The best way to prevent influenza is seasonal influenza vaccination, which is why the level of vaccination should be increased.
## References
1. Paules, Subbarao, Influenza (2017) *Lancet*
2. Escuyer, Gowie, George (2024) "Influenza virus surveillance from the 1918 influenza pandemic to the 2020 coronavirus pandemic" *Viruses*
3. Brydak (2008) "Influenza: Myth or real threat"
4. Ziegler, Mamahit, Cox (2018) "65 years of influenza surveillance by a World Health Organization-coordinated global network. Influenza Other Respir Viruses"
5. Cieślak, Kowalczyk, Szymański et al. (2017) "The Sentinel system as the main influenza surveillance tool" *Adv Exp Med Biol*
6. Byambasuren, Paradowska-Stankiewicz, Brydak (2020) "Epidemic influenza seasons from 2008 to 2018 in Poland: A focused review of virological characteristics" *Adv Exp Med Biol*
7. (2023) "Regulation of Ministry of Health from July 3, 2023 changing the regulation of the statistical research program of public statistics for"
8. Who; Available Online
9. Broberg, Svartström, Riess (2024) "Co-circulation of seasonal influenza A(H1N1)pdm09, A(H3N2) and B/Victoria lineage viruses with further genetic diversification, EU/EEA, 2022/23 influenza season" *Euro Surveill*
10. Ecdc (2022) "Flu News Europe bulletins -season"
11. Broberg, Vukovikj, Svartström (2024) "Antigenic changes in influenza A(H3N2) driven by genetic evolution: Insights from virological surveillance" *Euro Surveill*
12. "Regulation of Ministry of Health from January 5, 2023 changing the regulation on guaranteed benefits in the field of primary health care (Dz.U. 2023"
13. Pzh-Pib "Current Contents/Clinical Medicine" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12461958&blobtype=pdf | # Enhancement of lentiviral vectors gene delivery against VEGFR2 expressing cells by co-display of the binding and fusogenic moieties: a two molecules targeting approach
Farzin Roohvand, Roshanak Ahani, Mohammad Hossein Etemadzadeh, Nasir Mohajel, Mahdi Behdani, Reza Cohan, Mohamad Kalani, Navid Madani, Kayhan Azadmanesh
## Abstract
Background Pseudotyped lentiviral vectors (LVs) were used in cancer therapy as gene delivery systems for inhibiting tumor angiogenesis by targeting cells overexpressing "vascular endothelial growth factor receptor-2 (VEGFR2)". Herein, we report that switching from chimeric sindbis virus glycoprotein (SVG) harboring VEGFR2-specific nanobody (VEGFR2-Nb) to Two-Molecules Targeting Approach (TMTA: independent co-display of binding and fusogenic moieties) highly enhanced the transduction efficiency (TE) of the targeted LVs.Methods Several LVs co-displaying either the VEGFR2-Nb or the natural ligand (VEGF121) as targeting moiety, along with a de-targeted mutant form of SVG (as a binding deficient and fusion competent) fusogenic moiety were produced. LVs were constructed via various backbones and linkers (platelet-derived growth factor receptor "PDGFR" and CD28 as transmembrane domains, and HL and Fc as spacer domains).
ResultsExpression and incorporation of the VEGFR2-Nb and SVG onto lentiviral particles were confirmed by flowcytometry and Western blotting while their co-display was demonstrated by virus-capture ELISA and virus-cell binding assays. LVs co-enveloped with fusogen and either VEGFR2-Nb or VEGF121 showed higher TEs in VEGFR2expressing cells (72% and 91%, respectively) over LVs pseudotyped with chimeric fusogen containing the same nanobody (30%). In silico analyses indicated a direct correlation for the TE and the distance between the nanobody and the lipid bilayer.
ConclusionCompared to the chimeric strategy, the two-molecule targeting approach of LVs, due to its flexible and modular nature provides higher TE and thus great potentials for targeted gene delivery.
## Introduction
Gene therapy against a disseminated disease such as cancer is likely to achieve optimal results when delivered systemically to target metastatic lesions [1,2]. Nevertheless, the concept of "systemic gene therapy" faces significant challenges related to specificity, efficacy, and safety, which underscores the need for "targeted delivery" in gene therapies [3,4]. To address these issues, the engineering of targeted delivery vectors, such as Lentiviral vectors (LVs), to recognize cancer-specific or associated markers has been proposed as a promising solution [5]. Tumor-associated endothelial cells (TAECs) play a crucial role in tumor angiogenesis [6,7] and are accessible to vectors administered systemically [8,9]. Furthermore, the capacity of a single endothelial cell to support up to 100 tumor cells [8,10], provides an excellent opportunity for delivery of transgenes that can exert bystander effects [11]. Given that vascular endothelial growth factor Receptor 2 (VEGFR2), a key mediator of angiogenesis [12][13][14], is upregulated on TAECs across various tumor types [15], makes TAECs an intriguing candidate for targeted gene therapy via VEGFR2.
A prevalent method for achieving transductional targeting of LVs at the cell surface involves pseudotyping these vectors with heterologous viral glycoproteins (GPs) that are engineered to include a specific targeting moiety [16,17]. However, a significant limitation of this strategy is the potential impact of the targeting moiety's insertion on the fusion activity of the GP, which may reduce the infectivity of the viruses [18]. One strategy to bypass this obstacle takes advantage of sindbis virus GP (SVG). The SVG is composed of E1 and E2 GPs where E1 mediates fusion at low pH, independent of binding activity of E2 [19]. In previous work, we developed a binding-deficient SVG and created chimeric SVGs by inserting a VEGFR2specific nanobody (3VGR19) or VEGF121, a natural ligand of VEGFR2, between residues 71-74 of the sindbis virus E2 GP [20]. LVs pseudotyped with these chimeric glycoproteins (GPs) demonstrated a transduction efficiency of approximately 30% in VEGFR2-expressing, 293/ KDR cells; however, there remains a persistent need for improved transduction rates. Indeed, the insertion of a notably large targeting moiety into the sindbis GP presents significant challenges in preserving the structural integrity of both the targeting moiety and the envelope GPs in the constructed chimeras [21,22]. In this context, it is shown that insertion of avidin or streptavidin into E2 of the sindbis GP disrupted the production of infectious pseudotyped LVs likely due to the destabilization of trimers of E1/E2 heterodimers [23]. To address this concern, several groups have exploited adaptor-based approaches via bispecific antibodies [24] or insertion of monomeric biotin-binding domain [23], SpyTag [25], and PDZ1 peptide [26] into E2 molecule of detargeted sindbis GPs. While these strategies may alleviate the necessity for developing functional chimeric GPs for each targeting ligand, concerns regarding the potential instability of the vector-adaptor complex in vivo, particularly with bispecific antibodies, and the difficulties associated with the separate production of vector and adaptor molecules, including regulatory challenges, pose significant barriers to clinical implementation [1].
Another sophisticated strategy in transductional targeting takes advantage of the inherent ability of lentiviral virions to incorporate host cell plasma membrane proteins during the budding process. In this approach, the recombinant targeting moiety is co-expressed independently alongside a separate detargeted fusogenic GP on the surface of virus-producing cells, and therefore, both components are co-displayed on the resulting lentiviral particles [18]. This approach allows for the production of targeted lentiviral vectors as a unified product, which is advantageous for large-scale manufacturing and regulatory compliance. Furthermore, this strategy mitigates the difficulties associated with chimeric targeting methods, particularly in maintaining the correct functional conformation of the envelope glycoprotein and/or targeting component, thereby facilitating the creation of novel targeted LVs. Several studies adopted this approach and targeted LVs by co-displaying detargeted mutant forms of SVG that are binding-deficient yet retain fusogenic capabilities, in conjunction with various targeting moieties such as antibodies [27,28], natural ligands [29], and receptors [30]. This approach has proven to be flexible and readily optimized through modifications to transmembrane or spacer domains [31,32]. The targeting ligand plays a crucial role in the functionality of targeted LVs, as its specific and high-affinity binding to tumor cell-specific or associated markers represents the initial and vital step in the transduction of the intended cell types. Nanobodies, characterized by their small size, stability, and high-affinity antigen-binding properties, present a promising candidate as a targeting ligand [33]. As mentioned above, we previously showed that LVs pseudotyped with sindbis virus E2 GPs harboring VEGFR2-specific nanobody (3VGR19) or VEGF121 as a targeting ligand reached to a transduction efficiency of around 30% in VEGFR2-expressing, 293/KDR cells. In the current study, we investigated whether switching the targeting strategy from chimeric to two-molecule (i.e.: co-displaying targeting/binding and fusogenic moieties) will improve the transduction efficiency of targeted LVs. To this end, we examined various backbones/frameworks to display 3VGR19 nanobody [34] alongside the de-targeted mutant form of SVG. Additionally, to assess whether this strategy is independent of the targeting moiety, we used VEGF121 as a control for the targeting modality.
## Materials and methods
## Plasmids
Plasmids pLOX-CWgfp (LV transfer vector encoding GFP as the reporter gene; Addgene plasmid # 12241 [35]), psPAX2 (packaging plasmid; Addgene plasmid # 12260), and pMD2.G (encoding vesicular stomatitis virus glycoprotein G (VSV-G); Addgene plasmid # 12259) were gifts from Didier Trono (Department of Genetics and Microbiology, Faculty of Medicine, CMU, Geneva, Switzerland). Plasmids p2.2-L [20] and pDis-Nb [36] were previously described. p2.2-L encodes a binding deficient and fusion competent mutant form of SVG, in which HA-and His-tag linkers are inserted into E2 GP. pDis-Nb is a derivative of pDisplay (Invitrogen, USA) which expresses a membrane-anchored form of 3VGR19 nanobody. Plasmid pcDNA3.1-PS11-scFvFc-CD28-gp41 (706-713) [37], herein designated as pDF, was kindly provided by Dr. Wayne Marasco (Dana-Farber Cancer Institute, Boston, Massachusetts). This plasmid has been used to display antibody fragments on the surface of LVs and human cells. The pDis-Nb-HL vector was generated by insertion of a synthetic sequence encoding a helical linker (RGSGA(EAAAK) 7 ALGS) (Biomatik, Canada) [38], into the SalI recognition site of pDis-Nb. To construct pDF-Nb, the sequence corresponding to 3VGR19 nanobody was PCR amplified from pHEN6C-3VGR19 [34] and the resulting fragment was cloned at SfiI and NotI restriction sites, in frame with Fc domain of pDF to replace the ScFv. The pDF-Nb-ΔFc plasmid was constructed by cloning 3VGR19 nanobody into SfiI and XbaI restriction sites of pDF. The same strategies were used to construct different plasmids encoding membrane-bound VEGF121 (a natural isoform of VEGF, consisting of 121 amino acids) and the constructs were designated as pDis-VE, pDis-VE-HL, pDF-VE, and pDF-VE-ΔFc, respectively.
## Cells and transfections
293T cells were obtained from National cell bank of Iran (Pasteur Institute of Iran, Iran). 293/KDR cells were purchased from Sibtech (USA). Cells were grown in DMEM supplemented with 10% FBS, 1% v/v Glutamax (Life Technologies, USA), and Pen/Strep (100 units/mL of penicillin and 100 µg/mL of streptomycin). Transfections were performed by Turbofect transfection reagent (Thermoscientific, Lithuania) according to the manufacturer's instructions. Briefly, one night before transfection, 5 × 105 or 1 × 106 293T cells were seeded in 6-well plates or 6-cm cell culture dishes, respectively. Subsequently, cells were transfected with a total of 4 µg of plasmids and 6 µl of Turbofect transfection reagent for 6-well plates or 6 µg of plasmids and 12 µl of Turbofect transfection reagent for 6-cm cell culture dishes.
## LV production and titration
To produce LVs co-displaying the targeting motif (3VGR19 nanobody or VEGF121) and SVG, 293T cells were seeded in 6-cm culture dishes one night before transfection. The following day, cells were transfected with 2.6 µg pLOX-CWgfp, 1.7 µg psPAX2, 0.85 µg of p2.2-L, and 0.85 µg of either of the plasmids: pDis-Nb, pDis-Nb-HL, pDF-Nb, pDF-Nb-ΔFc, pDis-VE, pDis-VE-HL, pDF-VE or pDF-VE-ΔFc, using Turbofect transfection reagent. The resultant viruses were designated as Dis-Nb/2.2-L, Dis-Nb-HL/2.2-L, DF-Nb/2.2-L, DF-Nb-ΔFc/2.2-L, Dis-VE/2.2-L, Dis-VE-HL/2.2-L, DF-VE/2.2-L, and DF-VE-ΔFc/2.2-L, respectively. As controls, 293T cells were transfected with 3 µg pLOX-CWgfp, 2 µg psPAX2, and 1 µg of either pMD2.G (positive non-targeting controls, designated as VSV-G viruses) or p2.2-L (negative controls, designated as 2.2-L viruses). Viral supernatants were collected 48 and 72 h post-transfection. The supernatants were centrifuged at 3000 g for 15 min at 4 °C to remove cell debris. To concentrate viral vectors, centrifugation was performed at 48,000 g for 3 h at 4 °C and viral pellets were resuspended in cold PBS. Physical titration of viral vectors was performed by p24 (capsid) ELISA (Pasto Lentivirus HIV p24, Pasteur Institute of Iran, Iran) according to the manufacturer's instructions.
## SDS-PAGE and western blot analyses
Around 1 × 10 6 virus-producing cells were harvested and centrifuged at 300 g for 5 min. Cell pellets were resuspended in appropriate volumes of SDS loading buffer. In parallel, equal amounts of concentrated viruses (normalized by p24 content) were mixed with SDS loading buffer. Cell and virus suspensions were then placed in boiling water for 5 min. The denatured suspensions were separated by 12% SDS-PAGE and finally electrotransferred to a PVDF membrane. To detect the expression of Dis-Nb, Dis-Nb-HL proteins (expressed from pDis-Nb and pDis-Nb-HL, respectively) and 2.2-L (expressed from p2.2-L), the membranes were incubated with primary mouse anti-HA tag antibody (1:1000, Cell Signaling, USA) and secondary rabbit anti-mouse HRP conjugated antibody (1:2000, Cell Signaling, USA). Protein bands were finally visualized using Amersham ECL Western Blotting Detection Reagents (GE Healthcare Life Sciences, UK) and X-ray films. Intensities of bands were quantified by Fiji distribution of ImageJ software [39].
## Cell surface expression analyses of membrane-bound forms of nanobody (mNbs) by flow cytometry
Flow cytometry was used to detect the surface expression of mNbs. To this end, 293T cells were transfected with corresponding constructs and 48 h post-transfection, cells were dissociated by PBS containing 10 mM EDTA and 1 × 10 6 cells were stained for each group. For pDisplay-based construct (pDis-Nb or pDis-Nb-HL), mouse anti-c-myc-tag mAb IgG2a (1 µg/ml, Genscript, USA) and Anti-Mouse IgG2a PerCP-eFlour 710 (0.25 µg/test, eBioscience, USA) were used as primary and secondary antibodies, respectively. To detect pDF-Nb, cells were stained with FITC conjugated-anti-Fc antibody (Razibiotech, Iran). Mock-transfected cells were also stained as negative controls.
All flow cytometric analyses were performed by Cyflow SL (Partec, Germany) and the percentage of positive populations was determined by Super-Enhanced Dmax Subtraction (SED) algorithm of FlowJo software (FLOWJO LLC, USA).
## Virus capture ELISA
Virus capture ELISA was performed as previously described [16]. Briefly, a 96-well plate was coated with 2.5 µg/ml of extracellular domain of human VEGFR2 (G&P biosciences, USA) or 2.5 µg/ml bovine serum albumin (BSA) (Sigma, USA) using carbonate/bicarbonate buffer (pH 9.5). Wells were blocked with PBS-BSA 3%. Equal amounts of concentrated lentiviral particles (determined by p24 content) were added to each well and incubated for 1 h at 37 °C. Wells were then washed four times with PBS-FBS 2%. Captured viruses were lysed and the virus lysates were transferred to a p24 pre-coated ELISA plate (Pasto Lentivirus HIV p24, Pasteur Institute of Iran, Iran). Subsequently, p24 contents were determined according to the manufacturer's protocol.
## Virus-cell binding assay
0.5 × 10 6 293/KDR cells were incubated with equal amounts of concentrated LVs (normalized by p24 content) for 1 h at 4˚C. The virus-bound cells were detected by staining with His-probe mouse monoclonal antibody (1:50, Santa Cruz, USA) followed by Anti-mouse IgG, F(ab´)2 Fragment (PE Conjugate) (1:500, Cell Signaling, USA). As negative controls, 293/KDR cells or 293/KDR cells incubated with 2.2-L LVs were stained with the above-mentioned antibodies. Finally, cells were analyzed by Cyflow SL flow cytometer (Partec, Germany).
## Targeted transduction
293T or 293/KDR cells were seeded in 24-well cell culture plates. The next day (at 70% confluency), cell medium was removed and cells were incubated with equal amounts of each un-concentrated viral group (normalized by p24 content) for 8 h. Subsequently, viruses were replaced with fresh medium and expression of GFP was assessed 96 h later by flow cytometry using Cyflow SL (Partec, Germany).
## In Silico analyses
The molecular model for each transmembrane, linker, and main nanobody region was constructed and eventually assembled into a full-length molecule. In this regard, the secondary structure was predicted for each of designed molecule using PSIpred [40] and SPIDER2 [41] webservers. The transmembrane (TM) domain residues were predicted using TMHMM 2.0 webserver [42]. Molecular models in atomic scale were predicted by homology modeling, threading, and ab-initio methods using RaptorX [43], I-TASSER [44], QUARK [45], Swiss-model [46], and MODELLER [47] webservers. The best partial structure templates for each molecule, which matched more with the predicted secondary structures and transmembrane region prediction, were assembled and applied for threading one more time. The final molecular models were quality checked, energy minimized, and tested for stability by molecular dynamics (MD) simulations using NAMD version 2.12 software [48] and analyzed using VMD software [48]. A membrane patch of 50 Å dimensions, including all POPC (1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine) molecules was constructed using VMD. Four molecular models were implemented in separate membrane patches, solvated by explicit TIP3 water molecules, and ionized to the physiological concentration of NaCl. MD simulation runs were carried out for 500 to 1200 nanoseconds using NAMD version 2.12 to determine the behavior and the distance of the nanobody to the membrane in each model. All movies were generated by VMD software. Surface area accessibility (SASA) of the nanobody segment and the complementarity-determining regions (CDRs) were determined for each full-length molecule. Root mean square deviation (RMSD) and root mean square fluctuation (RMSF) were calculated and analyzed for the full-length molecules and nanobody segments after simulation runs.
## Results
## Construction of membrane-bound forms of 3VGR19 nanobody (mNbs) and membrane-bound forms of VEGF121 (mVEs) encoding plasmids
P2.2-L and pDF are presented in Fig. 1a and supplementary Fig. 1, respectively. Schematic images of plasmids encoding mNbs (pDis-Nb, pDis-Nb-HL, pDF-Nb, and pDF-Nb-ΔFc) are illustrated in Fig. 1b and images related to plasmids encoding mVEs (pDis-VE, pDis-VE-HL, pDF-VE, and pDF-VE-ΔFc) are shown in Fig. 1c. pDis-Nb-HL was generated by insertion of a helical linker between the nanobody and PDGFR transmembrane domain of pDis-Nb. pDF-Nb was generated by insertion of the nanobody into pDF upstream of the human IgG1 Fc domain. To remove the Fc domain, pDF-Nb-ΔFc was constructed by cloning 3VGR19 nanobody between human IgG1 signal peptide and CD28 transmembrane domain of pDF. The same procedures were taken for construction of mVE encoding counterparts. All constructs were verified by restriction enzyme digestion and sequencing analyses (data not shown).
## The mNbs were displayed on the surface of 293T cells
Efficient surface expression of proteins on producer cells is an essential prerequisite for incorporation of such proteins onto lentiviral particles [49]. Before investigating the surface presentation of mNbs on 293T cells, the expression of Dis-Nb and Dis-Nb-HL in virus-producing cells were assessed by western blotting using an anti-HA tag antibody. As depicted in Fig. 2a, protein bands with approximate sizes of 22 and 26 kDa (indicated by white arrows) corresponding to the expressed mNbs were detected in Dis-Nb/2.2-L and Dis-Nb-HL/2.2-L virusproducing cells, respectively. None of these protein bands were detected in 2.2-L or VSV-G virus-producing cells (used as negative control for nanobody expression).
Notably, the band intensity of Dis-Nb-HL was higher than that of the Dis-Nb (approximately a 1.5-fold increased intensity). Furthermore, in Dis-Nb/2.2-L, Dis-Nb-HL/2.2-L, and 2.2-L virus-producing cells, a band with the approximate size of 67 kDa (indicated by black arrows) corresponding to E2 GP was observed. Accordingly, the presence of mNbs on the surface of 293T cells was also assessed by flow cytometry. As depicted in Fig. 2b, about 97%, 98%, and 71% of cells were transfected with pDis-Nb, pDis-Nb-HL, and pDF-Nb, respectively, indicating surface expression of the corresponding mNbs. Of note, cells expressing Dis-Nb-HL demonstrated 50% increase in mean fluorescent intensity compared to Dis-Nb expressing cells (1.21 and 0.78, respectively).
## SVG (fusogen) and functional targeting molecules were co-incorporated onto the same LV virions
In addition to cell surface expression, it is important to have both functional targeting and fusogen molecules on the same virions. To investigate the incorporation of both As has been shown in Fig. 3a, western blot analysis demonstrated the presence of SVG (approximately 67 kDa; black arrows) along with Dis-Nb or Dis-Nb-HL (22 and 26 kDa, respectively; white arrows) in LV particles. Notably, in these virions, Dis-Nb-HL showed higher band intensity than Dis-Nb (approximately a 1.5-fold increase). To evaluate whether incorporated mNbs retained their ability to recognize their corresponding antigens, twostep capture ELISA was employed. As presented in Fig. 3b, Dis-Nb/2.2-L, Dis-Nb-HL/2.2-L, and DF-Nb/2.2-L LVs bound to VEGFR2 with high selectivity in comparison to BSA-coated wells (P value < 0.0001), while 2.2-LVs showed only background binding to VEGFR2, similar to that of BSA. Finally, to examine the co-incorporation of mNbs and SVG onto the same virions, a virus-cell binding assay was performed. 293/KDR cells were incubated with LVs at 4˚C for 1 h. Following the washing steps, virus-cell complexes were stained with an anti-His tag antibody that recognizes the His-tag present in SVG. In this condition, only mNb and SVG co-incorporating virions could bind to 293/KDR cells and be detected by antibody against SVG. The flow cytometric results of cells incubated with 2.2-L LVs revealed that they did not bind to 293/KDR cells and no histogram distribution shift was observed in comparison to mock treated cells (cells without any virus incubation) (Fig. 3c). On the other hand, as presented in Fig. 3d
## Two-molecule targeting system results in selective transduction of VEGFR2 expressing cells and is readily optimizable through modification of targeting module
To study the transduction efficiency and specificity of LVs co-displaying both 3VGR19 nanobody and SVG (Dis-Nb/2.2-L, Dis-Nb-HL/2.2-L and DF-Nb/2.2-L viruses), VEGFR2 expressing 293/KDR cells and 293T cells were used as target and negative control cells, respectively. VSV-G LVs and 2.2-L LVs served as the positive nontargeting and negative control viruses, respectively. Moreover, DF-Nb-ΔFc/2.2-L virus was used to evaluate the effect of Fc domain on transduction efficiency. As shown in Fig. 4, VSV-G LVs were able to transduce almost all cells in both 293/KDR and 293T cells, while low levels of GFP expression in both cell types (approximately 3 and 5%, respectively) was observed for that of the 2.2-L LVs. In the case of test groups, Dis-Nb/2.2-L and Dis-Nb-HL/2.2-L viruses had background infectivity on 293T cells similar to that of the 2.2-L LVs (approximately 4%); however, when incubated with 293/ KDR cells, transduction efficiency of around 14% and 37% were achieved, respectively. On the other hand, DF-Nb/2.2-L LVs reached a transduction efficiency of about 72% in 293/KDR cells and around 6% GFP expression in 293T cells. Additionally, as represented in supplementary Fig. 3, incubation of 293/KDR cells with DF-Nb-ΔFc/2.2-L virus resulted in transduction of approximately 5% of cells. Transduction assay was also performed for VEGF121-displaying LVs (Dis-VE/2.2-L, Dis-VE-HL/2.2-L, DF-VE/2.2-L, and DF-VE-ΔFc /2.2-L) (supplementary Fig. 4). Transduction efficiencies of Dis-VE/2.2-L, Dis-VE-HL/2.2-L, DF-VE-ΔFc /2.2-L, and DF-VE/2.2-L viruses on 293/KDR cells were approximately 56%, 77%, 88%, and 91%, respectively. In 293T cells, both Dis-VE/2.2-L and Dis-VE-HL/2.2-L viruses showed background transduction efficiency (≈ 3%), while a higher background infection of 293T cells was observed for DF-VE-ΔFc /2.2-L and DF-VE/2.2-L viruses (around 11% and 24%, respectively).
## In Silico studies
To obtain more insights into the structural features of the inserted nanobody in the constructed LVs and the impact of used linkers on nanobody display as well as their contribution to transduction efficiency, in silico studies were performed on Dis-Nb, Dis-Nb-HL, pDF-Nb, and pDF-Nb-ΔFc at the atomic level. The designed molecules contain regions from several known structures that may affect each other in the final molecular conformation. The transmembrane helix regions were predicted at residues 175 to 197 for Dis-Nb protein (total 202 residues), 221 to 243 for Dis-Nb-HL protein (total 248 residues), 178 to 200 for construct DF-Nb-ΔFc (total 224 residues), and 417 to 439 for DF-Nb protein (total 463 residues) (supplementary Fig. 5). The 3D models were successfully generated and Ramachandran analyses verified the quality of the full-length models, in which more than ∼ 97% residues in phi and psi plots were located in the favorite and allowed regions (supplementary Fig. 6). Following energy minimizations, the full-length molecules were simulated in POPC membranes at 310 K in 1 bar pressure constant condition. The SASA measurements either on the nanobody or the CDR segments revealed no meaningful difference between the designed constructs (supplementary Figs. 7, 8, 9, and 10). RMSD plot analyses of the fulllength molecules elucidated that all constructs reached a plateau during MD trajectories (supplementary Fig. 11 and supplementary movie files for: Dis-Nb, Dis-Nb-HL, DF-Nb). In addition, RMSD and RMSF calculations of full-length molecules, as well as nanobody segment could not explain the observed difference in the transduction results (supplementary Figs. 11, 12, 13, and 14). However, the measured distance between the targeting moiety and the lipid membrane was significantly different for each construct. The distance between the alpha carbon of the first residue of linkers and the phosphorous atom in the closest lipid molecule was obtained as 0, 28.26, 68.40, and 82.01 Å for DF-Nb-ΔFc, Dis-Nb, Dis-Nb-HL, and DF-Nb constructs, respectively (Fig. 5).
## Discussion
In the present study, we explored the potential benefits of transitioning from a chimeric to a two-molecule transductional targeting strategy in enhancing the efficiency of gene delivery via lentiviral vectors (LVs). Our findings indicated that both the 3VGR19 nanobody and VEGF121 (the natural ligand) exhibit improved performance with the two-molecule approach compared to the chimeric method. Nevertheless, the degree of enhancement is influenced by the specific context in which the targeting ligand is presented, including factors such as spacer and transmembrane (TM) domains, as well as the nature of the targeting ligand itself. Additionally, our results reveal that, unlike the 3VGR19 nanobody, the spacer domain is less significant for VEGF121, which targets a membranedistal region of its corresponding receptor. In this study, we adopted two backbones (pDisplay and pDF) to display nanobody on lentiviral particles. The pDisplay is a commercially available plasmid (containing murine Ig κ-chain leader sequence and platelet-derived growth factor receptor (PDGFR) TM domain) that was previously used to anchor targeting moieties into LVs' envelopes [50,51]. The pDF plasmid was constructed in an attempt to develop an optimal lentiviral display platform by combination of CD28 TM domain (an envelope incorporation motif derived from the membrane-proximal region of the HIV-1 gp41 cytoplasmic tail to enhance cell surface expression and protein incorporation onto LVs), along with the Fc region of the human IgG1 antibody to facilitate the display of scFv antibodies on the surfaces of LV particles [37].
In our study, the insertion of a helical linker between the transmembrane domain and the nanobody within the pDisplay-based constructs resulted in 2.5 fold increase in the transduction of 293/KDR cells (Fig. 4). This increase in transduction efficiency can be partially attributed to a 1.5-fold rise in surface expression and the integration of Dis-Nb-HL on both virus-producing cells and virions (Figs. 2 and3a, respectively). The noted enhancement in the expression of Dis-Nb-HL aligns with prior study, which demonstrated that the addition of a similar helical linker in Tf-fusion proteins led to elevated expression levels in HEK293 cells compared to the same fusion proteins lacking the helical linker [52]. The enhanced transduction efficiency may be elucidated by findings from Rasbach et al., who reported an approximate 5-fold increase in the transduction of Her2/neu positive cells through the use of the same linker [32]. They proposed that the insertion of the spacer domain (in their case, the helical linker) may have mitigated the distance between the Her2/neu-specific DARPin and its corresponding binding site located in a membrane-proximal domain of Her2/neu, thereby rendering the displayed DARPin more sterically favorable for binding to its target binding domain [53]. Following the same rationale, it can be inferred that the 3VGR19 nanobody may interact with one of the more membrane-proximal regions of the seven extracellular Ig-like domains of VEGFR2 [54]. Additionally, VEGF121 was employed as an alternative targeting moiety, which binds to the membrane-distal domains of VEGFR2 (domains II and III). The Dis-VE/2.2-L viruses demonstrated a 3.8-fold increase in transduction efficiency in KDR cells, along with a more pronounced histogram shift in the virus-cell binding assay when compared to Dis-Nb/2.2-L LVs. The observed results may be explained by the greater affinity of VEGF (Kd = 75-760 pM) [55][56][57] compared to 3VGR19 nanobody (Kd = 5.4 nM [34]) as well as potential differences in their accessibility to the respective binding sites. Conversely, the introduction of the helical linker had a negligible effect on the transduction efficiency of lentiviral vectors (LVs) presenting VEGF121, resulting in only a 1.38-fold increase (supplementary Fig. 4). This finding aligns with the membrane distal binding sites of VEGF121 within VEGFR2.
Utilizing DF-Nb, we observed an improved transduction rate relative to pDisplay-based mNbs (refer to Fig. 4). The enhanced transductional targeting efficacy of DF-Nb/2.2-L in comparison to Dis-Nb-HL/2.2-L viruses may be attributed to variations in the transmembrane domain (CD28 versus PDGFR), differences in the spacer domain (Fc domain versus HL), and/or the incorporation of the gp41 motif [37]. It has been previously established that the appropriate transmembrane domain and the inclusion of a gp41 incorporation motif play crucial roles in enhancing the cell surface display of heterologous proteins and their integration into lentiviral (LV) particles. Furthermore, it has been shown that the insertion The secondary structure of molecules were predicted for each of designed molecules using PSIpred [40] and SPIDER2 [41] webservers. The transmembrane (TM) domain residues were predicted using TMHMM 2.0 webserver [42]. Molecular models in atomic scale were predicted by homology modeling, threading and ab-initio methods using RaptorX [43], I-TASSER [44], QUARK [45], Swiss-model [46], and MODELLER [47] webservers. The distance between the alpha carbon of the first residue of linkers and the phosphorous atom in the closest lipid molecule was obtained as 0, 28.26, 68.40, and 82.01 Å for DF-Nb-ΔFc, Dis-Nb, Dis-Nb-HL, and DF-Nb constructs, respectively of Fc region from human IgG1 can significantly boost the cell surface expression of recombinant membranebound proteins [58]. Moreover, spacer domains such as IgG Fc region have been utilized to enhance antigen binding and T-cell signaling of chimeric antigen receptors (CARs), especially in the case of membrane-proximal epitopes [59,60]. This enhancement is likely linked to increased accessibility of the targeting moieties. While it is challenging to determine the specific contribution of each component to the observed higher transduction efficiency of DF-Nb/2.2-L compared to Dis-Nb-HL/2.2-L viruses, the IgG1 Fc domain appears to be a critical factor. Indeed, the removal of the Fc domain from DF-Nb resulted in a substantial decline in transduction efficiency, dropping from 72% in DF-Nb/2.2-L to 5% in DF-Nb-ΔFc/2.2-L (supplementary Fig. 3). On the contrary, no significant difference was observed in the transduction efficiency between DF-VE/2.2-L (88%) and DF-VE-ΔFc/2.2-L (91%) (supplementary Fig. 4). These findings are in agreement with the minimal impact of the helical linker on the transduction efficiency of LVs containing pDisplay-based mVEs (Dis-VE and Dis-VE-HL) and contrary to the findings related to pDisplay-based mNbs (Dis-Nb and Dis-Nb-HL), which can be attributed to the membrane-distal binding site of VEGF.
It is noteworthy that, unlike the chimeric approach, which demonstrated comparable transduction efficiency (approximately 30% of 293/KDR cells) for both 3VGR19 nanobody and VEGF121 [20], the present study revealed that the transduction efficiencies of all mVE-based lentiviral vectors (LVs) surpassed those of their mNb counterparts. One possible explanation for this observation might be structural limitations imposed on VEGF121 within the chimeric sindbis E2 GP, which are alleviated when VEGF121 is displayed independently in the twomolecule targeting approach. Conversely, the sindbis GP, which is used as a fusogenic molecule in the twomolecule targeting system (2.2-L), showed a higher level of non-specific transduction (≈ 3-6%) compared to the chimeric sindbis (≈ 1%). This suggests that the insertion of 3VGR19 nanobody or VEGF121 between residues 71-74 of E2 is more effective in detargeting 9reducing the targetting) of sindbis GP than the 20 amino acid linker used in the case of 2.2-L. Accordingly, the nonspecific transduction rates of all LVs using 2.2-L as fusogenic molecule were also in the range of 3-6%, with the exception of instances where VEGF121 was presented in pDF, which showed 11% for DF-VE-ΔFc/2.2-L and 24% for DF-VE/2.2 L LVs. The unexpectedly elevated nonspecific transduction observed in DF-VE/2.2 L and DF-VE-ΔFc/2.2-L LVs appears to be attributable to the nonspecific binding of mVEs displayed by pDF.
The results of our two-molecule targeting strategy for transduction demonstrated efficiencies that were comparable to, or even exceeded, those reported by Yang et al. [28] and Lei et al. [31]. Specifically, transduction efficiencies of 52% and 30% were achieved in 293T cells that stably expressed the target antigen, respectively. Of note, both of these studies used detargeted forms of SVG as fusogen and the full form of anti-CD20 antibody as targeting ligand which required additional accessory proteins such as Igα and Igβ in contrast to our display platform.
In our in silico study, we found no correlation between transduction efficiency and either solvent accessibility or the fluctuations of the nanobody within the constructs. This suggests that variations in the nanobody segment did not influence its accessibility to VEGFR2. However, the simulation results indicated a correlation between the distance of the nanobody from the lipid bilayer (measured at 0 Å, 28.26 Å, 68.40 Å, and 82.01 Å for DF-Nb-ΔFc, Dis-Nb, Dis-Nb-HL, and DF-Nb, respectively) and the observed transduction efficiencies (5%, 14%, 37%, and 72%, respectively) as illustrated in Fig. 5. This correlation implies that the distance between the targeting moiety and the surface of the enveloping lipid bilayer may serve as a significant predictor of ligand accessibility and binding efficiency for lentiviral vectors and their targets.
In our study, the factors affecting the transduction efficiency of targeted LVs were investigated in 293/KDR cells. To further expand these findings in clinical settings, it is essential to evaluate the identified factors in primary endothelial cells as well. Additionally, when using the Fc domain as a spacer, it is crucial to consider that Fc spacer domains may interact with Fc gamma receptors (FcγRs), potentially leading to non-specific activation and/or transduction of immune cells. This challenge must be resolved prior to advancing to in vivo studies, as previously indicated by the deletion or mutation of regions critical for Fc receptor binding in CARs [61][62][63]. Moreover, the improved transduction efficiency observed with the two-molecule strategy, coupled with the reduced non-specific transduction of chimeric SVGs [20] provides an opportunity to investigate the combination of both approaches (i.e., co-enveloping LVs with chimeric SVG and targeting molecule) to increase the avidity of LVs for the targeted cells, while benefiting from the diminished background transduction associated with chimeric SVG.
## Conclusion
Collectively, to our knowledge, we reported the first attempt to generate targeted LVs via independent co-display of detargeted SVG and nanobody. This dual-molecular targeting strategy surpassed the chimeric approach in terms of transduction efficiency. The extent of improvement in transduction efficiency appears to be influenced by both the context of the targeting ligand's display, including factors such as spacer and transmembrane (TM) domains, as well as the specific targeting ligand utilized. Notably, the removal of the Fc domain from nanobody-displaying LVs resulted in a significant reduction in transduction efficiency, decreasing from 72 to 5%, while the efficiency of VEGF121-containing LVs remained unaffected. Similarly, insertion of a helical linker between the transmembrane domain and nanobody into the pDisplay-based constructs led to 2.5 fold increase in the transduction of 293/KDR cells. We also found that higher transduction efficiencies for the two-molecule targeting system compared to chimeric were accompanied with slightly higher non-specific transduction (3-6% versus 1%). Finally, results of the in silico studies indicated a direct correlation for the transduction efficiency and the distance between the nanobody and lipid bilayer implying its role for the ligand accessibility and binding efficiency of LVs and their targets. Taken together, the modular nature of this strategy offers great possibilities to optimize transductional targeting of LVs via independent modification of targeting molecules which relieves the potential restrain on functionality of targeting ligands. In addition, it provides the possibility of individual or concurrent alteration in different domains of the display platform (such as targeting ligand, TM domain, and spacer). This strategy expedites the development of novel targeted LVs by allowing a diverse range of targeting ligands to be readily tested.
## References
1. Waehler, Russell, Curiel (2007) "Engineering targeted viral vectors for gene therapy" *Nat Rev Genet*
2. Hadi, Qutaiba, Allela et al. "Recent advances in various adeno-associated viruses (AAVs) as gene therapy agents in hepatocellular carcinoma" *Virol J*
3. Das, Menezes, Bhatia et al. (2015) "Gene therapies for cancer: strategies, challenges and successes" *J Cell Physiol*
4. Jones, Kovacevic, Ionescu et al. (2023) "The applications of targeted delivery for gene therapies in hearing loss" *J Drug Target*
5. Deng, Liang, Cui (2022) "Pseudotyped lentiviral vectors: ready for translation into targeted cancer gene therapy?" *Genes Dis*
6. Sanctis, Ugel, Facciponte et al. (2018) "The dark side of tumorassociated endothelial cells" *Semin Immunol*
7. Xu, Zhang, Xia et al. (2022) "Targeting the tumor stroma for cancer therapy" *Mol Cancer*
8. Aird (2012) "Endothelial cell heterogeneity. Cold Spring Harb Perspect Med"
9. Dong, Nor (2009) "Transcriptional targeting of tumor endothelial cells for gene therapy" *Adv Drug Deliv Rev*
10. Dhanabal, Ramchandran, Waterman et al. (1999) "Endostatin induces endothelial cell apoptosis" *J Biol Chem*
11. Trepel, Stoneham, Eleftherohorinou et al. (2009) "A heterotypic bystander effect for tumor cell killing after adenoassociated virus/phage-mediated, vascular-targeted suicide gene transfer" *Mol Cancer Ther*
12. Falcon, Chintharlapalli, Uhlik et al. (2016) "Antagonist antibodies to vascular endothelial growth factor receptor 2 (VEGFR-2) as anti-angiogenic agents" *Pharmacol Ther*
13. Ferrara, Adamis (2016) "Ten years of anti-vascular endothelial growth factor therapy" *Nat Rev Drug Discov*
14. Hida, Maishi, Sakurai et al. (2016) "Heterogeneity of tumor endothelial cells and drug delivery" *Adv Drug Deliv Rev*
15. Musumeci, Radi, Brullo et al. (2012) "Vascular endothelial growth factor (VEGF) receptors: drugs and new inhibitors" *J Med Chem*
16. Simmons, Whitehead, Kolokoltsov (2006) "Use of Recombinant lentivirus pseudotyped with vesicular stomatitis virus glycoprotein G for efficient generation of human anti-cancer chimeric T cells by transduction of human peripheral blood lymphocytes in vitro" *Virol J*
17. Patel, Giddings, Sechelski et al. (2013) "High efficiency gene transfer to airways of mice using influenza hemagglutinin pseudotyped lentiviral vectors" *J Gene Med*
18. Deng, Liang, Cui (2022) "Pseudotyped lentiviral vectors: Ready for translation into targeted cancer gene therapy? Genes & Diseases" *Genes Dis*
19. Leung, Ng, Chu (2011) "Replication of alphaviruses: a review on the entry process of alphaviruses into cells" *Adv Virol*
20. Ahani, Roohvand, Cohan et al. (2016) "Sindbis Virus-Pseudotyped lentiviral vectors carrying VEGFR2-Specific nanobody for potential transductional targeting of tumor vasculature" *Mol Biotechnol*
21. Michels, Ho, Buchholz (2022) "Precision medicine: in vivo CAR therapy as a showcase for receptor-targeted vector platforms" *Mol Ther*
22. Morizono, Pariente, Xie et al. (2009) "Redirecting lentiviral vectors by insertion of integrin-tageting peptides into envelope proteins" *J Gene Med*
23. Situ, Chua, Bae et al. (2018) "Versatile targeting system for lentiviral vectors involving biotinylated targeting molecules" *Virology*
24. Parker, Jacobs, Huckaby et al. (2020) "Efficient and Highly Specific Gene Transfer Using Mutated Lentiviral Vectors Redirected with Bispecific Antibodies. mBio"
25. Kasaraneni, Chamoun-Emanuelli, Wright et al. (2017) "Retargeting Lentiviruses via SpyCatcher-SpyTag Chemistry for Gene Delivery into Specific Cell Types. mBio"
26. Kasaraneni, Chamoun-Emanuelli, Wright et al. (2018) "A simple strategy for retargeting lentiviral vectors to desired cell types via a disulfide-bondforming protein-peptide pair" *Sci Rep*
27. Lei, Joo, Wang (2009) "Engineering fusogenic molecules to achieve targeted transduction of enveloped lentiviral vectors" *J Biol Eng*
28. Yang, Bailey, Baltimore et al. (2006) "Targeting lentiviral vectors to specific cell types in vivo" *Proc Natl Acad Sci*
29. Zhang, Kutner, Bialkowska et al. (2010) "Cellspecific targeting of lentiviral vectors mediated by fusion proteins derived from Sindbis virus, vesicular stomatitis virus, or avian sarcoma/leukosis virus" *Retrovirology*
30. Lee, Dang, Joo et al. (2011) "Engineered lentiviral vectors pseudotyped with a CD4 receptor and a fusogenic protein can target cells expressing HIV-1 envelope proteins" *Virus Res*
31. Lei, Joo, Zarzar et al. (2010) "Targeting lentiviral vector to specific cell types through surface displayed single chain antibody and fusogenic molecule" *Virol J*
32. Rasbach, Abel, Munch et al. (2013) "The receptor attachment function of measles virus hemagglutinin can be replaced with an autonomous protein that binds Her2/neu while maintaining its fusion-helper function" *J Virol*
33. Lecocq, Vlaeminck, Hanssens et al. (2019) "Theranostics in immuno-oncology using nanobody derivatives" *Theranostics*
34. Behdani, Zeinali, Khanahmad et al. (2012) "Generation and characterization of a functional nanobody against the vascular endothelial growth factor receptor-2; angiogenesis cell receptor" *Mol Immunol*
35. Salmon, Oberholzer, Occhiodoro et al. (2000) "Reversible immortalization of human primary cells by lentivector-mediated transfer of specific genes" *Mol Ther*
36. Ahani, Roohvand, Mohajel et al. (2016) "Surface display of vascular endothelial growth factor receptor-2 specific nanobody on 293T cells -A potential targeting moiety for lentiviral vector-based cancer therapy" *Romanian Biotechnol Lett*
37. Taube, Zhu, Xu et al. (2008) "Lentivirus display: stable expression of human antibodies on the surface of human cells and virus particles" *PLoS ONE*
38. Arai, Ueda, Kitayama et al. (2001) "Design of the linkers which effectively separate domains of a bifunctional fusion protein" *Protein Eng*
39. Schindelin, Arganda-Carreras, Frise et al. (2012) "Fiji: an open-source platform for biological-image analysis" *Nat Methods*
40. Ward, Mcguffin, Buxton et al. (2003) "Secondary structure prediction with support vector machines" *Bioinformatics*
41. Heffernan, Dehzangi, Lyons et al. (2016) "Highly accurate sequence-based prediction of half-sphere exposures of amino acid residues in proteins" *Bioinformatics*
42. Krogh, Larsson, Von Heijne et al. (2001) "Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes" *J Mol Biol*
43. Ma, Wang, Zhao et al. (2013) "Protein Threading using context-specific alignment potential" *Bioinformatics*
44. Yang, Yan, Roy et al. (2015) "The I-TASSER suite: protein structure and function prediction" *Nat Methods*
45. Xu, Zhang (2012) "Ab initio protein structure assembly using continuous structure fragments and optimized knowledge-based force field" *Proteins*
46. Bordoli, Kiefer, Arnold et al. (2009) "Protein structure homology modeling using SWISS-MODEL workspace" *Nat Protoc*
47. Webb, Sali (2014) "Protein structure modeling with MODELLER" *Methods Mol Biol*
48. Phillips, Braun, Wang et al. (2005) "Scalable molecular dynamics with NAMD" *J Comput Chem*
49. Munch, Muhlebach, Schaser et al. (2011) "DARPins: an efficient targeting domain for lentiviral vectors" *Mol Ther*
50. Goyvaerts, De Groeve, Dingemans et al. (2012) "Development of the nanobody display technology to target lentiviral vectors to antigen-presenting cells" *Gene Ther*
51. Goyvaerts, Dingemans, Groeve et al. (2013) "Targeting of human antigen-presenting cell subsets" *J Virol*
52. Amet, Lee, Shen (2009) "Insertion of the designed helical linker led to increased expression of tf-based fusion proteins" *Pharm Res*
53. Muth (2015) "Receptor-targeted viral vectors for basic and medical research"
54. Brozzo, Bjelic, Kisko et al. (2012) "Ballmer-Hofer K. Thermodynamic and structural description of allosterically regulated VEGFR-2 dimerization" *Blood*
55. Sawano, Takahashi, Yamaguchi et al. (1996) "Flt-1 but not KDR/Flk-1 tyrosine kinase is a receptor for placenta growth factor, which is related to vascular endothelial growth factor" *Cell Growth Differ*
56. Terman, Dougher-Vermazen, Carrion et al. (1992) "Identification of the KDR tyrosine kinase as a receptor for vascular endothelial cell growth factor" *Biochem Biophys Res Commun*
57. Waltenberger, Claesson-Welsh, Siegbahn et al. (1994) "Different signal transduction properties of KDR and Flt1, two receptors for vascular endothelial growth factor" *J Biol Chem*
58. Cheng, Roffler (2008) "Membrane-tethered proteins for basic research, imaging, and therapy" *Med Res Rev*
59. Abate-Daga, Davila (2016) "CAR models: next-generation CAR modifications for enhanced T-cell function" *Mol Therapy -Oncolytics*
60. Guest, Hawkins, Kirillova et al. (2005) "The role of extracellular spacer regions in the optimal design of chimeric immune receptors: evaluation of four different ScFvs and antigens" *J Immunother*
61. Hombach, Hombach, Abken (2010) "Adoptive immunotherapy with genetically engineered T cells: modification of the IgG1 Fc /`spacer/' domain in the extracellular moiety of chimeric antigen receptors avoids /`off-target/' activation and unintended initiation of an innate immune response" *Gene Ther*
62. Hudecek, Sommermeyer, Kosasih et al. (2015) "The nonsignaling extracellular spacer domain of chimeric antigen receptors is decisive for in vivo antitumor activity" *Cancer Immunol Res*
63. Jonnalagadda, Mardiros, Urak et al. (2015) "Chimeric antigen receptors with mutated IgG4 Fc spacer avoid Fc receptor binding and improve T cell persistence and antitumor efficacy" *Mol Ther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12753864&blobtype=pdf | # Editorial: RNA regulation mechanisms in microbialhost interactions
Wenxing Li, Sanqi An, Yousef Kwaik
## Introduction
The interaction between microbes and their hosts represents a dynamic and multifaceted evolutionary arms race, with both parties continually adapting their molecular arsenals. While traditional paradigms focused predominantly on proteinmediated signaling cascades, the past decade has witnessed a fundamental shift in focus, placing RNA biology at the center of this intricate cross-talk. It is now well-established that RNA molecules-encompassing messenger RNAs (mRNAs) and a diverse array of noncoding RNAs (ncRNAs), such as microRNAs (miRNAs), circular RNAs (circRNAs), and long non-coding RNAs (lncRNAs)-serve as critical modulators of cellular responses during infection. Pathogens have evolved sophisticated mechanisms to exploit these molecules and hijack host machinery, while hosts have co-evolved RNA-based surveillance systems to trigger innate and adaptive immunity.
This Research Topic "RNA Regulation Mechanisms in Microbial-Host Interactions" was curated to systematically explore this complex regulatory landscape, highlighting novel molecular mechanisms, technological advancements, and potential therapeutic applications. This Research Topic brought together six high-quality articles that collectively advance our understanding of how RNA regulation shapes the outcomes of microbial infections, spanning bacterial and viral pathogenesis, the development of innovative detection tools, and the emergence of RNA-based therapeutics.
## Epitranscriptomics and RNA profiling in infection
A major frontier in contemporary RNA biology is epitranscriptomics-the study of biochemical modifications on RNA molecules and their functional consequences. In their original research article, Zou et al. employed cutting-edge direct RNA sequencing technology to investigate the macrophage response to Mycobacterium tuberculosis (Mtb) infection. Their comprehensive study revealed multilayered epitranscriptomic remodeling in host cells and identified specific RNA modifications that correlate with distinct infection stages and immune activation states. These findings offer new molecular targets for understanding tuberculosis pathogenesis and may inform the development of host-directed therapies.
Complementing this focus on modified coding RNAs, Miao et al. shifted their attention to the emerging class of non-coding RNAs, specifically circular RNAs (circRNAs), during viral infection. Using Vesicular Stomatitis Virus (VSV) as a model system, they provided comprehensive profiling of both host-derived and virus-encoded circRNAs. Their findings highlight the dynamic differential expression of these exceptionally stable RNA molecules and suggest that they may function as molecular sponges for miRNAs or sequester RNA-binding proteins, thereby fine-tuning the host antiviral response. This study reinforces the emerging concept that circRNAs are functional players in the host-virus interactome rather than mere splicing byproducts or transcriptional noise.
## Endogenous viral elements and host pathology
The interplay between viral sequences and host physiology extends far beyond acute infection episodes. Li et al. presented intriguing findings on the unexpected impact of endogenous viral elements (EVEs) on human health, specifically in the context of glioma pathogenesis. They demonstrated that EVEs-remnants of ancient viral infections integrated into the human genome-can profoundly influence glioma clinical phenotypes by inducing the expression of the stem cell transcription factor OCT4 in host cells. This study elegantly bridges the traditionally separate fields of virology and oncology, suggesting that these "fossilized" viral remnants in our genome are not dormant relics, but rather, they can be reactivated in specific cellular contexts to drive pathological processes. This work provides a novel perspective on the long-term evolutionary and pathological consequences of microbial-host coevolution, raising important questions about the broader roles of EVEs in human disease.
## Advanced tools for detection and mechanism decoding
Advancing this rapidly evolving field requires robust and innovative methodologies. Two articles in this Research Topic directly addressed the technological needs of the research community. Ghosh et al. provided a comprehensive and timely review of emerging RNA-centric technologies specifically designed to probe RNA-protein interactions in viral systems. Focusing on positive-sense single-stranded RNA (+ssRNA) viruses-which represent major human pathogens-the authors systematically discussed how these cutting-edge tools are essential for decoding viral life cycles and identifying novel antiviral targets. Their review highlights techniques ranging from classic crosslinking and immunoprecipitation methods to state-of-the-art single-molecule imaging approaches.
On the diagnostic and surveillance front, Que et al. introduced the "HPD-Kit," a comprehensive and user-friendly toolkit for pathogen detection and molecular analysis. In an era where the rapid, accurate, and cost-effective identification of pathogens is critical for public health preparedness and outbreak response, this toolkit represents a practical application of molecular biology principles to provide actionable diagnostic solutions. The HPD-Kit aims to streamline the pathogen detection process and improve real-time surveillance of infectious agents, with particular relevance for resource-limited settings.
## Therapeutic implications and clinical translation
Understanding the fundamental mechanisms of RNA regulation in infection ultimately aims to foster the development of novel therapeutic interventions. He et al. conducted a systematic review and comprehensive bibliometric analysis to examine the global landscape of small interfering RNA (siRNA) therapeutic development. By mapping research trends, identifying knowledge hotspots, and analyzing collaborative networks in this field, the authors illustrated how foundational research into RNA interference mechanisms is progressively being translated into clinical strategies to combat microbial infections. Their analysis confirms that RNA-based therapeutics represent a rapidly maturing field with immense potential to address the growing threat of antimicrobial resistance, offering precision medicine approaches that could c omplement or e ven replace traditional antimicrobial agents.
## Future perspectives and research directions
The articles presented in this Research Topic collectively illustrate the remarkable breadth and depth of RNA regulation in microbial-host interactions. From the epitranscriptomic remodeling of immune cells and the circRNA landscapes of viral infections to the pathological roles of endogenous viral elements, these studies reveal that RNA serves as a central architect of the host-pathogen interface.
Looking forward, several exciting avenues warrant further investigation. Future research should increasingly integrate multiomics approaches, combining epitranscriptomics with proteomics and single-cell genomics to achieve a holistic understanding of hostpathogen dynamics. Advanced spatial transcriptomics and realtime imaging technologies promise to reveal how RNA regulation varies across tissue microenvironments and infection stages. The success of mRNA vaccines against COVID-19 has demonstrated the tremendous potential of RNA-based therapeutics, encouraging the exploration of circRNAs, self-amplifying RNAs, and host-directed therapies that manipulate RNA regulatory networks to enhance immunity while minimizing resistance development. Additionally, artificial intelligence and machine learning approaches will become increasingly important for predicting RNA structures and interactions based on high-throughput data, accelerating the discovery of novel therapeutic targets. Finally, evolutionary perspectives examining how RNA regulatory mechanisms have co-evolved in pathogens and hosts will provide insights into vulnerability points for therapeutic exploitation.
## Conclusion
We hope this Research Topic serves as both a comprehensive resource and a catalyst for future investigations, encouraging researchers worldwide to delve deeper into the "RNA world" and uncover hidden regulatory mechanisms that govern the delicate balance between health and disease. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12471917&blobtype=pdf | # Hepatitis E Virus in Croatian Liver Transplant Recipients: Seroprevalence and One-Year Post-Transplant Surveillance from a Combined Cohort Study (2019-2022)
Petra Dinjar Kujundžić, Tatjana Vilibić-Čavlek, Tomislav Kelava, Alan Ayoub, Jelena Prpić, Lorena Jemeršić, Adriana Vince, Anna Mrzljak
## Abstract
Hepatitis E virus (HEV) is a significant yet understudied cause of viral hepatitis among liver transplant (LT) recipients in Central and South-eastern Europe. We conducted a combined cross-sectional and prospective cohort study to determine the seroprevalence, incidence, and risk factors of HEV infection in Croatian LT recipients. A total of 766 adult LT recipients were analyzed for anti-HEV IgG and HEV RNA. Additionally, 152 patients listed for LT were followed prospectively with sampling before and up to 12 months after transplantation. Anti-HEV IgG seroprevalence in the cross-sectional cohort was 19.8%, with no active infections detected via HEV RNA. In the prospective cohort, baseline seroprevalence was 20.4%, declining post-transplant, while 3.9% of initially seronegative patients seroconverted without detectable HEV RNA. Older age and non-tertiary level of education were associated with seropositivity, while dietary and animal contact factors were not. These findings indicate a moderately high level of prior HEV exposure among Croatian LT recipients, comparable to some European regions, but a low incidence of post-transplant infection and no evidence of chronic HEV infection. Our results suggest that despite frequent exposure, clinically significant HEV infection is uncommon in this immunosuppressed population.
## 1. Introduction
Hepatitis E virus (HEV) is currently the predominant cause of acute viral hepatitis in humans. In particular, members of the species Paslahepevirus balayani, including HEV-1 to HEV-4, are recognized as the causes of major public healthcare concerns. While G1 and G2 were initially considered endemic only in developing countries with poor sanitation and are primarily transmitted via contaminated water, genotypes G3 and G4 are identified as major culprits in sporadic cases across industrialized nations showing zoonotic potential [1]. In Europe, HEV-3 is the predominant genotype, transmitted mainly via consumption of undercooked pork or direct animal contact with infected animals [2]. Importantly, it is also recognized as the cause of chronic hepatitis in immunocompromised individuals, including organ transplant recipients [3].
Liver transplant (LT) recipients are at particular risk of HEV infection due to lifelong immunosuppressive therapy and frequent exposure to blood products and donor organs. Chronic HEV infection in these subjects can lead to rapid progression of liver fibrosis and graft loss [3]. Although European data suggest varying levels of HEV seroprevalence among transplant populations, from as low as 1-3% in Greece and the Netherlands to over 25% in particular areas of Germany and France, regional and methodological differences limit generalizability [4][5][6][7].
In Croatia, HEV infection is an emerging public health concern. Studies confirm its endemicity in domestic pigs and wild boars, indicating a permanent risk of spread to humans [8,9]. Furthermore, previous reports have shown moderate to high HEV IgG seroprevalence in several subpopulations, including blood donors (21.5%), hemodialysis patients (27.9%), and individuals with chronic liver disease (15.1%) [10]. However, data on the post-transplant population have been limited to small cohorts [11].
We conducted a hybrid study at Croatia's national referral center comprising a crosssectional survey and a prospective cohort. The cross-sectional component estimated anti-HEV IgG seroprevalence and the point prevalence of acute infection at sampling (HEV RNA-positive). The prospective component analyzed 12-month post-transplant incidence of HEV infection (RNA-positive) and anti-HEV IgG kinetics. Together, these complementary designs provide a comprehensive epidemiologic profile of HEV in Croatian LT recipients.
Accordingly, our objectives were to (1) estimate anti-HEV IgG seroprevalence; (2) assess the point prevalence of RNA-confirmed acute HEV infections at the cross-sectional visit; (3) determine the 12-month post-transplant incidence of HEV RNA-positive infections; (4) evaluate the occurrence of chronic HEV infections; (5) identify demographic, clinical, and lifestyle factors associated with HEV exposure; and (6) analyze anti-HEV IgG dynamics over time.
## 2. Methods
## 2.1. Study Design
We conducted an observational hybrid study with two complementary components within a single national center. The cross-sectional survey estimated the cumulative burden of prior exposure and assessed the point prevalence of acute HEV infection at the study visit, defined strictly as HEV RNA positivity. The prospective cohort quantified the 12-month post-transplant incidence of RNA-confirmed HEV infection and characterized anti-HEV IgG kinetics (including seroconversion) under current practice. Seroconversion without positive HEV RNA was treated as a serologic outcome and not counted as a new infection. Both components were conducted at Merkur University Hospital, Zagreb, the national reference center for liver and multivisceral transplantation, which performed >90% of liver transplants in Croatia during the study period.
## 2.2. Study Period and Participants
Eligibility. All patients aged ≥18 years who underwent liver transplantation (LT) at Merkur University Hospital were eligible; no exclusion criteria were applied.
Cross-sectional cohort: We included 766 consecutive adult LT recipients transplanted 1994-2019, recruited during routine outpatient visits January-December 2022. At the visit, each participant provided a venous blood sample and completed a single structured questionnaire on putative HEV exposures. Specimens were stored and later batch-tested for anti-HEV IgG and HEV RNA.
Prospective cohort: We enrolled 152 consecutive patients listed for LT (September 2019-December 2022) who underwent transplantation and were scheduled for blood sampling at baseline (pre-LT) and at approximately 6 and 12 months post-LT. Follow-up occurred during scheduled transplant clinic visits and was supplemented by review of clinical records; blood from each time point was stored and later tested using the same assays as in the cross-sectional component. In this cohort, inclusion required completion of both follow-up visits, yielding an incidence denominator of 152; the exposure questionnaire was administered once per participant at either the 6-or 12-month visit, according to scheduling.
Outcomes: In the cross-sectional component, outcomes were (i) anti-HEV IgG seroprevalence and (ii) the point prevalence of acute HEV infection at the study visit, defined strictly as HEV RNA positivity. In the prospective component, the primary outcome was the 12-month incidence of RNA-positive HEV infection, and secondary outcomes were anti-HEV IgG kinetics (including seroconversion). Chronic HEV infection was defined a priori as persistent HEV RNA positivity for more than three months in an immunocompromised host.
Patient characteristics and potential exposures: We recorded demographics (age, sex); transplant-related variables (calendar year of LT and time since LT at sampling); immunosuppression (monotherapy, dual, or triple therapy; calcineurin inhibitor category: tacrolimus or cyclosporine); clinical factors (etiology of liver disease and comorbidities); and lifestyle/environmental factors (consumption of pork/offal, undercooked meat, animal contact, rural residence, blood transfusions, and travel). Laboratory variables included anti-HEV IgG indices, HEV RNA results, and liver biochemistry.
## 2.3. Laboratory Testing
For each sampling, 2 mL of venous blood was drawn. The blood was centrifuged, and the serum was separated and stored at -20 • C until testing. Serological testing for anti-HEV IgG antibodies was performed using a commercial enzyme immunoassay (recomWell HEV IgG, Mikrogen GmbH, Neuried, Germany). Samples that tested reactive in the initial screening were further confirmed using an immunoblot (IB) test (recomBlot IgG anti-HEV, Mikrogen GmbH, Neuried, Germany), using highly purified recombinant HEV antigens: O2N genotype 1/3, O2C genotype 1/3, O2M genotype 1, and O3 genotype 1/3. According to the manufacturer's specifications, the IB has a diagnostic sensitivity of 96.6% and a specificity of 97.1%.
In this study, automated extraction of viral RNA was used to detect HEV infection in liver transplant recipients. Total RNA was extracted from whole blood samples using the iPrep™ PureLink™ Total RNA Kit (Invitrogen, Life Technologies, Carlsbad, CA, USA), based on magnetic separation combined with ChargeSwitch ® technology. Samples were lysed under denaturing conditions to inactivate RNases, incubated, and centrifuged. The supernatant was then processed using the iPrep™ Purification Instrument. Final RNA volume was 50 µL, stored at -80 • C. HEV RNA detection was performed via reverse transcription-polymerase chain reaction (RT-qPCR) targeting the ORF3 gene, using 3 µL RNA in a 20 µL reaction with the 1Step RT-PCR Probe ROX L Kit (highQu). Primers and probes followed the protocol by Jothikumar et al. (2006) [12]. Amplification was performed on a Stratagene Mx3005P (Agilent Technologies, Santa Clara, CA, USA). Samples with Ct < 40 were considered positive. Positive and negative controls were included in each run to validate the results. This method ensures sensitive detection of HEV RNA, essential in immunocompromised patients where serology may be unreliable.
## 2.4. Data Collection
Demographic and clinical data were collected for each participant, including age, gender, date of transplantation, geographic region of residence, underlying liver disease, and details of immunosuppressive therapy. During each visit, standard laboratory parameters were assessed, including white blood cell count, lymphocyte count, platelet count, aspartate aminotransferase (AST), alanine aminotransferase (ALT), C-reactive protein (CRP), and blood levels of calcineurin inhibitors (CNIs). These analyses were performed using validated protocols in the certified clinical laboratory.
Participants completed a questionnaire on potential HEV risk factors, including education, household size, sanitation, domestic animals, diet, home-processed meat consumption, animal exposure, blood transfusion history, and travel abroad.
## 2.5. Study Size
The study size was determined based on the number of eligible patients during the study period. For the cross-sectional cohort (n = 766), the precision for a prevalence estimate of 20% was ±3% at the 95% confidence level. For the prospective cohort (n = 150), the study could detect an incidence of 2-3% with 95% confidence intervals of 0.7-5.7% and 1.4-7.6%, respectively.
## 2.6. Statistical Analysis
Categorical variables are presented as n (%) and continuous variables as median (interquartile range; IQR). Percentages were calculated using available case denominators; missing values were not imputed. Hypothesis tests excluded missing observations, and multivariable analyses used complete-case data. Group comparisons used Pearson's χ 2 (or Fisher's exact when any expected cell <5) and the Mann-Whitney U test. Overall anti-HEV IgG seroprevalence was estimated with 95% CIs (Wilson). Determinants of seropositivity were assessed with multivariable logistic regression; variables with p < 0.20 in preliminary bivariate comparisons were entered into the model. Assumptions were checked (logit linearity for continuous predictors; multicollinearity via VIF). The results are reported as odds ratios (Exp[B]) with 95% CIs, alongside coefficients (B), standard errors, and Wald χ 2 . Anti-HEV IgG kinetics (including seroconversion) were summarized descriptively and not treated as incident infection. All tests were two-sided (α = 0.05); analyses were performed in SPSS v19.0 (IBM Corporation, Armonk, NY, USA).
## 3. Results
## 3.1. Cross-Sectional Cohort
This cohort included 766 liver transplant recipients who underwent transplantation between 1994 and 2019. The majority of patients were male 68.6% (526/766), with a median age of 56 years (50.5-62.5) at the time of transplantation and 60 years at the time of sampling. Anti-HEV IgG antibodies were detected in 19.8% (152/766) of patients, while HEV RNA was not detected in any of the analyzed samples.
The predominant etiology of liver disease was alcohol-related liver disease (47.0% 360/766), followed by primary liver tumors (25.0%; 199/766), viral hepatitis (17.1% 131/766), and autoimmune liver diseases (9.9%; 76/766). Most patients (77.8%; 596/766) were receiving dual immunosuppressive therapy, primarily a calcineurin inhibitor (53.4%; 409/743 tacrolimus, 42.4%; 325/743 cyclosporine) in combination with an antimetabolite. Routine laboratory parameters showed stable post-transplant values.
Sociodemographic data showed that 61.2% (445/743) had secondary education, and 52.1% (387/743) lived in urban areas. Processed meat consumption was reported by 63.8% (489/766), of which 22.3% (171/766) products were homemade. Occupational animal exposure was noted in 14.8% (113/766), while 24.3% (186/766) owned a farm. Most used public water supplies (86,9%; 602/693) and public sanitation (58,2%; 430/738). Blood transfusions were reported by 65.4% (499/743), and 27.8% (213/766) reported international travel.
When comparing liver transplant recipients by anti-HEV IgG serostatus, seropositive patients were older at the time of sampling (median 61 vs. 59 years; p = 0.023). Education level also differed significantly (p = 0.017), with the lowest seroprevalence among those with tertiary education (4.6%) and higher proportions in those with secondary education or no formal education. No significant differences were observed by gender, residence, animal contact, dietary habits, animal exposure, water source, sanitation type, or recent travel abroad. Seroprevalence varied regionally, with the highest recorded in Slavonia (37.8%) and the lowest in Istria (7.4%), though these differences were not statistically significant (Table 1). Multivariable logistic regression confirmed that older age (OR = 1.03; 95% CI 1.01-1.05; p = 0.009) was independently associated with anti-HEV IgG seropositivity, while tertiary education was protective (OR = 0.33; 95% CI 0.15-0.74; p = 0.007). No significant independent associations were found for household size, vegetarian diet, occupational exposure to animals, or history of chronic hepatitis B or C (Table 2).
## 3.2. Prospective Cohort
This group included 152 subjects with end-stage liver disease listed for LT. HEV RNA was undetectable in all samples at all three time points. By the RNA-based definition, the 12-month incidence of HEV infection was 0/152 (0.0%; 95% CI 0.0-2.4%). The six post-LT seroconversions were analyzed as serologic kinetics only and were not classified as previous infection. The majority were male (66.0%; 100/152), with a median age of 59.5 years. The most common underlying cause of chronic liver disease was alcohol-related liver disease 36.8% (56/152). Tacrolimus was the most used calcineurin inhibitor (69.7% 106/152).
Demographic data showed that patients were almost equally distributed between rural and urban areas (49%; 75/152 vs. 51%; 77/152). Most participants lived in households with three to four members (59.0%; 90/152), while smaller households with fewer than two members and larger ones with more than five members each accounted for 20.0% (31/152) of the sample.
Regarding educational level, 20.0% (31/152) of participants were unskilled, while the majority had completed secondary education (59.0%; 90/152), with 9.0% (14/152) having a post-secondary degree and 11.0% (17/152) having completed university-level education.
A total of 26.0% (39/152) of patients reported owning a household farm. Consumption of processed meat products was reported by 45.0% (68/152) of participants, and 23.0% (35/152) were involved in the production of such products. Consumption of undercooked meat was documented in 15.0% (23/152) of patients, while 11.0% (16/152) reported occupational contact with animals.
Most participants obtained drinking water from a public supply system (84.0%; 128/152), while 9.0% (13/152) used well water, and 5.0% (8/152) used bottled water. Sanitary conditions also varied: 52.0% (79/152) of participants reported using public sewage systems, 46.0% (70/152) used septic tanks, and 2.0% (3/152) had pit latrines.
At baseline, 20.4% (31/152) were anti-HEV IgG-positive. At 6 and 12 months posttransplant, seroprevalence decreased to 19% (29/152) and 16.4% (25/152), respectively (Table 3).
Patients were stratified into three subgroups based on their serological status before transplantation to assess the dynamics of anti-HEV IgG antibodies over time:
1.
Seronegative patients (75.6%; 115/152)-Remained seronegative throughout follow-up.
## 2.
Persistently seropositive patients (20.4%; 31/152)-Patients were anti-HEV IgGpositive pre-transplant, with seroprevalence declining from 20.4% at baseline to 19.0% at 6 months and 16.4% at 12 months.
## 3.
IgG seroconverters (3.9%; 6/152)-These patients were initially seronegative but developed anti-HEV IgG positivity after transplantation. Four patients seroconverted within the first 6 months and two between the 6th and 12th months post-transplant.
All tested negative for HEV RNA. Among them, four were male, and three had alcoholrelated liver disease. Mildly elevated liver enzymes were noted in three patients, and one patient showed signs of mild acute cellular rejection. Immunosuppressive regimens included cyclosporine in four cases and tacrolimus in two (Table 4). [13], reinforcing that prior HEV exposure is common in the general population. Compared to other European cohorts, Croatian transplant recipients show higher seroprevalence than those in Spain (7.4%) [14], the Netherlands (2.1%) [5], and Greece (1.3%) [4] but comparable rates to Italy (19.2-33%) [15,16] and Germany (28.8%) [6]. Japan, by contrast, reports much lower prevalence (2.9%) [17], whereas southwestern France, a known hyperendemic region, records 29% [7]. These international comparisons position Croatia among countries with a moderately high HEV exposure burden in immunosuppressed patients.
Despite the elevated seroprevalence, no HEV RNA was detected in any of the 1070 posttransplant samples, regardless of collection time point. This suggests that HEV infection, while likely frequent, is mostly subclinical and self-limiting in this population, and does not progress to chronic infection. Such findings support a reassuring clinical outlook for LT recipients in this setting.
International variability in HEV seroprevalence may reflect differences in assay performance, regional dietary practices (e.g., pork consumption), sanitation standards, and HEV prevalence in animal reservoirs. In our cohort, older age was a consistent independent predictor of seropositivity, which supports the theory of cumulative lifetime exposure [7,18]. Gender was not associated with HEV exposure, mirroring previous studies [13].
Given that domestic pigs are the main HEV reservoir and that home-produced food is widely consumed in Croatia, it is notable that rural residency and consumption of homecured meat were not associated with seropositivity. Over 60% of participants consumed these products, and 20% reported avoiding undercooked meat. The absence of a significant association may be partly explained by recall bias, socially desirable reporting, or changes in dietary habits after transplantation, which could obscure true pre-transplant exposure patterns.
Drinking water source and sanitation (municipal vs. well; sewer vs. septic tank) also showed no association. On the other hand, non-tertiary level of education was significantly linked to seropositivity, likely serving as a proxy for lower health literacy, hygiene standards, and awareness [17].
HEV circulation in pigs has been documented in Croatia since 2010 [8,9], with animal seroprevalence reaching over 60%, especially during peak shedding periods around the third month of life [19]. Nevertheless, pig exposure variables did not correlate with human seropositivity, which may suggest that other routes of transmission, possibly blood transfusion or environmental contamination-merit further investigation.
Assay performance remains a critical issue in comparing HEV seroprevalence data across studies. Commercial ELISA kits vary widely in sensitivity and specificity, often producing divergent results. In the UK, estimates ranged from 3.6% to 16.2% depending on the test used, with sensitivity from 56% to 98% [20]. Our study employed the RecomWell ELISA (Mikrogen), validated in Croatian donor studies, which consistently shows seroprevalence between 17.8% and 20% [13].
We tracked anti-HEV IgG seropositivity at three intervals: pre-transplant (20.3%), six months post-transplant (19%), and twelve months post-transplant (16.4%). The gradual decline may be explained by waning immunity due to immunosuppressive therapy [21], and by the limited sensitivity of serological assays in immunocompromised patients [22], rather than true loss of immunity.
A small proportion (3.9%) of initially seronegative LT recipients seroconverted during follow-up without concomitant HEV RNA detection. These events cannot be classified as confirmed previous infections because (i) RNA negativity and the absence of anti-HEV IgM testing preclude establishing timing/acuity; (ii) passive transfer of anti-HEV antibodies via transfused blood products remains plausible post-transplant; (iii) donor HEV serology/RNA were unavailable; and (iv) given infrequent sampling and batch testing, transient viremia cannot be excluded. Although this proportion exceeds some prior reports in transplant populations (1-2.1%) [5,23] and aligns with observations from southwestern France [7], we interpret these as serologic kinetics of uncertain clinical significance, rather than definitive evidence of new HEV infection.
Interestingly, no cases of chronic HEV infection were identified. Immunosuppressive regimens may influence viral persistence-tacrolimus, for instance, has been linked to increased HEV replication and chronic infection risk [24], while cyclosporine, the predominant immunosuppressant used in this cohort, may limit viral replication (at least in hepatitis C) [25]. However, its impact on HEV remains unclear and warrants dedicated study.
This study has several limitations. The primarily cross-sectional design and routine sampling schedule may have missed transient or asymptomatic infections. In the prospective cohort, samples were collected at only three time points, stored, and tested at the end of this study rather than in real time, which may have resulted in missed short-lived viremia, delayed seroconversion, or imprecise timing of infection.
Anti-HEV IgM testing was not performed, and donor HEV serology/RNA were not assessed, limiting confirmation of acuity and evaluation of donor-derived transmission. Assay-related limitations, including lower sensitivity in immunosuppressed patients, may have led to underestimation of true prevalence; although RNA testing was performed at each time point, very brief periods of detectable viremia could still have been missed. Accordingly, IgG seroconversion without concomitant viremia cannot be interpreted as confirmed infection and may reflect passive antibody transfer from transfusions. We did not present univariable screens or apply multiplicity corrections; primary inference relied on a prespecified multivariable model (ORs, 95% CIs), acknowledging a residual risk of chance associations. Finally, the 12-month follow-up and rarity of events may have missed late-onset chronic infection and limited power for adjusted incidence analyses, and the single-center design may constrain generalizability.
## 5. Conclusions
Our study highlights that prior exposure to HEV is frequent among LT recipients in Croatia, with a seroprevalence rate comparable to the general population. However, active infection-especially chronic HEV-is exceptionally rare in this cohort.
Key risk factors for HEV seropositivity include older age and non-tertiary level of education, suggesting the role of cumulative exposure and health literacy in infection risk. In contrast, dietary habits, environmental conditions, and direct contact with HEV animal reservoirs showed no significant association.
Importantly, the lack of chronic HEV infections, even in an immunosuppressed population, supports a reassuring clinical outlook. Nonetheless, the potential for asymptomatic and transient infections, as well as the limitations of current serological assays, reinforces the need for continued epidemiological surveillance, enhanced risk stratification tools, and evaluation of preventive measures, including vaccination and donor screening.
Future research should explore transmission pathways beyond dietary and rural exposure and clarify the role of immunosuppression in HEV dynamics.
## References
1. Goel, Aggarwal (2016) "Advances in hepatitis E-II: Epidemiology, clinical manifestations, treatment and prevention" *Expert Rev. Gastroenterol. Hepatol*
2. Colson, Borentain, Queyriaux et al. (2010) "Pig liver sausage as a source of hepatitis E virus transmission to humans" *J. Infect. Dis*
3. Kamar, Selves, Mansuy et al. (2008) "Hepatitis E virus and chronic hepatitis in organ-transplant recipients" *N. Engl. J. Med*
4. Sinakos, Gioula, Liava et al. (2018) "Prevalence of hepatitis E in liver transplant recipients in Greece" *Epidemiol. Infect*
5. Haagsma, Niesters, Van Den Berg et al. (2009) "Prevalence of hepatitis E virus infection in liver transplant recipients" *Liver Transpl*
6. Darstein, Häuser, Mittler et al. (2020) "Hepatitis E Is a Rare Finding in Liver Transplant Patients With Chronic Elevated Liver Enzymes and Biopsy-Proven Acute Rejection" *Transplant Proc*
7. Buffaz, Scholtes, Dron et al. (2014) "Hepatitis E in liver transplant recipients in the Rhône-Alpes region in France" *Eur. J. Clin. Microbiol. Infect. Dis*
8. Prpić, Černi, Škorić et al. (2015) "Distribution and Molecular Characterization of Hepatitis E virus in Domestic Animals and Wildlife in Croatia" *Food Environ. Virol*
9. Jemeršić, Prpić, Brnić et al. (2019) "Genetic diversity of hepatitis E virus (HEV) strains derived from humans, swine and wild boars in Croatia from 2010 to 2017" *BMC Infect. Dis*
10. Mrzljak, Jemersic, Savic et al. (2021) "Vilibic-Cavlek, T. Hepatitis E Virus in Croatia in the" *One-Health" Context. Pathogens*
11. Mrzljak, Dinjar-Kujundzic, Vilibic-Cavlek et al. (2019) "Hepatitis E seroprevalence and associated risk factors in Croatian liver transplant recipients" *Rev. Soc. Bras. Med. Trop*
12. Jothikumar, Cromeans, Robertson et al. (2006) "A broadly reactive one-step real-time RT-PCR assay for rapid and sensitive detection of hepatitis E virus" *J. Virol. Methods*
13. Miletić, Vuk, Hećimović et al. (2019) "Estimation of the hepatitis E assay-dependent seroprevalence among Croatian blood donors" *Transfus. Clin. Biol*
14. Riveiro-Barciela, Buti, Homs et al. (2014) "Cirrhosis, liver transplantation and HIV infection are risk factors associated with hepatitis E virus infection" *PLoS ONE*
15. Binda, Picchi, Bruni et al. (2025) "The Prevalence, Risk Factors, and Outcomes of Hepatitis E Virus Infection in Solid Organ Transplant Recipients in a Highly Endemic Area of Italy" *Viruses*
16. Zanotto, Rittà, Pittaluga et al. (2023) "Seroprevalence of hepatitis E virus in liver transplant patients in Turin, Italy" *Panminerva Med*
17. Inagaki, Oshiro, Tanaka et al. (2015) "A nationwide survey of hepatitis E virus infection and chronic hepatitis E in liver transplant recipients in Japan" *EBioMedicine*
18. (2018) "European Association for the Study of the Liver. EASL Clinical Practice Guidelines on hepatitis E virus infection" *J. Hepatol*
19. Dremsek, Joel, Baechlein et al. (2013) "Hepatitis E virus seroprevalence of domestic pigs in Germany determined by a novel in-house and two reference ELISAs" *J. Virol. Methods*
20. Bendall, Ellis, Ijaz et al. (2010) "A comparison of two commercially available anti-HEV IgG kits and a re-evaluation of anti-HEV IgG seroprevalence data in developed countries" *J. Med. Virol*
21. Zhang, Zhang, Zhou et al. (2014) "Protection against hepatitis E virus infection by naturally acquired and vaccine-induced immunity" *Clin. Microbiol. Infect*
22. Rossi-Tamisier, Moal, Gerolami et al. (2013) "Discrepancy between anti-hepatitis E virus immunoglobulin G prevalence assessed by two assays in kidney and liver transplant recipients" *J. Clin. Virol*
23. Abravanel, Lhomme, Chapuy-Regaud et al. (2014) "Hepatitis E virus reinfections in solid-organ-transplant recipients can evolve into chronic infections" *J. Infect. Dis*
24. Kamar, Garrouste, Haagsma et al. (2011) "Factors associated with chronic hepatitis in patients with hepatitis E virus infection who have received solid organ transplants" *Gastroenterology*
25. Ciesek, Steinmann, Wedemeyer et al. (2009) "Cyclosporine A inhibits hepatitis C virus nonstructural protein 2 through cyclophilin A" *Hepatology*
26. "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=PMC12326983&blobtype=pdf | # Correction for Raines et al., "Nanoparticle delivery of Tat synergizes with classical latency reversal agents to express HIV antigen targets"
Samuel Raines, Shane Falcinelli, Jackson Peterson, Ellen Van Gulck, Brigitte Allard, Jennifer Kirchherr, Jerel Vega, Isabel Najera, Daniel Boden, Nancie Archin, David Margolis
## Abstract
Acknowledgments, paragraph 1: "…(CARE)] to D.M.M." should read "…(CARE)] and NIH UM1AI169633 to D. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12836990&blobtype=pdf | # Begomoviral pre-coat protein boosts potato virus X in mixed infection through interfering with antiviral RNAi response
Supriya Chakraborty, Dibyendu Ghosh, M Malavika, Pratibha Rawat
## Abstract
Background Plant-infecting viruses cause severe crop losses throughout the world. The frequent occurrence of mixed infection (co-existence of multiple viruses in a single infected plant) in the field makes the situation more alarming. Mixed infection of two viruses often boosts the fitness of either of the viruses and thus increases disease severity. Maize lethal necrosis disease and rice tungro disease are some devastating examples. This study focused on the mixed viral infection between begomovirus, a ssDNA virus, and potato virus X (PVX), a ssRNA virus.
ResultsV2 protein encoded by a monopartite begomovirus, pepper leaf curl Bangladesh virus (PepLCBV), promotes the accumulation of PVX. When Nicotiana benthamiana plants were infected with both PVX and PepLCBV, aggravated PVX-induced symptoms and increased PVX titre were detected compared to plants infected with PVX alone. This observation suggests that PepLCBV promotes the viral titre of PVX during mixed infection. Next, when PepLCBVΔV2 infectious clone was co-infiltrated with PVX, no increase in PVX-associated symptoms was observed in the plants, which confirmed that the V2 protein plays a pivotal role in mediating the interaction between these two viruses. PepLCBV-V2 was experimentally identified as a potent suppressor of post-transcriptional gene silencing. Its interaction with suppressor of gene silencing 3 (NbSGS3) was confirmed through yeast two-hybrid and BiFC. Silencing of NbSGS3 facilitated PVX pathogenesis. Next, specific amino acids responsible for the suppressor activity of the V2 protein were identified, and the silencing suppressor mutant V2 failed to boost the PVX titre. A similar observation was recorded when the association of PVX with a bipartite begomovirus, tomato leaf curl New Delhi virus, was checked in both N. benthamiana and tomato.
ConclusionThe present study, for the first time, reports that begomoviruses promote the pathogenesis of potato virus X during mixed infection and identifies begomoviral pre-coat protein (V2/AV2) as a key player in mediating this virus-virus interaction. We further demonstrated that the silencing suppression activity of V2 is important for regulating PVX titre in mixed-infected plants.
## Background
Viruses can infect a wide range of economically important crops, vegetables and ornamental plants and are considered a global threat to agro-economy [1]. As per an estimate from 2014, plant diseases caused by viral pathogens cost almost USD30 billion a year [2]. Mixed viral infection, where more than one virus infects a plant, often makes the disease more severe [3] due to synergistic viral association, where one virus can positively influence replication, movement, tissue tropism, host range, and even transmission of another virus, leading to extensive yield losses and creating a serious threat to food security [4]. The association of maize chlorotic mottle virus (MCMV), a machlomovirus, with its potyvirus partner, such as sugarcane mosaic virus (SCMV), maize dwarf mosaic virus (MDMV), johnsongrass mosaic virus (JGMV), synergistically caused the severe pandemic of maize lethal necrosis disease (MLND). The total estimated annual loss due to MLND, in sub-Saharan East Africa alone, is about 20.5 million USD [5]. Rice tungro disease is another major pandemic that occurs in ricegrowing countries due to mixed infection between two viruses, rice tungro bacilliform virus (RTBV) and rice tungro spherical virus (RTSV). The infected plants would appear stunted, yellowing and mottling of symptomatic leaves and decreased tillers with a reduction in grain quantity and the emergence of sterile grains. Estimated yield losses were up to 90% with annual production losses of 1.5 billion USD [6].
In this study, we have demonstrated a mixed viral association of potato virus X (PVX) and begomoviruses. PVX is a positive-sense ssRNA virus belonging to the genus potexvirus of the family Alphaflexiviridae. The genomic RNA of PVX is 6.4 kb in size, with a methyl guanosine cap at the 5' end and a 3'-poly (A) tail. PVX can infect several economically important plants like potato, tobacco, tomato, pepper and can be transmitted via mechanical contact and vegetative propagation. PVX encodes only five proteins. The RNA-dependent RNA polymerase (RdRp) of size 166 kDa is the first protein to be expressed from the viral genomic RNA, while the remaining four proteins are expressed from sub-genomic RNAs (sgRNA). The movement of PVX is carried out with the help of the triple gene block (TGB) proteins. The sgRNA1 codes for the TGB1 protein (also denoted as p25 because of its size of 25 kDa) while the other TGB2 (12 kDa) and TGB3 (8 kDa) are expressed from sgRNA2. The coat protein (CP) is expressed from sgRNA3 [7]. PVXencoded TGB1 mediates 26 S proteasomal degradation of Argonaute 1 and thus acts as a viral suppressor of RNA silencing (VSR) [8].
On the other hand, begomoviruses are ssDNA viruses belonging to the largest family of plant-infecting viruses, Geminiviridae. They are transmitted by whiteflies (Bemisia tabaci) and can infect a wide range of weeds, crops and vegetables such as tomato, pepper, okra, beans, cucurbits, and radish [9]. Begomoviruses are classified into two groups based on their genome organization: monopartite begomoviruses, which consist of one ssDNA component as their genetic material, and bipartite begomoviruses, which include two genetic components, DNA-A and DNA-B, within their genome structure [10]. Begomoviruses encode a few multitasking proteins and newly identified small proteins, which mediate the infection cycle of these viruses [11,12]. While the replication of begomoviruses is mediated by the replication initiator protein (Rep/C1/AC1) and replication enhancer protein (REn/C3/AC3) [13], the trans-activator protein (TrAP/ C2/AC2) facilitates expression of some viral genes. C4/ AC4 protein acts as a potent VSR and pathogenicity determinant [14]. Virion sense strand encodes two proteins, coat protein (V1/AV1) for encapsidation and precoat protein (V2/AV2). Begomoviral pre-coat protein is a multifunctional protein. In the case of tomato yellow leaf curl virus (TYLCV), the V2 protein interacts with SGS3 and inhibits post-transcriptional gene silencing (PTGS) [15,16]. Recently, V2 encoded by cotton leaf curl Multan virus has been reported to interact with calmodulin (CAM) 3 and breach the CAM3-CAMTA3 module of PTGS [17]. In addition, it also interacts and inhibits AGO4, which is involved in the RNA-dependent DNA methylation pathway [18]. V2 can also interact with histone deacetylase 6 (HDA6) to block transcriptional gene silencing (TGS) [19]. V2 plays a key role in the movement of monopartite begomovirus [20]. However, bipartite begomoviruses encode movement protein (BC1) and nuclear shuttle protein (BV1) for intra-and inter-cellular movement [21].
In this current study, we report the association of PVX and begomovirus, where the latter promotes the accumulation of PVX in mixed infection. In addition, we have also demonstrated the role of begomoviral precoat protein in mediating the interaction between PVX and begomoviruses such as pepper leaf curl Bangladesh virus (PepLCBV) and tomato leaf curl New Delhi virus (ToLCNDV).
## Material & methods
## Plant growth conditions and virus infection
The seeds of Nicotiana benthamiana, Nicotiana tabacum cv. Xanthi, and Solanum lycopersicum cv. Pusa Early Dwarf were germinated and grown in soil mix. The germinated seedlings were transferred to individual pots and maintained at a temperature of 25 ± 2 °C in 16 h photoperiod with 60% relative humidity in an insect-proof greenhouse. Four-week-old plants were used for virus infection. Infectious clones of potato virus X (GenBank accession no. AY297843), tomato leaf curl New Delhi virus DNA-A and DNA-B (ToLCNDV) (GenBank accession no. U15015 and U15017) and pepper leaf curl Bangladesh virus (GenBank accession no. JN663864) were used to infect plants through agro-infiltration method.
## Plasmid construction
The coding sequence of V2 was amplified from Pep-LCBV DNA and initially cloned into the pJET1.2 vector (Thermo Fisher Scientific, Massachusetts, USA), followed by cloning into the pGR106 vector for developing Pep-LCBV-V2-PVX construct and into the pGADT7 vector for Yeast-two-hybrid (Y2H) assay. V2 was also cloned into pENTR/D TOPO (Invitrogen) vector, followed by into pSITE1648 and pGWB414 destination vectors for BiFC and silencing suppression study, respectively. For developing V2ΔVSR, the required mutations were introduced in primers (Table S1) and base overlapping extension PCR was performed. The mutation was confirmed by sequencing. The coding sequence of ToLCNDV encoded AV2 was amplified and initially cloned into the pJET1.2 vector, followed by cloning into the pRT101 vector and finally into the binary vector pCAMBIA2300 for developing transgenic lines. NbSGS3 was amplified from cDNA of N. benthamiana and cloned into pGBKT7 vector. NbSGS3 was also cloned into pENTR/D TOPO vector, followed by into the pSITE1651 destination vector for the BiFC study.
## Y2H assay
Y2H assay was carried out using Saccharomyces cerevisiae strain AH109 having Gal4 system according to manufacturer's protocol (CloneTech). After co-transformation of combinations of constructs, the transformed yeast colonies were allowed to grow in two drop-out (2DO) media plates [SD/-leu-trp] at 28 °C for 48 h. Random yeast colonies were selected from 2DO plate and serial dilutions were made using sterile double distilled water and dropped on 3DO (SD/-his-leu-trp) plates, 3DO plates supplemented with 1mM and 2.5mM 3-aminotriazole and allowed to grow for 48 h at 28 °C [22].
## Bi-molecular fluorescence complementation (BiFC) assay
A. tumefaciens strain EHA105 harbouring the confirmed clones of V2-pSITE1648 (V2-nYFP) and NbSGS3-pSITE1651 (SGS3-cYFP) (OD 600 of 0.6 each) were agro-infiltrated on N. benthamiana leaves. Wet-mounted epidermal cells were visualized under a confocal microscope (Leica Microsystem, Wetzlar, Germany) 72 and 96 h post-infiltration (hpi) [23]. Leaves agro-infiltrated with (i) SGS3-cYFP and pSITE1648 empty vector (nYFP), (ii) V2-nYFP and pSITE1651 empty vector (cYFP), were also visualized as negative controls. Similarly, the interaction between V2ΔVSR-nYFP and SGS3-cYFP was also checked.
## Virus-induced gene Silencing (VIGS) assay
Tobacco rattle virus (TRV) based VIGS was performed according to Singh, et al. [24]with minor modifications. Briefly, the 300 bp long specific coding sequence region of NbSGS3 was cloned into pTRV2 vector in anti-sense orientation to develop the pTRV2-NbSGS3 construct, followed by transformation into Agrobacterium tumefaciens strain GV3101. A. tumefaciens harboring either pTRV2 empty vector (OD 600 of 0.4) or pTRV2-NbSGS3 construct (OD 600 of 0.4) was agroinfiltrated along with pTRV1(OD 600 of 0.4) to perform VIGS. PDS was silenced as a positive control to detect the onset of silencing. After seven days post-infiltration, photobleaching was observed in the PDS-silenced plants. Next, both nonsilenced and NbSGS3-silenced plants were agro-infiltrated with an infectious clone of PVX (OD 600 of 0.4). After 10 days post-infiltration, two uppermost symptomatic leaves were harvested for analyzing PVX titre.
## Silencing suppression assay
For the silencing suppression assay, A. tumefaciens strain EHA105 harbouring V2 (35 S::V2, cloned in pGWB414 vector) was infiltrated along with 35 S::YFP (pSITE3CA empty vector). OD 600 of 0.5 for each construct was used during agro-infiltration. For comparison, pGWB414 empty vector was agro-infiltrated along with 35 S::YFP. As a positive control, tombusviral silencing suppressor p19 was used. Agro-infiltrated leaf was visualised under UV lamp 72 h post-infiltration. Intense yellowish green patches (indicative of more YFP accumulation) suggest the silencing suppression activity of the protein [17].
## Development of stable Transgenic tobacco plants overexpressing ToLCNDV-encoded AV2 protein
The leaf explants of N. tabacum cv. Xanthi were transformed with A. tumefaciens EHA 105 strain carrying ToLCNDV encoded AV2 protein for 15 min following the procedure described by Horsch et al. 1985 [25]. The transformed explants were co-cultivated in Murashige Skoog media supplemented with benzyl amino purine (BAP) and indole acetic acid (IAA) with their stock concentrations as 100 mg/L. The explants were incubated for 48 h in the dark at 25 °C. Next, the explants were selected in MS media supplemented with Kanamycin (50 mg/L) and Cefotaxime (200 mg/L) and kept at 16 h photoperiod at 25 °C. After root development, the transgenic tobacco plants were transferred to soil for hardening. The plants expressing the AV2 protein were later confirmed by PCR. Transgenic plants were maintained in glass house at 16 h photoperiod at 25 ± 2 °C at 60% relative humidity until T2 seeds were collected. T2 generation plants were grown on soilrite, and expression of AV2 was confirmed by sqRT-PCR.
## Isolation of genomic DNA and RNA
Genomic DNA and total RNA were isolated from upper most two systemic leaves as described previously [26]. The quality of isolated RNA was checked using a spectrophotometer (Model ND2000, Nano Drop, Thermo Fisher Scientific, Massachusetts, USA).
## Quantitative real-time (qRT) PCR, semi-quantitative reverse transcription (sqRT) PCR and quantitative PCR (qPCR)
cDNA was prepared from one µg DNase-treated RNA using reverse transcriptase (M-MuLV RT RNase H-minus, Thermo Fisher Scientific, Massachusetts USA) and oligo (dT 18 ) primer. cDNA was then diluted five times with nuclease-free water. This diluted cDNA has been used as a template for qRT-PCR and sqRT-PCR. Both qRT-PCR and sqRT-PCR were performed as mentioned previously [26]. Elongation factor 1α has been used as an internal control. For qRT-PCR, three independent biological replicates with three technical replicates have been used. qPCR was performed following the procedures described previously [26]. 25 S rRNA has been used as the internal control for normalization of qPCR data. Fold change was quantified using 2 -ΔΔCt method during the analysis of qRT-PCR and qPCR data [27]. All the primers used for these studies are mentioned in Table S1.
## Statistical analysis
Data obtained from qRT-PCR were denoted as mean ± standard deviation. Statistical significances were determined by student t-test using GraphPad Prism 8 (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001, ****P ≤ 0.0001).
## Result
## PepLCBV and PVX influence each other's pathogenesis during mixed infection
To check the interaction between PepLCBV and PVX, we infected N. benthamiana plants with (i) PVX, (ii) PepLCBV and (iii) PVX & PepLCBV, and assayed their infectivity. PVX-associated symptoms (Fig S1 ) appeared on the systemic leaves on 4th days post infection (dpi) and symptom severity got maximized (severe mosaic and necrotic spots) on 7th -9th dpi. However, the severity of PVX-associated symptoms decreased gradually on/after 10 dpi, and plants almost recovered from PVX infection at 21 dpi. Interestingly, the plants co-infected with PVX and PepLCBV, failed to recover from PVX infection even at 21 dpi (Fig S2 ,1A, Table S2). Next, we quantified PVX titre and found a significant rise (~ 3.4 fold) in the accumulation of PVX in the presence of Pep-LCBV compared to only PVX-infected plants (Fig. 1B). Surprisingly, we observed an opposite phenomenon when we examined the impact of PVX on PepLCBV accumulation. PepLCBV-associated symptoms (upward leaf curling and severe stunting) were significantly less in the plants co-infected with PVX and PepLCBV as compared to only PepLCBV infected plants (Fig. 1A, S2A, S2C, Table S3). While analyzing viral titre through qPCR, a sharp decrease (~ 85%) in PepLCBV titre was recorded in the co-infected plants as compared to only PepLCBVinfected plants (Fig. 1C). However, we were keen to decipher how the presence of PepLCBV increases PVX titre during mixed infection.
## PepLCBV boosts PVX accumulation in mixed infection through its V2 protein
Pre-coat protein (V2/AV2) is reported to play significant roles in the pathogenesis of several begomoviruses [15][16][17][18][19][20]. We previously reported the role of ToLCGVencoded V2 protein in mediating asymmetric synergism between ToLCGV and ToLCNDV [28] and highlighted the involvement of ToLCNDV-AV2 protein in inhibiting symptom recovery in N. tabacum [29]. Keeping these observations in mind, we hypothesised the involvement of PepLCBV-encoded V2 in boosting PVX pathogenesis during mixed infection. To check this, PepLCBV-V2 was over-expressed through PVX-based expression vector, pGR106, in N. benthamiana, and we observed more severe PVX-associated mosaic symptoms on V2-pGR106 infiltrated-plants as compared to only pGR106 infiltrated plants (Fig. 2A, Table S4). We confirmed the presence of V2 transcripts in these over-expressed plants through sqRT-PCR (Fig. 2B). While quantifying the PVX titre through qRT-PCR we found that PVX accumulation gets almost two-fold increase in the presence of PepLCBV-V2 (Fig. 2C). This observation suggested that V2 could promote PVX accumulation. To further confirm the involvement of V2 protein in boosting PVX infection, PepLCBVΔV2, an infectious clone of Pep-LCBV having a premature stop codon in the open reading frame of V2 was used (Fig S3). Next, we co-infected N. benthamiana plants with (i) PVX & PepLCBV, (ii) PVX & PepLCBVΔV2, and (iii) PVX and empty vector pCAMBIA2300 and compared PVX infectivity. After 21days post-infection, we observed severe PVX-associated mosaic symptoms in plants co-infected with PVX and PepLCBV, while the plants (i) infiltrated with PVX along with pCAMBIA2300 and (ii) infiltrated with PVX and PepLCBVΔV2, showed mild PVX-associated symptoms (Fig. 2D). While analyzing the PVX titre through qRT-PCR, we observed a significant rise (~ 3.3 fold) in the accumulation of PVX in the presence of PepLCBV as compared to only PVX-infected plants (Fig. 2E). Interestingly, co-infection of PepLCBVΔV2 failed to boost PVX titre (Fig. 2E). This clearly suggests that PepLCBV promotes PVX accumulation through V2.
## PepLCBV encoded V2 protein interacts with NbSGS3 and acts as a VSR
Pre-coat proteins encoded by begomoviruses such as tomato yellow leaf curl virus (TYLCV), tomato leaf curl New Delhi virus (ToLCNDV) and tomato leaf curl Gujarat virus (ToLCGV) have been reported as viral suppressors of RNAi (VSR) [15,28]. TYLCV (Israeli isolate) encoded V2 interacts with AtSGS3 and blocks PTGS [15]. To check whether PepLCBV encoded V2 targets SGS3, N. benthamiana homolog of AtSGS3 was cloned and protein-protein interaction between PepLCBV-V2 and NbSGS3 was checked through a yeast two-hybrid assay. We found that yeast cells co-transformed with V2 and NbSGS3 can easily grow on three drop-out selection media (SD-LTH) supplemented with 1mM or even 2.5mM 3AT (Fig. 3A). Next, the interaction was confirmed in planta through BiFC, and we observed reconstitution of fluorescence in V2 and NbSGS3 overexpressed epidermal cells (Fig. 3B,S4). After that, the ability of V2 in suppressing PTGS was demonstrated through silencing suppression assay. When the leaves were seen under UV-light, we observed a more intense yellowish green patch in the presence of PepLCBV-V2 as compared to the empty vector control (Fig. 3C). Next, for relative quantification of YFP transcripts, sqRT-PCR was performed, and we detected more accumulation of YFP transcripts in the presence of PepLCBV-V2 compared to the empty vector control (Fig. 3D). P19 was used as a positive control. These observations demonstrate that PepLCBV-encoded V2 protein interacts with NbSGS3 and inhibits PTGS.
## NbSGS3 negatively regulates PVX pathogenesis
SGS3 interacts with RDR6 [30] and SGS3-RDR6 complex mediates the production of secondary vsiRNA [31]. As the PepLCBV-V2 protein interacts with NbSGS3, we tested the impact of NbSGS3 on PVX pathogenesis. We opted for a TRV-based VIGS approach to silence NbSGS3 and infected both silenced and non-silenced plants with PVX. Interestingly, more PVX-associated symptoms (Fig. 4A, Table S5) and viral titre (~ 2 fold) (Fig. 4C) were recorded in NbSGS3-silenced plants as compared to non-silenced plants. Reduced accumulation of NbSGS3 transcripts was observed in the silenced plants (Fig. 4B). These observations emphasize the positive impact of NbSGS3 in regulating anti-viral response against PVX.
## Silencing suppression activity of V2 is important for boosting PVX titre
Cysteine residues at 84th and 86th positions of TYLCV-V2 were mapped to be necessary for its interaction with AtSGS3 [15]. We analysed the amino acid sequences of pre-coat proteins encoded by begomoviruses and observed that 'C 84 -X-C 86 ' motif is quite conserved among begomoviral pre-coat proteins (Fig S5A). We attempted to develop a V2 mutant that cannot interact with NbSGS3 and hence fails to act as a VSR. For this, we performed site-directed mutagenesis to replace C 84 and C 86 with S 84 and S 86 . We developed the mutant and confirmed it through sequencing (Fig S5B ,S6). We confirmed that the V2 mutant cannot interact with NbSGS3 through BiFC (Fig S5E). In addition, we also demonstrated that this mutant V2 failed to block PTGS through silencing suppression assay followed by relative ,S5D) and hence named it as V2ΔVSR. Next, to check the impact of V2ΔVSR in PVX pathogenesis, we infected N. benthamiana plants with (i) PVX, (ii) V2-PVX, and (iii) V2ΔVSR-PVX. As noticed previously, more PVX-associated symptoms (Fig. 5A) and viral titre (almost two-fold increase) were recorded in V2-PVX-infected plants as compared to only PVX-infected plants (Fig. 5C). Interestingly, V2ΔVSR-PVX-infected plants showed mild mosaic symptoms comparable to only PVX-infected plants (Fig. 5A). Similarly, PVX accumulation was significantly reduced (~ 34%) in V2ΔVSR-PVX-infected plants when compared with V2-PVX-infected plants (Fig. 5C, Table S4). The presence of V2 transcripts were detected through sqRT-PCR (Fig. 5B). This suggests that the ability of V2 in suppressing PTGS is important for the observed interaction between PVX and PepLCBV.
## ToLCNDV boosts PVX accumulation in mixed infection through its AV2 protein
We observed that a monopartite begomovirus, Pep-LCBV, promotes PVX accumulation in mixed infection. To validate whether similar phenomena occur in the case of bipartite begomoviruses or not, we took the tomato leaf curl New Delhi virus and co-infected N. benthamiana plants along with PVX (Fig. 6A,S7). An infectious ToLCNDVΔAV2 clone having a premature stop codon in the open reading frame of AV2 was also used [29]. We observed more PVX-associated symptoms (Fig. 6A) and higher accumulation of PVX in the presence of ToL-CNDV as compared with only PVX-infected plants (Fig. 6B, Table S6). However, ToLCNDVΔAV2 failed to boost PVX accumulation (Fig. 6A andB). As tomato serves as a natural host for ToLCNDV, we co-infected tomato plants with PVX and ToLCNDV. Severe mosaic symptoms (Fig. 6C) and a significant, rise in PVX titre (Fig. 6D) were recorded in tomato plants co-infected with PVX and ToLCNDV as compared to only PVX-infected tomato plants. Next, to demonstrate the effect of ToLCNDV-AV2 in PVX pathogenesis, we overexpress ToLCNDV-AV2 (Fig. 6F) along with PVX and found that, similarly to PepLCBV-V2, ToLCNDV-AV2 also promotes PVX accumulation (Fig. 6E andG). To further confirm our observation, we raised stable transgenic tobacco plants overexpressing ToLCNDV-AV2. The transgenic plants were validated by detecting AV2 transcripts through sqRT-PCR (Fig. 7B). Two confirmed independent T2 lines, namely 14b and 50b (Fig S8 ), were challenged with PVX (Table S7). After 10 days post-infection, severe mosaic symptoms were observed in these PVX-infected transgenic lines compared to PVX-infected wild-type tobacco plants (Fig. 7A). While quantifying PVX accumulation, we found around 5.6-and 3.6-fold increase in PVX titre in 14b and 50b lines, respectively, compared to wild type plants (Fig. 7C). Along with the previous observations, hyper-susceptibility of ToLCNDV-AV2 overexpressing transgenic lines to PVX infection further reiterates that ToLCNDV promotes accumulation of PVX in mixed infection through its pre-coat protein.
## Discussion
When plant viruses infect a host, the plant cell recognises the presence of foreign viral genetic material and activates the gene silencing machinery to degrade the viral genome and prevent its spread in the neighbouring cell. Dicer-like (DCLs) proteins cleave the viral dsRNA intermediates into small interfering RNAs (siRNA). Argonaute (AGOs) proteins bind to these virus-derived siRNAs and mediate the formation of the RISC complex, which finally targets complementary viral mRNAs and cleaves them. Next, RNA-dependent RNA polymerases (RDRs) further form dsRNA, leading to the generation of secondary virus-derived siRNAs (2° vsiRNA) and thus amplify the host's RNAi response [32]. SGS3 interacts with RDR6 and co-localises in SGS3/RDR6 cytoplasmic bodies [30]. This RNA-binding protein facilitates RDR6-mediated production of 2° vsiRNA with specificity. Different variants of DCLs, AGOs and RDRs have been reported to participate in anti-viral RNAi response against different plant-infecting viruses. In the case of PVX, silencing of DCL2 and DCL4 aggravates the viral pathogenesis in tomato [33]. AGO2 significantly hampers the infectivity of PVX in Arabidopsis thaliana and plays a key role in determining the host range [34,35]. However, the role of host machineries involved in 2° vsiRNA production was not characterised in the case of PVX infection. Here, we have shown that silencing of SGS3, a co-factor of RDR6, significantly promotes the accumulation of PVX in N. benthamiana (Fig. 4A andC).
Plant-infecting viruses encode effectors that interact with SGS3 and hamper the SGS3-RDR6 module of PTGS. Citrus tristeza virus encoded p20 protein, Sri Lankan cassava mosaic virus encoded AC4 protein, and turnip mosaic virus encoded VpG protein interact with SGS3 and promote its degradation through autophagy [36][37][38]. TYLCV-encoded V2 protein was also reported to interact with AtSGS3 [15]. In this study, we found that PepLCBV encoded V2 interacts with NbSGS3 and thus inhibits PTGS (Fig. 3). This observation reiterates the conserveness of V2-SGS3 interaction in different 'hostbegomovirus' pathosystems. In addition, we have also demonstrated that the ability of V2 to act as a VSR is crucial for PVX-PepLCBV interaction (Fig. 5).
The synergistic association of PVX with several potyviral partners like potato virus A (PVA), potato virus Y (PVY), tobacco etch virus, tobacco vein mottling virus, pepper mottle virus was already reported [39,40]. In this association of potyvirus-PVX, both the viral titre and symptom severity of PVX get significantly increased, and potyviral silencing suppressor HC-Pro plays the pivotal role in mediating synergism [41]. While delineating the molecular mechanism of PVX-potyviral synergism, Gonzalez-Jara et al. further demonstrated that the VSR activity of HC-Pro is directly linked to its ability to boost PVX titre during the mixed infection [42]. Recently, De et al. further added that HC-Pro also reduced the glutathione accumulation in plant cells, which helps PVX to propagate [40]. In the case of other mixed viral associations, VSRs play the key roles in promoting synergism [4]. During the co-evolution of host and virus, viruses evolved In the case of rice tungro disease, RTSV encoded polymerase and coat protein promote the VSR activity of RTBV encoded ORF-IV, leading to a synergism between RTSV and RTBV [44]. In the case of the begomovirusbetasatellite complex, betasatellite-encoded βC1 protein is a potent suppressor of both PTGS and TGS and thus promotes the accumulation of helper begomovirus [45,46]. In the case of tomato leaf curl disease in India, asymmetric synergism is observed between tomato leaf curl Gujarat virus (ToLCGV) and ToLCNDV, where ToLCGV-encoded VSRs, such as TrAP and pre-coat proteins, play pivotal roles in mediating the interaction [28]. All the above-mentioned reports clearly emphasise the impact of VSRs in regulating virus-virus interactions. In this current study, we established V2 as a central player in promoting PVX accumulation in PVX-PepLCBV mixed infection through over-expression and mutant study. However, chances of other begomoviral protein/s having an ancillary role in boosting PVX titre during mixed infection cannot be overlooked.
Another interesting finding of our study is that PVX inhibits the pathogenesis of PepLCBV in mixed infection (Fig. 1A andC, S2C). Similarly, the severity of ToLCNDV-associated leaf curling symptoms was less in N. benthamiana plants co-infected with PVX and ToL-CNDV compared to only ToLCNDV-infected plants (Table S8). These observations clearly pinpoint that PVX negatively regulates begomovirus pathogenesis in mixed infection. A similar observation was recorded in PVX-PVA infected N. benthamiana plants, where PVA positively regulates the accumulation of PVX; however, PVX inhibits the propagation of PVA [40]. We are currently looking into the mechanism behind PVX-mediated inhibition of begomoviral pathogenesis.
## Conclusion
In a nutshell, our study demonstrates that begomoviruses promote the pathogenesis of PVX during mixed infection in both N. benthamiana and tomato. We found that begomoviral pre-coat proteins (V2/AV2) play the pivotal role in mediating this begomovirus-potexvirus interaction. In addition, we have demonstrated that SGS3 negatively regulates PVX pathogenesis. Furthermore, begomoviral pre-coat protein interacts with SGS3 and inhibits the SGS3-RDR6 module of RNAi, which facilitates more PVX accumulation during mixed infection.
## References
1. Tatineni, Hein (2023) "Plant viruses of agricultural importance: current and future perspectives of virus disease management strategies" *Phytopathol-ogy®*
2. Nicaise (0660) "Crop immunity against viruses: outcomes and future challenges" *Front Plant Sci*
3. Moreno, Jj (2020) "When viruses play team sports: mixed infections in plants" *Phytopathology®*
4. Ghosh, Chakraborty (2021) "Impact of viral silencing suppressors on plant viral synergism: a global agro-economic concern" *Appl Microbiol Biotechnol*
5. Redinbaugh, Stewart (2018) "Maize Lethal Necrosis: An Emerging, Synergistic Viral Disease"
6. Jones (2021) "Global Plant Virus Disease Pandemics and Epidemics" *Plants*
7. Verchot (2022) "Potato virus X: A global potato-infecting virus and type member of thePotexvirusgenus" *Mol Plant Pathol*
8. Chiu, Chen, Baulcombe et al. (2010) "The silencing suppressor P25 of potato virus X interacts with Argonaute1 and mediates its degradation through the proteasome pathway" *Mol Plant Pathol*
9. Fiallo-Olivé, Lett, Martin et al. (2021) "ICTV virus taxonomy profile: Geminiviridae 2021" *J Gen Virol*
10. Devendran, Kumar, Ghosh et al. (2021) "Capsicum-infecting begomoviruses as global pathogens: host-virus interplay, pathogenesis, and management" *Trends Microbiol*
11. Devendran, Namgial, Reddy et al. (2022) "Insights into the multifunctional roles of geminivirus-encoded proteins in pathogenesis" *Arch Virol*
12. Gong, Tan, Zhao et al. (2021) "Geminiviruses encode additional small proteins with specific subcellular localizations and virulence function" *Nat Commun*
13. Ghosh, Chakraborty (2022) "Selective recruitment of plant DNA polymerases by geminivirus" *Trends Genet*
14. Medina-Puche, Orílio, Zerbini et al. (2021) "Small but mighty: functional landscape of the versatile geminivirus-encoded C4 protein" *PLoS Pathog*
15. Glick, Zrachya, Levy et al. (2008) "Interaction with host SGS3 is required for suppression of RNA silencing by tomato yellow leaf curl virus V2 protein" *Proc Natl Acad Sci U S A*
16. Fukunaga, Doudna (2009) "DsRNA with 5′ overhangs contributes to endogenous and antiviral RNA silencing pathways in plants" *EMBO J*
18. Wang, Gong, Wu et al. (2021) "A calmodulin-binding transcription factor links calcium signaling to antiviral RNAi defense in plants" *Cell Host Microbe*
19. Wang, Wu, Gong et al. (2019) "Geminiviral V2 protein suppresses transcriptional gene Silencing through interaction with AGO4" *J Virol*
20. Wang, Yang, Wang et al. (2018) "Tomato yellow leaf curl virus V2 interacts with host histone deacetylase 6 to suppress Methylation-Mediated transcriptional gene Silencing in plants" *J Virol*
21. Zhao, Gong, Liu et al. (2023) "Geminivirus C5 proteins mediate formation of virus complexes at plasmodesmata for viral intercellular movement" *Plant Physiol*
22. Fondong (2013) "Geminivirus protein structure and function" *Mol Plant Pathol*
23. Gnanasekaran, Ponnusamy, Chakraborty (2019) "A geminivirus betasatellite encoded βC1 protein interacts with PsbP and subverts PsbP-mediated antiviral defence in plants" *Mol Plant Pathol*
24. Sharma, Sahu, Prasad et al. (2021) "The Sw5a gene confers resistance to ToLCNDV and triggers an HR response after direct AC4 effector recognition" *Proc Natl Acad Sci U S A*
25. Singh, Ghosh, Chakraborty (2022) "Optimization of Tobacco Rattle Virus (TRV)-Based Virus-Induced Gene Silencing (VIGS) in Tomato"
26. Horsch, Fry, Hoffmann et al. (1985) "A simple and general method for transferring genes into plants" *Science*
27. Ghosh, Devendran, Sharma et al. (2025) "A glycine rich protein (NbRBGA) confers broad-spectrum tolerance to begomovirus-betasatellite complex through activation of salicylic acid signalling pathway" *Plant Physiol Biochem*
28. Livak, Schmittgen (2001) "Analysis of relative gene expression data using realtime quantitative PCR and the 2 -∆∆CT method" *Methods*
29. Basu, Singh, Singh et al. (2021) "Role of viral suppressors governing asymmetric synergism between tomato-infecting begomoviruses" *Appl Microbiol Biotechnol*
30. Basu, Kushwaha, Singh et al. (2018) "Dynamics of a geminivirus-encoded pre-coat protein and host RNA-dependent RNA polymerase 1 in regulating symptom recovery in tobacco" *J Exp Bot*
31. Kumakura, Takeda, Fujioka et al. (2009) "SGS3 and RDR6 interact and colocalize in cytoplasmic SGS3/RDR6-bodies" *FEBS Lett*
32. Li, Wang, Zhou (2017) "SGS3 cooperates with RDR6 in triggering geminivirusinduced gene silencing and in suppressing geminivirus infection in Nicotiana benthamiana" *Viruses*
33. Jin, Zhao, Guo (2021) "Recent advances in understanding plant antiviral RNAi and viral suppressors of RNAi" *Curr Opin Virol*
34. Kwon, Kasai, Maoka et al. (2020) "RNA silencingrelated genes contribute to tolerance of infection with potato virus X and Y in a susceptible tomato plant" *Virol J*
35. Jaubert, Bhattacharjee, Mello et al. (2011) "Argonaute2 mediates RNA-silencing antiviral defenses against potato virus X in Arabidopsis" *Plant Physiol*
36. Brosseau, Bolaji, Roussin-Léveillée et al. (2020) "Natural variation in the Arabidopsis AGO2 gene is associated with susceptibility to potato virus X" *New Phytol*
38. Zhang, Yang, Zhang et al. (2025) "Citrus Tristeza virus p20 suppresses antiviral RNA silencing by co-opting autophagy-related protein 8 to mediate the autophagic degradation of SGS3" *PLoS Pathog*
39. Liu, Kong, Liu et al. (2025) "Sri Lankan cassava mosaic virus Silencing suppressor AC4 mediates autophagic degradation of SGS3/RDR6 bodies in Plants" *Plant Cell Environ*
40. Cheng, Wang (2017) "The potyvirus Silencing suppressor protein VPg mediates degradation of SGS3 via ubiquitination and autophagy pathways" *J Virol*
41. Vance, Berger, Carrington et al. (1995) "Ming Shi X. 5′ proximal potyviral sequences mediate potato virus X/potyviral synergistic disease in transgenic tobacco" *Virology*
42. De, Chavez-Calvillo, Wahlsten et al. (2018) "Disruption of the methionine cycle and reduced cellular gluthathione levels underlie potex-potyvirus synergism in Nicotiana benthamiana" *Mol Plant Pathol*
43. Pruss, Ge, Shi et al. (1997) "Plant viral synergism: the potyviral genome encodes a broad-range pathogenicity enhancer that transactivates replication of heterologous viruses" *Plant Cell*
44. González-Jara, Atencio, Martínez-García et al. (0894) "A single amino acid mutation in the Plum pox virus helper componentproteinase gene abolishes both synergistic and RNA silencing suppression activities" *Phytopathology®*
45. Csorba, Kontra, Burgyán (2015) "Viral Silencing suppressors: tools forged to finetune host-pathogen coexistence" *Virology*
46. Anand, Pinninti, Tripathi et al. (2022) "Coordinated action of RTBV and RTSV proteins suppress host RNA silencing machinery" *Microorganisms*
47. Kumar, Gupta, Chakraborty (2023) "Geminiviral betasatellites: critical viral ammunition to conquer plant immunity" *Arch Virol*
48. Gnanasekaran, Kishorekumar, Bhattacharyya et al. (2019) "Multifaceted role of geminivirus associated betasatellite in pathogenesis" *Mol Plant Pathol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584166&blobtype=pdf | # Distinct memory CD4 + T cell subset tropism of two CCR5-tropic HIV-1 in a rapid progressor
Manukumar Honnayakanahalli Marichannegowda, Yasmine Farah, Meera Bose, Eric Sanders-Buell, David King, Leilani Francisco, Anne Leigh, Eller, Abdur Rashid, Sodsai Tovanabutra, Nelson Michael, Merlin Robb, Hongshuo Song
## Abstract
Background Low HIV-1 infection level in the central memory CD4 + T cell subset is a hallmark of both non-progressive HIV infection and non-pathogenetic SIV infection in the natural hosts. However, an important gap in knowledge is whether CCR5-tropic HIV-1 variants have different memory CD4 + T cell subset preferences.Case Summary Here, we identified clear compartmentalization of two CCR5-tropic HIV-1 in different memory CD4 + T cell subsets in a rapid progressor. Participant 40512 was identified in the RV217 cohort. While the transmitted/founder (T/F) virus in 40512 was compartmentalized in the central memory CD4 + T cells, the superinfecting virus was compartmentalized in the effector memory CD4 + T cells. Both viruses rely on CCR5 to infect primary CD4 + T cells. The T/F virus is more than 100-fold more resistant to the CCR5 inhibitor Maraviroc than the superinfecting virus.
ConclusionThis case report demonstrates that CCR5 HIV-1 variants have distinct memory CD4 + T cell subset preferences in vivo. Because CD4 + T cell subset targeting is highly relevant for HIV-1 pathogenesis, understanding the underlying molecular mechanisms may provide deeper insights into HIV-1 therapeutics and functional cure. KEYWORDS HIV-1, CD4 + T cell subset, CCR5, tropism, pathogenesis C CR5-tropic HIV-1 mainly infects memory CD4 + T cells. Memory CD4 + T cells are heterogeneous and comprise central memory (CM), transitional memory (TM), and effector memory (EM) subsets based on cell differentiation stages (1). These memory cell subsets differ especially in their location in the body and the migratory pheno type. Previous studies demonstrate that low HIV-1 viral burden in the CM CD4 + T cells correlates with long-term non-progressive HIV-1 infection (2, 3) and is a hallmark of non-pathogenetic SIV infection (4, 5). These findings indicate the importance of CD4 + T cell subset tropism of HIV/SIV in determining viral pathogenesis and disease progression. Of note, previous studies including ours showed that CXCR4 tropic HIV-1, which is more virulent than CCR5-tropic virus, preferentially infect the naïve and CM CD4 + T cells in vivo (6-8). However, an important, yet unanswered question is whether CCR5-tropic HIV-1 variants have different memory CD4 + T cell subset preferences, which could determine the rate of disease progression. Here, we demonstrate clear compartmentalization of two CCR5-tropic HIV-1 in different memory CD4 + T cell subsets in a rapid progressor.
last HIV negative test, thus allowing investigation of HIV evolution and host immune response from the earliest days of infection (9). Longitudinal virus sequencing demon strated that participant 40512 was initially infected by a single transmitted/founder (T/F) virus and was superinfected on day 401 (Fig. 1). Both the T/F virus and the superinfecting virus are CRF01_AE.
In our previous study on HIV-1 coreceptor switch, we investigated virus coreceptor usage for 21 RV217 Thailand participants at 2 years after HIV-1 transmission, by both virus isolation and deep sequencing (7). While participant 40512 did not harbor CXCR4 virus, this participant had the fastest CD4 decline rate among participants who harbored CCR5 virus alone (0.76 cells/µL/day vs 0.23 cells/µL/day) (7). The viral load (VL) set point in 40512 was relatively high (Fig. 2A) (detailed information on CD4 and VL dynamics for RV217 Thailand participants was reported in our previous studies [6,7]). Of note, the rapid CD4 loss in 40512 occurred before superinfection (Fig. 2A), suggesting that it was caused by the T/F virus rather than the superinfecting (SI) strain. After superinfection, the superinfecting virus was predominant in plasma, while the original T/F virus became a minor lineage (Fig. 1).
Like other RV217 participants who harbored CCR5 virus alone, participant 40512 had undetectable cell-associated HIV-1 RNA in the naïve CD4 + T cells, but cell-associated HIV-1 RNA was detected in all memory CD4 + T cell subsets as shown in our previous study (7). Sequencing of the HIV-1 RNA in each memory CD4 + T cell subset identified clear compartmentalization of the T/F virus and the superinfecting strain (Fig. 2B). While the T/F lineage was replicating in the CM and TM CD4 + T cells, the superinfecting lineage was replicating in the EM CD4 + T cells (Fig. 2B). Phylogenetic analysis showed that the superinfecting strain, which was predominant in plasma, originated in the EM CD4 + T cells, while the T/F lineage originated in the CM and TM CD4 + T cells (Fig. 2B). The compartmentalization was confirmed by two independent cell sorting and HIV-1 RNA sequencing experiments (Fig. 2B). These data demonstrate that the T/F virus and the superinfecting strain preferentially infect different memory CD4 + T cell subsets in vivo. The EM CD4 + T cells, which were preferentially infected by the superinfecting strain, released more virions into the plasma than the CM and TM CD4 + T cells. While longitudi nal PBMC samples were not available for this study, longitudinal HIV-1 sequencing using plasma samples showed that the superinfecting strain remained predominant in plasma for all subsequent time points (data not shown).
Coreceptor assay confirmed that both viruses are CCR5-tropic (Fig. 3A). The superin fecting virus could also use CCR3 with low efficiency (Fig. 3A). In NP-2 CCR5 cells, the infectivity of both viruses can be completely inhibited by 1 µM of the CCR5 inhibi tor Maraviroc. However, the T/F virus had a 146-fold higher Maraviroc IC 50 than the superinfecting strain (45.5 nM vs 0.31 nM) (Fig. 3B). These data suggest that these two viruses might use CCR5 in different ways (e.g., use different CCR5 conformations). In primary CD4 + T cells, the infectivity of both viruses can be nearly completely inhibited by 10 µM Maraviroc (Fig. 3C). Therefore, both viruses rely on CCR5 to enter primary CD4 + T cells, while the contributions of other coreceptors are very minimal, if at all present.
We next determined whether the two viruses preferentially infect different CD4 + T cell subsets in vitro. Purified CD4 + T cells were infected by pseudoviruses containing the GFP reporter. Among the cells infected by the T/F virus, 73.6% were CM, 8.5% were TM, and 17.9% were EM. Among the cells infected by the SI strain, 62% were CM, 10.6% were TM, and 27.4% were EM (Fig. 3D). This single-round infection assay showed that the T/F virus had an advantage in infecting the CM CD4 + T cells over the SI strain (P < 0.0001, chi-squared test), while the SI strain had an advantage in infecting the EM CD4 + T cells (P < 0.0001, chi-squared test). Therefore, their compartmentalization in vivo could be determined, at least in part, at the entry level. It is likely that even a small difference in virus entry ability could be amplified after multiple rounds of viral replication in vivo, consequently leading to the compartmentalization in different memory CD4 + T cell subsets as observed in participant 40512. However, because in vitro stimulation may alter the susceptibility of each CD4 + T cell subset to HIV-1 infection, the in vitro data may not accurately reflect a virus CD4 + T cell subset targeting in vivo.
## DISCUSSION
The current study shows that CCR5-tropic HIV-1 comprises diverse variants with distinct memory CD4 + T cell subset preferences. This finding has implications for better understanding HIV-1 pathogenesis and transmissibility. Multiple lines of evidence suggest that the CD4 + T cell subset targeting during HIV and SIV infection could be an important determinant for disease progression. First, low HIV-1 infection burden in the CM CD4 + T cells is a hallmark of non-pathogenetic SIV infection in the natural hosts (4,10,11). Second, in HIV-1 infection, long-term viremic non-progressors have significantly lower HIV burden in their CM CD4 + T cells than progressors (2). Furthermore, CXCR4 tropic HIV-1, which is in general more pathogenic than CCR5 virus, has replication advantage in the CM CD4 + T cells in vivo in comparison to co-existing CCR5 virus (7). Because the CM CD4 + T cells primarily locate in the lymph nodes and are critical for CD4 homeostasis, an important question to address in the future is whether CCR5 variants preferentially targeting the CM CD4 + T cells can cause faster disease progression (e.g., the highly virulent subtype B HIV-1 variant recently identified in the Netherlands [12]), and whether long-term viremic non-progressors were infected by viruses with low ability to infect the CM CD4 + T cells. Regarding HIV-1 transmissibility, because the EM CD4 + T cells are mainly distributed in the mucosal sites where most HIV-1 transmissions occur, whether CCR5 variants preferentially infecting the EM CD4 + T cells have higher transmis sibility requires further study.
Another important question to address in the future is the molecular mechanisms for the distinct memory CD4 + T cell subset tropism of the CCR5 HIV-1. While previous studies indicate that CCR5 molecules exist as different conformational states on the cell surface (13), it remains unknown whether different memory CD4 + T cell subsets express CCR5 as different conformational forms, which is responsible for their unequal susceptibility to different CCR5 variants. The in vitro experiment using pseudovirus showed different CD4 + T cell subset preferences at the entry level. Several reasons might explain why both the T/F and the SI virus could infect all memory CD4 + T cell subsets in vitro. First, in vitro stimulation may alter the phenotype of each cell subset. For example, the CM CD4 + T cells, which are at resting state in vivo, may become more susceptible to HIV-1 infection after in vitro stimulation. Second, a modest difference in virus entry ability, as observed in the single-round infection assay, will be amplified after multiple rounds of viral replication cycles in vivo, leading to a clear compartmentalization as observed in participant 40512. This possibility needs to be further determined by virus competition assay using infectious molecular clones. Additionally, potential post-entry mechanisms could also exist. In summary, characterization of the underlying molecular mechanisms is expected to provide deeper insights into HIV-1 prevention, treatment, and functional cure.
## References
1. Pepper, Jenkins (2011) "Origins of CD4 + effector and central memory T cells" *Nat Immunol*
2. Klatt, Bosinger, Peck et al. (2014) "Limited HIV infection of central memory and stem cell memory CD4+ T cells is associated with lack of progression in viremic individuals" *PLoS Pathog*
3. Descours, Avettand-Fenoel, Blanc et al. (2012) "Immune responses driven by protective human leukocyte antigen alleles from long-term nonprogressors are associated with low HIV reservoir in central memory CD4 T cells" *Clin Infect Dis*
4. Paiardini, Cervasi, Reyes-Aviles et al. (2011) "Low levels of SIV infection in sooty mangabey central memory CD4 + T cells are associated with limited CCR5 expression" *Nat Med*
5. Letvin, Mascola, Sun et al. (2006) "Preserved CD4+ central memory T cells and survival in vaccinated SIV-challenged monkeys" *Science*
6. Marichannegowda, Setua, Bose et al. (2024) "Transmission of highly virulent CXCR4 tropic HIV-1 through the mucosal route in an individual with a wild-type CCR5 genotype" *EBioMedicine*
7. Marichannegowda, Zemil, Wieczorek et al. (2023) "Tracking coreceptor switch of the transmitted/founder HIV-1 identifies coevolution of HIV-1 antigenicity, coreceptor usage and CD4 subset targeting: the RV217 acute infection cohort study" *EBioMedicine*
8. Roche, Tumpach, Symons et al. (2020) "CXCR4-Using HIV strains predominate in naive and central memory CD4 + T cells in people living with HIV on antiretroviral therapy: implications for how latency is established and maintained" *J Virol*
9. Robb, Eller, Kibuuka et al. (2016) "Prospective study of acute HIV-1 infection in adults in east Africa and Thailand" *N Engl J Med*
10. Brenchley, Silvestri, Douek (2010) "Nonprogressive and progressive primate immunodeficiency lentivirus infections" *Immunity*
11. Chahroudi, Bosinger, Vanderford et al. (2012) "Natural SIV hosts: showing AIDS the door" *Science*
12. Wymant, Bezemer, Blanquart et al. (2022) "A highly virulent variant of HIV-1 circulating in the Netherlands" *Science*
13. Lee, Sharron, Blanpain et al. (1999) "Epitope mapping of CCR5 reveals multiple conformational states and distinct but overlapping structures involved in chemokine and coreceptor function" *J Biol Chem* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12780830&blobtype=pdf | # Impact of a molecular syndromic panel on Clostridioides difficile detection and clinical interpretation
Nancy Matic, Shayan Mph, Jennifer Bilawka, Leah Gowland, Willson Jang, Colin Pharmd, Victor Leung, Michael Payne, Aleksandra Stefanovic, Christopher Lowe, Marc Romney
## Abstract
After implementation of a molecular syndromic panel for infectious diarrhea, a significantly greater proportion of C. difficile results were classified as colonization rather than infection compared to the pre-implementation period. Routine C. difficile reporting from multiplex panels should be re-evaluated to minimize diagnostic uncertainty in some patients.
## Introduction
Diagnosis of Clostridioides difficile infection (CDI) is challenging, with multiple testing algorithms proposed 1,2 . With increased availability of molecular syndromic panels in clinical laboratories, many patients without typical risk factors are routinely tested for C. difficile using multiplex assays, adding further complexity to interpretation of results. Some groups have suggested suppressing C. difficile results from panels 3 ; performing additional confirmatory testing with enzyme immunoassays (EIAs) or secondary molecular tests 4,5 ; or, sending panel results for clinical review. Furthermore, the impact of panels on C. difficile positivity rates and clinical outcomes has not been fully characterized.
Our laboratory implemented an infectious diarrhea panel (IDP) in late 2023. This multiplex panel is performed on all stool samples submitted from outpatients and inpatients (<72 hours of admission) presenting with acute gastroenteritis. To assess the impact of IDP on CDI rates, we investigated C. difficile laboratory results and clinical interpretation for a 6-month period pre-and post-IDP implementation.
## Methods
Prior to IDP implementation, patients presenting to two acute tertiary care hospitals (St. Paul's Hospital and Mount Saint Joseph Hospital, Vancouver, BC) and nearby long-term care sites and outpatient clinics with clinical suspicion for CDI had stool samples submitted for dedicated two-step C. difficile testing. Stool samples first underwent molecular detection of the tcdB gene (Xpert® C. difficile/Epi, Cepheid). Positive samples underwent further analysis by enzyme immunoassay (EIA) for direct detection of C. difficile toxin A/B and glutamate dehydrogenase (GDH) antigen (C. Diff QuikChek Complete, Techlab). If toxin was not detected by EIA, final results were reported as "Indeterminate" rather than "Positive." Clinical review of each inpatient case (including patients admitted from the Emergency Department [ED]) was conducted by the hospital's antimicrobial stewardship program (ASP), determining whether patients were infected (new-onset ≥3 loose stools in 24 hours without an alternate diagnosis) or colonized (alternate diagnosis identified based on clinical review of stool charts, medications [e.g., laxatives or other medications associated with diarrhea], laboratory results, underlying conditions, and final discussion with the patient's attending physician) 6,7 .
After implementation of IDP, dedicated two-step C. difficile testing remained available for inpatients when clinically suspected; however, all patients presenting with acute gastroenteritis for any reason had stool samples tested by IDP (BioFire® FilmArray® Gastrointestinal [GI] Panel, bioMérieux), replacing traditional stool bacterial culture and ova & parasite examination. Samples with C. difficile incidentally detected by IDP underwent further testing by EIA. Clinical review by infection prevention and control (IPAC) and ASP was conducted for all inpatient cases, as previously described 6,7 .
Results were retrospectively reviewed for a 6-month period after implementation of IDP (February-July 2024). Rates of positivity and colonization versus CDI were compared to the same 6-month period before IDP implementation (February-July 2023). Fisher's exact test (GraphPad QuickCalcs) was used for statistical analysis where applicable, with p < 0.05 considered significant.
## Results
In the post-IDP period, C. difficile was the most frequently detected pathogen on IDP in our patient population (10.8% detection rate). A higher number of stool samples underwent C. difficile testing compared to the pre-IDP period (1,661 vs 1,049), with a greater proportion submitted from ED and outpatients with a younger median age (Table 1). Of note, the number of orders for dedicated C. difficile testing decreased by nearly half compared to the pre-IDP period. Of samples testing positive for C. difficile by IDP, a significantly higher proportion (27%) tested negative for both GDH and toxin EIA compared to dedicated C. difficile testing in the same post-IDP period (6%, p < 0.001) and the pre-IDP period (11%, p < 0.001).
Clinical review of inpatient cases revealed a significantly greater proportion of patients tested by IDP was interpreted as "Colonized" compared to patients in the pre-IDP period (46.9% vs 27.5%, p = 0.01). Even when comparing to patients who had dedicated C. difficile tests performed within the same post-IDP period, colonization rates were higher among patients tested by IDP only, although this difference did not quite reach statistical significance (46.9% vs 37.3%, p = 0.30). Clinical outcomes including critical care admission, surgical intervention, and 30day all-cause mortality did not significantly differ between the preand post-IDP periods (Table 2).
## Discussion
Molecular syndromic panels have several advantages including improved efficiency and turnaround time; however, routine testing for C. difficile regardless of patients' pre-test probability (or prevalence of the condition in the population being tested) may not be optimal. Patients with CDI typically have risk factors and clinical presentations that differ from those with foodborne or community-acquired infectious diarrhea, and molecular assays for C. difficile toxin genes may be detecting asymptomatic carriers of C. difficile rather than those with CDI 8 . Our study demonstrates two different patient populations being tested for C. difficile in the pre-and post-IDP periods, with an impact on clinical interpretation of the results.
The detection rate of C. difficile by IDP in our study was similar to what has been previously described in other centres using the BioFire GI panel (9.7-16.3%) [3][4][5] . The majority of these samples tested negative for toxin EIA in our study, consistent with previous reports (57-78%) 4,9,10 . Our laboratory previously observed negative EIA toxin in 60% of samples when using a laboratorydeveloped test 7 and 67% when using Xpert 6 as the initial molecular assay in a two-step algorithm, which increased to 73% using IDP in this study. A key finding was the significant increase in the proportion of samples testing negative for both GDH and toxin EIA compared to our pre-IDP period. No samples with both negative GDH and toxin EIA were observed in our centre's previous study using a laboratory-developed assay as the initial molecular test 7 . Clinical review was conducted on only a subset of the cases (inpatients), which demonstrated a higher proportion being interpreted as "Colonization" when tested by IDP. This increase may be driven by the higher rate of GDH and toxin EIA negative samples in this cohort, and also potentially reflects the lower CDI pretest probability in patients undergoing IDP. The significant decrease in dedicated C. difficile orders during the post-IDP period suggests many clinicians ordered IDP instead of dedicated C. difficile testing. This may be concerning from an ASP perspective, as previous studies have demonstrated patients with positive C. difficile results by a molecular assay are likely to receive treatment regardless of their pretest probability or EIA results 10 . Unfortunately, IPAC and ASP surveillance teams in our centre are not able to review C. difficile results from outpatients and those discharged from ED; it is unclear how clinicians in the community may be interpreting and managing indeterminate C. difficile IDP results.
This study has additional limitations, including potential confounding factors during the pre-and post-IDP periods that may have affected C. difficile positivity rates and clinical interpretation; however, this study design was necessary to evaluate real-world data after implementation of a new testing method. Additional outcomes of interest including antibiotic usage and symptom resolution are not routinely collected by IPAC and ASP teams and were not available for analysis. Sample size was limited due to C. difficile positivity rates in our patient population.
With a two-step algorithm in use, the IDP did not significantly alter C. difficile "Positive" and "Indeterminate" rates, although a greater proportion of inpatient cases was interpreted as colonization after clinical review. To prevent potential over-treatment of C. difficile IDP results, it would be important to continue the two-step algorithm and clinical review, and consider suppression of C. difficile results from molecular syndromic panels in populations where clinical consultation is not available or for which colonization rates are high.
## References
1. Mcdonald, Gerding, Johnson (2018) "Clinical practice guidelines for Clostridium difficile infection in adults and children: 2017 update by the infectious diseases society of America (IDSA) and society for healthcare epidemiology of America (SHEA)" *Clin Infect Dis*
2. Gateau, Couturier, Coia et al. (2018) "How to: diagnose infection caused by Clostridium difficile" *Clin Microbiol Infect*
3. Park, Hitchcock, Gomez et al. (2017) "Is follow-up testing with the filmarray gastrointestinal multiplex PCR panel necessary?" *J Clin Microbiol*
4. Murphy, Fowler, Iwen et al. (2017) "Evaluation of the biofire FilmArray® gastrointestinal panel in a midwestern academic hospital" *Eur J Clin Microbiol Infect Dis*
5. O'neal, Murray, Dash et al. (2020) "Evaluating appropriateness and diagnostic stewardship opportunities of multiplex polymerase chain reaction gastrointestinal testing within a hospital system" *Ther Adv Infect Dis*
6. Lowe, Shakeraneh, Lee et al. (2022) "Optimizing the interpretation of Clostridioides difficile two-step diagnostic algorithm results through antimicrobial stewardship" *Antimicrob Steward Healthc Epidemiol*
7. Zou, Leung, Champagne (2018) "Clinical heterogeneity of patients with stool samples testing PCRþ/Tox-from a two-step Clostridium difficile diagnostic algorithm" *Eur J Clin Microbiol Infect Dis*
8. Polage, Gyorke, Kennedy (2015) "Overdiagnosis of clostridium difficile infection in the molecular test era" *JAMA Intern Med*
9. Wadskier, Cowman, Szymczak (2020) "Diagnostic stewardship of Clostridioides difficile polymerase chain reaction results from syndromic diarrhea panel and implications for patient outcomes" *Diagn Microbiol Infect Dis*
10. Pender, Throneberry, Grisel et al. (2023) "Syndromic panel testing among patients with infectious diarrhea: the challenge of interpreting Clostridioides difficile positivity on a multiplex molecular panel" *Open Forum Infect Dis* |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.